Abstract: In order to optimize the inventory cost of power materials and ensure the balance of inventory structure, based on the application of multi population parallel genetic algorithm, a multi-level inventory optimization model for materials was designed and studied using a certain power enterprise as an example. Calculate the total cost of multi-level inventory in the power supply chain based on the cumulative relationship of supply chain levels; based on the calculation results, establish an optimization model with the goal of minimizing the annual total cost; introducing a multi subpopulation collaborative evolution mechanism for genetic algorithm optimization, using the optimized algorithm to solve the model, output the optimal solution, and complete the optimization of multi-level inventory management of power materials. The comparative experimental results show that the designed method performs the best in multi-level inventory optimization of power materials.