Research Of Model Predictive Control In Buck-Boost Converter

With the continuous development of science and technology,the DC/DC converter in the production and lives is playing an increasingly important role,and the higher requirement about the performance of DC/DC converter is proposed.Therefore,a better control algorithm is needed to make DC/DC converter have better performance.Model predictive control(MPC)has been widely applied because of the excellent handling constraints capacity and the excellent control effect.With the development of the chip technology,the large online computational burden no longer limits the spread of MPC.Researching on the Buck-Boost converter,combining the MPC which is applied to the converter output voltage control strategy,focusing on the target converter fast response,a method of time optimal MPC is proposed.Based on the analysis of Buck-Boost converter topology and operating principle,the nonlinear model of converter is obtained by using state-space averaging method.And then a linear model is obtained by linearization near the stability point of the converter.With this linearization model as a model of MPC,and selecting the appropriate target function and constraints,the Buck-Boost converter MPC algorithm is proposed obtained.For the demand of convert needs to be a new stable state when the output voltage changes,a time optimal MPC based on polyhedral reachable set is proposed,which introduces a new method of reducing the offline calculation of a series of reachable set as well as one step MPC to reduce the amount of online calculation.Through the simulation on Matlab/simulink,the MPC and time optimal MPC can be proved to be correct and feasible.