Improved Particle Swarm Optimization Algorithm And Data Rectification In Chemical Process

In the process of chemical industry, aging and distortion of the equipments and instruments lead to non-compliance with the principle of material conservation and energy conservation of logistics components. It requires the detection information and prior knowledge of the process, using best method to adjust the measured values of known variables, correcting and processing the industrial process data. It can estimates unmeasured data, completes data information and provides reliable data for process modeling, control and optimization.Therefore, in this paper the study of improved particle swarm optimization algorithm is carried out. To solve the problem of "premature" phenomenon of PSO in dealing with optimization process, an improved particle swarm optimization algorithm with the combination of chaos and harmony search was proposed. CPSO is able to make it back to the random motion state from local optimum, and then introduce the harmony search algorithm for global optimization again. Four standard test functions were selected to make a simulation of comparison, and the results showed that the improved CPSO-HS has strong ability to jump out of local optimal position, and fast convergence. Improved CPSO-HS is applied successfully to parameter estimation of the model of heavy oil thermal cracking. Simulation results demonstrated that improved CPSO-HS has better calculation accuracy compared with CPSO and GA.In the application of depropanizing column data rectification, the errors of instruments and equipments lead it not to meet the principle of material balance. In this paper, the rectification of the 10 components'molar flow rate of 4 logistics in depropanizing column is made by improved CPSO-HS and from the results we can see that the rectification data can meet the material balance. It also verifies that the present method is more accurate compared with the CPSO.