Target Tracking and Path Planning of Mobile Sensor Based on Deep Reinforcement Learning

Abstract

Path planning is a classical problem of artificial intelligence, with a wide range of applications in defense and military, road traffic, and robotics simulation. However, most of the existing path planning algorithms have the problems of a single environment, discrete action space, and manual modeling. As a machine learning method that does not require artificially providing training data to interact with the environment, the deep reinforcement learning obtained by reinforcement learning has further enhanced the ability to solve practical problems. This paper proposes to use the DDPG (Deep Deterministic Policy Gradient) algorithm on the mobile sensor to achieve path planning on the target. The DDPG algorithm combines strategies such as DQN, ActorCritic, and PolicyGrient, which introduce deep reinforcement learning to continuous action space and further enable decision-making judgments in complex continuous environments.