Filename
stringlengths 22
64
| Paragraph
stringlengths 8
5.57k
|
|---|---|
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | All the PNe should have thermal spectra in the radio continuum and the difference between flux densities of an optically thin PN at these two frequencies is about 5% (after allowance for the Gaunt factor). If the PN are optically thin, then this change is smaller than the uncertainties in the radio flux determinations. If the nebulae were fully optically thick in the radio domain (S ∝ να), then the flux ratios between 1.4 GHz and 843 MHz would be 2.8. Taking even this most extreme case, the range of values for either one of the frequencies is more than any global change in the ratios due to any mild optical depth effects. Furthermore, the significant difference between the present PN sample and the previous estimate using bright PNe from the Acker et al. (1992) database remains. For the present analysis, we do not distinguish between results from these two surveys. The reason for the small number of radio detections found partly reflects the confusion from high source density along the Galactic Plane, but is also due to the intrinsic radio weakness of the MASH PNe. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Ten PNe detected by MSX at 8.3 µm also have radio fluxes. Fifteen detected by IRAC at 8.0 µm also have radio fluxes. We have examined the relationships between these two pairs of flux densities. The regression lines for MSX and radio fluxes have a roughly lin- ear proportionality with an unweighted mean of 6±1.4 and a median ratio of 5±1.5. The corresponding regression between IRAC and radio fluxes has a mean of 9±2 and a median of 5±2.5. Given the factor of 0.74 to be applied to IRAC at 8.0 µm to correct its diffuse calibration, the IRAC mean and median would become 6.5±1.5 and 4±2, in the validated MSX calibration basis. Within their uncertainties, the MSX/radio and IRAC/radio ratios agree at the 1σ level. Combining them with inverse-variance weighting yields an overall MIR/radio ratio of 4.6±1.2 for the MASH PNe. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Cohen & Green (2001) took a set of 21 PNe from the Acker et al. (1992) catalog and derived a median ratio of MSX/radio of 12. This is half the median value of 25±5 they obtained for Hii regions, which has been confirmed by more detailed analysis (Cohen et al. 2007). The values for the MASH PNe are more than a factor of 2 smaller than these selected known PNe from Acker et al. To find an explanation, firstly the MSX/radio flux ratio and PN diameters for the 21 previously known PNe, together with our 10 MASH nebulae, were compared. No trends are seen and the plotted points for the MASH PNe overlap the distribution for the sample taken from Acker et al. (1992). Then we examined the two populations to see whether the samples were comparable in angular size. The 21 PNe have a median diameter of 12±5′′ while the 10 MASH PNe for which we have MSX and radio detections have a median of 20±3′′. As reported by Parker et al. (2006), the typical MASH PN is significantly larger (and hence generally more evolved) than those previously listed in the previously known PN compilations (the average diameter of all 905 MASH PNe is 51′′ compared with <10′′ for the ∼1500 previously known PNe in Acker et al. (1992)). |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | are past the ultra-compact phase, PN evolution is marked by a progressively diminishing It is not MIR/radio ratio as nebulae expand, their stars cool, and their PDRs dwindle. possible to probe this evolution at the level of individual PNe using the present data, but the ensemble averages provide adequate evidence of this phenomenon. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | For 17 of the 58 PNe there is simply no hint of either a MIR central star nor of diffuse MIR emission associated with the Hα object. The remaining 41 PNe (71%) are extended IR objects with a MIR morphology that either matches or complements that in Hα. The most frequent type is that of a PDR in which 8.0-µm emission occurs around parts of the nebular rim but is displaced to the outside of the ionized gas. In 11 of the 41 PNe two or three of the shorter IRAC bands appear to follow the distribution of Hα emission across the PN. In 5 of the 41 nebulae with MIR extension, all four IRAC bands trace the ionized gas. We have noted the possible presence of a candidate central star in the IRAC images for 11 of the 41 PNe. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Forty-one PNe of our sample of 58 MASH PNe observed by GLIMPSE have MIR coun- terparts. For many of these, it is not possible to extract meaningful quantitative spatially- integrated fluxes because of the complex structured background in the MIR. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | The ratio of integrated diffuse 8-µm and radio fluxes is a discriminator between thermal and nonthermal emission regions (Cohen & Green 2001). A ratio of MIR/radio fluxes of about 25 implies thermal emission, while very small values around 0.06 indicate nonthermal processes (Cohen et al. 2007). PNe are thermal emitters and we have compared the MASH sample with the set of bright PNe for which a ratio of 12 was calculated (Cohen & Green 2001). The difference in evolutionary state between a typical MASH PN and a nebula drawn from old catalogs causes the median MIR/radio ratio for our sample of MASH PNe to be about 5. Evolved PNe are of lower density, lower UV optical depth, and have only weak or negligible PDRs, reducing the contribution of PAH emission in bands near 8 µm, and hence their MIR detectability with IRAC. The recognition of this evolutionary trend in MIR properties emphasizes the importance of the MASH catalog of newly-discovered PNe. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | The fraction of bipolar nebulae within our subset of 58 MASH PNe is 28%, more than twice the fraction found for the entire MASH catalog. This reflects the low scale heights and short lives of the higher-mass stars that are the progenitors of Type I PNe. These are preferentially observed by the GLIMPSE survey due to its restriction to Galactic latitudes of only ±1◦. The MIR emission in Type I nebulae is likely to come from the warm dust in the associated circumstellar disks. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | False-color IRAC imagery (encoding bands at 4.5, 5.8, 8.0 µm as blue, green, and red, respectively) reveals MIR counterparts of PNe that are often clearly distinct from the color of their surroundings. The three types of false color found in our survey appear to differentiate between the degree of excitation of the optical nebular spectra as determined by [Nii]/Hα. The most frequently encountered false color is violet, which represents the weakest optical excitation. This probably corresponds most frequently to PNe that contain only H recom- bination lines. By combining false color MIR information with the relationship between MIR and Hα morphologies of PNe (i.e. whether MIR emission in IRAC bands mimics or is found outside the Hα distribution), one could assess the excitation of the nebula and decide whether fine-structure lines produce the emission in the IRAC bands. For example, the [Arii] and [Ariii] lines at 6.99 and 8.99 µm, respectively, would contribute to IRAC’s 8.0-µm band as would the 7.64-µm [NeVI] in very high-excitation PNe. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | The ratio of IRAC and MSX flux densities in PNe confirms that a multiplicative cor- rection factor of ∼0.74 should be applied to IRAC 8.0-µm diffuse emission fluxes to match the absolutely validated calibration of MSX. These MASH PNe probe the diffuse calibration on a spatial scale up to 1.3′, complementing our work on Hii regions that explored IRAC calibration up to a scale of 24′ (Cohen et al. 2007). This factor is already recommended by the Spitzer Science Center, and has been derived from knowledge of the instrument and from elliptical galaxies. We independently confirm this value for the 8.0-µm correction factor on a spatial scale between slightly resolved PNe and well-resolved Hii regions. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Legacy Science Program, was provided by NASA through contracts 1224653 (Univ. of Wisconsin, Madison), 1224988 (Space Science Institute), 1259516 (UC Berkeley), with the Jet Propulsion Laboratory, California Institute of Technology under NASA contract 1407. RI acknowledges his funding as a Spitzer Fellow. This work made use of data products from the Midcourse Space eXperiment. Processing of the data was funded by the Ballistic Missile Defense Organization with additional support from NASA’s Office of Space Science. This research has also made use of the NASA/IPAC Infrared Science Archive, which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. This research made use of Montage, funded by the National Aeronautics and Space Administration’s Earth Science Technology Office, Computational Technnologies Project, under Cooperative Agreement Number NCC5- 626 between NASA and the California Institute of Technology. This research made use of SAOImage ds9, developed by Smithsonian Astrophysical Observatory. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Frew D.J. & Parker Q.A., 2005, in Szczerba R., Stasinska G., Gorny S., eds, AIP Conf. Proc. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 1.— Left: MIR false color morphology of the IRAC counterpart to PHR1517−5751 (blue, green, and red represent emission at 4.5, 5.8, and 8.0 µm, respectively). Image is 4×4′. The object consists of an almost complete ring which has the appearance of an Hii region rather than a PN. Center of the small circle (12′′ radius) is the MASH centroid of the nebula. Right: Hα image of the same PN covering the same area as shown in the IRAC image, overlaid by white contours of 8.0-µm emission at levels of 50, 70, 100, 200, 300 MJy sr−1. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 2.— Red spectrum of PHR1517−5751 showing the weakness of the [Nii] lines relative to Hα that define this object as an Hii region. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 4.— Optical spectra of PHR1253−6350. Top: blue spectrum showing [Feii] lines at 4575 and 4665˚A, together with nebular lines of Hβ and [Oiii] on an obvious continuum. Bottom: red spectrum showing a faint red continuum with faint [Feii] lines at 6474 and 7094˚A; Hei lines (5876, 6678, 7065˚A); Hα; the red nebular lines of [Nii], [Sii], and [Ariii] at 7136˚A. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 5.— MIR image at 8.0 µm of a probable sinuous PDR showing the location of the PN candidate, PHR1841−0503, just off the ridge. The cross has 9′′ arms; the large circle has a 72′′ radius). |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 6.— A pair of 8.0-µm IRAC images of the true PNe PHR1246−6324 and PHR1457−5812 (top), compared with their respective Hα images (below). Dimensions of the small cross and large circle as in Fig. 5. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 7.— False color images of PNe from GLIMPSE mosaics (not residual images). PHR designations appear below each PN image. In the bottom right corner is the field of PHR1843−0232, illustrating graphically the extreme difficulty of extracting any object in the green circle against the very bright and variable 8.0-µm sky background (red) emission of PAHs. All images are of 4 arcmin square fields and any small green circle that guides the eye to the location of a PN’s optical centroid has a 12′′ radius. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 9.— Quartet of IRAC bands overlaid on Hα for PHR1246−6324. The MIR images suggest dominantly thermal emission by warm dust in a tilted disk, although the greater extent at 8.0 µm might also signify a PDR. Contour levels at 3.6, 4.5, 6.8, 8.0 µm are: 2, 3, 4, 5, 6, 8, 10, 20, 30 MJy sr−1; 1.5, 2, 3, 4, 5, 6, 8, 10, 30; 9, 11, 13, 15, 20, 25, 30; and 25, 30, 40, 55, 65, respectively. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 10.— Quartet of IRAC bands overlaid on Hα for PHR1457−5812. We suggest the MIR structure is due to H2 line emission. Contour levels at 3.6, 4.5, 5.8, 8.0 µm are: 4, 5, 6, 8, 10, 15, 40, 60, 80 MJy sr−1; 4, 5, 6, 8, 10, 15, 30, 50; 10, 11, 12, 15, 20, 30, 40; and 32, 36, 38, 41, 44, 48, respectively. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 11.— Quartet of IRAC bands overlaid on Hα for PHR1408−6229. This large high- excitation bipolar PN shows a diffuse waist terminated to east and west by bright MIR PDRs. Contour levels at 3.6, 4.5, 6.8, 8.0 µm are: 4, 6, 7, 10, 20, 70 MJy sr−1; 2.5, 4, 7, 20, 50; 9, 10, 25, 50; and 20, 21, 22, 30, 50, respectively. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 12.— White contours of 8.0-µm emission (20, 40, 100 MJy sr−1) overlaid on the gresycale Hα image of PHR1244−6231. The horizontal line of 8.0-µm patches near 0.317 longitude is an artifact due to a very bright star outside the field presented, to the south-east. . |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 13.— Comparison of spatially integrated IRAC 8.0 and MSX 8.3-µm fluxes for PNe. The solid line is the regression based on errors in both variables. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 14.— Ratio of IRAC to MSX fluxes for PNe plotted against diameters of the nebulae. Solid line is the regression line flanked (dash-dotted) by the ±1σ extreme fits. Error bars are shown both for the ratios and for diameters. Note the slope of the regression line that suggests the correctness of the IRAC calibration for a point source, and the steady increase in ratio to about 1.5 with increasing nebula size. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Fig. 15.— Diagnostic IRAC color-color plot ([3.6]-[4.5] vs. [5.8]-[8.0]) comprising 87 types of IR point source. Key: pluses - normal dwarfs, giants, supergiants; asterisks - AGB M stars; diamonds - AGB visible C stars; triangles - AGB deeply embedded IR C stars; squares - hyperluminous objects (these objects include deeply embedded OH/IR stars and early-type hypergiants (Cohen 1993); a small number are required to reproduce MIR source counts at low latitude: Wainscoat et al. 1992); crosses - exotica (T Tau stars, reflection nebulae); larger filled circles - planetary nebulae; small filled circle - bright compact Hii regions. The reddening vector corresponding to an AV of 10 mag is shown by the shaft of the arrow in the upper left corner. The three rectangles are described in the text. The solid one is that occupied by the MASH PNe in this paper (±2 standard errors of the median). The July 2004 relative spectral response files were used to represent the IRAC bands. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Table 1. Attributes of PNe studied in Hα, MIR, and radio. Listed are: PN name (col.1); PN status as T(rue), L(ikely), or P(ossible) PN (col.2); coordinates (cols.3-6); apparent size (col.7); optical morphology type (col.8); whether a PN can be distinguished by false color from its environs (col.9); whether optical and IR morphologies are related (col.10); the existence of an MSX 8.3-µm detection (col.11) or a radio counterpart (cols.12-13); and whether we have identified an optical or IR central star. |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | n n n y n n n y n n n n n n n n n y? n n n n n y ... ... ... ... ... ... ... ... ... ... ... |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Table 2. Radio and MIR flux densities for PNe with two or more detections of these. Listed are: PN name (col.1); PN status (col.2); representative diameter (col.3); optical morphology (col.4); radio detection by M(GPS2) or N(VSS) (col.5); radio flux density with parenthesized uncertainty (col.6); MSX and IRAC flux densities (cols. 7-8); ratio of IRAC to MSX flux (col.9); ratio of MSX to radio flux (col.10); ratio of IRAC to radio flux (col.11). |
Processed_Spitzer_IRAC_observations_of_newly-discovered_plan.txt | Table 3. Observed and corrected median colors (in magnitudes) of MASH PNe in the four IRAC bands, compared with colors synthesized from ISO SWS spectra of a sample of well-known Acker et al. (1992) PNe. |
Processed_Experimental_set-up_of_the_LUNASKA_lunar_Cherenkov.txt | technique to enable an instrument such as the SKA to be used at its full sensitivity for a UHE particle search. To this end, we have been using the Australia Telescope Compact Array (ATCA), a radio interfer- ometer of six 22 m antenna located in New South Wales, Australia, as a test-bed to develop experimental methods scalable to giant, broad-bandwidth radio arrays such as the SKA. In particular, we target the ‘high- frequency’ (& 1 GHz) regime, both because ATCA and the likely high-frequency component of the SKA are/will be arrays of small dishes, and because the NuMoon/LOFAR collaboration [6] is conducting similar development for low-frequency observations. Here, we describe our experimental set-up and equipment, as used for two three-day observation periods in February and May 2008, which allowed as to use an unprecedented 600 MHz bandwidth in the one data channel to search for these elusive particles (results of this search are presented in Ref. [9]). |
Processed_Experimental_set-up_of_the_LUNASKA_lunar_Cherenkov.txt | The lunar Cherenkov technique, proposed by Dagke- samanskii and Zheleznykh [1], is a method to detect ultra-high energy cosmic-rays (CR) and neutrinos (ν). The technique uses radio-telescopes to observe Earth’s Moon, and search for the coherent Cherenkov (radio) radiation produced – via the Askaryan Effect [2] – from a high-energy particle cascade in the Moon’s outer layers. This enables the entire visible surface of the Moon to be used as detector of particles at energies where the radiation (which arrives as a short-duration pulse) is sufficiently strong to be detectable. Although past experiments, first at Parkes [3], and subsequently at Goldstone (GLUE) [4], Kalyazin [5], and Westerbork [6], have not recorded any confirmed detections, sim- ulations indicate that a next-generation radio-telescope such as the SKA (Square Kilometre Array [7]) could probe the ‘guaranteed’ flux of UHE ν, and would likely be sensitive to the known flux of UHE CR [8]. |
Processed_Experimental_set-up_of_the_LUNASKA_lunar_Cherenkov.txt | The ATCA is a suitable test-bed for the SKA in the few-GHz frequency range, where the SKA’s collecting area will probably be a large number of small dishes. We used the ATCA’s L-band receiver, with a nominal range of 1.2–1.8 GHz, to give us 600 MHz of bandwidth in dual linear ‘A’ and ‘B’ polarisations. At the time of the experiment, the cable to the control room could handle only 128 MHz at our 8-bit digitisation. Therefore, we used the sampler boards for the ATCA CABB (Compact Array BroadBand) upgrade to implement an FPGA- based digitisation and pulse detection algorithm at each of three of the six ATCA antennas. A diagram of the signal path at each antenna is given in Fig. 1. |
Processed_Experimental_set-up_of_the_LUNASKA_lunar_Cherenkov.txt | logic was necessarily simple, due to the difficulties of time. A trigger implementing complex logic in real occurred if the voltage magnitude on either polarisation exceeded an adjustable threshold, upon which both A and B buffers from that antenna were returned to the control room and recorded to disk, along with a time- stamp at sampling accuracy. |
Processed_Experimental_set-up_of_the_LUNASKA_lunar_Cherenkov.txt | Since communication on short timescales between the antennas was impossible, each operated independently, which meant that the sensitivity of the experiment was limited by the threshold on each individual antenna. To compensate, we set the threshold on each antenna to be as low as possible. Since the instrumental dead-time was approximately 8 ms per 1 µs of buffer length, we reduced the length of the buffers returned to 256 samples (125 ns) per polarisation band. Setting the threshold between 5.5 and 6 σ kept the trigger rates to 40 − 50 Hz per band, for a mean dead-time per antenna of 5%. Our 8-bit precision was thus barely enough to allow small (< 0.1 σ) adjustments of the threshold values while keeping a reasonable dynamic range between the trigger threshold and the ADC saturation point. We recommend 10-bit precision if possible for future experiments. |
Processed_Experimental_set-up_of_the_LUNASKA_lunar_Cherenkov.txt | in Our simple detection algorithm was inefficient two important ways. Firstly, by only using polarisation channels individually (as opposed to in combination, e.g. X 2 + Y 2) to form a trigger, the full signal strength would be seen only if the signal polarisation happened to align with one of the channels. Since the signal paths within the antenna for the two polarisation bands were similar but not identical, extra cable would have had to be inserted to adjust for path differences to allow the more optimal real-time trigger algorithm to be used, which was not possible for the observations reported here. Secondly, our non-infinite sampling rate meant that we would inevitably miss the narrow peak of a real pulse. Although we would have sufficient information to reconstruct the true signal in the 1.2–1.8 GHz range to arbitrary accuracy in post-processing (sampling greater than Nyquist), doing so was impossible in real time. Therefore our sensitivity to narrow pulses was reduced, although our over-sampling partially compensated for this effect. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | New candidate Planetary Nebulae in the IPHAS survey: the case of PNe with ISM interaction. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Abstract: We present the results of the search for candidate Planetary Nebulae interacting with the interstellar medium (PN-ISM) in the framework of the INT Photometric Hα Survey (IPHAS) and located in the right ascension range 18h-20h. The detection capability of this new Northern survey, in terms of depth and imaging resolution, has allowed us to overcome the detection problem generally associated to the low surface brightness inherent to PNe-ISM. We discuss the detection of 21 IPHAS PN-ISM candidates. Thus, different stages of interaction were observed, implying various morphologies i.e. from the unaffected to totally disrupted shapes. The majority of the sources belong to the so-called WZO2 stage which main characteristic is a brightening of the nebula’s shell in the direction of motion. The new findings are encouraging as they would be a first step into the reduction of the scarcity of observational data and they would provide new insights into the physical processes occurring in the rather evolved PNe. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Moreover, excluding these objects from global studies (morphology, abundances, luminosity function...etc) may bias our understanding of planetary nebulae. As an il- lustration, few PNe are described in the literature as “PNe with ISM interaction”, which is the step before the complete dilution of the nebulae in the interstel- lar medium (Borkowski et al. (1990), Ali et al. (2000), Xilouris et al. (1996) and Tweedy et al. (1996)). The study of the interaction process would give new in- sights into several aspects of the PN evolution. Indeed, the density difference between ISM and PNe will affect their shape. This is expected to be observable in old objects where the nebular density declines sufficiently to be overcome by the ISM density. Other phenom- ena like the flux and brightness enhancement following the compression of the external shell, the increase of the recombination rate in the PN Rauch et al. (2000), the occurrence of turbulent Rayleigh-Taylor instabili- ties and the implication of magnetic fields Dgani et al. (1998) are among the physical processes which need to be addressed not only from a theoretical but also observational point of view. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | The low surface brightness generally associated to PNe-ISM has for a long time prevented any deeper ob- servation and good statistical study of these interac- tions, where only the interacting rim is well seen. New generations of Hα surveys have overcome this prob- lem. A perfect example is the discovery of PFP 1 by Pierce et al. (2004) in the framework of the AAO/UKST SuperCOSMOS Hα survey (SHS) (Parker et al. 2005). This PN, starting to interact with the ISM at the rim, is very large (radius = 1.5 ±0.6 pc) and very faint (logarithm of the Hα surface brightness equal to -6.05 erg cm−2.s−1.sr−1). In order to unveil and study this “missing PN population” in the Northern hemisphere we need surveys providing the necessary observing depth: the Isaac Newton Telescope (INT) Photometric Hα Survey (IPHAS) is one of them and will complete the work done in the South by the SHS. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | IPHAS is a new fully photometric CCD survey of the Northern Galactic Plane, started in 2003 (Drew et al. (2005), Gonzalez-Solares et al (2008)) and which has now been completed 1. Using the 2.5m Isaac Newton Telescope (INT) in La Palma (Canary Islands, SPAIN) and the Wide Field Camera (WFC) offering a field of view of 34.2×34.2 arcmin2, IPHAS targets the Galac- tic plane in the Northern hemisphere, at a latitude and covers 1800 deg2. This range of -5 international survey is conducted not only in Hα but also makes use of two continuum filters, respectively the Sloan r’ and i’. IPHAS is viewed as an enhance- ment to former narrow-band surveys, first due to the use of CCD and the particularly small pixel scale al- −1 but also (and lowed by the WFC with 0.33 arcsec pix mainly) due to the depth reached for point sources de- tection. Thus sources with a r’ magnitude between 13 and 19.5-20 could be detected with a very good pho- tometric accuracy. The most interesting characteristic for our purpose is the ability to detect resolved ex- tended emissions with an Hα surface brightness down to 2×10−17 erg cm−2 s−1 arcsec−2 . |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | In this paper we will focus on extended (candi- date) PNe (i.e. objects with a size greater than 5 arc- sec). They were searched for via a visual inspection of 2 deg2 Hα-r (continuum removal) mosaics made from the different IPHAS observations. And in order to al- low the detection of objects of multiple size and bright- ness level, the mosaics were binned at respectively 15 pixels×15 pixels (5 arcsec) and 5 pixels×5 pixels (1.7 arcsec). The first binning level, which is of particu- lar interest to us, helps to detect resolved, low surface brightness objects (down to the IPHAS limit) and to accentuate the contours/shape of the nebulae (this is particularly useful to see, for example, the full extent of an outflow or a tail). The second set, is used to de- tect intermediate size nebulae i.e. smaller than ∼15-20 arcsec in diameter. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | The first area that has been fully investigated is the re- gion between RA=18h and RA=20h. We detected 233 candidate PNe among which other nebulosities may be found e.g. small HII regions (Sabin, PhD thesis, to be published). Around 20% of this sample have been so far spectroscopically confirmed as PNe (Sabin et al., in preparation). If we look at the particularities of the PNe and candidate PNe uncovered, we observe that from the point of view of the size, large objects (greater than 20 arcsec) constitute the main new group (Fig. 1). As large objects are generally considered as more evolved, we are confident in finding in this group new old PNe and by extension new cases of PNe interacting with the surrounding ISM (PNe/ISM). |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | PNe mixing with the ISM can also be filamentary struc- tures from old SNRs. A spectroscopic analysis is the only way to have a clear identification. The work presented here is based on the classification from Wareing et al. (2007) (WZO 1-4 called after the authors’ names) and will allow us to establish the de- gree of interaction for each nebula. Their classifica- tion is the result of the first extensive investigation of the applicable parameter space, varying stellar pa- rameters, relative velocities through the ISM and ISM densities. The depth reached by the IPHAS survey combined with the binning detection method allowed us to iden- tify 21 cases of interacting candidate PNe. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | The first group of PNe/ISM concerns those where the main PN is still unaffected and which may display a distant bow shock. In our area of study (18h-20h), the majority of candidates answer the first condition, but none show the outer bow shock. Outside this area, the nicknamed “Ear Nebula” or IPHASX J205013.7+465518 with a 6 arcmin size may be coincident with a WZO1 description as this object is a confirmed bipolar PN (Fig. 3) surrounded by a shell which may be an AGB remnant shell or would indicate a multiple shell nebula (Fig. 2). |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | This category concerns PNe showing a bright rim in the direction of motion. This is the most common fea- ture found in our sample and 17 objects out of 21 fall under this classification. Fig. 4 presents 3 examples with different angular sizes, although they all display a diameter on the order of a few arcmin (we consid- ered the assumed full extent of the round nebulae). We point out in Fig. 4-Top the difficulty to determine the true direction of motion regarding the CS position and off-axis bow shock. Such a geometry could be ex- plained by an ISM gradient from high on the left to low on the right. We also notice a particularly low observed surface brightness (SB) which may explain previous non detections. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | This type is exemplified by PNe whose geometric cen- tres are shifted away from the central star (CS): both are no longer coincident. An example, is the ancient PN Sh 2-188 around which IPHAS has uncovered an extended structure (Wareing et al. 2006). We identi- fied 3 candidate PNe coincident with this description. The most probing WZO3 type in our sample is pre- sented in figure 5 and corresponds, according to hy- drodynamical models, to a PN with a CS velocity of about 100 km/s. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | The WZO4 corresponds to the most difficult types of the CS has left the vicinity of PN to be detected: the now totally disrupted PN, leaving an amorphous structure. The challenge does not lie in the detection ability (it enters in the IPHAS range of detection) but more in the selection of the objects as possible PNe due to the total lack of symmetry or axi-symmetry. This type of interaction is also discussed in more detail by Wareing et al. in these proceedings. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | We identified 1 candidate PN which could fit the given description. Fig. 6 presents the selected can- didate in the top panel. We suggest the the nebular material has been moved from the front to the rear leaving a remnant “wall of material”. We also notice that some features may be linked to turbulence effects. The comparison with the hydrodynamical model (bot- tom panel) seems to support this hypothesis. Nev- ertheless a spectroscopic confirmation of the nebula’s nature will be needed. The model implies a velocity relative to the ISM of 100 Km/s and an evolution in the post-AGB phase of 10 000 years. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Fig. 7-top shows that the majority of the WZO2 nebu- lae types are located in zones of relatively low ISM den- sity (compared to the Galactic Centre). The low stress exerted on the nebulae may explain why they still keep their quasi circular shape. The ISM is more dense in the Galactic Plane than in the zone towards the anti- centre or the zone above a height of 100 pc (from obser- vation of neutral hydrogen gas, Dickey et al. (1990)). We therefore expected a greater influence of the in- teraction process in this area. Indeed, we observed that the most advanced stages of interaction, namely WZO3 and WZO4, are detected in areas of high ISM density, where PN are more likely to be affected by such densities. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | The size distribution, Fig. 7-bottom, indicates that although most of the detected candidate PNe are large2, i.e with a size greater than 100 arcsec, or of medium size i.e. between 20 and 100 arcsec, small nebulae also show signs of interaction. This confirms that the ISM interaction process does not “a priori” only im- ply “old” nebulae. We also observe that large objects mainly lie at higher latitudes than smaller nebulae but it is also interesting to notice that we detect large ob- jects in zones of high extinction; large PNe seem to sur- vive at relatively low latitudes. They would undergo strong alteration by the ISM and would display more advanced stages of interaction. Those disruptions tend to affect them more than smaller size nebulae at the same latitude range. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | IPHAS appears to be an excellent tool to study PNe in- teracting with the ISM. Indeed the survey contributes to the detection of nebulae so far hidden mainly due to their faintness. Thus, 21 objects have been identified as possible planetary nebulae interacting with the ISM. They show diverse sizes (although the majority display a diameter greater than 100 arcsec) and morphologies corresponding to the four different cases of interaction commonly defined going from the unaffected to the to- tally disrupted nebula. The most common stage is the WZO2 corresponding to nebulae showing a brightening of their rim in the direction of motion. This is coinci- dent with the observations made by Wareing and al (in these proceedings) crossing different Hα surveys. We were also able to reach those targets at low latitudes and found that some could survive in those environ- ments although they would be strongly affected by the ISM. The total lack of PNe/ISM at the highest point of ISM density (b=±0.5 deg and 30 deg <l < 50 deg) can either be due to the limitation of IPHAS or be- cause they have been totally destroyed by the effects of ISM interaction. The next logical step is the spectroscopic identification of these sources, their central star study and physical size determination. The low surface brightness implies the use of particular means such as integral field spec- troscopy to be able to retrieve the maximum infor- mation. Therefore a new programme of IPHAS PN candidate follow-up spectroscopy led by Q. Parker, A. Zijlstra and R. Corradi is now underway. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Acker, A., Marcout, J., Ochsenbein, F., Stenholm, B. , Tylenda, R.,1992, Garching: European Southern Observatory,Strasbourg - ESO catalogue of galactic planetary nebulae. Part 1; Part 2. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Drew, J. E., Greimel, R., Irwin, M. J., Aungwero- jwit, A., Barlow, M. J., Corradi, R. L. M., Drake, J. J.., G¨ansicke, B. T., Groot, P. and 26 co- authors,2005,MNRAS,362,753. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Gonzalez-Solares, E. A., Walton, N. A., Greimel, R. and Drew, J. E., Irwin, M. J., Sale, S. E., Andrews, K., Aungwerojwit, A., Barlow, M. J. and 41 co- authors, 2008,MNRAS,388,89. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Parker, Q. A., Phillipps, S., Pierce, M. J., Hartley, M., Hambly, N. C., Read, M. A., MacGillivray, H. T., Tritton, S. B., Cass, C. P., Cannon, R. D., Cohen, M., Drew, J. E., Frew, D. J., Hopewell, E., Mader, S., Malin, D. F., Masheder, M. R. W., Morgan, D. H., Morris, R. A. H., Russeil, D., Russell, K. S. and Walker, R. N. F., 2005, MNRAS, 362, 689. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Parker,Q. A., Acker, A., Frew, D. J., Hartley, M., Peyaud, A. E. J., Ochsenbein, F., Phillipps, S., Russeil, D., Beaulieu, S. F., Cohen, M., K¨oppen, J., Miszalski, B., Morgan, D. H., Mor- ris, R. A. H., Pierce, M. J. and Vaughan, A. E.,2006,MNRAS,373,79. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Wareing, C. J., O’Brien, T. J., Zijlstra, A. A., Kwitter, K. B., Irwin, J., Wright, N., Greimel, R. and Drew , J. E.,2006,MNRAS,366,387. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 2: An example of WZO1 type: The “Ear Nebula” IPHAS PN. North on the top and East on the left. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 3: WHT spectra of the “Ear Nebula” using the R300B and R158R gratings. This nebula, for which we show some of the “strongest” emission lines useful for an identification, presents a clear [NII] over- intensity and it has been confirmed as true PN using the revised diagnostic diagram from Riesgo et al. (2006) (particularly the log [Hα/[SII]] vs log [Hα/[NII]] diagram). |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 4: Examples of WZO2 types. Top: Size=1.2 arcmin and SB=3.4e−17 erg cm−2 s−1 arcsec−2. Middle: Size= 8.5 arcmin and SB=1.1e−16 erg cm−2 s−1 arcsec−2, Bottom: 4.3 arcmin and SB=2.7e−16 erg cm−2 s−1 arcsec−2. North on the top and East on the left. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 5: WZO3 type of ISM interaction in a IPHAS candidate PNe. North on the top and East on the left. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 6: A possible example of WZO4 ISM interaction in one IPHAS PN candidate (top: North on the top and East on the left) with the corresponding hydrodynamical model (bottom) [reproduction of figure 5(d) from Wareing et al. (2007)]. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 7: Galactic distribution of the candidate PNe/ISM according to their stage of interaction and their size. |
Processed_New_candidate_Planetary_Nebulae_in_the_IPHAS_surve.txt | Figure 8: Example of candidates with sizes greater than 100 arcsec (respectively 7.7 and 2.9 arcmin) and located at b=±1 deg. These objects present a WZO2 stage of interaction and only their (very) faint interacting rim are seen. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | This paper presents a model for dynamic adjustment of the mo tivation degree, using a reinforcement learning approach, in an action selection mechanism previously developed by the authors. The learning takes place in the modification of a parameter of the model of combination of internal and external stimuli. Experiments the claimed properties are presented, using a VR simulation developed for such purposes. The importance of adaptation by learning in action selection is also discussed. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | In the context of the beh aviour-based systems (Bro oks, 1986; Bro oks , 1989 ; Maes, 1990 ; Maes, 1994 ), the action selection problem refers to what actio n the agent (robot, animat or artificial creature) must select every mom ent in time; whereas an action selection mechanism (A SM ) specifies how these actions are selected. An A SM is a comp utational mechanism that must produce as an output a selected action, when different external and/or internal stimuli have been given as inputs. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The great majority of the ASM p roposed in the literature (Tyrrell, 1993), include an additive mod el for stimuli combination. The stimuli ad dition could be interpreted in many ways, but the essence is that this type of stimuli combination mod els a be haviour with a more reactive tendency than a motivated one. This can be explained in the following terms: for high values of the external inputs, the additive comb ination could surp ass some value thresho ld previously established and, therefore, trigger a certa in external action, still for non significant values of the internal inputs. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | An alternative to eliminate the undesired effects of the pure additive and m ultiplicative com binatio ns is to co nsider a model that incorp orates the mo st valuab le elements of bo th types of combination. This is achieved by the combination model of an ASM developed by us (Go nzález, 2000). |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | Adaptation is one o f the desirable characteristics in the action selection of an auto nom ous agent. T he ad aptatio n in an ASM can b e obtained from three main approaches: preprogrammed adaptive behaviours, learned adaptive behaviours and evolved adaptive behaviours (Meyer and Guillot, 1990). Adap tation by learning is a proc ess by which an autonom ous agent is able to a djust its behaviour to different types of environmental changes. An autonomous agent is adap tive by lea rning if it has skills, which allo w to improve its performance in time (M cFarland, 1990). |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | In this paper we discuss how the adapta tion by learning improves the action selection in an autonomous agent, from an A SM developed by the authors (González, 2000; González et. al., 2000; G ershenson et. al., 2000). Two types of adaptatio n by learning are present in this ASM: (1) classical conditioning, which allows that new behaviours and emergents properties to arise in the action selection, increasing its adaptive level, and (2) reinforcement learning approach (Humph rys, 1996), which allows the motivation degree in the action selection to be adjusted dynam ically. This paper discusses the second type of adaptation by learning. This type of learning takes place over a parameter belonging to the model for combination of external and internal stimuli deve loped by us (Go nzález et. al., 2000) and used by the ASM. T his parameter allows to regulate the dependence degree of the external behaviour executed by the agent from its internal states, and its current value is learned dynamically. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | functional unit of the ASM, the internal behaviour. Section 3 presents the intero/extero/drive congruence internal behaviour, where the combination of external and internal signals is carried out. Section 4 discusses the mod el for co mbination o f external and internal stimuli used by the ASM, and the role of the parameter that allows to regulate the motivation degree in this combination. In section 5 we present and discuss the learning process by which the mo tivation degree in the action selection can be adjusted dynamically. Finally, section 6 presents the simulation and experim ents de velop ed to verify the role of this type of adaption by learning in the action selection. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | network of bla ckboard nod es developed by the authors (González and N egrete, 199 7; Negrete and Gonz ález, 1998). A blackboard nod e the following comp onents: a set of independent modules called knowledge sources, which have specific knowledge about the problem domain; the blackboard, a shared data structure through which the kno wledge sources com munic ate to each other by means of the blackboard; the communication mechanisms, which establish the interface between the nodes and a node and the external and internal mediums; the activation state registers of the knowledge sources; and a control mechanism, which determines the ord er in that the know ledge sources will operate on the blackboard. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | extension of the Internal Behaviour Network (IBeNet) (González, 2000). As it can be appreciated in Figure 1, the actual architecture of the AS M exhibits two blackboard nodes: the cognitive node and the motivational node. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The basic functional unit of the BeCA is the “internal behaviour”. The term “internal behaviour” has been used to describe the pro cesses of creation and modification of solution elements that occur on the blackboard of the node. Structurally, an internal behaviour is a package of elemental behavio urs, also known as production rules (if <condition> then <action>). Elemental behaviours of different kinds are organized forming “behavio ural columns” that cross vertically different blackboard levels. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The structure of an elemental behaviour has three basic elements: a parameters list, a condition part, and an action part. The parameters list spec ifies whic h are the condition elements, the action elements, and the coupling strengths related with the elemental behaviour. The condition of an elementary beha viour describes the configuration of solution elements on the blackboard that is necessary, so that the elementary behaviour contributes to the solution processes of the pro blem. The way in w hich an elemen tal behaviour contributes to the solution of the problem is specified in its action, which can consist of the creation or modification of solution elements in certain blackboard levels. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | A coupling strength is represented by a vector Fa = (Fai1, Fai2, ..., Fain) of n real components, where each o f these com ponents represent the efficacy with which the elemental behaviour i can satisfy the condition j. Depend ing on the nature of the elemental behaviour, the components of the vector Fa may have fixed or modifiable values. The existence of mo difiable coupling strengths allows the incorporation of learning processes to improve the action selection, making it more adaptive. The vecto r Fa of coupling strengths is to an elemental behaviour the same as a weight vector to an artificial neuron. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The structure of a blackboard node in BeCA consists of five basic elements: the internal behaviours, the blackboard, the activity state registers of the internal behaviours, the interface/communication mechanisms, and the competition mechanism. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | between the cognitive node and external medium; and the receptor and transmitter mechanisms, which establish communication with the motivational node. The role of the cognitive node includes the processes of representation of the perceptual signals, integration of internal and external signals, inhibition of reactive responses, and the selection of the external behaviour that adjusts better to the actual external cond itions and internal need s. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The blackboard of the motivational node is structured in four abstraction levels: internal perceptions, external percep tions, intero/extero/drive congruents, and drive. The internal behaviours that operate in the motivational node are the a n d f o l l o w i n g : consu mm atory preferences selector. The communication of this node is established from the interocep tors, which define the interface between the node and the internal medium; and the receptor and transmitter mechanisms, which establish the communication with the cognitive node. The role of the motivational node includes the processes of representation of internal signals, combination of internal and external signals and the more app ropriate consummatory preference. A full discussion of the model for combination of external and internal stimuli, used by the motivational node in BeC A, can be found in (Go nzález et. al., 2000). |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | At the level of the intero/extero/drive congruence internal behaviour of the motivational node is carried out the combination of external and internal signals to determine the action that must be executed by the entity. These signals are internal registered perceptions levels, respectively, of the motivational node. The solution elements registered in the external perceptions level are originated by unconditional stimuli (US), conditional stimuli (CS), or both; all of them initially projected in the external perceptions level and later recovered from the perceptual persistents level of the cognitive node. On the othe r hand , the solutio n elements the registered motivational node represent the values of the internal states that are able to satisfy part of the conditions of certain intero/extero/drive congruence elementary behaviours that ope rate on this level. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The model for combination of internal and external stimuli used by intero/extero/drive congruence internal behaviour is the one given by expression (1). |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | C is the new activation level correspondent to the where Ai intero/extero/drive congruents level of the motivational E is the signal from the internal perceptions level and node; O i S is the signal from the external Fai D is the perceptions level and Faij D its coupling strength; signal from the drive level, and Fai and (cid:34)i regulates the combination of the internal and external signals. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | the cognitive node) and the drive signal (solution eleme nt in the consumm atory p referents level of the cognitive node) coincides. On the other hand, although a very strong internal signal exists, corresponding to an urgent necessity of the entity, if the usable external signal is not present, the intero/extero/drive congruence elemental behaviour will not activate. Then, when (cid:97)i = 0, b oth factors (external signal and internal signal) are necessary so that the intero/extero/drive congruence internal b ehav iour w ill be activated. Physio logically this can be interpreted, when thinking that the inputs that come from the internal states sensitize this internal behaviour with the signals that come from the external medium. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | W hen (cid:97)i is different of zero, more weight (o r impo rtance) is being granted to the internal state than to the external inputs. the total absence of external So, still intero/extero/drive congruence elemental behaviour could be activated for a very strong value of the internal signals (when the value of (cid:97)i is near o ne). T hus, when the internal ne ed is extrem ely high and external signals do not exist, the intero/extero/drive congruence elemental behaviour can be activated, and therefore to cause a type of "preactivation" of the conditions of the attention to preferences elementary behav iours (in particular, only one of these elementary behaviours will be the one that will receive greater preactivation). This it is a typical example of an internal behaviour whose actions are d irected to the internal medium. This mechanism constitutes the underlying base of the exploratory behaviour oriented to a specific objective, which is another of the distinguishing properties of the BeCA W hen there is a total satiety of the internal necessity that caused the activation of the intero/extero/drive congruence E = 0), then the level of activation of elemental behaviour (O i C = 0), this elemental behaviour will decrease to zero (Ai indep endently of the existence of usable external signals. For activation levels of zero or very near to zero the elementary behaviour will not be activated. In this case, the observed external behaviour always will be adaptive. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | E is the value of the internal signal from the internal where O i S represents the value perceptions blackboa rd leve l, (cid:51)jFaij of all the external signals that are asso ciated to internal state i, and (cid:50) is a threshold value. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The increment for the parameter (cid:97)i is determined by expression (3). This increment can be seen as a hyperb olic divergence from (cid:97)min, as shown in Figure 3. In expression (3), (cid:42) determines the length of the d ivergence (how much time it will take to (cid:34)i to go from (cid:34)min to (cid:34)max), and (cid:68) determines the speed of the divergence. This smo oth modification beha viour simulates an historic memory of the environment (remembered scenario), so that the value of (cid:97)i is increased only after severa l iterations in a certain enviro nment. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The decrement of the parameter (cid:97)i is determined by expression (4). T his is similar to the increment described by expression (3), o nly that it hyperbolically diverges from (cid:34)max, as can be seen in Figure 4. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | In Figure 5, we can appreciate an example of the behaviour of the parameter (cid:97)i, as it is increased, decreased, or remains constant, in dependence of the perceived scenario and the internal need s. No te that the inc reme nt is faster than the decreme nt, because the values of (cid:97)i are closer to (cid:34)max than to (cid:34)min. The graphics of this section were taken from the simulation described in the next section. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The type of learning described in this section can be seen as a particular type of reinforcement learning. This is, the reward or the p unishm ent will be applied to the motivation degree of the action selection, in dependency of how abundant or limited are the signals required in the enviro nment, and how strong or weak are the internal states. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | In order to verify the role of the reinfo rcem ent learning in the action selection, a programme was written implementing the behavioural columns architecture, being this used by anima ts (artificial anim als) simulated in virtual reality. The anima ts were inspired by an original idea developed by Negrete (Negrete and M artinez, 1996). |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The graphical represe ntation o f the animat is a cone with a semicircumscript sphere. He has p rimitive motor and perceptual systems developed specially for this simulation. The animat has internal states to satisfy, such as hunger, thirst and fatigue, and the qualities of stren gth and lucidity; which are affected by the diverse behaviours that he can execute. The implemented behaviours are: wander, explore, approach a specific signal, avoid obstacles, rest, eat, drink, and runaway from a specific signal. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The simulation allows to modify the different parameters of the animat, for example, the speed of conditioning or reinforce ment. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | W e considered an initial state as shown in Figure 6. We had two animats in separated environments: one with abundant sources of food, and the other one with none. Other stimuli were created in both environments. Both animats were initialized with the same param eters: very high hunger, and no other internal ne ed, with an initial value of (cid:34) for the hunge r colum n of 0.7 . |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | Figure 7 shows what occurred after some time of execution of the simulation. T he anim at in the scarce environment explored in search of foo d, but found none. This remembered scenario lead to the incremental adjustment of his respective (cid:34)i until it reached (cid:34)max. On the other hand, the abundant environment allowed the animat in it to satisfy his hunger quick ly. Once his hunger was satiated, the animat wandered, while his respective (cid:34)i decreased. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | will be difficult to find, will begin to explore. Meanwhile, the animat with an (cid:34)i equa l to (cid:34)min will wander, because he learned that he will find abundant food with low effort, and thus allows himself to wander, and w ill not exp lore un til his hunger is higher. These behaviours can be seen in Figure 8. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | W ith these experiments we could corroborate that with the dynamic adjustment of the motivation degree in the combination of external and internal signals, an autonomous agent will learn certain qualities of its actual dynamic environment, making its behaviour more adaptive. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | In this paper we have discussed and shown how the action selection in an autonomou s agent has been imp roved with adaptation by learning. The type of learning discussed here, reinforcement learning, was used to dynamically adjust the motivation degree in a model for combination of external and internal stimuli incorporated in an ASM previously develop ed by us. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | The dynam ic variation of p aram eter (cid:97)i in the combination model (expression 1) does the action selection more adaptive, allowing the autono mous agen t to develop in an environment from w hich he possesses certain knowledge. It is impo rtant to note that the representation of this knowledge lies in one parameter, not in symbolic representation of knowledge. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | Brooks, R. A. (198 6) A robust layered control system for a mob ile robot. IEE E Jo urnal of R obo tics and Automation. RA-2, April, pp. 14-23. |
Processed_Dynamic_Adjustment_of_the_Motivation_Degree_in_an_.txt | Brooks , R. A. (1989) A rob ot that walks: Emergent behaviour from a carefully evolved network. Neural Computation, 1, 253-262. |
Subsets and Splits