In this paper, multi-step backtracking in reinforcement learning is introduced into the traditional transmission grid planning, and the α-Q(λ) algorithm based on multi-step backtracking is proposed ...