中文摘要：

针对非线性控制系统难以建立精确数学模型的情况,以焦炉集气管压力系统为对象,提出了一种基于多核线性支持向量机的辨识建模方法.通过分析多种核函数对于数据的分类精度,利用训练数据对系统进行辨识建模.仿真结果表明,基于多核线性支持向量机的辨识建模算法具有相当高的辨识精度,能够解决一类非线性系统的建模问题.

英文摘要：

It is difficult for nonlinear control system to establish a precise mathematical model. Taking pressure system of coke oven gas collector as object, an identification modeling was presented based on multi kernel function supported vector machine. The classification accuracy of the data was analyzed, and the system was modeled by the training data. The simulation results show that the identification algorithm based on linear support vector machine has high identification accuracy, and can solve the modeling problem of a class of nonlinear systems.