No. 1151
	Advances and industrial applications of soft computing
	Investigating R&D committee on advances and industrial applications of soft computing
	Keyword : soft computing, meta-heuristics, optimization, industry applications
	Soft computing includes evolutionary computing such as Genetic Algorithms and learning methods such as Neural Networks, and has attracted attention recently as a new optimization paradigm. Although soft computing provides efficient approximate optimization solvers, it had been considered to be a time-consuming method. With significant performance advances of computers in recent years, soft computing has been tried to apply to various industrial optimization problems such as manufacturing systems, energy systems, logistic systems, and information systems. �@Moreover, many researchers have been studied the theoretical aspect of soft computing in terms of convergence performance, search efficiency, and quality of obtained solutions. In this article, we survey advances and industrial applications of soft computing, and show the current situation and the future work of soft computing. Specifically, we report the latest move of various soft computing methods such as Particle Swarm Optimization, Genetic Algorithms, Tabu Search, Chaotic Optimization, Pulsed Neural Networks, Reinforcement Learning, and Fussy Deduction. We also deal with industry applications such as robotics, PID controllers, vehicle routing, energy distribution systems, water conveyance, diagnosis support systems, road transportation systems, and nuclear power generation.
	©2007. The Institute of Electrical Engineers of Japan