Over the past few decades, as the main tool of intelligent material transportation, automatic guided vehicles (AGVs) have been widely used in modern production systems, logistics, transportation, industry, and commerce to further improve productivity, reduce labor costs, raise energy efficiency, and enhance safety. Path planning is a key issue in the field of AGVs to ensure that they do not collide with obstacles during movement and reach the destination as fast as possible to complete the assigned task. We propose two different and crucial operating environments in this paper. More specifically, in a static environment, a multi-objective mathematical model is established with the shortest path and the maximum smoothness, and the improved Levy random quantum particle swarm optimization (LRQPSO) algorithm is used to solve the proposed model and screen the AGV’s driving path. In a dynamic environment, an inductive steering algorithm (ISA) that considers the movement of obstacles is proposed to make the AGV avoid obstacles rationally. By combining the steering characteristics of the two environments, AGV speed control rules are set and applied to the steering process in complex environments to ensure that the AGV can travel more smoothly and quickly. Simulation results show that the proposed method can ensure the obstacle avoidance and flexible steering of an AGV, and improve the driving speed and work efficiency in the two environments. In addition, compared with the conventional algorithm, the smoothness, operation speed, and work efficiency of the AGV are significantly increased using the improved LRQPSO algorithm and ISA.