When unmanned ground equipment travels autonomously in a complex environment with multiple obstacles, it is required to maintain a certain safety distance from the obstacles in addition to planning the path effectively and fastly. For the traditional planning algorithms, there are problems such as easy to fall into the minimal value or the planning path is close to the obstacles. In this paper, a new path planning method is proposed, which integrates the Flow Field Pathfinding (FFPF) algorithm and the Artificial Potential Field (APF) method, replaces the APF endpoint attractive vector with the cost difference between points and points in the FFPF as a vector, combines it with the obstacle repulsion vector in the APF to form a new potential field, and then plans the path according to the direction of the gradient descent. On this basis, in order to avoid local potential energy equilibrium, the potential energy field function is improved by making the obstacle combined force vertical vector, which effectively solves the local minima problem. By comparing with traditional path planning algorithms in complex environments such as concave obstacle, S-shaped obstacle, multi-obstacle and complex S-shaped obstacle environments, the method not only can plan paths quickly, but also generates paths that maintain a certain safety distance from obstacles.