Article

Optimal and stable fuzzy controllers for nonlinear systems based on an improved genetic algorithm

Dept. of Electron. & Inf. Eng., Hong Kong Polytech. Univ., Kowloon, China
IEEE Transactions on Industrial Electronics (Impact Factor: 6.5). 03/2004; DOI: 10.1109/TIE.2003.821898
Source: IEEE Xplore

ABSTRACT This paper addresses the optimization and stabilization problems of nonlinear systems subject to parameter uncertainties. The methodology is based on a fuzzy logic approach and an improved genetic algorithm (GA). The TSK fuzzy plant model is employed to describe the dynamics of the uncertain nonlinear plant. A fuzzy controller is then obtained to close the feedback loop. The stability conditions are derived. The feedback gains of the fuzzy controller and the solution for meeting the stability conditions are determined using the improved GA. In order to obtain the optimal fuzzy controller, the membership functions are further tuned by minimizing a defined fitness function using the improved GA. An application example on stabilizing a two-link robot arm will be given.

0 Bookmarks
 · 
46 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Understanding of the considerable variation in action potential (AP) shape throughout the heart is necessary to explain normal and pathological cardiac function. Existing mathematical models reproduce typical APs, but not all measured APs, as fitting the sets of non-linear equations is a tedious process. The study describes the integration of a pre-existing mathematical model of an atrial cell AP with a genetic algorithm to provide an automated tool to generate APs for arbitrary cells by fitting ionic channel conductances. Using the Nygren model as the base, the technique was first verified by starting with random values and fitting the Nygren model to itself with an error of only 0.03%. The Courtemanche model, which has a different morphology from that of the Nygren model, was successfully fitted. The AP duration restitution curve generated by the fit matched that of the target model very well. Finally, experimentally recorded APs were reproduced. To match AP duration restitution behaviour properly, it was necessary simultaneously to fit over several stimulation frequencies. Also, fitting of the upstroke was better if the stimulating current pulse replicated that found in situ as opposed to a rectangular pulse. In conclusion, the modelled parameters were successfully able to reproduce any given atrial AP. This tool can be useful for determining parameters in new AP models, reproducing specific APs, as well as determining the locus of drug action by examining changes in conductance values.
    Medical & Biological Engineering & Computing 10/2005; 43(5):561-71. · 1.79 Impact Factor
  • Source
    ICEIS 2005, Proceedings of the Seventh International Conference on Enterprise Information Systems, Miami, USA, May 25-28, 2005; 01/2005
  • [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents a real-coded genetic algorithm (RCGA) with new genetic operations (crossover and mutation). They are called the average-bound crossover and wavelet mutation. By introducing the proposed genetic operations, both the solution quality and stability are better than the RCGA with conventional genetic operations. A suite of benchmark test functions are used to evaluate the performance of the proposed algorithm. Application examples on economic load dispatch and tuning an associative-memory neural network are used to show the performance of the proposed RCGA.
    Soft Computing 12/2006; 11(1):7-31. · 1.30 Impact Factor