论文部分内容阅读
改进和实现了支持向量机用于函数回归估计的算法,并将支持向量机和粒子群优化算法运用于压气机特性计算,以寻找压气机进口换算空气流量和效率随增压比、转子转速、压气机进气角度变化的非线性函数关系,将进气角度作为一个影响因子列入考虑范围,改进了以往只考虑压气机特性随转子转速和增压比变化的计算方法.用支持向量机回归估计压气机特性随各因子变化的非线性函数,用粒子群优化算法对非线性函数逼近度进行全局优化,计算结果表明,设计的算法能准确地回归估计出描述压气机特性的非线性函数,算法是有效的.
The algorithm of support vector machine (SVM) for function regression estimation is improved and implemented. The support vector machine and particle swarm optimization algorithm are applied to compressor characteristics calculation to find the ratio of air flow rate and efficiency of compressor inlet with turbocharger ratio, rotor speed, Compressor intake angle change of non-linear function of the relationship between the intake angle as an impact factor to be taken into account, improved in the past only consider the compressor characteristics with the rotor speed and turbocharger ratio change calculation method using support vector machine regression The particle swarm optimization algorithm is used to globally optimize the approximation degree of nonlinear function. The results show that the designed algorithm can accurately regress and estimate the non-linear function describing the characteristics of the compressor, The algorithm is valid.