Research on Autonomous Driving Decision-making Strategies based Deep Reinforcement Learning

Zixiang Wang,Hao Yan,Changsong Wei,Junyu Wang,Shi Bo,Minheng Xiao

2024-08-06

Abstract:The behavior decision-making subsystem is a key component of the autonomous driving system, which reflects the decision-making ability of the vehicle and the driver, and is an important symbol of the high-level intelligence of the vehicle. However, the existing rule-based decision-making schemes are limited by the prior knowledge of designers, and it is difficult to cope with complex and changeable traffic scenarios. In this work, an advanced deep reinforcement learning model is adopted, which can autonomously learn and optimize driving strategies in a complex and changeable traffic environment by modeling the driving decision-making process as a reinforcement learning problem. Specifically, we used Deep Q-Network (DQN) and Proximal Policy Optimization (PPO) for comparative experiments. DQN guides the agent to choose the best action by approximating the state-action value function, while PPO improves the decision-making quality by optimizing the policy function. We also introduce improvements in the design of the reward function to promote the robustness and adaptability of the model in real-world driving situations. Experimental results show that the decision-making strategy based on deep reinforcement learning has better performance than the traditional rule-based method in a variety of driving tasks.

Machine Learning

What problem does this paper attempt to address?

This paper aims to solve the problem of efficient and safe decision - making for self - driving cars in complex and changeable traffic environments. Traditional rule - based decision - making methods have limitations because they rely on the prior knowledge of designers and are difficult to deal with complex traffic scenarios. This paper proposes to use advanced deep reinforcement learning models to independently learn and optimize driving strategies. By modeling the driving decision - making process as a reinforcement learning problem, the adaptability and decision - making ability of self - driving cars in complex traffic environments can be improved.

Specifically, in the paper, **Deep Q - Network (DQN)** and **Proximal Policy Optimization (PPO)** are used for comparative experiments. DQN guides the agent to select the best action by approximating the state - action value function, while PPO improves the decision - making quality by optimizing the policy function. In addition, the design of the reward function in the paper has also been improved to promote the robustness and adaptability of the model in actual driving situations. The experimental results show that the decision - making strategies based on deep reinforcement learning perform better than traditional rule - based methods in various driving tasks.