In this paper, we investigate how to apply industrial unmanned aerial vehicles (UAVs) for autonomous power line inspection in smart grid from an energy efficiency perspective. Firstly, the energy consumption minimization problem is formulated as a joint optimization problem, which involves both the large-timescale optimization and the small-timescale optimization. Then, the NP-hard joint optimization problem is transformed to a two- stage optimization problem based on energy consumption magnitude and optimization timescale differences. Next, the first-stage and second-stage problems are solved by exploring dynamic programming (DP) and auction matching, respectively. Finally, the proposed algorithm is verified based on realistic power grid topology. Simulation results demonstrate that the proposed scheme achieves significant energy consumption reduction.