考虑风电接入的大电网多目标动态优化调度研究

发布时间：2018-03-22 23:28

本文选题：大电网　切入点：风电接入　出处：《华南理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着全球能源和环境问题的日益突出，风能等可再生资源得到迅速发展，风力发电越来越多地接入电网，但是风力发电的强间歇性和随机波动性对系统的优化调度带来一定的影响。本文针对考虑风电接入并且兼顾经济性和环保性的大电网多目标动态优化调度问题展开研究。针对风电接入的大型电力系统，提出了以发电总燃料耗量、污染气体排放量和购电费用最小为目标的多目标动态优化调度模型，引入正/负旋转备用容量应对风电随机性的影响，并考虑了抽水蓄能机组的实际运行特性对优化调度的影响，最后对风电场的有功出力进行修正，从而保证在常规机组可承受备用的范围内，实现风电的最大化利用。采用法线边界交叉(Normal Boundary Intersection, NBI)法和原对偶内点法对上述优化模型进行求解，首先根据NBI法的基本原理将多目标优化问题转换为一系列单目标优化问题，并采用原对偶内点法对这一系列单目标优化问题进行求解，从而获得均匀分布的Pareto最优解集。然后，采用熵权双基点法从Pareto最优解集中决策出折衷最优解，为运行人员提供决策指导。最后，对某省级电网进行仿真计算，计算结果表明，NBI法、原对偶内点法和熵权双基点法的结合能够有效地求解电力系统多目标动态优化调度问题，获得的折衷最优解在兼顾三个目标最优的基础上具有明显较高的综合效益。上述算法在计算大电网动态优化调度模型时，存在模型规模大、矩阵维数高的问题，，计算耗时很大。为了提高计算速度，本文又提出了基于NBI法和原对偶内点法的多目标解耦算法及并行计算。该算法首先采用NBI法将多目标优化问题转换为一系列单目标优化问题，再根据原对偶内点法求解这类单目标优化问题时形成的修正方程系数矩阵的特殊结构，对修正方程进行解耦降阶处理，实现静态变量和动态变量的解耦分离，最后借助Matlab并行计算平台实现并行计算，快速有效地得到一系列均匀分布的Pareto最优解。最后，对某省级电网进行仿真计算，计算结果表明，所提算法能够快速、有效地获取多目标动态优化调度问题的Pareto最优解集，具有很好的实用价值。
[Abstract]:With the increasingly prominent global energy and environmental problems, renewable resources such as wind energy have developed rapidly, and wind power is increasingly connected to the power grid. However, the strong intermittent and stochastic volatility of wind power generation has a certain impact on the optimal scheduling of the system. This paper focuses on the multi-objective dynamic optimal scheduling problem of large power grid considering wind power access and taking into account the economic and environmental protection. For large scale power system connected by wind power, a multi-objective dynamic optimal scheduling model is proposed, which aims at minimizing the total fuel consumption, polluting gas emissions and the cost of electricity purchase. The influence of positive / negative rotating reserve capacity on the randomness of wind power is considered, and the effect of actual operation characteristics of pumped storage unit on optimal dispatching is considered. Finally, the active power output of wind farm is modified. Thus, the maximum utilization of wind power can be realized within the range of bearing reserve of conventional units. The above optimization model is solved by normal Boundary intersection (NBI) method and original dual interior point method. Firstly, according to the basic principle of NBI method, the multi-objective optimization problem is transformed into a series of single-objective optimization problems. The original dual interior point method is used to solve this series of single-objective optimization problems, and the uniformly distributed Pareto optimal solution set is obtained. Then, a compromise solution is obtained from the Pareto optimal solution set by using the entropy weight double-base point method. Finally, a provincial power network is simulated and calculated. The results show that the combination of NBI method, dual interior point method and entropy weight double-point method can effectively solve the multi-objective dynamic optimal dispatching problem of power system. The obtained compromise optimal solution has obvious comprehensive benefit on the basis of considering the three objectives optimality. In the calculation of dynamic optimal dispatching model of large power network, the algorithm has the problems of large scale and high dimension of matrix, and it takes a lot of time to calculate. In this paper, the multi-objective decoupling algorithm based on NBI method and the original dual interior point method and parallel computation are proposed. Firstly, the NBI method is used to transform the multi-objective optimization problem into a series of single-objective optimization problems. Then according to the special structure of the coefficient matrix of the modified equation which is formed when the original dual interior point method is used to solve this kind of single-objective optimization problem, the modified equation is decoupled and reduced to realize the decoupling separation of static and dynamic variables. Finally, a series of uniformly distributed Pareto optimal solutions are obtained quickly and effectively by using Matlab parallel computing platform. Finally, the simulation results of a provincial power network show that the proposed algorithm is fast. It is of great practical value to obtain the Pareto optimal solution set of multi-objective dynamic optimal scheduling problem.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM614

【参考文献】