【摘要】 提出了柴油机燃烧系统参数优化的一种新方法——神经网络与正交试验结合法。这种方法以较小规模正交试验数据为神经网络的训练样本，利用神经网络良好的非线性映射能力来进行燃烧系统优化性能预测，达到了与大规模试验相同的优化效果，具有试验量少、效果好的优点更多还原

【Abstract】 This paper puts forward a new method to optimize the parameters of diesel engine combustion system——the combination of neural network and orthogonal experiment.With small scale orthogonal experiment data as the training samples of neural network,this method predicts the optimum property of combustion system by using the excellent nonlinear mapping capacity of neural network,so that it can achieve the same optimum effectiveness as with large scale experiments.This method has the advantage of small scale experiments and wonderful effectiveness.更多还原