Abstract: In photovoltaic and hybrid energy storage system，variational mode decomposition（VMD）is

commonly used to realize the distribution of system residual power，but the number of decomposed modes and

secondary penalty factor value affect the error of power reconstruction directly，and the inertia weight error of

commonly used concave and S-type functions is large. Taking Pearson correlation coefficient as fitness function，an

exponential function inertia weight particle swarm optimization algorithm was proposed to optimize VMD

parameters，obtain the [K，α] in the VMD algorithm respectively，decompose the Phess and distribute the Phess to the battery and supercapacitor reasonably. Then，a hybrid energy storage power complementary optimal control strategy considering the state of charge（SOC）of supercapacitor was proposed to make its SOC operate in a stable region.In the example analysis，the symmetrical mean absolute percentage error（SMAPE）was used to compare and analyze the power reconstruction error under three inertia weight function optimization algorithms，the SOC of

supercapacitor and optimization control were considered. The results show that the power reconstruction error

obtained by using the exponential function inertia weight algorithm is the smallest，the SOC of supercapacitor is

controlled in the stable region to avoid overcharge and overdischarge and prolong its service life.

      Key words: power reconfiguration；exponential function inertia weight；variational modal decomposition

（VMD）；power complementary control