Study Of Active And Reactive Power Load Forecasting And Optimization

According to the depletion of exhaustible resources and deterioration of the ecological environment, it is imperative to save energy and reduce emission. And it's extremely crucial to deal with this situation for the power grid. So the grid interconnection of large-scale renewable resources generation and the increasing responsive load prompt the coordinated development of backbone power grid, Distributed Generation(DG) and Microgrid in future power grid.Against this background and under the precondition of safe operation of power grid, it is very important to realize the refinement decision-making to utilize the initiative of the power grid scheduling and control for dealing with the passive variation. Based on the previous studies, it has important theoretical and practical value for improving the operation efficiency of the future power grid for how to improve the forecast accuracy, analysis the regularity of voltage support and solve the unit commitment problem with AC power flow.From the perspective of short-term operation optimization, this dissertation focuses on the load forecasting of the nodal active and reactive power, the research of reactive power optimization in AVC and the unit commitment considering AC power flow constraints. The main works and achievements of the dissertation are as follows:1. Considering the weakness of separated nodal load forecasting and based on the correlation analysis, this paper proposes a multi-dimensional nodal load forecast system and corresponding method for effective prediction of any nodal load of the grid. This system includes topology partitioning of the grid energy flow according to layers and regions, basic forecasting unit composed of each layer's total amount of load and its nodal loads, and combination forecasting to any node. The forecasting method is based on techniques including the least square support vector machine, Kalman filtering and weighted recursive least squares. Multi-dimensional load forecasting can realize any nodal load forecasting with the help of energy flow topology, which achieve the organic combination of nodal load variation regularity and association relationship between nodes. The proposed method can meet the demand of security analysis and decision-making in power grid.2. Corresponding to the hierachical voltage control mode of AVC, the reactive power optimization in the tertiary voltage control adopts two-stage processing method. The dynamic reactive power optimization aims to determine the discrete variables in advance(the next24hours) with the full use of continuous variables. And based on this, the advanced reactive power optimization only considers the continuous variables, which will help to decide the base point of the secondary voltage control. A new method that combination of genetic algorithm and the interior point method is proposed to solve the coupling of time intervals caused by the action number constraint of the discrete variables:In the outer circle, the improved genetic algorithm is employed to solve the optimal arrangement of discrete variables. In the inner part, the nonlinear primal-dual interior point method is adopted to solve the continuous variables optimization. And this method can solve the dynamic reactive power optimization effectively.3. Considering the effects on unit commitment made by voltage and reactive power constraints, the model of unit commitment considering ac power flow constraint is established, which includes the limits for safe operation of the generator. According to the Benders decomposition, the primal problem is decomposed into a master problem and a subproblem. The master problem solves unit commitment without AC constraints, and the subproblem checks the AC constraints for the result of the master problem. Benders cuts which formed by subproblem are introduced to the master problem to reschedule unit commitment. The mechanism of voltage and reactive power characteristic affecting the combination states of the units is analyzed, and the essence of the decision is the compromise and coordination of the economic and security operation of the power grid.4. A security-constrained unit commitment method considering corrective control is proposed. It helps to improve the security of power grid operation for introducing the unit states to the pre-contingency dispatch. Base on the Benders decomposition, the optimization problem is decomposed into a master problem and a subproblem. Unit commitment considering corrective ability of the power system can not only reduce the impact of the pre-contingency on normal condition and improve the operation efficiency, but also improve the adaptability of the decision. Meanwhile, ac power flow constraint is introduced to avoid the infeasible solution of the unit commitment problem.