In this paper we present CBC (context based clearing), a procedure for solving the niching problem. CBC is a clearing technique governed by the amount of heterogeneity in a subpopulation as measured by the standard deviation. CBC was tested using the M7 function, a massively multimodal deceptive optimization function typically used for testing the efficiency of finding global optima in a search space. The results are compared with a standard clearing procedure. Results show that CBC reaches global optima several generations earlier than in the standard clearing procedure. In this work the target was to test the effectiveness of context information in controlling clearing. A subpopulation includes a fixed number of candidates rather than a fixed radius. Each subpopulation is then cleared either totally or partially according to the heterogeneity of its candidates. This automatically regulates the radius size of the area cleared around the pivot of the subpopulation.

System modeling and simulation are widely used in robotic vehicle engineering to reduce development time, improve the vehicle design, and miniaturize complex systems. This paper focuses on the steering control of a four-wheel steering vehicle. Optimally-tuned PID controllers are designed for autonomous ground vehicles. The required mission path and stability of steering are achieved by reducing the error between the vehicle heading and the target point. Predefined waypoints are set for the trajectory tracking system. An optimal genetic algorithm (GA) is proposed to tune the vehicle controllers. The proposed GA-tuned PID controller exhibits satisfactory simulation results and promising experimental results. This paper reports the current status of an ongoing research project for a fully autonomous unmanned ground vehicle (UGV).

No abstract

No abstract

No abstract