Designing a genetic neural controller of differential braking system for vehicle based on model reference

Abstract EN

n this paper, the structure of the controller is consists of a Modified Elman Neural Networks MENN model that is learned on-line by using genetic algorithm teachings in order to achieve required yaw rate and reduce lateral velocity in a short period of time to prevent vehicle from sliding out the curvature.

By using differential braking system and front wheel steering angle has automatically controlled the vehicle lateral motion when the vehicle rotates the curvatures.

The robust feedback neural controller is achieving the excellent transient state output of the system by minimizing the error between the model reference output and the model output of the system.

Where the model of the system is also MENN that learned by two stages off-line and on-line, in order to guarantee that the model output accurately represents the actual output of the system by using dynamic Back Propagation Algorithm (BPA).