基于混沌优化和SVM的发动机燃油消耗最低点软测量方法  

Soft sensor method for estimating engine minimum fuel consumption of based on chaos optimization and SVM

作　　者：周大为[1] 高翔[1] 夏长高[1] 

机构地区：[1]江苏大学汽车与交通工程学院,镇江212013

出　　处：《仪器仪表学报》2011年第2期463-468,共6页Chinese Journal of Scientific Instrument

基　　金：江苏省科技厅(BA2007089)资助项目

摘　　要：为了达到节省燃油的目的,通常需要利用发动机控制程序对不同工况下发动机燃油消耗量进行测算,以确定最佳的发动机工作点。由于发动机各个参数(如节气门开度、发动机转矩、发动机转速、输出功率和燃油消耗量等)之间存在复杂的高度非线性关系,常用算法难以快速得到发动机的最佳工作点。以某发动机的测试数据为例,采用预测精度高、泛化能力强的支持向量机算法拟合出该发动机主要运行参数之间的非线性关系,进而运用变尺度混沌优化算法分别测算出该发动机在给定功率、给定转速和给定转矩条件下,燃油消耗最低工作点所对应的发动机运转参数。最后,利用测功机对真实发动机进行的实验验证所提方法的有效性。In order to reduce fuel consumption,engine control program needs to continuously estimate the fuel consumption under different operating situations and adjust the engine to suitable running state.Because of the highly nonlinear relationship among the engine parameters,such as opening proportion of throttle,engine torque,engine speed,output power and fuel consumption and etc,traditional algorithms have difficulties in searching the optimal working state quickly.Taking the test data from a real engine as an example,support vector machine（SVM） algorithm that has high estimating accuracy and good generalization ability is used to fit the nonlinear relationships among different engine parameters.A mutative scale chaos optimization algorithm is employed to calculate the running parameters for minimum engine fuel consumption under fixed power,fixed torque and fixed speed.Verification experiment was carried out using a dynamometer;and experiment result shows the effectiveness of the proposed method.

关 键 词：混沌优化 支持向量机 发动机 燃油消耗 

分 类 号：U462.34[机械工程—车辆工程]