A Neurodynamic Optimization Approach to Robust Pole Assignment for Synthesizing Linear Control Systems Based on a Convex Feasibility Problem Reformulation
A neurodynamic optimization approach to robust pole assignment for synthesizing linear control systems is presented in this paper. The problem is reformulated from a quasi-convex optimization problem into a convex feasibility problem with the spectral condition number as the robustness measure. Two coupled globally convergent recurrent neural networks are applied for solving the reformulated problem in real time. Robust parametric configuration and exact pole assignment of feedback control systems can be achieved. Simulation results of the proposed neurodynamic approach are reported to demonstrate its effectiveness.
KeywordsRobust pole assignment recurrent neural networks state feedback control global convergence
Unable to display preview. Download preview PDF.
- 12.Le, X., Wang, J.: Neurodynamic optimization approaches to robust pole assignment based on alternative robustness measures. In: Proc. Interntional Joint Conference on Neural Networks (2013)Google Scholar
- 13.Le, X., Wang, J.: Robust pole assignment for synthesizing feedback control systems using recurrent neural networks. IEEE Trans. Neural Networks and Learning Systems (in press, 2013)Google Scholar