Optimal generation scheduling in power system using frequency prediction through ANN under ABT environment

In a competitive and deregulated power scenario, the utilities try to maintain their real electric power generation in balance with the load demand, which creates a need for the precise real time generation scheduling （GS）. In this paper, the GS problem is solved to perform the unit commitment （UC） based on frequency prediction by using artificial neural network （ANN） with the objective to minimize the overall system cost of the state utility. The introduction of availability-based tariff （ABT） signifies the importance of frequency in GS. Under- prediction or over-prediction will result in an unnecessary commitment of generating units or buying power from central generating units at a higher cost. Therefore, an accurate frequency prediction is the first step toward optimal GS. The dependency of frequency on various parameters such as actual generation, load demand, wind power and power deficit has been considered in this paper. The proposed technique provides a reliable solution for the input parameter different from the one presented in the training data. The performance of the frequency predictor model has been evaluated based on the absolute percentage error （APE） and the mean absolute percentage error （MAPE）. The proposed predicted frequency sensitive GS model is applied to the system of Indian state of Tamilnadu, which reduces the overall system cost of the state utility by keeping off the dearer units selected based on the predicted frequency.

Learning to branch in the generation maintenance scheduling problem

To maximize the reliability index of a power system,this study modeled a generation maintenance scheduling problem that considers the network security constraints and rationality constraints of the generation maintenance practice in a power system.In view of the computational complexity of the generation maintenance scheduling model,a variable selection method based on a support vector machine(SVM)is proposed to solve the 0-1 mixed integer programming problem(MIP).The algorithm observes and collects data from the decisions made by strong branching(SB)and then learns a surrogate function that mimics the SB strategy using a support vector machine.The learned ranking function is then used for variable branching during the solution process of the model.The test case showed that the proposed variable selection algorithm-based on the features of the proposed generation maintenance scheduling problem during branch-and-bound-can increase the solution efficiency of the generation-scheduling model on the premise of guaranteed accuracy.

Generation maintenance scheduling based on multiple objectives and their relationship analysis

In a market environment of power systems, each producer pursues its maximal profit while the independent system operator is in charge of the system reliability and the minimization of the total generation cost when generating the generation maintenance scheduling(GMS). Thus, the GMS is inherently a multi-objective optimization problem as its objectives usually conflict with each other. This paper proposes a multi-objective GMS model in a market environment which includes three types of objectives, i.e., each producer's profit, the system reliability, and the total generation cost. The GMS model has been solved by the group search optimizer with multiple producers(GSOMP) on two test systems. The simulation results show that the model is well solved by the GSOMP with a set of evenly distributed Pareto-optimal solutions obtained. The simulation results also illustrate that one producer's profit conflicts with another one's, that the total generation cost does not conflict with the profit of the producer possessing the cheapest units while the total generation cost conflicts with the other producers' profits, and that the reliability objective conflicts with the other objectives.