Study Of Fuzzy Programming Of Production Planning For Rare Earth Plants

Because of the questions of the joint cost of production that is hard to allocate and the imbalance between the product demand and the partition and so on, the production planning model has become the hot topic in the Rare Earth(RE) plants research field in the recent years. In the assignment of the cost of joint production, some scholars provided to establish the joint product model based on activity-based costing. The fuzzy decision-making has proposed since 70 years of last century, it has made a series of progress in theoretically and applied in many fields.A model of RE products mix (M1) was established in the paper. It has two main characteristics: First, all weights were measured according to theirs oxide, the material ingredients were representing by the grade and the partition pattern, the activiy were measured by the output weight. So, the product mix model was simplified .Second, It improves the activity-based costing, making the sub-product-group containing all elements of RE as the cost-assignment object. This model has been used in the Baogang Rare Earth Company for more than half year; It shows model is simple and feasible. It increases profits, by comparing the result of the model optimization and the enterprise original product mix, while reduced the output and consumption of material.The Rare-Earth element is paragenetic mineral ,the fluctuation of the market demand of various elements will lead to the fluctuation of the RE concentrates and products price. And the downstream overseas enterprise may easily disturb the market price, thus it is difficult to forecast the marke. In the paper, we introduce linear fuzzy program and the production planning fuzzy optimization model (M2) was established in RE plants. The model uses the triangle fuzzy membership function that expressed indefinite price and the demand of the materials. The solution set of the most optimistic linear programming was solved by the Buckley model and the solution set of the most pessimistic linear programming was solved by the Julien model. The solution set of the model provide simple and practical choice plan for the decision-makers under fuzzy decision-making.When the RE material partition (weight ratio of some kind of ReO in all ReO) were not balanced with the product demand, the material which easily preserved must became backlog, so that it can be satisfied with the big demand order. In view of this, one production planning model took backlog into account (M3) was established in RE plants. The objective function of the model is piecewise continuous function. The model was solved by the iterative linear program. Its optimized result show that the output value of the produce plan model which took backlog into account was improved than the products mix model without backlogging.Finally, the production plan optimization software which based on the Web was developed in this paper, Java Web call component which generated by Matlab Builder for java ,the software realize the parameter input, the result output and other request function by graphic user interface. This software brings the operation of enterprise staff more convenience.