Objective Regarding the modulation of cascaded H-bridge inverter the selective harmonic elimination technology has the advantages of low switching loss elimination of specific subharmonics and high transformation efficiency. However the traditional particle swarm optimization PSO algorithm has poor convergence and is easy to be locally optimized when solving harmonic elimination equations. A modified particle swarm optimization MPSO algorithm was proposed. Methods In this algorithm the nonlinear inertia weight was used to replace the linear inertia weight and chaotic mapping was introduced into the nonlinear inertia weight to produce the random quantity with better randomness. The new inertia weight can balance the global search ability and local search ability of particles so that particles have the ability to jump out of the local optimum in the late stage. In addition the algorithm optimized the updating mechanism of speed and position to enhance the convergence speed of the algorithm and to ensure that the particles still have a certain population diversity advantage in the later stage. Results According to the nonlinear harmonic elimination equations of cascaded H-bridge inverter the fitness function was established to ensure the maximum reduction of the target harmonic under the premise of the output voltage base wave. When applied to SHEPWM of cascaded H-bridge inverter MPSO algorithm obtained optimized switching angle in the range of 1 ~ 1. 2 adjustment system. Compared with the PSO algorithm MPSO can improve the convergence accuracy and solving success rate. The seven-level CHB inverter simulation platform verified that the optimized switching angle obtained by MPSO algorithm can effectively eliminate the 5th and 7th harmonics. Conclusion By using the nonlinear inertia weight optimizing the velocity of particles and applying position update mechanism of particles the success rate of switch angle solution can be increased and the obtained solution can effectively eliminate the target subharmonics.