机构地区: 中国农业大学工学院
出 处: 《计量学报》 2008年第4期329-333,共5页
摘 要: 在阐述支持向量机(SVM)和最小二乘支持向量机(LS-SVM)的原理和算法并对两者的特点进行比较后,为装载机载重测量建立了基于LS-SVM的软测量模型,并从核函数选择以及核参数确定两方面阐述了LS-SVM软测量建模的过程,最后与RBF函数网络以及BP网络的软测量建模结果进行对比。仿真分析结果表明,LS-SVM同时兼顾了精度和泛化能力两方面的性能,其最大泛化误差仅为6.863 8×10-6,是适合装载机载重软测量的建模方法。 After the nonlinear regression algorithm of support vector machineis (SVM) and least square support vector machines (LS-SVM) are introduced and their characters are compared, the soft sensor modeling method for dynamic weighing of wheel loader based on LS-SVM is given. Through choosing the kernel function and selecting its parameter, the process to set up the nonlinear model is discussed in detail. Finally the soft sensor modelings based on BP and RBF neural network are compared with that of the LS-SVM. The emulation result indicates that the LS-SVM has the best integrative character, and its generation error is only 6.8638 × 10^-6, so LS-SVM is an effective method for dynamic weighing of wheel loader.