With the construction and development of new power system, grid business presents high reliability, high security protection, high flexibility, massive access level characteristics, 5G as the frontier technology of wireless network access, with high speed, wide connectivity, low latency features and advantages, and diversified grid business communication needs are highly compatible. Based on the characteristics of 5G communication technology, this paper analyzes its practicality in the power system. The main two protocols of the current autonomous network routing protocol for power system are proposed, and the inter-cluster routing optimization of OLSR is carried out by using AO algorithm. Simulate the predation behavior of skyhawk, develop the search strategy in the optimization process of AO algorithm, and construct the mathematical model of AO optimization algorithm. A quasi-inverse solution is used on the basis of the inverse solution to further increase the population diversity and convergence speed of the AO algorithm, while an adaptive weight factor strategy is used to balance the global search and local exploration capabilities of the AO algorithm. Simulation experiments are utilized to investigate the performance of the IAO algorithm as well as the PDR and delay in the mobile scenario of the power system. Comparing the PDR of the three protocols at different expected delivery distances, IOLSR still maintains a delivery rate of about 28% at a distance of 350m-500m. The optimized IOLSR shows further reduction in delay compared to OLSR in most of the cases with an average delay of 10829.43ns.