The application of neural techniques to the problem of torque ripple minimization in a switched reluctance motor (SRM) is presented. More conventional techniques for torque linearization and decoupling are reviewed, after which the application of a neural network to the problem is described. Results obtained experimentally and by simulation of a 4 kW IGBT converter and 4-phase SRM are used to illustrate the approach. The networks used have been implemented using both digital signal processor (DSP) and field programmable gate array (FPGA) technologies.
|Title of host publication||Drives II|
|Number of pages||6|
|Publication status||Published - 1993|
|Event||Proceedings of the 5th European Conference on Power Electronics and Applications - Brighton, Engl|
Duration: 13 Sept 1993 → 16 Sept 1993
|Conference||Proceedings of the 5th European Conference on Power Electronics and Applications|
|Period||13/09/93 → 16/09/93|