Operation Optimization Algorithm And Systems For Rolling Production Process Based On Evolutionary Computation

Production process operation optimization lies between production scheduling layer and process control layer in the integrated automation system in process industry, and its main task is to determine the optimal values for control variables (rolling pressure, temperature, etc.) based on the production task determined by the scheduling layer and the process constriants and objective. These values are then set as the targets for the process control layer. Because the values of control variables directly determine product quality and energy consumption, the production process optimization becomes very important for process industry enterprises.This thesis takes the hot rolling and cold rolling in iron and steel industry as background, and focuses on the modeling and algorithms for the production process operation optimization problem. Based on the model and algorithms, the corresponding operation optimization systems are then developed. The main contents of this thesis are as follows:(1) In the modeling of the operation optimization problem in hot rolling and cold rolling production process, the classical mechanism model is used to calculate the values of process parameters. According to practical demands, two models are constructed, i.e., the models for hot rolling production process operation optimization with the objective to minimize the convexity of rolled strips, and the models for cold rolling production process operation optimization with the objective to minimize the total energy consumption.(2) For the operation optimization of hot rolling process, the differential evolution (DE) algorithm incorporating the opposition-based learning is proposed, and the computational results based on practical production data illustrate the efficiency of this algorithm.(3) For the operation optimization of cold rolling process, the particle swarm optimization (PSO) that incorporates the clustering-based population update strategy is developed. The results of computational experiments using both benchmark problems and practical production data illustrate the efficiency of the improvement strategy and the proposed algorithm.(4) Based on the constructed models and algorithms, the operation optimization systems for the hot rolling and cold rolling processes are developed, respectively. The computational results based on practical production data show that the models and algorithms are correct and efficient.