基于BP神经网络的节能车弯道降速数学模型分析

Curve Deceleration Analysis of Energy-saving Vehicle Based on BP Neural Network Model

针对节能车为有效降低燃油消耗率,在整车经过弯道时,发动机需怠速滑行的特点,对GPS车载数据采集系统得到的数据进行弯道部分数据提取。基于提取出的北京金港国际赛车场150余组数据,通过引入BP神经网络算法的方式,建立了节能车弯道降速的BP神经网络预测模型。在经遗传算法优化后,拟合优度的结果显示,BP神经网络的预测较为接近真实情况,预测效果较为良好,可以用于节能车滑行过弯速度变化情况预测,并为不同赛道不同工况下整车速度分配提供参考依据。

Aiming at the characteristic that the engine needs to idle in order to effectively reduce the fuel consumption rate of the energy-saving vehicle when the vehicle passes through the curve, the data obtained by the GPS vehicle data acquisition system are extracted from the curve parts. Based on over 150 sets of data extracted from Beijing Jingang International Circuit, a BP network prediction model of deceleration in the curve for energy-saving vehicle is established by employing the BP neural network algorithm. After optimized by genetic algorithm, the results of goodness of fit indicate that the prediction of BP neural network is relatively approximated to the reality, and the prediction effect relatively good. And the BP network model can be used for predicting the speed change for energy-saving vehicle sliding over curves, providing a reference basis for vehicle speed distribution at different situation on different racing tracks.