公开(公告)号：US20180261085A1

公开(公告)日：2018-09-13

申请号：US15641434

申请日：2017-07-05

Applicant: FUJITSU LIMITED

Inventor： Lei LIU ,  Wei-Peng CHEN ,  Ying LIU

IPC: G08G1/08 ,  G06N99/00

CPC classification number: G08G1/08 ,  F21W2111/02 ,  F21W2111/023 ,  G06F17/30598 ,  G06K9/00785 ,  G06K9/4638 ,  G06N7/005 ,  G06N99/005 ,  G08G1/005 ,  G08G1/0129

Abstract: Technologies are described to adjust a learning rate of Q-learning being used to control traffic signals at an intersection. In some examples, a method may include generating control actions for traffic signals at an intersection based on Q-learning, determining a frequency of change in traffic pattern of the intersection, and adjusting a learning rate of the Q-learning based on the determined frequency of change in traffic pattern of the intersection. The Q-learning may determine the generated control actions based on at least a portion of historical traffic data of the intersection, and the change in traffic pattern may be a change from a first traffic pattern of the intersection to a second traffic pattern of the intersection.

公开(公告)号：US20190139405A1

公开(公告)日：2019-05-09

申请号：US16003017

申请日：2018-06-07

Applicant: FUJITSU LIMITED

Inventor： Ying LIU ,  Lei LIU ,  Wei-Peng CHEN

IPC: G08G1/08 ,  G06K9/00 ,  G08G1/005 ,  G08G1/01 ,  G06N20/00 ,  G06F16/28 ,  G06K9/46 ,  G06N7/00 ,  F21W111/023 ,  F21W111/02

CPC classification number: G08G1/08 ,  F21W2111/02 ,  F21W2111/023 ,  G06F16/285 ,  G06K9/00785 ,  G06K9/4638 ,  G06N7/005 ,  G06N20/00 ,  G08G1/005 ,  G08G1/0129

Abstract: Technologies are described to provide control of traffic signals based at least in part on multiple Q-learning categories. In some examples, a method may include clustering historical traffic data into multiple traffic pattern clusters, and generating multiple Q-learning categories, where each Q-learning category corresponds to a traffic pattern cluster of the multiple traffic pattern clusters. The method may also include determining a first Q-learning category of the multiple Q-learning categories to use in controlling traffic signals at an intersection based at least in part on a first traffic data of the intersection, where the first Q-learning category corresponds to a first traffic pattern cluster, and the first traffic data corresponds to the first traffic pattern cluster. The method may additionally include generating a first control action for the traffic signals at the intersection based at least in part on the first Q-learning category.