Real Time Neural Network Path Planning Algorithm for Robot

Abstract

The emergence of robots not only changed the traditional industrial production mode, but also greatly promoted the progress of social civilization. Whether in daily life or in industrial production practice, the technical level of robots is improving every day, which emphasizes the high level of national science and technology. Robot path planning technology is an important part of robot research. The purpose of this paper is to focus on the research of robot path planning algorithm, learning and in-depth development. This paper introduces the discovery and communication of robot real-time neural network planning algorithm in machine learning, and analyzes and studies it. In order to study the robot real-time neural network path planning algorithm, through the experimental comparison of different algorithms, focus on exploring the effect of different algorithms on path planning. The research results show that the algorithm speed of robot on real-time network path is 24% higher than that of normal network path algorithm, and can be increased to 30% after deep mining, and the efficiency of machine algorithm can be increased to 35% under more complete algorithm. Therefore, the robot real-time neural network path planning algorithm can complete the task more efficiently.

Advancements in Learning-Based Navigation Systems for Robotic Applications in MRO Hangar: Review

“…The work in [ 44 ] comparatively analysed heuristic neural network algorithms for path planning and obstacle avoidance. Du et al [ 45 ] further explored the application of real-time neural network algorithms for generating collision-free routes to the destination, aiming to improve the motion control and obstacle detection accuracy of robot systems. Overall, this paper leveraged many learning-based navigation method studies, providing a broader understanding and recommendation on the application and suitability of recent models based on their functionalities in complex environments.…”

Research on the Local Path Planning for Mobile Robots based on PRO-Dueling Deep Q-Network (DQN) Algorithm

“…There are many effective path planning methods for obstacle avoidance at present. The traditional methods mainly include A * algorithm [1], Dijkstra algorithm [2], fuzzy control algorithm [3], genetic algorithm [4], artificial potential field method [5] and neural network [6]. Reinforcement learning algorithm [7] has received widespread attention because it can solve the shortcomings of traditional algorithms such as strong dependence on environment in robot path planning.…”