Process real-time optimization using Clonalg algorithm

Based on machine learning and immune mechanism, a useful framework for process real-time optimization is put forward in this paper. The proposed framework using real time evolutionary (RTE) approach with Clonalg algorithm (RTE-Clonalg) need not wait for the steady-state which is involved by the traditional Real-Time Optimization (RTO). RTE-Clonalg can adjust the set point continuously to approach the optimal points, so that the system can deal with the higher frequency disturbance, since the steady-state shouldn´t be waited for to execute RTE system. The performance of the proposed method was successfully proved by Tennessee Eastman process.