Research On Minimum-Cast-Based Raw Material Assurance System And Its Application

The main errand of nonferrous metallurgy enterprise, which has continuous processing and belong to flow enterprise, is refine nonferrous through complex physical and chemical process. Raw material supplying plays an important role in enterprise's products. There are too many problems in raw material supplying, such as, too many sources and types, complex constituent, variation price, large amount of needing, occupied funds, and safe inventory needed, et al. Raw material assurance system is established under these circumstances. The main research of this thesis concentrates on the following parts: some optimization algorithms, raw material purchase, raw material inventory and real time early alarm of raw material purchase.The conception of raw material assurance (RMA) and raw material assurance system (RMAS) were proposed in this thesis. As a new idea of raw material management in enterprise, RMA integrated technologies in system engineering, automation and information science, assured the right production, aimed at minimum raw material cost of enterprise. The RMAS is the functional system under the ideal of RMA. There are 5 functional sub systems and 3 supports sub system in RMAS, and their systematic architecture and functions are given in this paper.The conception of intelligent integrated raw material assurance (IIRMA) was proposed in the thesis, according to the analysis of raw material supplying in the nonferrous metallurgy enterprise, which has continuous processing.In the process of research, some optimization algorithms will be used. Due to the limitation of the traditional optimization algorithm and the fact that intelligent optimization is gaining wider range of application over time, the thesis concentrates on the research of chaotic neural network and hybrid genetic algorithm. An improved chaotic neural network has been proposed to integrate chaotic dynamics and convergence dynamics in the purpose of conquering the shortcoming of neural network optimization that tends to converge at local optima in linear optimal problem. Besides, the thesis puts forward a hybrid algorithm which combines simulated annealing and genetic algorithm. The hybrid algorithm uses the process of simulated annealing to improve the result of genetic algorithm and to help solving the possible local optima convergence in traditional genetic optimization. The later can solve nonlinear and large-scale combination optimal problems The purchase of raw materials plays crucial role in nonferrous metallurgy enterprise that characterizes continuous processing. It is the very beginning of an enterprise's production activity. Not only does the purchase of raw materials occupy large amount of capital, but also the quantity and quality of the raw materials purchased decides the smoothness and continuity of the production. It's not exaggerating to say that the supply department controls the heart beat of theenterprise. On the basis of analysis and summing up the raw material purchase behavior of this kind of enterprises, the thesis puts forward an intelligent integrated optimization strategy for raw material purchase in the purpose of minimizing the cost on raw material purchase without affecting the quality and quantity of the enterprise's production, so as to maximize the profit. According to the real raw material purchase situation of a certain metallurgy enterprise, a purchase model is established and the chaotic neural network is applied to the model to obtain an optimized purchase strategy. Results show that, with this optimized strategy, a considerable amount of capital has been saved.The importance of the raw material purchase in nonferrous metallurgy enterprise makes the purchase early warning loom large and the effectiveness of the early warning has direct influence over the profit of the enterprise. The thesis proposes the concept of intelligent integrated quality-price real-time early warning and gives a general framework. The warning system combination the purchase strategy and market price of the enterprise's p...