Cultural Algorithm And Its Application In Chemical Dynamic Optimization

As the process industry asking for the increasing quanlity of optimal economic indicator and control performance, the current static optimization methods based on the static models of chemical processes have not applied, and for the actual chemical process, the state variables are dynamic, changing with time or space. The dynamic process model owns the characteristics of high dimensions, strong nonlinear and complex reaction mechanism, therefore, it is necessary to develop dynamic optimal methods for the dynamic process analysis. Nowadays, the application of intelligent optimization algorithms in the chemical dynamic optimization problems is rarely present, therefore this work takes the Improved Knowledge-Based Cultural Algorithm(IKBCA) as the research key pointIn view of the characteristics of chemical dynamic optimization, this paper proposes a new chromosome encoding with the characteristics of "soft constraints", and also proposes a new method in combination with the idea of "orthogonal design method" to acquire the hidden information of the evolution from the global viewpoint. After illustrating the rationality and validity, the studying results realize a valid solving for the dynamic optimization with the terminal time given and the state variables unconstrainted.For illustrating the universality of this algorithm, the paper selected several application cases of chemical dynamic optimization with typical control characteristics, such as the optimal temperature control of Batch Reactor, the optimal catalyst distribution of Tubular Reactor, the optimal flow rate of protein in Park-Ramirez Bioreactor and Continuous Stirred Tank Reactor (CSTR).The results of the application cases illustrate that the IKBCA is able to co-ordinates the diversity of individuals and the searching performance durning the evolution, and suitable to varies of Chemical dynamic optimization. In the final, the thesis draws some conclusions of whole study and prospected the future intelligent algorithms of Chemical dynamic optimization.