A new method of the performance evaluation for the multivariable control system

The traditional multi-variable filter correlation algorithm must know the transfer function, or the first few terms of the Markov matrix. However, the actual industrial systems are often unable to get the exact model-like system, limiting the practical application of this approach. In this paper, a new method based on the principal component analysis (PCA) was proposed to evaluate the performance of the multivariable processes. The properties of de-coupling and lowing orders of PCA were used to solve the problems in the performance evaluation for the multi-variable processes. The simulation results show that the new method is effective.