Evolutionary linearisation in the frequency domain
Few solutions have been found so far to frequency domain MIMO system linearisation problems owing to practical difficulties in obtaining derivatives of the nonlinear model. The authors develop an evolutionary linearisation technique to achieve an automated linearisation goal, directly using the plant I/O data. It allows linearisation for an entire operating region and for the entire interested frequency range. the benefits of which cannot be matched by existing methods.
- Franklin, G.F., Powell, J.D., Emami-Naeini, A.: Feedback control of dynamic systems, 1991 (Addison-Wesley)
- Renders, J.M., Bersini, H.: `Hybridizing evolution algorithms with hill-climbing methods for globaloptimisation: Two possible ways', Proc. 1st IEEE Int. Conf. Evolutionary Computing, IEEE WorldCong. Comp. Intel., 1994, 1, Orlando, p. 312–317
- Tan, K.C., Li, Y., Murray-Smith, D.J., Sharman, K.C.: `System identification and linearisation using evolution algorithms withsimulated annealing', Proc. 1st IEE/IEEE Int. Conf. Genetic Algorithms in Eng. Syst.:Innovations and Appl., 1995, Sheffield, United Kingdom, p. 164–169