鲁棒优化方法在输电系统规划与运行中的应用研究

发布时间：2018-05-05 02:19

本文选题：鲁棒优化 + 不确定性　；参考：《南京邮电大学》2017年硕士论文

【摘要】：近年来,可再生能源得到迅速发展,其中以风电尤为突出。可再生能源能有效地改善环境污染问题,缓解能源危机。风电预测技术的局限性,气候、地形分布的多样性,风电本身间歇性、不确定性以及部分可预测性的特点,对电力系统规划与运行带来挑战。风电的概率分布函数本身就是不确定的,如果将其作为已知量进行处理不能反映风电出力的真实情况,无法从根本上保证电力系统规划与运行方法的有效性。鲁棒优化方法可解决上述随机变量的概率分布函数不完全已知的情况,本文讨论两种鲁棒优化方法,其一是盒式鲁棒优化方法,该方法将风速以及风机输出功率设定为随机变量,使用区间不确定集刻画风速,采用随机变量的分布情况刻画风电输出功率的随机特性,把随机问题改写成确定型问题。另一种是概率分布鲁棒优化方法,已知不确定量的一阶和二阶矩,采用条件风险价值(CVaR,Conditional value at risk),推演出包含不确定量的CVaR约束模型,再采用对偶优化理论、Schur补和S-lemma,把CVaR约束转变为双线性矩阵不等式(BMI,Bi-linear Matrix Inequality)约束,把基于BMI约束的不确定模型化为确定模型。最后,确定性模型的求解使用基于BMI的免疫粒子群算法。本文研究了上述鲁棒优化方法在输电系统柔性负荷调度、输电系统无功规划和输电系统可用输电容量评估问题中的应用。本文把风电作为新能源的代表进行研究,新能源出力的不确定性,可使用柔性负荷调度予以平衡,然而柔性负荷的互动具有自主性、随机性与无序性,将会进一步增加电力系统中的不确定特性,本文在上述背景下提出基于鲁棒优化方法的输电系统柔性负荷调度策略。采用概率分布鲁棒优化方法解决输电网无功规划问题,可有效应对于风电函数概率分布集合中的任意可能分布,在给定的概率约束下确保输电网的运行安全,并且将输电网的网损和无功设备的投资成本之和最小化。采用鲁棒优化方法解决输电系统可用输电容量评估问题,可应对于风电概率分布不完全可知的情况,该方法可以保证网络安全运行约束前提下,最大化可用传输容量。仿真结果以及算例分析表明所提方法在输电系统规划与运行中应用的可行性和有效性。
[Abstract]:In recent years, renewable energy has been developed rapidly, especially wind power. Renewable energy can effectively improve the problem of environmental pollution and alleviate the energy crisis. The limitation of wind power forecasting technology, the diversity of climate and terrain distribution, the intermittent, uncertainty and partial predictability of wind power bring challenges to power system planning and operation. The probability distribution function of wind power is itself uncertain. If it is treated as a known quantity can not reflect the true situation of wind power generation, it can not fundamentally guarantee the effectiveness of power system planning and operation methods. Robust optimization method can solve the problem that the probability distribution function of random variables is not completely known. In this paper, we discuss two robust optimization methods, one is boxed robust optimization method, and the other is boxed robust optimization method. In this method, wind speed and fan output power are set as random variables, interval uncertainty sets are used to describe wind speed, and the distribution of random variables is used to describe the stochastic characteristics of wind power output power. The other is the probabilistic distribution robust optimization method, in which the first and second moments of the uncertainty are known, and the conditional risk value (Cvar) conditional value at riskn is used to deduce the CVaR constraint model with uncertainty. Then by using the dual optimization theories such as Schur complement and S-lemma, the CVaR constraint is transformed into a bilinear matrix inequality constraint, and the uncertain model based on BMI constraint is transformed into a deterministic model. Finally, the BMI-based immune particle swarm optimization algorithm is used to solve the deterministic model. In this paper, the application of the robust optimization method to the problems of flexible load scheduling, reactive power planning and the evaluation of the available transmission capacity of transmission systems is studied. This paper studies wind power as the representative of new energy. The uncertainty of new energy output can be balanced by flexible load dispatching. However, the interaction of flexible load has autonomy, randomness and disorder. It will further increase the uncertainty of power system. In this paper, a robust optimization method based flexible load scheduling strategy is proposed. Using the robust optimization method of probability distribution to solve the problem of reactive power planning in transmission network can effectively deal with any possible distribution in the set of probability distribution of wind power functions and ensure the safety of transmission network under given probability constraints. The sum of network loss and investment cost of reactive power equipment is minimized. The robust optimization method is used to solve the problem of evaluation of the available transmission capacity of transmission systems. It can be used to maximize the available transmission capacity under the condition that the probability distribution of wind power is not completely known. Simulation results and numerical examples show that the proposed method is feasible and effective in transmission system planning and operation.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM715;TM732

【参考文献】