基于自适应变异PSO-SVM的APU性能参数预测模型

APU performance parameter prediction model based on adaptive variation PSO-SVM

为提高辅助动力装置(auxiliary power unit,APU)性能参数预测的精度,针对支持向量机(support vector machine,SVM)模型在实际使用中遇到的参数选择问题,采用自适应变异粒子群优化(particle swarm optimization,PSO)算法实现对SVM惩罚参数和核参数的优化选择,提出一种基于自适应变异PSO算法优化SVM的APU性能参数预测模型。进一步分析了预测模型不同预测步长对短期预测精度的影响。利用某型APU性能参数数据进行了验证,并与多种预测模型进行了对比实验。实验结果表明,对于排气温度(exhaust gas temperature,EGT)的预测,自适应变异PSO-SVM模型的平均绝对百分比误差(mean absolute percentage error,MAPE)比标准PSO-SVM模型低47%;对于滑油温度(oil temperrature,OT)的预测,自适应变异PSO-SVM模型的MAPE比标准PSO-SVM低29%,为短期APU性能变化趋势预测提供了一定的参考。

In order to improve the accuracy of performance parameter prediction of auxiliary power unit(APU),and solve the problem of parameter selection of support vector machine(SVM)model in practical use,adaptive mutation particle swarm optimization(PSO)is adopted to optimize the selection of penalty parameters and kernel parameters of SVM.And a prediction model of APU performance parameters based on adaptive mutation PSO algorithm is proposed.Furthermore,the influence of different prediction steps on short-term prediction accuracy is analyzed.The performance parameters of an APU are verified and compared with other prediction models.The experimental results show that for the prediction of exhaust gas temperature(EGT),the mean absolute percentage error(MAPE)of adaptive mutation PSO-SVM model is 47%lower than that of standard PSO-SVM model;for the prediction of oil temperature(OT),the MAPE of adaptive mutation PSO-SVM model is 29%lower than that of standard PSO-SVM,which provides some reference for short-term prediction of APU performance trend.