The Material Balance Calculation With The Bayer Process Based On The Particle Swarm Optimization

The Bayer process is the most main technique to produce Alumina. The material balance calculation is the dynamical equilibrium computation process, which keeps to conservation of mass, achieves the material information integration about each technologic node, including the whole chemic reaction process which is from Alumina ore input to Alumina output. It not only has the theory guidance to construct the production line, choose production equipment and confirm key craft parameter, but also provides the basis to improvement production craft, enhancing the product yield rate and product quality.In order to establish the precise mathematical model of the actual production process, this thesis embarked from the basic principle of Alumina production, introduced a typical Bayer process and the characteristic of main procedure, analyzed in the Alumina production the calculational characteristic and difficulty, proposed the principle for system modeling and established the mathematical model about actual flow. This model is a non-linear, multi-feedback and complex production systemic mathematics abstract, which contains the whole information of all the production process.To solve the mathematical model precisely, the thesis embarked from Particle Swarm Optimization (PSO) algorithmic background, analyzed in with emphasis the parameters'significance of the algorithm and the establishment rule, moreover, given some improved algorithms. Introduced several kinds of optimized algorithm, analyzed their basic principle and compared them with PSO algorithm, which demonstrated that the later was superiority in this kind of project problem.Using the MATLAB tool established the whole systemic software platform, which was advantageous for the user. This thesis proposed the software's principle of design, given the systemic main structural framing, completed the application software's establishment.Taken the ShanXi Aluminum Manufacturer's Condition as the example, we could confirm the software's validity. The computed result indicated that, the new algorithm was better than the traditional PSO with both a better stability and a steady convergence, which was more feasible and efficient in practical application. This software has completely applied in the actual projects.