Social emotional optimization algorithm (SEOA) is a novel swarm intelligent population-based optimization algorithm by simulating the human social behaviors. However, it's diversity is decreased increased when solving high-dimensional multi-modal optimization problems. Therefore, in this paper, a new hybrid SEOA with Metropolis rule is introduced to enhance the exploration capability. To test the performance, five famous benchmarks are selected, and compared with the standard version with different dimensions. Simulation results show this hybrid algorithm can increase the global search capability significantly.