The research in the field of the localization and navigation of mobile robot is becoming more and more significant. Among them, the local real-time path planning for robot autonomous navigation in complex environment is particularly important. However, the current local path planning algorithms are all planning the local path in the same plan-scope. Sometimes, this situation leads to a waste of time and computing resources. Therefore, in this article, a new path planning method called the variable planning size (VPS) for robot in static workspace is presented. In this paper, the graph theory is the basic algorithm of path planning. In addition, the mobile robot using this algorithm could autonomous to decide the planning scope according to the complexity of the surrounding environment. Therefore, in a complex environment, the robot is more efficient to find a local optimal path. As a result, the VPS path planning algorithm can plan out of the local optimal path preferably.