Digitization of the optimization method by the particle swarm governed by chaos dynamics

2016 Fiscal Year Final Research Report

• Project/Area Number: 26630172
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Communication/Network engineering
• Research Institution: Nagaoka University of Technology
• Principal Investigator: Tsubone Tadasho 長岡技術科学大学, 工学研究科, 准教授 (50334694)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: 粒子群最適化 / 非線形理論 / 非線形回路 / カオス

Outline of Final Research Achievements

This study has been intended to realize optimization method by the swarm intelligence in digital system.　Using the deterministic model based on artificial neural circuit in previous study, we have aimed at the theory construction of hardware of the swarm intelligence.　 As a result, we have realized two models and showed a correlation between parameters and performance quantitatively from the neural-chaos oscillator which was basic dynamics, and we have suggested design guidance. In addition, implemented circuits have been given to realize the movement of the particles on a programmable circuit, and we could show the behavior of the circuit and relations of the optimization performance and was able to establish the design guidance of an effective parameter. We have also been able to show that suggestion technique was effective in a dynamic problem.

Free Research Field

工学