Research On Optimization Of Photovoltaic System Control Method Based On Particle Swarm Optimization

As the most promising energy source,the solar photovoltaic(PV)power generation has developed rapidly with the support of financial subsidy policies in many countries.At present,the solar photovoltaic power generation industry presents two development trends.Firstly,the financial subsidies of many countries are continuously reduced or even cancellation,which leads to the cost performance of PV power generation will become the most important factor to promote the sustainable development of the industry.Secondly,the newly installed installations are mainly distributed PV systems,which will become the main utilization form of solar photovoltaic power generation.PV systems are divided into grid-connected and offgrid types,both of which have mismatching problems and harmonic optimization control problems.The mismatching problem is mainly caused by partial shade and module parameter mismatch,which will not only reduce the efficiency of the PV system,but also cause multiple peak points on the output p-v curve,making it difficult to track the maximum power point quickly and accurately.PV power generation is intermittent and random.With the increase of installed capacity of distributed PV system,the harmonic injected into the power grid increases rapidly,and reducing the output harmonic of PV system becomes very urgent and important.In this paper,the control method optimization problem of PV system based on particle swarm optimization is systematically studied.The main work and innovations are summarized as follows:(1)For the mismatching problem of PV system,a simulation model is established to further study the efficiency reduction of PV system caused by mismatching and the multi-peak curve problem of PV array output caused by mismatching.Research shows that the efficiency of grouped MPPT PV system is less affected by partial shade,and the affected degree is obviously less than that of entralized MPPT PV system.The efficiency of grouped MPPT and centralized MPPT PV system is affected by the parameter mismatch of modules in a similar way,and the influence degree increases rapidly with the increase of parameter mismatch.By analyzing the characteristics of PV array output multi-peak curve,it is found that there may be a situation that the power of multiple peak points is very close to each other and it is difficult to accurately locate the region of global peak point directly.Therefore,the rapid scanning strategy of p-v curve is the core of multi-peak MPPT method.(2)The optimization accuracy and optimization reliability of PSO algorithm deteriorate rapidly with the increase of spatial dimension.To solve this problem,a multi-swarm and multi-velocity particle swarm optimization algorithm is proposed to solve complex high-dimensional problems such as harmonic optimization of PV systems.It consists three particle swarms and three speed update modes.Information sharing is realized between the particle swarms.Different particle swarms adopt different speed updating methods.The main particle swarm is the basic particle swarm,which constitutes the rest of the particle swarms;the global auxiliary particle swarm uses two speed update methods to increase the diversity of the optimization trend and prevent it from falling into partial optimum;and the local auxiliary particle swarm is composed of particles with larger fitness to enhance the local search ability.Eight improved particle swarm optimization algorithms are selected as comparison objects,and 14 benchmark functions are tested in 10-dimensional,30-dimensional and 100-dimensional problem spaces.The simulation results show that when the dimension of the problem rises from 10 to 30,the optimization results of most comparison algorithms become worse obviously;when the dimension of the problem rises to 100,the optimization results of all comparison algorithms become worse obviously,and the optimization effects of some comparison algorithms are very poor.However,the optimization results of the multi-swarm and multi-velocity particle swarm optimization algorithm are almost constant when the problem dimension rises from 10 to 30,100,and is significantly better than the comparison algorithms.(3)To solve the optimization problem of PV inverter harmonic control,a genetic algorithm–particle swarm optimization(GA-PSO)is proposed to optimize SVPWM control sequence to reduce the output harmonic content of three-phase PV inverter system.The optimization freedom degree of the SVPWM is analyzed.Three SVPWM control sequence optimization strategies with different optimization potentials are proposed for these degrees of freedom,and corresponding optimization simulation models are established for simulation analysis The experimental platform is built to verify the optimal SVPWM control sequence obtained by the optimization algorithm.Simulation verification and experimental results show that the average generation number of GA-PSO is only 1/500 of the genetic algorithm and immune algorithm,and the weighted total harmonic distortion is reduced by more than 20% compared to these two conventional algorithms.(4)For the maximum power point tracking problem in multi-peak cases,a multipeak MPPT control method based on the improved particle swarm optimization algorithm is proposed to improve the tracking speed and precision in the multi-peak case.The performance characteristics of existing single-peak and multi-peak MPPT methods are analyzed.The speed updating formula of PSO is modified by the I-V and P-V curve characteristics of PV array outputs.The co-evolution principle of the improved PSO multi-peak MPPT method is proposed.The simulation model of PV system is established and the performance of various MPPT methods is compared and analyzed.The simulation results show that the improved PSO multi-peak MPPT method can still track the global maximum power point quickly and accurately under the extreme multi-peak P-V curve.The precision of the improved PSO multi-peak MPPT method is significantly better than Perturbation and Observation(P&O)and PSO method.Its tracking accuracy is better than P&O and its tracking speed is better than PSO-method,which help improve the MPPT efficiency.