【作者】 杨明；

【导师】 韩学山；

【作者基本信息】 山东大学 ， 电力系统及其自动化， 2009， 博士

【摘要】 随着电网规模的扩大,市场竞争机制的引入,用户对供电可靠性要求的提高以及环境保护与能源问题的日益突出,电力系统的超前调度理论与算法面临着新的挑战。这其中,旋转备用的配置问题因与系统运行的可靠性、经济性紧密关联而成为研究的焦点。对于超前调度来说,较多的旋转备用意味着较高的供电可靠性,但这必然会使系统的供电经济性受到影响,反之亦然。此外,对于相同的备用容量,分布在不同的机组上其效果也是不同的。以往研究中,旋转备用所采用的确定性配置方法,虽然容易理解且实现简单,却无法量化备用配置与系统运行可靠性间的对应关系,难以找到备用产生的根源,从而造成备用配置不当、责任归属不明的后果,无法适应电力工业市场化改革的需要。而随机性、波动性更强的间歇式可再生电源在电力系统的应用激化了系统运行可靠性与经济性间的矛盾。因此,在当前形势下,以超前调度为线索,以旋转备用配置为焦点,对电力系统运行可靠性与经济性间的协调问题进行探讨,进行适应电力系统运行趋势变化、符合市场机制操作的超前调度模型和方法的研究是迫切而又具有重要的理论与实践意义的。在上述背景下,本文围绕超前调度及其旋转备用配置问题,在回顾研究历程、总结存在问题的基础上,以经济原则为导向,以优化数学为工具,对其进行深入、细致地研究,其主要工作和创新性成果体现如下。(1)针对确定性备用配置方法的不足,提出了基于等响应风险约束的超前调度方法。该方法以概率性手段进行超前调度的旋转备用配置,可以将系统的响应风险维持在给定水平;以电力不足期望(Expected Demand NotSupplied,EDNS)指标表征系统的响应风险,通过引入{0,1}变量对EDNS表达的适当变化使其限值可以作为约束并入超前调度模型,形成{0,1}混合整数优化问题,实现了调度与响应风险评估的统一;通过引入针对整数变量的附加约束,将{0,1}混合整数优化问题转变为连续变量的非线性优化问题,并利用内点法进行求解,加快了模型的求解速度。(2)针对基于等响应风险约束超前调度方法中系统响应风险水平的选择问题,提出了计及用户停电损失的超前调度方法。该方法通过用户停电损失评价率(Interrupted Energy Assessment Rate,IEAR)指标和系统的电量不足期望(Expected Energy Not Supplied,EENS)指标将调度所对应的用户停电损失期望(Expected Customer Interruption Cost,ECOST)体现于目标函数中,在寻求扩展目标函数最小的同时刚好确定系统需维持的响应风险水平及备用配置方案;延续了基于等响应风险约束超前调度中对响应风险指标的解析化处理方法,构建模型形成{0,1}混合整数优化问题;将超前调度物理模型上的解耦方法与算法上的分解计算方法相结合,成功地求解了由混合整数优化问题转化而成的二次优化问题,提高了模型的计算效率。(3)在计及用户停电损失超前调度方法的单母线模型的基础上提出了计及网络安全约束及用户停电损失的超前调度方法。该方法将网络安全约束纳入到计及用户停电损失的超前调度模型中,同时考虑了时间耦合约束、空间关联约束以及状态耦合约束,使调度在概率、安全与时间耦联上有机结合,更加符合电力系统的运行实际;模型中将事故后机组的输出功率以及切负荷功率作为独立变量进行决策,旋转备用配置由事故前后机组输出功率的最大调整量确定,蕴含信息更加全面;在采用原-对偶内点法进行求解时,通过将耦合变量与非耦合变量的分离,构建了KKT条件牛顿修正方程的快速求解方法,使内点法的迭代效率提高,从而加快了模型的求解速度,增强了模型的可行性。(4)在深入分析温室气体配额-交易排放控制策略的基础上提出了温室气体配额-交易框架下的超前调度方法。该方法在超前调度中计及了发电过程所伴随的温室气体排放问题,能够适应当前电力系统节能减排工作的需要;充分利用燃料市场与排放配额交易市场的价格信息,导出了温室气体排放影响因子,对机组的发电成本曲线进行修正;方法实现简单,无需增加任何附加约束,仅需在调度目标中引入环境影响因子即可完成,故对以往超前调度算法具有很好的继承性。总之,本文工作在继承前人成果的基础上,在优化模型和算法上、风险与经济的折中上、可靠性与决策的关联上,对电力系统超前调度的研究取得一定的进展。尽管如此,面对当今复杂电力系统,如何研究电力系统运行中各元件的概率规律、经济规律,以及能源开发应用的策略,使本文的研究能在实际中得以实现,都需要进一步深入地开展工作。更多还原

【Abstract】 With the expansion of grid scale,the introduction of market competition mechanism, the customer’s higher requirement of power supply reliability and the increasing concerns about environment and energy issues,the power system advanced dispatch theories and algorithms are facing new challenges.Among these challenges,the spinning reserve allocation becomes the research focus because of its direct connection to power system operating reliability and economy.For advanced dispatch,more spinning reserve means higher operating reliability,but the operating economic objective will be inhibited at the same time,and vice versa.Further,the operating reliability level will be different when the same spinning reserve capability is allocated to different generators.In previous studies,the deterministic methods for spinning reserve allocation are easy to understand and implement.However,it is hard to quantify the relationship between the spinning reserve and system operating reliability as well as to find out the cause of spinning reserve.All these will result in improper spinning reserve allocation and spinning reserve responsibility allocation.Therefore,the traditional deterministic spinning allocation methods cannot satisfy the requirements of the power system deregulation.Especially,the increase of intermittent renewable power sources in power system intensifies the conflict between system operating reliability and economy. Therefore,in the current situation,taking spinning reserve allocation into account and focusing on the coordination of power system operating reliability and economy,to propose new advanced dispatch model and algorithm,which can follow power system changing trend and be suitable for electrical market operation,has significant theoretical and practical meanings.As mentioned above,this thesis around spinning reserve allocation problem, reviewing the history of advanced dispatch,aiming at the shortage of current advanced dispatch method,based on economic theory,using optimization skill, does thorough research on advanced dispatch.The main contribution of this thesis can be extracted as:(1) Aiming at the shortage of traditional deterministic advanced dispatch,an advanced dispatch method named advanced dispatch with equal response risk is proposed in this thesis.Probabilistic spinning reserve allocation method is used in this method.The response risk can be kept in the pre-defined level using this method.The expected demand not supplied(EDNS) index is used here to quantify the relation between the spinning reserve allocation and system reliability level.In the context,the EDNS is reformulated to be suitable for being incorporated into traditional advanced dispatch as a constraint through introducing {0,1} variables to its express.In this way,the dispatch process and the response risk estimate process can be incorporated.The resulted model is a mixed integer optimization problem.Additional constraints for {0,1} variables are then bounded to the original problem to converse mixed integer optimization problem to quadratic optimization problem to use traditional available quadratic optimization programs.And this accelerates the calculation of the model.Focusing on how to specify the system operating reliability level in advanced dispatch with equal response risk method,an advanced dispatch method considering customer interruption costs is proposed.Expected energy not supplied(EENS) and interrupted energy assessment rate(lEAR) indices are introduced here to assess the expected customer interruption costs(ECIC).In the proposed method,the expected customer interruption costs because of spinning reserve insufficiency are added to the objective function.Through this way,the predetermined operating reliability level is no longer necessary.Instead,the optimal reliability level as well as spinning reserve allocation can be decided during the optimizing progress.Further,physical and algorithmic decomposed methods are cooperated to solve the quadratic optimal problem which results from original {0,1} mixed integer optimization.Based on advanced dispatch considering customer interruption costs method, an advanced dispatch method considering network security constraints and customer interruption costs is proposed.The time-coupled constraints, space-coupled constraints and state-coupled constraints are included in the model which is more practical for power system operation.The pre-contingency active power outputs,post-contingency active power outputs and post-contingency load shedding are treated here as independent variables in this method.Because of the large size of the model,a decomposed method for solving primal-dual interior point KKT condition,which makes full use of weak coupled nature in time-coupled and state-coupled constraints,is proposed.This method can speed up the calculation,and enhance the effectiveness of the proposed method.Based on analysis of cap-and-trade greenhouse gases emission control policies,a compromising advanced dispatch method under cap-and-trade framework is proposed.The proposed method takes greenhouse gases emission into account,and this is suitable for power system energy conservation and emission reduction requirements.An emission affection factor is deduced using fuel price,emission allowance price and emission character of generators.The emission affection factor can be used to modify the generator’s generation-cost curves.For the proposed method,no additional constraints are needed,and the only change of advanced dispatch model is to modify the objective function with emission allowance factors,so the method is easy to implement.To sum up,absorbing the advantages of current researches,this thesis develops advanced dispatch in model and algorithm,tradeoff between response risk and operating economy and relationship between operating reliability and decision.However,facing the quick development of current complex power system,there are still many works to do,such as the calculation of power system components fault rates in real-time,power system operation economic rules,as well as energy utilization strategies.更多还原