关键词:神经网络; 遗传算法; 近似模型; 悬架
Application of feedforward NN in design of vehicle suspension
REN Yuan, BAI Guang-chen
(School of Jet Propulsion, Beijing University of Aeronautics & Astronautics, Beijing 100083, China)
The GA-BP Bayesian algorithm was used to establish the kinematics analysis approximation model for suspension. This algorithm is a new feedforward NN training algorithm developed by the authors. Its aim is to improve the generalization ability of networks, and it trains a network with the purpose of obtaining the weights corresponding with the maximum posterior probability. Taking a wishbone type independent front suspension for example, the GA-BP Bayesian algorithm was used to establish the kinematics analysis approximation model whose output was the maximum sideways displacement, and the comparison with three other methods was made. They were L-M backpropagation, polynomial regression, and general regression neural network. The results show that the approximation models based on GA-BP Bayesian algorithm have higher prediction accuracy than the ones based on three other methods, and that the prediction accuracy they have can hardly be affected by random factors. ......