Speeding Up GenOPT optimisation

edit

Hey fellow researchers

I use a combination of dymola and GenOPT. Dymola generates some data and GenOPT optimises it. GenOPT is used to perform parameter identification with multiple random start up points in order to increase the chances of meeting the global optimum.

Presently to get the results it takes about 2-3 days( courtesy 10 random runs and high accuracy ). The quality of the results is excellent and comparable to conventional softwares. Are there techniques, steps by which the time span of the whole procedure can be reduced for e.g.: A faster result can be arrived to by providing stricter constraints for the optimisation process.

Regards