Research On The Optimal Control Of Single-Phase Photovoltaic Systems Based On Adaptive Dynamic Programming

Energy is the basic material for survival and development of human and society,and it is also the blood of the world economy.In expected future,more than 60%of the worldwide energy will be indirectly consumed by converting into electrical energy.Along with the environmental problems caused by fossil fuels and the growing demand for electricity,the demand for using the new clean energy generation is increasingly urgent.As one of the most clean and huge renewable energy resources,solar energy has become focus of the world.Therefore,the study of photovoltaic grid-connected control methods is a hot topic for the industry and academia in the current domestic and international.Combing with the current research and development of control strategies in photovoltaic grid-connected systems,the control methods of single-phase photovoltaic grid-connected system has been researched deeply in this thesis.The main content is the research on the application of optimal control in single-phase photovoltaic systems,and the main contributions are:Firstly,the background and significance of photovoltaic generation has been elaborated further firstly,and the control methods of the two-stage photovoltaic generation system have been discussed in detail.For the application of optimal control in single-phase photovoltaic grid-connected system,the basic principle of optimal control has been presented.Then we have told about the mathematical derivation of dual heuristic programming and zero-sum two-person differential game which laid the foundation for the design of optimal controller.For the single-phase photovoltaic grid-connected system,a current continuous mode(CCM)Boost Converter based small-signal model and a LC-filtered based single-phase photovoltaic grid-inverter model have been established.Secondly,to achieve the maximum power point tracking,during the converter MPPT controlled,a solution to the zero-sum two-person differential game containing control constraints has been presented.Then we have used neural networks to solve Hamilton-Jacobi-Isaacs(HJI)equation to get the optimal controller.When the photovoltaic array works at the maximum power point,a solution to affine nonlinear systems with control constraints has been presented.We have also used dual heuristic programming to get the optimal controller.Simulation has verified the effectiveness and good characteristics of the proposed method for the boost chopper converter.Finally,in the design of optimal controller for single-phase grid connected PV inverter,considering the switch duty cycle with saturation constraints,the mathematical model of Single-phase grid connected photovoltaic power generation system inverter has been established.Then we have put forward a optimal controller which used a neural network(NN)-based online simultaneous policy update algorithm(SPUA).And we have also proved the convergence of this algorithm.Finally,simulations have been provided to illustrate the effectiveness and the robustness for the external disturbances of the proposed method.