The Flexible Purchasing Policy Research Of Electric Power Under Uncertainty

Electric Power industry is one of the most important basic energy industries which connected with the national welfare and people’s livelihood. With the reform of electric power system, its traditional vertical integration electric power system had changed to electric power supply chain. Traditional vertical electric power system included plant grid separation and competitive pricing online. Electric power supply chain involved generation, transmission, distribution and sales which are independent to each other. In the electric power supply ch ain, electric power is perishable and face to the dynamic uncertain outside environment, like uncertain demand and operate fuel cost. How to make the best purchasing policy with goal of minimum gross cost under uncertain scenarios is very important. Based on this circumstance, this paper is focus on following problems.Firstly, study the electric power supply chain. Electric power system is a complex system, and owing to the product of electric power ’s real-time capability, it’s meaningful to study electric power supply chain and compared it with traditional supply chain. Result shows that tradition SCM concerns the good flow throughout the logistics network in the physical distribution. However, electric power SCM focused real-time balancing of power supply and demand in the grid dispatching which is more important.Secondly, the model research under one uncertain variable(only fuel cost f). Introduce the traditional stochastic programming model under uncertain f, and then put forward the new robust optimization model. Make the numerical experiments and compare these two methods. The research concluded that the stronger the robustness, the higher the gross cost in robust optimization model. Robust optimization method is more flexible than stochastic programming method in worst scenarios. It’s better to selected the best uncertain scope and best optimization method according to the real uncertain scope and robustness request.Thirdly, research the robust decision under the uncertain variable demand d and fuel cost f, and they are independent to each other. Use the robust optimization method to study the electric purchasing decision. Supposed two uncertain variables are independently in order to prove the useful of experimental algorithm. In conclusion, when the uncertain scope of d and f are changed proportionally, the minimum gross cost is positive proportional to the step of uncertain scope. When uncertain scope of d and f are changed proportional independently, gross cost is higher with the uncertain scope enlarged, but gross cost and uncertain scope are not positive proportionally.Finally, research the robust decision under the uncertain variable demand d and fuel cost f which are associated with each other. Analyze the policy relation between d and f, then crated the robust optimization model in which d and f are associated to each other. Make numerical experiments and use DOE method to analyze which variable is more affected to the gross cost. Got a conclusion that compare with uncertain variable demand d, fuel cost f have more effects to the gross cost. So Grid Company should more consider the fuel cost when making purchasing decision. And this conclusion is also suitable to other industry especially related oil industry.This paper is focus on the study of electric power purchasing decision under uncertainty. Based on the negative relationship between fuel cost f and demand d, robust optimization is created. This model and its conclusion not only used is electric industry but also useful in other industries especially related oil industry, like auto industry.