Research On The Relationship Model Between Raw Block Quality And Calcined Production Process Parameters

The production process of the prebaked anode is a manufacturing process with complicated process and unclear reaction mechanism between raw materials.Most of the current domestic and foreign counterparts control the quality through methods such as controlling charts or establishing general mathematical models,but the effect in practical applications is still not ideal.Neural network algorithm is a mathematical model developed by simulating human brain structure.It can establish a very complex mapping relationship between input and output.Its complex relationship between calcined process parameters and raw block quality in prebaked anode production can be adaptively and self-learning strongly nonlinear.Therefore,the application of the neural network algorithm in the prebaked anode production process is very necessary.According to the research requirements of the relationship model between the calcined process parameters and the quality evaluation parameters of the raw block,the GASA algorithm was re-inducted and designed.It not only possesses the strong global search ability of genetic algorithm,but also has the strong local search ability of simulated annealing algorithm.At the same time,we use the GASA algorithm to optimize the initial weights and thresholds of the BP neural network algorithm to form the GASA-BP algorithm.Finally,we use the calcined production process parameters as input,and the mass parameters of the raw block as output,and apply the GASA-BP algorithm to form the prediction model between the raw block quality and the calcination process parameters.The specific work and content of this article are as follows:(1)Analyze the relationship between different processes in the production process of prebaked anodes and the role of raw materials,select appropriate characteristic parameters and conduct data collection and data preprocessing.(2)Study simulated annealing algorithm and genetic algorithm and BP neural network algorithm,according to their respective characteristics to optimize,form GASA-BP algorithm.Firstly,this paper obtains the initial progeny population through the related operations of the genetic algorithm,and then receives or performs the Metropolis criterion according to the individual fitness value in turn,and treats the progeny individual group obtained after the treatment as the final progeny population.At the same time,this process is repeated according to the number of iterations and termination conditions.This is the GASA algorithm.Then,because the initial weights and thresholds in the BP neural network algorithm have a negative impact on the final prediction result,the GASA algorithm is used to search for the optimal solution in its solution space,and the prediction accuracy rate of the BP neural network model is improved,ie GASA-BP algorithm.(3)Using the design idea of software engineering,the SSM integrated architecture and GASA-BP algorithm are used to design and implement a prediction system for the quality of raw materials and calcined process parameters.The purpose of the system design is to analyze the relationship between the calcining process parameters and the quality of the raw block,and to display the results in a tabular and graphical format to provide a reference for the decision-making of parameters in the production process.