Unit commitment using quadratic programming and unit decommitment

We have developed an innovative power generation scheduling method using quadratic programming and unit decommitment (QPUD). The advantage of using our method is that it can simultaneously solve unit commitment and economic load dispatch problems. We relax the binary variables of the unit state into continuous variables to apply QP to the problem. The initial unit commitment is made if a variable of more than zero is set to one and the other is set to zero. Then continuous variables are set at all time horizons for only one unit and in this condition a generation schedule is made using QP. If this unit continuous variable is equal to zero, it can be decommitted and other variables set to one. These procedures converge the value of the variables to zero or one and reduce the objective function of the problem, which is the sum of fuel and start-up costs. We applied this method to a test system of daily generation scheduling that consisted of 29 thermal units. The schedule satisfied all constraints, including load-power balance, operation reserve, minimum up/down-times, and fuel consumption. Generation costs were reduced by 0.088% compared with the conventional priority method. This demonstrates the effectiveness of the proposed method.