PSO-BP神经网络在露天矿卡车油耗预测中的应用

PSO and BP neural network application in open pit mine truck fuel consumption forecast

针对露天矿燃油消耗问题,利用粒子群优化算法对BP网络的权值和偏置进行优化,建立了基于粒子群优化BP神经网络的露天矿卡车油耗量预测网络模型.该方法使用由PSO优化的BP模型来拟合影响露天矿卡车油耗众多因素与油耗值之间的复杂关系.仿真结果表明:模型具有预测精度高、稳定性好等特点,适用于露天矿卡车油耗的预测,在露天矿燃油消耗预测中具一定的实用价值.

According to the problems existing in the fuel consumption of open pit enterprise, a mathematical model for the open pit truck fuel consumption forecast is set up by using PSO algorithm to optimize the weights and bias of BP neural network. The method uses the PSO to optimize the BP model which is used to do the fitting of the complicated relationship between fuel consumption and a number of important influences upon the open-pit mine truck fuel consumption. The simulation results show that the model has high prediction accuracy, good stability, and suitable for open pit truck fuel consumption forecast, which provides some practical value in the forecast of open pit fuel consumption.