Way point based deliberative path planner for navigation

We present and implement a genetic algorithm (GA) based deliberative path planning algorithm using waypoints for navigation of robots and AGV´s. The waypoint navigation system is advantageous as it does not require a complete knowledge of the environment and can replace traditional robot navigation systems. The proposed GA for robot navigation problem is NP-hard and multiobjective. The performance of the implemented GA is compared with the results obtained from two popular algorithms namely the Dijikstra´s algorithm and A* algorithms. We propose a new representation that reduces the size of the chromosome array for large number of waypoints; the proposed GA has faster convergence and obtains near optimal solutions. The system has been developed for a near real-time experiment using a Matlab engine located on an online server. An integrated system complete with the required hardware was constructed for the purpose of testing the performance of the various algorithms using a real world geographical information system (GIS) and actual node data (waypoints) of Kuala Lumpur city. The proposed new system has been tested and is found to be suitable for adaptation in robot navigation, path planning and intelligent traffic guidance systems.