Study Of Multivariable Generalized Predictive Control

Predictive control as an available controlling algorism has been proved by a lot of emulations and practical applications. The domestic and overseas learners have done a lot of studies in single variable system, but the industry processes are always multivariable systems, GPC being an optimized controlling algorism, one of its merits is that it can handle multivariable controlling problems. This paper aims at studying complicated systems.Firstly, the paper mainly puts the important on studying multivariable generalized predictive control algorism, it has deduced this algorism based on existent theory, its peculiar decoupling capability can overcome the complexities and shortcomings of traditional separated and decoupling control. By simulation, it achieves a good result in tracing property, robustness and anti-interference, it can effectively handle the instability and non- minimum phase of the process. Aiming at its great calculating quantity, the paper has studied another algorithm which calls multivariable single generalized predictive control algorithm.Secondly, the paper has studied the time delay problem of the practical system. The traditional approximation method arises errors when it handles time delay system. So a farther study has been made, it add time delay to the controlling system and can control some time delay systems inerrably,...