Robot Navigation with a Polar Neural Map

Center for Advanced Computer StudiesUniversity of Southwestern LouisianaLafayette, LA 70504maida©cacs, usl. eduNeural maps have been recently proposed as an alter-native method for mobile robot path planning (Glasius,Komoda, and Gielen 1995). However, these proposalsare mostly theoretical and are primarily concerned withbiological plausibility. Our purpose is to investigatetheir applicability on real robots.Information about the environment is mapped on atopologically ordered neural population. The diffusiondynamics force the network into a unique equilibriumstate that defines the navigation landscape for the giventarget. A path from any initial position to the target(corresponding to the peak of the activation surface)is derived by a steepest ascent procedure. The figuresbelow show an example on a 50 x 50 rectangular map(a. Environment, b. Contours of activation, c. Path).We attempted to implement the approach on a No-mad 200 mobile robot for sonar-based navigation. How-ever, we found that the neural map requires reorgani-zation in a polar topology that reflects the distributionof the sonar data points, the only source of informationabout the environment. The polar map covers the localcircular area around at the robot. Sonar data points aremapped scaled to the physical robot size. At each stepof the control loop, the dynamics of the map is usedto derive the angular and radial displacement requiredto reach the target from the current configuration. Asimplified example is shown below (bird’s eye view).Sensor uncertainty and noise is handled by a sonarshort-term memory and appropriate coordinate map-ping for reuse. Motion control is based on an optimiza-tion procedure that combines ideas from Fox, Burgard,and Thrun (1997) and Hong et al (1996), and takes account the kinematic and dynamic constraints of therobot. The complete architecture of the resulting local(sensor-based) navigation system is shown below.