Research On Single-phase Inverters Based On Sliding-mode Control

The use of green energy is advocated in this century,and utilizing new energy to solve the problem of increasing demand for electricity with insufficient power supply capacity,and the pollution of people’s living environment is getting worse,which is favored by scientists all over the world with the world’s energy consumption rising sharply.People began to realize the importance of the research on renewable energy grid power generation technology.In the process of converting renewable energy into electric energy,as the core of the inverter has become the focus of the researchers.The current study of inverter is mainly the design of the control algorithm and topology to improve system stability and power quality which is from inverter.In this dissertation,the single-phase full-bridge inverter with voltage source is taken as the research object,and multiple feedback control loops where voltage loop design is master are adopted for inverter.Two sliding mode controllers optimized for disturbance are designed to improve the output performance of inverter.The working principle of the inverter is researched and analyzed,and a simple mathematical model based on that is established.After introducing the reasoning process of the basic control principle of sliding mode,an ordinary sliding mode controller is designed according to the mathematical model of the inverter.At the same time,the system stability is proved and deduced.The experimental results are compared with PID algorithm,and it shows that the sliding mode control improves the dynamic and robust performance of the system.The controller is improved for reducing the influence of chattering on the system and the total harmonic distortion(THD)of the output voltage waveform further,which is combined with the disturbance observer and the neural network algorithm.The disturbance observer is a method to estimate system interference,which introduces equivalent compensation in control and does not need to establish accurate mathematical model for disturbance signal.The neural network can fit arbitrary nonlinear function well and reduce the influence of unknown disturbance effectively.The neural network adopts RBF network and extreme learning machine,which can approximate the disturbance term in the second order dynamic equation of inverter.And an adaptive law is designed to automatically correct the weight of network.In order to verify the performance of the optimized sliding mode controller,the comparison experiment is made with PID controller and ordinary sliding mode controller,and the results are analyzed.The results show that the THD of output voltage is about 2.24% and 3.88% at linear and nonlinear loads,respectively.The sliding mode adaptive controller can improve the output voltage waveform quality and system dynamic response performance,which can reduce the impact of chattering and disturbance on the system also.