| Fuel management plays an important role in coal-fired power plant. A reasonable fuel planning will contribute to the security and economy of the plant. Presently, most of commercial Management Information Systems are still wandering at the transaction processing stage and scarcely provide effective decision-making support.According to the facts above, considering the objective reality of fuel plans for power plants, we use the optimization theory such as linear programming,multi-objective programming, integer programming, stochastic programming and genetic algorithm to establish the optimal fuel-planning model. It provides a set of practical solutions for the decision problem of semi-structure, multi-layer and multi-object in the fuel planning. In this article, the optimal problem of fuel planning is divided into two parts that are the purchasing model and the blending model. Using mix-integer programming and genetic algorithm methods, we establish the purchasing model that consists of the critical factors including coal mine, transportation and inventory, solve the decision problems such as multi-point purchasing, purchasing batch and the arrival time. Considering the complexity and the randomization of the parameters for coal quality, to ensure the plant production and the requirements of the environment, we establish the blending model, determine the reasonable strategy for coal allocation and provide our solutions for the decision problem of multi-object.We apply the established model into Yoyang power plant of Hua Neng Inc. We use the model-base technology to denote and manage the optimal model. Combined with Work Flow flat and Data Warehouse technology, we perform the system design. The concrete implementation of the Fuel Management System brings the obvious economic efficiency to Yoyang power plant of Hua Neng Inc. It proves the feasibility and validity of our model. |
