STRATEGY NETWORK MODEL GENERATION METHOD AND APPARATUS FOR AUTOMATIC VEHICLE DRIVING

WIPO Patent Application WO/2018/205245

A1

A strategy network model generation method and apparatus for automatic vehicle driving. The method comprises: forming vehicle state information collected at each test moment into a state information set, and processing the state information set, so as to obtain a feature state information set (S101); searching, according to a pre-set reward value function and a pre-constructed vehicle action set, for a vehicle action obtaining the maximum reward value from each piece of state information in the feature state information set (S102); training a network model of an extreme learning machine according to the feature state information set, the maximum reward value corresponding to each piece of state information in the feature state information set, and the vehicle action obtaining the maximum reward value (S103); and generating, according to a training result of the network model of the extreme learning machine, a strategy network model for automatic vehicle driving (S104), thereby effectively reducing the consumption of computation resources, and effectively improving the generation efficiency of the strategy network model for automatic vehicle driving.

LI HUIYUN (CN)
WANG ZHENG (CN)
LIU FENFEN (CN)

PCT/CN2017/084081

November 15, 2018

May 12, 2017

Click for automatic bibliography generation  

SHENZHEN INST OF ADV TECH CAS (CN)

G06N5/02; G06N99/00

CN107229973A	2017-10-03
CN106080590A	2016-11-09
CN106428000A	2017-02-22
CN103605285A	2014-02-26
US6493614B1	2002-12-10

SHENZHEN YOUTH PATENT AND TRADEMARK AGENCY LTD. (CN)

View/Download PDF      PDF Help

Previous Patent: PROCESS FOR THE EPOXIDATION OF PROPENE

Next Patent: PARALLEL COMPUTATION ENGINE FOR GRAPH DATA