Intelligence Modeling And Optimal Control Of The Synthetic Ammonia Decarbornization

Modeling of complex industrial process control and optimization has been a hot issue in the field of industrial process control. As the complex industrial process is affected by many factors and the system has lag and the random, fuzzy, gray and so many uncertainties, model is quite difficult to be built up. The BP neural network algorithm used in approximation of function has two shortcomings: the one is its slower convergence rate and the other is its falling into the local minimum. In order to solve the problems, a new method is proposed in this paper and is complicated on a typical complex system---the synthetic ammonia decarbornization industrial process. And some simulation studies with calcination section have been done. This paper studies the optimization and modeling of the synthetic ammonia decarbornization and neural network theory applied to particle filtering.1) Summing up the traditional methods of complex industrial process in modeling and optimization, and pointing out their flaws and shortcomings. As neural network has a very strong solving function of dealing with complex industrial processes faced by non-linear and uncertainty problem, a neural network control program based on particle filter is proposed.2) The synthetic ammonia decarbornization is comprehensive analyzed. To the industrial process control status and development trend, RBF neural network control model to be established.3) The present situation and development trend of the particle filter and neural network theory are detailed introduced. The learning capability of RBF neural network algorithm and the advantage of particle filtering algorithm on processing nonlinear system are used. A method based on RBF neural network and particle filtering algorithm is proposed to deal with complex industrial process control and optimization.4) By Simulation experiments, the combination of RBF neural network and particle filter combined has the feasibility and advantages. From the result, the combination of RBF neural network and particle filter has a brilliant and broad prospect of development.5) The model is used to a class of complex industrial process and its application (calcining process). The feasibility and correctness of the model is validated from the perspective of experimental by using of MATLAB. It provides an efficient way for the complex system modeling and optimization control research. Finally, concluding the main content of this paper. The future direction of studying is indicated as well.