Endpoint Prediction And Multi-furnace Optimal Scheduling Method For Copper Melting Production

In copper melting production,converter endpoint prediction and multi-furnace optimal scheduling are of great significance for improving production efficiency and stabilizing product quality.Among them,the accurate blowing cycle is an important basis for the optimal scheduling of matte between the flash furnace and the converter.However,due to the fact that the blowing endpoint of matte in the converter cannot be accurately detected,and the time delay and randomness exist in the operation of the flash furnace and the converter,it is very difficult to predict the endpoint and optimize the scheduling of the production process.Therefore,on the basis of analyzing the process mechanism of copper melting production,this paper proposes a research program for the endpoint prediction and multi-furnace optimal scheduling of copper melting production.The main research work and innovative achievements are as follows:(1)Aiming at the multivariate coupling,nonlinearity and uncertainty of the copper converter blowing process,a blowing endpoint prediction and dynamic calibration method for the time-varying air supply flow is proposed.First,a hybrid Regularized Extreme Learning Machine model is established to predict the blowing endpoint;then,the blowing time correction rule oriented to the time-varying air supply flow of the converter is extracted,and the predicted value of the endpoint is dynamically calibrated.The experimental results show that the proposed algorithm can accurately predict the blowing endpoint of the copper converter.(2)Aiming at the unique production characteristics and large-scale production constraints of the copper melting production process,with the optimization objectives of minimizing the waiting time and reducing the flow change,a mathematical description method for scheduling constraints such as the semi-continuous operation of the flash furnace,the batch process of the converter and the operation time of the process is proposed.A time-slot-based mixed integer linear programming model is established as a multi-furnace scheduling model,and the effectiveness of the model is verified by a practical example.(3)Aiming at the discrete combinatorial optimization problem of multi-furnace scheduling in copper melting production,a Hybrid Discrete Particle Swarm Optimization algorithm is proposed.First,the Particle Swarm Optimization algorithm is discretized,and a two-layer coding scheme based on cycle sorting and equipment selection is designed according to the characteristics of the converter operation;then,the particles are updated by operations such as crossover and mutation,and Simulated Annealing algorithm is introduced as a local search strategy to improve the solution quality and efficiency of the algorithm.The experimental results of production examples show that the proposed algorithm can effectively solve the multi-furnace scheduling model and realize the multi-furnace optimal scheduling of copper melting production.