Deep and reinforcement learning-based real-time online path planning method of

Abstract

The present invention provides a deep and reinforcement learning-based real-time online path planning method. According to the method, the high-level semantic information of an image is obtained through using a deep learning method, the path planning of the end-to-end real-time scenes of an environment can be completed through using a reinforcement learning method. In a training process, image information collected in the environment is brought into a scene analysis network as a current state, so that an analytical result can be obtained; the analytical result is inputted into a designed deep cyclic neural network; and the decision-making action of each step of an intelligent body in a specific scene can be obtained through training, so that an optimal complete path can be obtained. In an actual application process, image information collected by a camera is inputted into a trained deep and reinforcement learning network, so that the direction information of the walking of the intelligent body can be obtained. With the method of the invention, obtained image information can be utilized to the greatest extent under a premise that the robustness of the method is ensured and the method slightly depends on the environment, and real-time scene walking information path planning can be realized.

Description

technical field [0001] The invention relates to the fields of computer image processing and machine learning, and specifically relates to a real-time online path planning method of deep reinforcement learning, which implements path planning of real-time scenes by applying deep learning and reinforcement learning. Background technique [0002] Traditional path planning methods include simulated annealing algorithm, artificial potential field method, fuzzy logic algorithm, tabu search algorithm, etc.; intelligent bionics methods include ant colony algorithm, neural network algorithm, particle swarm algorithm, genetic algorithm, etc.; there are also some artificially invented Algorithms are widely used because of their excellent characteristics. These algorithms generally have strong path search capabilities and can play a good role in discrete path topology networks, including: A* algorithm, Dijkstra algorithm, Floyd algorithm, etc. With the continuous development of science a...

AI Technical Summary

Problems solved by technology

[0013] Neural network and reinforcement learning each have certain problems. Neural network has excellent learning ability, but poor generalization ability is its fatal shortcoming. For reinforcement learning, when the system becomes complex, a large number of parameters are required to describe it. This causes a combinatorial explosion of state-space-to-action-space mappings, which in turn affects the optimization problem of action decisions

There are no research reports on path planning through deep reinforcement learning using image parsing information

Images