【作者】 宋晓喆；

【导师】 甘德强； 汪震；

【作者基本信息】 浙江大学 ， 电气工程， 2013， 博士

【摘要】 保证电力系统的安全稳定运行一直是电力工作者们研究的重要内容,其对国家经济发展和人民日常生活具有重要意义。近年来随着新能源的大规模接入和极端气象灾害的频繁发生,关于电力系统安全稳定不确定性的研究正逐渐引起学术界的重视。相对于确定性方法,基于风险的分析方法综合考虑了各种扰动发生的概率和严重程度,不但能够客观、全面地评估电力系统的安全稳定水平,并且通过基于风险的优化控制可以实现电力系统经济性与安全稳定性的协调。本文将基于风险的分析方法引入到电力系统的暂态稳定评估与协调控制问题中,在基于风险的暂态稳定评估、暂态稳定预防与紧急协调控制和低频减载(Under Frequency Load Shedding, UFLS)3个方面进行了一定的研究。第一,在基于风险的电力系统暂态稳定评估方面,为了分析灾害天气对电力系统安全稳定水平的影响,提出了一种台风天气条件下的电网暂态稳定风险评估方法。该方法通过建立台风天气条件下的输电线路故障率模型,得到未来短期内输电线路发生故障的概率。对于故障后不满足暂态稳定约束或静态安全约束的情况,建立了以切机切负荷为控制措施的紧急控制优化模型,并以最小紧急控制成本作为故障严重程度的量化指标。通过新英格兰10机39节点系统算例验证了所提风险评估方法的有效性。第二,在基于风险的暂态稳定预防和紧急协调控制方面,指出该协调控制属于大规模的混合整数非线性优化问题,其优化求解具有一定的挑战。本文分别从优化问题分解和优化问题转化的角度提出了求解该协调控制问题的两种新方法：1)从优化问题分解的角度,提出了基于暂态安全风险指标的协调控制问题求解方法。该方法定义了一个系统暂态安全风险指标,并将其作为约束引入到协调控制问题中,进而将该问题分解为相对容易求解的暂态安全风险约束预防控制子问题和后续各故障下的暂态稳定约束紧急控制子问题。在此基础上建立了一个二层优化模型,其中上层通过引入风险协调参数来调整系统暂态安全风险约束的门槛值,进而改变下层预防控制子问题和后续各紧急控制子问题的最优解,以优化协调控制总成本。提出了一种黄金分割法和逐次线性规划法相结合的混合方法求解该二层优化模型。通过新英格兰10机39节点系统和一个实际系统算例验证了所提方法的有效性,并分析了风险协调参数与预防控制成本和紧急控制总期望成本之间的关系。2)从优化问题转化的角度,提出了求解协调控制问题的连续化方法。通过引入非线性互补函数,将离散的紧急控制变量转化为等价的连续变量,从而将协调控制问题转化为一般的非线性优化问题,并采用逐次线性规划法求解转化后的优化问题。在迭代求解过程中,针对预防控制变量和紧急控制变量之间的耦合关系,进一步引入Benders分解算法以加快线性优化模型的求解速度。最后通过新英格兰10机39节点系统算例验证了所提方法的有效性。第三,在基于风险的UFLS参数优化方面,指出UFLS的切负荷量和系统频率的暂态过程有着紧密联系,在传统的控制代价最小模型基础上,提出了一种协调切负荷量和暂态频率恢复性能的综合性价比优化模型,同时考虑了各种预设场景的风险概率因素。在采用改进差分进化算法求解该问题的过程中,针对优化问题的稳态约束和超调约束提出分段惩罚函数法的处理策略来提高算法的执行效率；为了防止算法收敛到局部最优解,当算法出现停滞现象时引入混沌变异来增强种群的多样性。通过一个实际电力系统的算例验证了所提模型和算法的可行性和有效性,并表明所提模型比控制代价最小模型具有更好的综合表现。更多还原

【Abstract】 Power system stability and security are concerned with the system’s capability to withstand all kinds of potential disturbances and they have always been focuses of power system operation and control. In recent years, with the penetration of large-scale renewable energy generation and the frequent occurrence of extreme disaster weathers, the research on the impact of the embedded uncertainties on power system operation and the corresponding measures to improve system performance under these circumstances have become greater concern than ever. In essence, risk-based method can take into account both the occurrence probability and severity of all probable disturbances. The purpose of this dissertation is to introduce a risk-based analysis framework into the transient stability assessment and coordination control. The methods proposed here can not only assess the security level of system operation, but also can consider security and economy tradeoff through a risk-based security control.This dissertation is organized into three parts. In the first part, a new method to assess the transient stability risk of power systems under typhoon weather conditions is proposed, so as to analyze the influence of disaster weather on the security level of power system operation. This method models the contingency rate of transmission lines under typhoon weather, then obtain corresponding short-term contingency probabilities. For the contingency which could not satisfy transient stability constraints or post-contingency static security constraints, an optimization model for emergency control is formulated, and the minimal control cost is used to quantify the severity of the contingency. The New England10-generation39-bus system is used to demonstrate the effectiveness of the proposed method.The second part of this dissertation focuses on the coordination of preventive and emergency controls for transient stability enhancement, which is is a large-scale mixed integer nonlinear programming problem. Two new methods are proposed to solve this problem.1) The first method solves the coordination problem by decomposing it into two less complicated sub-problems. Firstly, a transient security risk index is defined and introduced as a constraint into the coordination problem, in which case the coordination problem can be decomposed into a transient security risk constrained preventive control sub-problem and transient stability constrained emergency control sub-problems. Furthermore, a bi-level optimization model of coordination control is developed, in which on the upper level a risk coordination parameter is adjusted to minimized the total coordination cost; the lower level includes preventive and emergency control sub-problems. Finally, a hybrid method that combines the golden section search method and the successive linear programming method is proposed to solve the bi-level optimization model. The effectiveness of the proposed method is demonstrated by using the New England10-generation39-bus system and a real power system.2) The second method transforms the coordination problem into a continuous problem that can be solved by traditional nonlinear optimization algorithms. By introducing a set of nonlinear complementary problems, discrete emergency control variables are equivalently transformed into continuous variables, and the coordination problem is subsequently transformed into an ordinary nonlinear programming problem, which is solved by the successive linear programming method. Furthermore, Benders decomposition is introduced to speed up the computation efficiency of linear optimization model. The test on New England10-generator39-bus system demonstrates the effectiveness of the proposed method.In the third part of this dissertation, a new UFLS parameter optimization model is developed, which considers both the load shedding amount and the frequency recovery performance during the transient period, and the risk of different operating scenarios is synthesized in addition. Further, a solution approach based on improved differential evolution algorithm is proposed, in which both steady frequency constraints and frequency overshoot constraints in UFLS are judiciously treated. A chaotic behavior is introduced into the algorithm to enhance the population diversity when the stagnation symptom occurs. The feasibility and effectiveness of the proposed model and algorithm are validated by the simulation study based on the UFLS optimization problem in a real power system.更多还原