This paper deals with the auto-tuning of dc motor drives. To improve the performance of self-commissioning on currently available on industrial drives, an on-line auto-tuning based on hybrid EAs is proposed. This strategy integrates a local searcher, simplex method, in an evolutionary framework, global searcher, in order to speed up the convergence. Moreover it is very reliable because experimentally tests each possible solution, consequently the final result is not affected by the accuracy of the motor model. The test-bed is based on Matlab solutions, in particular, the Hybrid EA code is implemented by means of Genetic Algorithm and Direct Search Toolbox, and the rapid-prototyping system is based on xPC Target and xPC TargetBox Matlab. Experimental results prove the effectiveness of the proposed approach not only in comparison with conventional commissioning, but also when compared with further accurate hand-calibration.
On-line Hybrid EAs for Auto-tuning of DC Motor Drives / Busto, E.; Cascella, G. L.; Salvatore, N.; Gassi, D. A.. - CD-ROM. - (2005). (Intervento presentato al convegno International Conference on CAE and Computational Technology for Industry, TCN-CAE 2005 tenutosi a Lecce nel October 5-8, 2005).
On-line Hybrid EAs for Auto-tuning of DC Motor Drives
G. L. Cascella
;
2005-01-01
Abstract
This paper deals with the auto-tuning of dc motor drives. To improve the performance of self-commissioning on currently available on industrial drives, an on-line auto-tuning based on hybrid EAs is proposed. This strategy integrates a local searcher, simplex method, in an evolutionary framework, global searcher, in order to speed up the convergence. Moreover it is very reliable because experimentally tests each possible solution, consequently the final result is not affected by the accuracy of the motor model. The test-bed is based on Matlab solutions, in particular, the Hybrid EA code is implemented by means of Genetic Algorithm and Direct Search Toolbox, and the rapid-prototyping system is based on xPC Target and xPC TargetBox Matlab. Experimental results prove the effectiveness of the proposed approach not only in comparison with conventional commissioning, but also when compared with further accurate hand-calibration.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
