Automobile longitudinal multi-state control method based on deep reinforcement learning priority extraction

Abstract

The invention discloses an automobile longitudinal multi-state control method based on deep reinforcement learning priority extraction. The method comprises the following steps: 1, defining a state parameter set s and a control parameter set a of automobile driving; 2, initializing deep reinforcement learning parameters, and constructing a deep neural network; 3, defining a deep reinforcement learning reward function and a priority extraction rule; 4, training the deep neural network and obtaining an optimal network model; and 5, obtaining a t moment state parameter st of the automobile, inputting the t moment state parameter st into the optimal network model to obtain an output at, and executing the output at by the automobile. According to the method, the multi-state driving of the automobile in the longitudinal direction is completed by combining the priority extraction algorithm and the deep reinforcement learning control method, so that the safety of the automobile in the driving process is higher, and traffic accidents are reduced.

Description

Technical field[0001]The present invention relates to the field of longitudinal polymorphic control of smart vehicles, and is specifically a vehicle longitudinal polymorphic control method based on deep strengthening learning priority.Background technique[0002]With the rapid development of urban economy and the continuous improvement of people's living standards, the number of urban motor vehicles has also increased significantly, and the car has become an indispensable step-by-step tool, bringing fast and convenient, and also brings a series Security Question. Due to the limited number of technical capabilities of the driver or other uncontrollable external factors, traffic issues such as two cars or multi-cars often occur, while bringing the loss of life and property, it also caused great difficulties to the road. With the continuous development of automotive related technologies, many car companies have introduced adaptive cruise systems and emergency braking systems. The adaptiv...

AI Technical Summary

Problems solved by technology

The adaptive cruise system uses radar and other sensors to obtain road data ahead, and maintains a certain distance and speed from the vehicle in front according to the corresponding algorithm, but the adaptive cruise system is often turned on at a higher speed, such as above 25km / h, below this speed It requires manual control by the driver; the emergency braking system refers to a technology that can actively brake to avoid accidents when a car is driving in a non-adaptive cruise state and encounters an emergency in front of it, such as an emergency stop of the car in front or a sudden encounter with a pedestrian. , but there are related reasons such as sensor misjudgment and environmental error, it cannot be applied to a variety of driving environments, resulting in dangerous accidents

Images