Intelligent Optimization Based On The Decomposition Rate Of Prediction Methods Of Operation Research

In the process of alumina production by sintering , the process of the continuous carbonation decomposition (PCCD) is one of the key procedures, in which the resolution ratio gradient and the last resolution ratio directly affect the output and quality of product. Due to the complexity of the PCCD, it's hard to manipulate the condition of processing precisely, and this causes great fluctuation in the PCCD. Therefore, it's necessary to develop the work of modeling and optimize the PCCD for energy consumption reduction and increase production increasment purposes.The process of the continuous carbonation decomposition firstly is introduced as researched objects in the paper and the factors which affect the ratio of decomposition of the PCCD are analyzed. Because of the process with complex mechanism, the method of Rough Set is applied to build the model of the PCCD. With the complexity character of attribute reduction in Rough Set, an improved algorithm which can decrease the complexity of attribute reduction is proposed to build the predictive model of the PCCD by dealing with amount of historical data related to the process of the PCCD. The incremental method is adopted to update the knowledge basement of the PCCD for solving the alteration of steady point of the PCCD. The predictive module of the PCCD is verified its effectiveness by the factual production data.Aiming at status that production is unstable caused by the great fluctuation of the ratio of decompsition, genetic algorithm is adopted to optimize the distribution of resolution ratio gradient. At the same time, in genetic algorithm the operators are improved so as to improve the ability of global optimization. An improved genetic algorithm is adopted to optimize the distribution of resolution ratio gradient to each sub-space which divided by the predictive model of the PCCD, and then optimization results are treated as setting value of fuzzy control system with the PCCD. The result show that the optimized distribution of resolution ratio gradient not only stabilize the process, but also improve the ratio of decomposition.