Modified Vector Field Histogram with a Neural Network Learning Model for Mobile Robot Path Planning and Obstacle Avoidance

In this work, a Modified Vector Field Histogram (MVFH) has been developed to improve path planning and obstacle avoidance for a wheeled driven mobile robot. It permits the detection of unknown obstacle to avoid collisions by simultaneously a steering the mobile robot toward the target; a regular grid map representation for a work space environment is caurried out. A Neural Network (NN) model is used to learn many critical situations of environment during robot navigation among obstacles using MVFH. Also, digital filter has been utilized for improving the robustness of obstacle avoidance trajectory of mobile robot. The proposed MVFH-NN has been implemented and tested by using MobotSim program simulation and MATLAB . The developed algorithm showed good navigation properties and can be used in complex real world (maze-like environment), it also shows good ability to overcome limitations of the traditional VFH algorithm (like wide candidate valley, narrow hallway and target distance limitation).