Improve the unit commitment scheduling by using the neural-network-based short-term load forecasting

Unit commitment scheduling of the utility company relies upon the forecast of the demand, demand pattern, availability and capacity of the generators, minimum/maximum up and down time of the generators, and heat rate. According to the experiences of a local utility company, the difference of the fuel cost can reach a million dollars per day with different unit commitment scheduling. Accurate hour-ahead and day-ahead demand forecasting play important roles for proper unit commitment scheduling. This paper describes the procedure to improve the unit commitment scheduling by using the hour-ahead and day-ahead results from the newly developed neural network based short-term load forecasting program in the supervisory control and data acquisition and energy management system. Comparison of field records is also provided.     

A solution method using neural networks for the generator commitment problem

This paper studies the feasibility of applying the Hopfield-type neural network to unit commitment problems in a large power system. The unit commitment problem is to determine an optimal schedule of what thermal generation units must be started or shut off to meet the anticipated demand; it can be formulated as a complicated mixed integer programming problem with a number of equality and inequality constraints. In our approach, the neural network gives the on/off states of thermal units at each period and then the output power of each unit is adjusted to meet the total demand. Another feature of our approach is that an ad hoc neural network is installed to satisfy inequality constraints which take into account standby reserve constraints and minimum up/down time constraints. The proposed neural network approach has been applied to solve a generator scheduling problem involving 30 units and 24 time periods; results obtained were close to those obtained using the Lagrange relaxation method.     

Unit commitment using hybrid models: a comparative study for dynamic programming, expert system, fuzzy system and genetic algorithms

Hybrid models for solving unit commitment problem have been proposed in this paper. To incorporate the changes due to the addition of new constraints automatically, an expert system (ES) has been proposed. The ES combines both schedules of units to be committed based on any classical or traditional algorithms and the knowledge of experienced power system operators. A solution database, i.e. information contained in the previous schedule is used to facilitate the current solution process. The proposed ES receives the input, i.e. the unit commitment solutions from a fuzzy-neural network. The unit commitment solutions from the artificial neural network cannot offer good performance if the load patterns are dissimilar to those of the trained data. Hence, the load demands, i.e. the input to the fuzzy-neural network is considered as fuzzy variables. To take into account the uncertainty in load demands, a fuzzy decision making approach has also been developed to solve the unit commitment problem and to train the artificial neural network. Due to the mathematical complexity of traditional techniques for solving unit commitment problem and also to facilitate comparison genetic algorithm, a non-traditional optimization technique has also been proposed. To demonstrate the effectiveness of the models proposed, extensive studies have been performed for different power systems consisting of 10, 26 and 34 generating units. The generation cost obtained and the computational time required by the proposed model has been compared with the existing traditional techniques such as dynamic programming (DP), ES, fuzzy system (FS) and genetic algorithms (GA).     

考虑风电机组无功特性的安全约束机组组合方法

随着风能的广泛使用,安排发电计划时更多的风电机组将会被引入,这对传统的机组组合提出了新要求。风电出力具有很强的波动性,将风电出力按一个区间放入原模型中更显合理。另外,异步风电机组的结构与普通火电机组不同,异步电机发电的同时要吸收一定的无功功率,因此模型用交流潮流约束更合理。由此建立的是一个非线性混合整数问题模型,为了提高计算效率,将问题分解为2层优化子问题,第1层为无网络约束的机组组合问题,第2层为以网损最小为目标函数的交流网络约束最优潮流问题,对于最优潮流算完后仍有电压或线路潮流越限的,将形成一些新的约束返回原问题。考虑到普通异步风电机组的大量使用,在处理约束问题时对风电机组采用无功功率—电压模型,避免出现无功不足而导致电压越限。以添加了风电机组的IEEE 57节点测试系统为算例,验证了该方法的可行性。     

Stochastic Hopfield artificial neural network for unit commitment and economic power dispatch

This paper presents a Hopfield artificial neural network for unit commitment and economic power dispatch. The dual problem of unit commitment and economic power dispatch is an example of a constrained mixed-integer combinatorial optimization. Because of uncertainties in both the system load demand and unit availability, the unit commitment and economic power dispatch problem is stochastic. In this paper we model forced unit outages as independent Markov processes, and load demand as a normal Gaussian random variable. The (0,1) unit commitment-status variables and the hourly unit loading are modelled as sample functions of appropriate random processes. The problem variables over which the optimization is done are modelled as sample functions of random processes which are described by Ito stochastic differential equations. The method is illustrated by a simple example of a power system having three machines which are committed and dispatched over a four-hour period. In the method, unit commitment and economic dispatch are done simultaneously.     

A dynamic programming based fast computation Hopfield neural network for unit commitment and economic dispatch

This paper develops a new dynamic programming based direct computation Hopfield method for solving short term unit commitment (UC) problems of thermal generators. The proposed two step process uses a direct computation Hopfield neural network to generate economic dispatch (ED). Then using dynamic programming (DP) the generator schedule is produced. The method employs a linear input–output model for neurons. Formulations for solving the UC problems are explored. Through the application of these formulations, direct computation instead of iterations for solving the problems becomes possible. However, it has been found that the UC problem cannot be tackled accurately within the framework of the conventional Hopfield network. Unlike the usual Hopfield methods which select the weighting factors of the energy function by trials, the proposed method determines the corresponding factor using formulation calculation. Hence, it is relatively easy to apply the proposed method. The Neyveli Thermal Power Station (NTPS) unit II in India with three units having prohibited operating zone has been considered as a case study and extensive study has also been performed for power system consisting of 10 generating units.     

考虑小水电启停策略的准入容量计算

随着分布式电源的大力发展,其对原有系统产生了巨大影响.文中以小水电为例,首先分析分布式电源的启停对无功电压控制的影响,然后给出考虑单个分布式电源启停的准入极限功率计算模型.接着分析了多个分布式电源接入系统后,原有的计算模型结果可能过于乐观,提出了多分布式电源接入的至少准入容量的概念,它反映了"最恶劣"机组组合情况下的准入极限功率,并给出了基于主从决策的双层优化模型和求解算法.最后用IEEE 112节点系统的仿真结果证明了模型、算法的有效性.     

A combination of the genetic algorithm and lagrangian relaxation decomposition techniques for the generation unit commitment problem

Unit commitment involves the scheduling of generators in a power system in order to meet the requirements of a given load profile. An analysis of the basis for combining the genetic algorithm (GA) and Lagrangian relaxation (LR) methods for the unit commitment problem is presented. It is shown that a robust unit commitment algorithm can be obtained by combining the global search property of the genetic algorithm with the ability of the Lagrangian decomposition technique to handle all kinds of constraints such as pollution, unit ramping and transmission security.     

考虑机群出力约束的机组优化组合模型

实际电网运行中,存在朵群出力受约束的情况,针对此问题,建立了考虑机群出力约束的机组组合模型,将问题转化为两层多下级无关联的主从递阶单目标决策问题,主问题是经典的机组组合问题,从问题是以机组出力上下限修正量最小为目标的局部优化问题,进而提出了主从问题的协调算法,吉林电网实际应用表明了该方法的有效性和实用性。     

考虑网络安全约束的机组组合新算法

市场机制驱使电网运行于安全极限的边缘,考虑网络安全约束的机组组合问题变得尤为重要,基于对偶原理的拉格朗日松弛法是解决这一问题的有效途径。文章提出了一种解决网络安全约束下的机组组合问题的新算法,在拉格朗日对偶分解的基础上结合变量复制技术,通过引入附加人工约束将网络约束嵌入单机子问题中,实现在机组组合中考虑网络安全约束。该算法摆脱了现有各种处理手段在解决网络安全约束的机组组合问题时将网络安全约束与机组启停相分离的不足,揭示了安全经济调度和安全约束下的机组组合在概念上的区别和联系。     

Application of genetic-based neural networks to thermal unitcommitment

A new approach using genetic algorithms based neural networks and dynamic programming (GANN-DP) to solve power system unit commitment problems is proposed in this paper. A set of feasible generator commitment schedules is first formulated by genetic-enhanced neural networks. These pre-committed schedules are then optimized by the dynamic programming technique. By the proposed approach, learning stagnation is avoided. The neural network stability and accuracy are significantly increased. The computational performance of unit commitment in a power system is therefore highly improved. The proposed method has been tested on a practical Taiwan Power (Taipower) thermal system through the utility data. The results demonstrate the feasibility and practicality of this approach     

考虑输电网络影响的火电机组优化组合的一种实用算法

提出一种考虑输电网络损耗及线路过负荷的火电机组优化组合的实用算法。用动态规划法建立一个初始解,运用启发式手法对初始解进行修正,使之逐个满足各约束条件,得到运行可能解,并通过更新发电机起动优先顺序使此过程反复进行直至得到(准)最佳解。在求解过程中引入最优潮流计算,使考虑输电网络损耗及线路过负荷等网络因素对发电机组优化组合的影响成为可能,并提出一种调节发电机出力和改变发电机组合相结合的消除线路过负荷的方法。在IEEE-118母线(36机)系统上对所提出的算法进行了各种条件下的仿真计算,考察了网络损耗及线路过负荷对发电机组优化组合的影响,验证了所提算法对解决考虑输电网络因素影响的发电机组优化组合问题的有效性。     

考虑同步机调差系数灵敏度与频率约束的机组组合

为解决高比例新能源并网带来的系统惯量水平降低及频率安全问题,有必要从同步机的角度出发,进一步发挥同步机组的调频能力,提高系统的频率稳定。考虑系统中各同步机组频率支撑能力的不同,基于灵敏度的方法分析不同机组调差系数的改变对最大频率偏差的影响程度。综合考虑同步机与风机参与调频,推导最大频率偏差的解析表达式。在考虑频率安全约束的基础上,提出考虑同步机调差系数灵敏度的多目标机组组合模型,并采用快速非支配多目标优化算法(non-dominated sorting genetic algorithms-II, NSGA-II)进行模型求解。仿真结果表明,所提模型在考虑频率约束的机组组合模型基础上,进一步发挥了同步机组的调频能力,抑制了最低点频率的跌落,改善了系统的频率响应。     

一种求解环境经济负荷调度问题的人工神经网络新方法

提出了一种求解火力发电系统的环境经济负荷调度（EELD）问题的新的人工神经网络方法。该方法可以克服传统神经网络方法引起的参数问题，而且不论其初始点是否位于可行域内，神经网络模型所表示的动态系统的状态都会快速收敛于对应的EELD问题的全局最优解。文中还利用一个3机测试系统证实了此方法的有效性。     

电力系统中机组组合的现代智能优化方法综述

在深入探讨电力系统机组组合的各种现代智能优化算法的基础上，加以分类总结，详细评述了各种方法所取得的研究成果和存在的不足之处。具体表现在：由于模拟进化算法的随机性，不能保证每次计算都能收敛到全局最优解，同时还存在“早熟”现象；模拟退火算法存在收敛速度慢的缺点；禁忌搜索算法存在对初始解依赖性强和搜索过程只是单对单的操作；人工神经网络的学习训练易陷入局部极值区，同时指出不同的具体问题，网络适合的隐含层数目和节点数目较难确定；模糊优化算法中隶属函数的确定及专家系统中专家的知识、经验和规则的获取都是棘手的问题。     

Binary fireworks algorithm for profit based unit commitment (PBUC) problem

The deregulation of electricity markets has transformed the unit commitment and economic dispatch problem in power systems from cost minimization approach to profit maximization approach in which generation company (GENCO)/independent power producer (IPP) would schedule the available generators to maximize the profit for the forecasted prices in day ahead market (DAM). The PBUC is a highly complex optimization problem with equal, in equal and bound constraints which allocates scheduling of thermal generators in energy and reserve markets with no obligation to load and reserve satisfaction. The quality of the solution is important in deciding the commitment status and there by affecting profit incurred by GENCO/IPPs. This paper proposes a binary coded fireworks algorithm through mimicking spectacular display of glorious fireworks explosion in sky. In deregulated market GENCO/IPP has the freedom to schedule its generators in one or more market(s) based on the profit. The proposed algorithm is tested on thermal unit system for different participation scenarios namely with and without reserve market participation. Results demonstrate the superiority of the proposed algorithm in solving PBUC compared to some existing benchmark algorithms in terms of profit and number of iterations.     

Benders decomposition applied to security constrained unit commitment: Initialization of the algorithm

Benders decomposition has been broadly used for security constrained unit commitment problems, despite the fact that it may present convergence difficulties due to instabilities and to the mixed integer nature of the unit commitment problem. The initialization of Benders decomposition has been recognized as a prominent feature for the algorithm enhancement. In this work, a new Benders decomposition initialization methodology is proposed. The objective of the initialization is to include inexpensive network signals that can be added during the initial unit commitment master problem. Numerical simulations using the IEEE-118 and RTS-96 systems are performed to illustrate the benefits of the proposed initialization methodology. Results suggest that the initialization of Benders decomposition applied to security constrained unit commitment problems improves the overall convergence of the algorithm.     

机组优化组合的人工神经网络拉格朗日混合方法

本文提出了一种求解电力系统组合优化问题的混合神经网络－拉格朗日方法，至今，拉格朗日枪驰法－直被记是机组优化组合近解的实用方法，这样，基于神经网络的监督学习和自适应识别概念，我们用神经网络来推测负荷需求与拉格朗日乘子的非线性关系，并且采用了优化的学习速率和势态项来加速网络的收敛，数值计算的结果表明本文的方法是可行的。     

PSO演化神经网络集成的边际电价预测新方法

为了克服神经网络模型结构和参数难以设置,学习算法收敛速度慢等缺点,提出了一种基于粒子群优化的演化神经网络集成新模型对日前交易电力市场的边际电价进行预测。该模型将边际电价预测问题转化为神经网络实际输出与预测输出误差最小化问题,首先采用粒子群优化算法把神经网络的结构和权重映射成问题空间中的粒子,通过粒子速度和位置更新方程进行粗学习,获得多个相对占优的神经网络结构和初始权重并构成神经网络集成预测模型,然后采用梯度学习算法和交叉验证对神经网络集成单元的权重进行细学习,并以误差最小的神经网络集成单元的输出作为神经网络集成预测模型的输出。运用此方法对加州日前交易电力市场的边际电价进行了日预测,结果表明其优于三层BP神经网络预测方法。     

含多感应发电机配电网的暂态稳定研究

随着新能源的开发和利用,小型和微型分布式发电机(distributed generator,DG)开始接入配电网,在配电网中可能会出现暂态稳定问题.鼠笼式感应发电机(induction generator,IG)由于成本低、维护方便而在分布式发电中广泛应用.研究了多感应发电机配电网的暂态稳定问题,在单台感应发电机稳定判别的基础上,定义了多感应发电机配电网的暂态稳定域和临界切除时间(critical clearing time,CCT).利用转子运动方程的静等值电路,计及配电网网络方程约束和故障后感应发电机转子转速的变化,提出一种计算配电网多感应发电机临界转速和CCT的解析法.该方法与采用动态模型的仿真方法相比,计算量小,且能够揭示出多感应发电机之间暂态稳定性的内在联系.采用PSCAD/EMTDC仿真软件中的5阶动态模型仿真验证了该方法的正确性,并且分析了静态等值计算结果小于仿真结果的原因.