Modeling management systems using multiple KALMAN filter

In this paper, we propose a method of modeling and prediction of management systems using multiple model Kalman filter. A management system may be very complicated and nonlinear. A complicatedly nonlinear system may be linearized responding to several key operating points, for instance N operating points. Then N linearly simple models are obtained to represent the complicatedly nonlinear system. The multiple Kalman filter is applied to modeling the complicated management system. Simulation studies are conducted using MATLAB language. Simulation results show that the method is effective.