Bagher Aalizadeh

In this study, two adaptive neural network and neurofuzzy identification models are proposed to identify vehicle handling under uncertainties. These models are used to identify vehicle handling in different road friction coefficients and velocities. These two identification models modify their weights to cope with uncertainties using back propagation of error as a learning algorithm. However, an adaptive model has some limitations to identify real systems. The ability of adaptation is not the same for all identification models; some models are more robust to cope with a specific uncertainty or a wider range of uncertainties. In this study, adaptiveness of two identification models are compared under two different uncertainties. First, a precise model in CARSIM software is simulated and a set of input/output data of vehicle response are collected. Then an initial three-layer neural network is trained in MATLAB software. In addition, a Neurofuzzy model is also trained in ANFIS (adaptive neurofuzzy inference system) toolbox of MATLAB software. Then this trained model is applied to the vehicle in different maneuvers, velocities and road friction coefficients. Results show that proposed neural network identifies the vehicle handling more efficiently than neurofuzzy model in conditions that are away from training condition. However, proposed neurofuzzy model is more precise and accurate than neural network in the training condition.

In this study, a neurofuzzy controller is proposed to improve vehicle handling in different road friction coefficients. This controller adapts itself to neutralize the effects of unpredictable changes in road friction coefficient on vehicle handling. This adaptive neurofuzzy controller can improve vehicle handling, manoeuvrability and path tracking. First a proportional-integral-derivative (PID) controller is proposed and tuned by using PSO (particle swarm optimization). Then, this tuned PID controller is applied to the vehicle system and training data is gathered. The next step is to train a fuzzy controller by importing thistraining data tothe ANFIS (adaptive neurofuzzy inference system) toolbox of MATLAB software. Then the trained fuzzy controller is applied to a vehicle that exploits AFS (active front steering system). This controller is able to adapt itself during manoeuvres, by using back propagation of error as a learning algorithm. Results show that neurofuzzy controller can improve handling of the vehicle in different road conditions, because neurofuzzy conteroller can adapt itself in unpredictable situations.

In this paper, a new optimal adaptive controller for the active front steering control of a vehicle is proposed. Due to the availability and applicability of proportional-integrator-derivative (PID) controllers, this controller is picked up; but, to overcome its limitations, two optimization and adaptation schemes are employed. The reference transfer function between the yaw rate of a typical vehicle and its steering angle is derived. The actual dynamics is simulated using CarSim toolbox of MATLAB. Best vehicle handling was aimed to be reached for three famous driving manoeuvres by proposing an efficient but economic controller. An optimization is done on the PID coefficients for a specific road condition using the honey bees' algorithm, and then a two-layer artificial neural network is trained using the back propagation learning rule to adjust the controller coefficients for any arbitrary road conditions. The uncontrolled, desired, optimized controlled, and optimized plus trained controlled yaw rate of vehicles are drawn for three manoeuvers and three road conditions. The integral of squared error between the desired and actual trajectories for different manoeuvers and road conditions are evaluated and compared between each other. The performance of the proposed PID controller that was optimized by Bees Algorithm and trained by a neural network was proved. It is noted that the optimized PID controller is good for all road conditions but not excellent. The more we move away from the reference road (in which, the optimization is done), the error will be larger. Thus for other conditions, using the artificial neural network can help decrease the error significantly.