基于最小二乘支持向量机的超磁致伸缩执行器磁滞非线性模型

本资料包含pdf文件1个，下载需要1积分

The model of giant magnetostrictive actuator based on least squares support vector machine and the experimental data was developed to describe the hysteresis nonlinearity of giant magnetostrictive actuator. The results show that least squares support vec- tor machine can fit experimental data accurately. In order to solve the model quickly, only a half of experimental data were used for modeling, and the model based neural network was also built to make comparison with the model based on least squares support vector machine. The results show that the displacement error of model based on least squares support vector machine is less than 1.2μm, the displacement error of model based on neural network is more than 1.5 μm.为准确模拟超磁致伸缩执行器的磁滞非线性特性，基于最小二乘支持向量机与实验数据建立了超磁致伸缩执行器的模型，结果显示所建模型都能够充分逼近非线性实验数据点。为提高模型的求解速度，仅选用一半的实验数据建立了小样本模型，并与神经网络所建小样本模型对比，结果表明：基于最小二乘支持向量机所建模型的位移预测误差小于1．2μm，而基于神经网络所建模型的位移预测误差大于1．5μm。