Aiming at the problems of low solution accuracy and efficiency in the optimal capacity allocation of energy storage system in wind farms with traditional methods, an improved multi-objective grasshopper optimization algorithm (IMOGOA) is proposed. Three strategies including Fuch chaos mapping, cosine adaptive parameters, and Levy flight are adopted for improvement, which makes the initial solution distribution of the algorithm more uniform, global exploration and local development more coordinated, and enhances the ability for algorithm to jump out of the local optimum. Performance tests are conducted to compare the improved algorithm with multiple algorithms such as multi-objective particle swarm optimization and et al. Experimental results show that the improved algorithm has better optimization accuracy and stability. When applied to the optimal capacity allocation for hybrid energy storage system in wind farms, compared with other algorithms, the improved algorithm can quickly find the Pareto optimal solution set. While meeting the system requirements, it minimizes the cost of the hybrid energy storage system, which verifies the effectiveness of the algorithm on improving strategy and its applicability to practical optimization problems.