摘要
针对高速列车动力建模问题,提出了基于最小二乘支持向量机(LS-SVM)的高速列车广义非线性模型子空间辨识方法。先给出描述高速列车单质点力学行为的随机离散非线性状态空间模型,并进一步构建了高速列车广义非线性模型;采用LSSVM回归方法构造广义非线性函数,并运用子空间辨识方法,直接由增广输入、输出数据得到高速列车广义非线性模型参数矩阵。最后对上述模型进行了数值仿真。结果表明:所提出的基于LS-SVM的子空间辨识方法比常规LS-SVM方法、线性子空间方法对列车模型具有更高的预报性能,用于高速列车的建模是有效的,可用于具有非线性、强耦合的高速列车运行过程数学模型的辨识。
Subspace identification based on least squares support vector machine （LS-SVM） is proposed for modelling general non- linear system of high-speed train. Firstly, a general nonlinear state-spaee model is established to describe the dynamic behavior of high-speed train as a single-point-mass object. Then, by using the LS-SVM regression method, the classical subspace identifica- tion algorithms are extended to the general nonlinear model of high-speed train. Finally, the numerical simulation is implemented and the results show that the proposed LSSVM subspace identification method has better predictive performance than LSSVM method and subspace identification method. Therefore, the proposed method is effective in the modeling for high-speed train.
出处
《中国科技论文》
CAS
北大核心
2015年第19期2225-2231,2241,共8页
China Sciencepaper
基金
国家自然科学基金资助项目(61263010
60904049)
江西省青年科学基金资助项目(20114BAB211014)
江西省教育厅研究项目(GJJ14399)
关键词
高速列车
广义非线性模型
最小二乘支持向量机
子空间辨识
high-speed train
general nonlinear model
least squares support vector machine
subspace identification
