Industrial vehicles are universally applied in logistics warehouses, where the workflow includes loading goods from container lorries, transporting to specific shelves, and unloading them. To handle the surge of logistics quantity along with rapid development of e-commerce, increasingly number of automated guide vehicles (AGVs) replaced the traditional ones to handle transporting missions, where automated path planning is one of the core technologies to lead this change. This work researches on principles and properties of algorithms applied in industrial path planning, especially focusing on algorithms based on graph searching, sampling, and reinforcement learning. Two scales of simulated warehouses maps and three typical algorithms are implemented in this work. By analyzing experimental data from several planning path missions, this work concludes that D* lite algorithm finds out the shortest paths all the times, RRT* algorithm always has highest computing speed, but SARSA algorithm does not perform as well as the others on both feasibility and efficiency.