A Layered Fuzzy Controller For Nonholonomic Car-Like Robot Motion Planning

A system for real time navigation of a nonholonomic car-like robot in a dynamic environment consists of two layers is described: a Sugeno-type fuzzy motion planner; and a modified proportional navigation based fuzzy controller. The system philosophy is inspired by human routing when moving between obstacles based on visual information including right and left views to identify the next step to the goal. A Sugeno-type fuzzy motion planner of four inputs one output is introduced to give a clear direction to the robot controller. The second stage is a modified proportional navigation based fuzzy controller based on the proportional navigation guidance law and able to optimize the robot's behavior in real time, i.e. to avoid stationary and moving obstacles in its local environment obeying kinematics constraints. The system has an intelligent combination of two behaviors to cope with obstacle avoidance as well as approaching a target using a proportional navigation path. The system was simulated and tested on different environments with various obstacle distributions. The simulation reveals that the system gives good results for various simple environments