The Chemical Production Process Optimization Based On Extend Continuous Ant Colony Optimization Algorithm

Optimization is an effective technique for improving the performance of chemical production process system which is a typical complex system. It can remarkably improve the efficiency, reduce the use of energy, utilize the resource reasonably, and boost the economic benefit. But as the scale of object problem becomes more and more largeer, of which the model structure becomes more and more complicated too, thus makes it more and more difficult to optimize the object, and has been a great challenge to the exiting optimizing method. So the require for high efficiency intelligent optimization techniques becoming more and more necessary.In this paper, a class of newly intelligent optimization method named Ant Colony Optimization (ACO) be used to optimize chemical process. Its predominant distributed pattern of problem solving achieves great success in combinational problems, but to many practical chemical process problems, they always are expressed as continuous optimization problems. It is an imperative challenge on how to apply the basic ant colony algorithm idea to the problems solving in continuous space, which is the major of this paper.First of all, this paper carries out an in-depth analysis on the basic principle of ant colony optimization algorithm from the perspective of both biology and mathematics, then establishes the mathematical model and the procedure framework of ACO.Then this paper discusses the discrete nature of basic ACO , and the main ideas for solving continuous domain optimization problems with it. Under these ideas, this paper puts forward one kind of continuous ACO(CACO) algorithm named Extend CACO(ECACO). Then to compare it's validity with Gridding Partition CACO, this paper chooses a typical continuous domain optimization problem for them to solve. Subsequently, the ECACO is combined with neural network to optimize the value of ANN's weight and threshold ,which makes up for the shortage of BPNN easily falling into the local minimum. After that a example is used to verify the feasibility of new ECACO BP network.Finally, the ECACO and the ECACO BPNN are applied to optimize the Alkylation process and thermally coupled distillation respectively by MATLAB programming. The ECACO comes up by this paper originally is used to solve the continuous domain problem with nonrestraint, but this paper uses it to Alkylation process optimization with restriction. The simulation results show that the stability and good solving global optimization performance. Moreover, the new ECACO BP network also shows a good simulation result and sufficient accuracy in thermally coupled distillation process modeling.