Investigation On Shepwm Technology Based On Fuzzy Adaptive PSO

Power electronic equipment not only improves the production efficiency,but also brings a lot of harmonics.The frequent interruption of power electronic devices also increases the power loss.In order to improve the efficiency of power utilization and eliminate harmonics,PWM technology is a good solution.However,in high-power applications,due to electromagnetic interference,heat dissipation,switching loss and other factors,the switching frequency of high-power switching devices is often limited,usually only a few hundred Hz.At this time,the traditional PWM technology will make the output waveform contain a lot of low order harmonics.If SHEPWM is used,the harmonics can be eliminated precisely at the same switching frequency.The technical principle of SHEPWM is introduced firstly in this paper,and the mathematical model of bipolar three-phase SHEPWM is established.Then,the basic principles of traditional Newton iterative algorithm and basic PSO algorithm for solving nonlinear transcendental equations are introduced.In order to solve the problem of slow convergence and fall into local optimum of PSO easily,a fuzzy adaptive particle swarm optimization intelligent algorithm is proposed which uses the fuzzy control theory to dynamically adjust the inertia weightω,learning factors c₁and c₂of PSO,based on the three parameters of particle fitness coefficient,particle similarity and particle fitness error.Then,in the simulation model of bipolar three-phase voltage source inverter circuit,through the comparative study of SHEPWM technology,SPWM technology and SVPWM technology,and fuzzy adaptive PSO algorithm,Newton iterative algorithm,traditional optimization algorithm with constraints,GA and basic PSO algorithm are used to solve the nonlinear transcendental equations established by SHEPWM technology.We proved that the switching angle obtained by the SHEPWM technology based on fuzzy adaptive PSO can not only effectively eliminate the specific number of harmonics,but also has the highest probability of convergence and the best performance,which effectively improves the defects of traditional intelligent algorithm.By building a physical experiment platform,the feasibility of applying fuzzy adaptive PSO in SHEPWM technology is verified.Finally,through statistical analysis,the selection principle of three key parameters in fuzzy adaptive PSO is analyzed which contained adaptation threshold F₁,the threshold of the best fitness difference between two generations of particles error,Threshold of difference between individual optimal fitness and global optimal fitness in contemporary PSO F₂.The ranges of three key parameters are determined,and the rationality and effectiveness of the parameter determination method are proved by simulation.