Intelligent Satisfaction Of The Alumina Raw Slurry Deployment Process Optimization And The Application

Owing to the indefiniteness of supplied mine source, the uncertainty of the composition of returned lye and liquid waste, there exists much composition fluctuation of raw slurry. In order to improve the quality of slurry, the qualified slurry has to be gained through mixing of full-filled tanks for the production process of alumina with sintering. Strong coupling is in existence among the index of raw slurry with numerous effect factors. Because of the interrelation between the indexes, the performance of other indexes will be changed when adjustment of a certain one. In the current technology environment, much fluctuation composition of the raw mix slurry caused by the intensity of arrangement computing and the difficulty of obtaining an optimization scheme..In order to solve these problems, an optimal model for raw slurry arrangement is designed in this paper, which is based on the material balance principle. This is a multi-constraint and multi-target optimization problem with the existence of strong coupling. Due to limitation of time-consuming of the traditional optimal methods to deal with combinatorial optimization problem and the fuzziness of the objectives and constraints resulted by the restriction of industrial environment and technology condition, multi-objective satisfactory optimization theory is studied and is introduced into the raw slurry optimal arrangement. The satisfactory optimal model for the raw slurry arrangement is given and the satisfactory degree functions are built, which can reflect the relationship between the loosened quantity of soft constraints and the desire of operators. A new method suitable for multi-objective satisfactory optimization is established based on self-adaptive genetic algorithm with population ranking technique, niche technique and pareto solution set filter. The intelligent method based on satisfactory theory is used to get the optimum solution of the raw slurry arrangement process and the optimization results show that the computational efficiency is increased greatly, the composition fluctuation in mixed raw slurry is decreased a lot, and more satisfactory performances are achieved than that of traditional methods. The computation results based on the industrial database demonstrate the proposed approach's effectiveness and practicality.