Chemical Process Optimization Based On Improved Artificial Fish Swarm Algorithm

Many problems which are fairly complex in chemical process need to be optimized, In addition, as the scale of chemical process expands, the objective function becomes more and more complicated and the number of independent variables and constraints increases rapidly. Then it becomes difficult to solve these problems using traditional optimization algorithms. Artificial Fish Swarm Algorithm (AFSA) is a new direction in the research of intelligence optimization algorithms, which not only provides a brand new theory to solve complex system problems, but also offers a new idea to chemical process problems.The following works are carried out with AFSA in this paper:(1) First of all, a general history of chemical process optimization including the challenges faced is introduced. Some concepts, classification and the development process of optimization are involved next. Besides, several common intelligent optimization algorithms are concerned.(2) Then the basic Artificial Fish Swarm Algorithm is presented, which included the background, artificial fish structure, basic behavioral description, the principle of algorithm, behavioral choice and so on. Some research overview is made moreover.(3) Afterwards, the paper analyzes the influences to the optimization results of variable parameters with several experiments. To the deficiencies of AFSA, some improvements are put forward. The improved Artificial Fish Swarm Algorithm (IAFSA) is proved to be effective and practical with some classic functions.(4) Finally, the IAFSA is applied in chemical process optimization. It shows good results with the mathematical models of Heat Exchanger Network (HEN) and Butylene Alkylation (BA) process.