Generalized Predictive Control Method Based On Improved Fuzzy Logic And Its Application

Model predictive control(MPC)is an advanced control strategy based on multiobjective optimization,which is extensively used in industrial process control systems such as petrochemical,steel metallurgy,and so on.As a subclass of MPC,generalized predictive control(GPC)integrates adaptive control features on the basis of traditional MPC,and has better robustness to external disturbances in industrial control systems.However,as a control strategy based on optimization problem solving,the GPC controller contains a series of coupled design parameters such as the softening factor and the weighting coefficient of the control variable.Therefore,how to efficiently tune these parameters in accordance with the controlled system is a pivotal component of GPC methodology design.In addition,the existing GPC cost function generally only contains the input and output signal errors of the system,which is difficult to describe the dynamic performance of the controlled system accurately and thereby failing to full exploit the performance advantages of the GPC controller.To solve the above problems,a series of GPC parameter tuning algorithms based on improved fuzzy logic and a new GPC control method based on improved cost function are proposed in this thesis,which specifically includes the following contents:Firstly,a GPC parameter tuning method based on improved fuzzy logic and event-triggered mechanism(ETM)is proposed for multiple-input-multiple-output(MIMO)industrial process control systems.The fuzzy system model is established by using novel fuzzy logic target parameters such as slope and the Gaussian bilateral membership function,which improves the fit of membership function to more accurately representing the current state of the system.Then,an on-line segmentally tuning strategy of GPC parameters based on improved fuzzy logic is proposed,and further combined with ETM to optimize the sampling and updating frequency of the on-line tuning algorithm to reduce the computing burden of the GPC controller.Moreover,the correctness and effectiveness of the proposed algorithm are verified through its application to a variable air volume air conditioning system.Secondly,for the process control system with model uncertainty,a GPC parameter tuning algorithm based on improved type-2 fuzzy logic is proposed to cope with the model mismatch between the controller model and the actual controlled object by using the uncertainty coverage of type-2 fuzzy logic to ensure the robust performance of the GPC controller.In addition,the sparrow search algorithm(SSA)is used to construct the membership degree function to improve the fuzzy effect of the type-2fuzzy logic system,and the ETM is used to optimize the operational efficiency of SSA optimization algorithm to avoid the problem of long convergence time when the fitness function is relatively complex.Moreover,a GPC parameter tuning algorithm based on improved type-2 fuzzy logic is proposed,and the correctness and effectiveness of the proposed algorithm are verified by simulation experimental.Finally,to address the problem that the cost function of the existing GPC controller only contains system input and output errors,which is difficult to accurately describe the dynamic performance of the controlled system,an improved GPC cost function containing time domain performance indexes is proposed,and the corresponding GPC control algorithm is designed to further improve the dynamic performance of the system.In addition,the convergence and stability of the proposed GPC algorithm are strictly proved by theoretical analysis.Moreover,the correctness and effectiveness of the proposed algorithm are verified by a semi-physical experiment platform of variable air volume air conditioning system.