A method of simplified modeling based on kernel function principal component analysis

As the complexity of system increases, the calculation in the control process has grown in index. It would effect the stability and control precision of system. In this paper input character vectors are extracted based on kernel function principal component analysis, input space dimension is simplified and input vector space is reconstructed. Linear regression is completed by support vector machine and simplified model of control system is built. By controlling beam and ball control system, the result indicates the complexity of system based on kernel function principal component analysis has decreased, also control precision and general ability are improved. The experimental results show that the method is very effective.