Skip to Main Content
Distribution utility purchases power in multi-periods, and the decision-making at risk before this period must affect that the distribution utility makes decision to this period. For this reason, Multi-period conditional value at risk (CVaR) is used to measure the dynamic risk when distribution utility purchases power and the decision-making of power purchase are dynamically revised by Bayesian methods. In multi-markets, the optimal model of power purchase is simulated in multi-periods to minimize the risks and achieve the optimal power allocation by genetic algorithm at the given confidence level. As the simulation shows the power purchase model of dynamically revision can better reflect the risk distribution utility faces and the practically revenue rates at minimum risk.