Application Of Improved Adaptive Genetic Algorithm In The Design Of Zero Sequence Filter Optimization

In the low-voltage distribution network with the three-phase four wire circuit,the existence of a large number of non-linear load,resulting in serious harmonic pollution,one of them is called zero-sequence currents and each phase of them has the same size and same phase,causing the neutral line overload.Passive filters and active filters cannot meet the requirements of voltage source nonlinear harmonic suppression due to their inherent limitations,the flux-compensated zero-sequence filter without capacitor topology has the characteristics of not being influenced by the power grid parameters,and zero-sequence filter has no harmonic amplification and no detuning,and it can achieve the harmonic low-impedance channel of voltage-type non-linear load and have obvious balance function to the serious unbalance caused by single-phase loads.The leakage reactance of the zero-sequence filter is an important parameter that affects the filtering effect.Reducing the equipment loss is also an important measure to improve the reliability of the filter.Both are closely related to the electromagnetic and structural parameters of the zero-sequence filter.Therefore,the optimal design of zero-sequence filter is proposed.A 150 kVA/400V zero sequence filter is as the research object,an improved adaptive genetic algorithm based on individual concentration is proposed and applied to the zero sequence filter optimization design with the object functions of zero-sequence leakage resistance and no-load loss.The crossover probability and mutation probability of this algorithm can be adjusted automatically according to the individual concentration.At the same time,the premature convergence of the solution in the evolutionary process is avoided by adjusting the initial population fitness and adding immigration strategy,making the optimization variable jump out of the local feasible solution domain.The genetic algorithm,the adaptive genetic algorithm and the improved adaptive genetic algorithm are used to optimize the structural parameters of the zero sequence filter.The simulation results show that the improved adaptive genetic algorithm has the ability of global optimization.Under the same constraints,it can reduce the zero-sequence leakage reactance and no-load loss to a greater extent,improved adaptive genetic algorithm helps to improve the electromagnetic and structural parameters of the zero sequence filter.Taking a set of 150kVA/400V flux-compensated zero-sequence filter as the test object,the filter effect,no-load current,equipment efficiency,voltage drop and other performance parameters are tested.The test results show that the neutral current drops below 10%before the zero-sequence filter is put into operation,the no-load current,equipment efficiency and voltage drop deviation are all within the engineering requirements.