基于混沌算法的工程车辆自动变速神经网络控制

Study on Neural Network Control of Automatic Transmission for Construction Vehicle Based on Chaotic Algorithm

为了克服 BP神经网络收敛速度慢和易陷入局部极小点的不足 ,提出了一种基于混沌算法的工程车辆自动变速神经网络控制方法 ,并利用 ZL 5 0型装载机传动系统换挡控制试验的数据进行了验证性仿真试验。仿真结果表明 :该方法对于车辆自动换挡智能控制是可行的 。

A method of neural network control of automatic transmission for construction vehicle based on chaotic algorithm has been put forward in the paper. A validating, simulating experiment was carried out by using the experimental data acquired from the shifting control experiment done previously on powertrain system of ZL50 wheel loader. In fact, the chaotic optimization method of neural network is an improved algorithm proposed to overcome the defects of BP neural network, which has a slow convergence speed and can not jump out of the local minima. The results of simulation showed that the presented shift decision method was effective for intelligent control on automatic shift of construction vehicle and could realize gear selection of transmission box correctly according to the working condition.