ν-gap Model Reduction by Means of Particle Swarm Optimization
- Resource Type
- Conference
- Authors
- Masaaki Kanno; Tae-Hyoung Kim (김태형)
- Source
- 대한기계학회 춘추학술대회. 2010-11 2010(11):1227-1231
- Subject
- Model reduction(모델 축소)
ν-gap
particle swarm optimization(개체 집단 최적화)
- Language
- Korean
This paper considers the model reduction problem of linear time-invariant systems in the ν-gap and attempts to solve the problem by using particle swarm optimization (PSG). The ν-gap is an effective metric to measure the closeness of two systems when put in the feedback loop. Therefore finding a low-order system that approximates a high-order system in the ν-gap sense can be of practical significance. Nevertheless this model reduction problem is difficult to solve and little attempt has been made so far, partially due to the non-convex nature of the problem. Recently the PSG algorithm, whose mechanism was inspired by flocks of birds and shoals of fish, has attracted a lot of attention. The empirical evidence indicates its superiority in solving various non-convex problems. The paper thus applies the PSG algorithm to the ν-gap model reduction problem and its feasibility is investigated based on some numerical examples.