Traffic control method based on Categorical-DQN optimistic exploration

A traffic control, optimistic technology, applied in the field of optimistic exploration, which can solve problems such as uncontrollable signal phase

Active Publication Date: 2021-09-14
DALIAN MARITIME UNIVERSITY
View PDF6 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The invention provides a traffic control method based on Categorical-DQN optimistic exploration, which sol

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Traffic control method based on Categorical-DQN optimistic exploration
  • Traffic control method based on Categorical-DQN optimistic exploration
  • Traffic control method based on Categorical-DQN optimistic exploration

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0036] It should be noted that, in the case of no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and examples.

[0037] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments It is only some embodiments of the present invention, but not all embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and in no way taken as limiting the invention, its application or uses. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordina...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a traffic control method based on Categorical-DQN optimistic exploration, and relates to the technical field of optimistic exploration. The method comprises the following steps of: S1, modeling intersections as agents, and initializing a current network Z (o, a; theta) and a target network Z' (o, a; theta'), o being local observation, a being a signal phase to be selected at the next moment, theta being a current network parameter, and theta' being a target network parameter; S2, initializing an empirical replay memory (ERM); S3, setting a greedy factor epsilon and an optimistic factor tau as 1; and S4, setting the number of training times M, and repeatedly training the agents in S1 for M times. The method can be applied to a traffic environment. Intersections are modeled as agents and trained in a multi-agent environment, cooperation between the intersections is achieved, and traffic congestion is effectively relieved. The improvement of the effect of the method also leads to improvement of the effect in the traffic environment.

Description

technical field [0001] The invention relates to the technical field of optimistic exploration, in particular to a traffic control method based on Categorical-DQN optimistic exploration. Background technique [0002] Traditional RL research on independent learning (IL) MARL is mainly based on the "optimistic" principle, and the agent selects and evaluates an action according to the maximum expected return (MER) or the weighted value of MER and expected return. These agents optimistically assume that all other agents take actions that maximize their rewards. Therefore, they update the action's evaluation only if the new evaluation is better than the previous one. While for Deep Reinforcement Learning (DRL) algorithms, it exhibits inherent flaws in deep reinforcement learning collaborative problems, such as low sampling efficiency due to storing outdated experience in Experience Replay Memory (ERM), when other intelligent When the strategy of the agent changes, the effect of ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G08G1/01G08G1/08G08G1/081G06N3/00
CPCG08G1/0125G08G1/0137G08G1/08G08G1/081G06N3/006
Inventor 张程伟田宇房迪娜
Owner DALIAN MARITIME UNIVERSITY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap