Application research on case-based reasoning combined with adaptation model for dynamic scheduling of steel production
Chunguang Chang, Dingwei Wang, Kunyuan Hu, B. Zbeng, Huiying Li, Chunguang Chang · 2004
The case-based reasoning system combined with adaptation model (CBRSCWAM) is proposed to optimize dynamic scheduling problem for the iron & steel production. The model function, model establishing and model resolving are presented in detail. To resolve easily, the nonlinear adaptation model is converted into linear adaptation model, then, the two-stage algorithm is utilized to improve resolving efficiency. On the first stage, genetic algorithm (GA) that introduces heuristic to generate population is adopted to deal with the problem to reduce dimension. On the second stage, simplex method is used to resolve the simplified model. The simulation result obtains tests that the CBRSCWAM is much fitful to resolve complex problem such as iron & steel dynamic scheduling than single model approach or conventional CBR approach. With the perfect of the prototype system, it is put into application, and much profit may appear.