Data reconciliation for crude oil distillation units based on self-adaptive probability of mutation in genetic algorithm

Self-adaptive probability of mutation was adopted in genetic algorithms. In this self-adaptive genetic algorithm (SAGA), the probability of mutation was varied depending on both the evolved generations and the individual fitness value. The probability of finding the global optimal solution is improved. Based on SAGA, a synchronous arithmetic about data reconciliation and lapse error was employed. The measured and lapse errors were supposed to be normal distribution. Then this arithmetic was used in data reconciliation of steady-state flows along crude oil distillation units. And the results show that more accurate data are obtained.