Abstract: To solve the problem that the path planning efficiency of the traditional one-way rapid exploring random tree (RRT) algorithm is low and easy to fall into the local minimum point, a path planning algorithm of adaptive heuristic multiple rapid exploring random tree (AHMRRT) was proposed. On one hand, the AHMRRT algorithm based on the multi-random tree construction strategy can simultaneously generate four random trees from the starting point, target point, and sub-target point for extended search for improving path planning efficiency. On the other hand, the algorithm adaptively changed the generation strategy of new node according to the conditions of obstacles in the environment by adding an adaptive heuristic bias factor during the growth of a single random tree. When exploring free space, the algorithm quickly expanded to the target point under the influence of the bias factor to improve the search efficiency. When exploring multiple obstacle spaces, the algorithm called the random sampling function to prevent it from falling into the local optimum. In the simulation experiment, the comparative experiments of AHMRRT and PGRRT and BRRT algorithms were designed in four different environments. The simulation experiment results prove the feasibility and efficiency of the algorithm.