AUTONOMOUS ROBOT FAILURE RECOGNITION DESIGN USING MULTI-OBJECTIVE GENETIC PROGRAMMING
An evolutionary autonomous failure recognition approach is presented using multi-objective genetic programming in this paper. It is compared with the conventional robot failure classification algorithm. Detailed analysis of the evolved feature extractors is tempted on investigated problems. We conclude MOGP is an effective and practical way to automate the process of failure recognition system design with better recognition accuracy and more flexibility via optimizing feature extraction stage.
Feature Extraction Autonomous robot Failure recognition Multi-objective Genetic Programming
YANG ZHANG
Electronic and Electrical Engineering Department, The University of Sheffield, S1 3JD, UK
国际会议
2006 International Conference on Machine Learning and Cybernetics(IEEE第五届机器学习与控制论坛)
大连
英文
4563-4568
2006-08-13（万方平台首次上网日期，不代表论文的发表时间）