论文部分内容阅读
分析了无人机用电控活塞发动机试验特点以及试验中存在的难点,针对电控发动机高海拔标定试验中进气歧管压力(manifold air pressure,简称MAP)传感器数据的传统线性插值方法不能完全表述电控发动机非线性特性的缺陷,提出采用BP(back propagation)神经网络模型的解决方案.为避免目前应用神经网络方法中所存在的不足,通过采用原始数据分组方法进行网络训练误差的实时反馈和控制,较好地解决了神经网络训练过程中容易陷入“局部最优”和“过拟合”状态,并对BP神经网络预测结果给予了详细研究,训练误差和预测误差分析结果表明了该方法的可行性和计算结果的可信性.
The characteristics of the experiment and the difficulties in the experiment are analyzed. The traditional linear interpolation method for the manifold air pressure (MAP) sensor data in the high altitude calibration of electronically controlled engine can not be completely This paper presents a solution to BP neural network model with BP (back propagation) neural network model.In order to avoid the shortcomings of the current neural network method, this paper presents a real-time feedback of network training error by using the original data grouping method And control, which solves the problem of “local optimum” and “overfitting” easily trapped in the neural network training process, and gives a detailed study of the prediction results of BP neural network, training error and prediction error analysis The results show the feasibility of the method and the credibility of the calculation results.