Fast estimation of transient stability limits by combining direct method with least squares technique

This paper proposes a novel method for calculating transient stability limits by combining the direct method with the least squares technique. In this paper, the nonlinear relationship between the transient energy margin (ΔV) and the generator mechanical power input (P_m) at different fault clearing times (T_cl) has been studied. The functional relation of ΔV=f(P_m,T_cl) has been established by using the least squares technique. In this way, the transient stability limits at different fault clearing times under a given large disturbance can be quickly obtained. In order to make the proposed method suitable for handling large scale power systems, dynamic equivalent technique has also been used. The proposed method has then been tested on the 59-machine Northeast China power system, and the results have been compared with other methods.     

基于相量测量技术和模糊径向基网络暂态稳定性预测

提出一种新的基于模糊聚类的径向基神经网络及其训练算法,利用同步相量测量装置获得的故障后短时间内各发电机的功角,经简单运算后作为神经网络的输入,其输出为多机电力系统稳定性的分类结果。对４９机实际系统在不同接线方式和故障位置条件下,进行了有无切机控制两种情况下的数值仿真实验,结果表明所提出的方法对系统的失稳预测和切机控制决策是有效的,神经网络训练时间短,分类精度高。     

A new intelligent algorithm for online voltage stability assessment and monitoring

This paper presents a new approach for assessing power system voltage stability based on artificial feed forward neural network (FFNN). The approach uses real and reactive power, as well as voltage vectors for generators and load buses to train the neural net (NN). The input properties of the NN are generated from offline training data with various simulated loading conditions using a conventional voltage stability algorithm based on the L-index. The performance of the trained NN is investigated on two systems under various voltage stability assessment conditions. Main advantage is that the proposed approach is fast, robust, accurate and can be used online for predicting the L-indices of all the power system buses simultaneously. The method can also be effectively used to determining local and global stability margin for further improvement measures.     

Voltage stability monitoring of power systems using reduced network and artificial neural network

This paper presents network reduction based methodologies to monitor voltage stability of power systems using limited number of measurements. In a multi-area power system, artificial neural networks (ANNs) are used to estimate the loading margin of the overall system, based on measurements from the internal area only. Information regarding the important measurements from the external areas is considered in measurement transformation through the network reduction process, to enhance the estimation accuracy of the ANNs. A Z-score based bad or missing data processing algorithm is implemented to make the methodologies robust. To account for changing operating conditions, adaptive training of the ANNs is also suggested. The proposed methods are successfully implemented on IEEE 14-bus and 118-bus test systems.     

A new wavelet network based method to estimate the lightning-related risk of failure of power system apparatus

This paper proposes a new and fast wavelet network based method for estimating the risk of failure caused by lightning overvoltages in arrester protected networks. First, failure risks obtained by simulations are used as the training data for training the wavelet network. The trained wavelet network is then used for accurate and fast estimating of the lightning-related risk of failure of power system apparatus for all possible conditions. The accuracy of the proposed method has been tested and verified under various conditions in the 230 kV network of Sistan–Baluchestan. Performance of the new method has also been compared with several existing methods under same conditions, and the test results show better accuracy of the proposed method. The proposed method not only does not have the restriction of conventional methods, but also it does not have the limitations associated with traditional neural networks based algorithms such as convergence to local optimum points, over-fit and/or under-fit problems. The main contribution of the paper is an accurate (due to proper selection of the training data set based on the k-fold cross validation technique and using wavelet network for estimation), fast (mean calculation time for the network risk of failure computation is 54 s) and simple wavelet network-based algorithm (as compared to the conventional algorithms) for estimating the lightning-related risk of failure of power system apparatus.     

人工神经网络在电力系统暂态稳定分析中的应用

人工神经网络以其可学习的特性和其高度的并行结构在电力系统暂态稳定分析中具有极大的应用潜力。本文阐述了应用较为广泛的逆导（ＢＰ）人工神经网络模型求解电力系统故障后的稳定平衡点和确定暂态极限功率。实例证明，本文提出的这种新方法计算速度快、精度高。具有很强的自适应性和工程实用的广阔前景。     

A new power system transient stability assessment method based on Type-2 fuzzy neural network estimation

Transient stability assessment (TSA) of large power systems by the conventional method is a time consuming task. For each disturbance many nonlinear equations should be solved that makes the problem too complex and will lead to delayed decisions in providing the necessary control signals for controlling the system. Nowadays new methods which are devise artificial intelligence techniques are frequently used for TSA problem instead of traditional methods. Unfortunately these methods are suffering from uncertainty in input measurements. Therefore, there is a necessity to develop a reliable and fast online TSA to analyze the stability status of power systems when exposed to credible disturbances. We propose a direct method based on Type-2 fuzzy neural network for TSA problem. The Type-2 fuzzy logic can properly handle the uncertainty which is exist in the measurement of power system parameters. On the other hand a multilayer perceptron (MLP) neural network (NN) has expert knowledge and learning capability. The proposed hybrid method combines both of these capabilities to achieve an accurate estimation of critical clearing time (CCT). The CCT is an index of TSA in power systems. The Type-2 fuzzy NN is trained by fast resilient back-propagation algorithm. Also, in order to the proposed approach become scalable in a large power system, a NN based sensitivity analysis method is employed to select more effective input data. Moreover, In order to verify the performance of the proposed Type-2 fuzzy NN based method, it has been compared with a MLP NN method. Both of the methods are applied to the IEEE standard New England 10-machine 39-bus test system. The simulation results show the effectiveness of the proposed method in compare to the frequently used MLP NN based method in terms of accuracy and computational cost of CCT estimation for sample fault scenarios.     

On-line voltage stability assessment using radial basis function network model with reduced input features

In recent years, voltage instability has become a major threat for the operation of many power systems. This paper presents an artificial neural network (ANN)-based approach for on-line voltage security assessment. The proposed approach uses radial basis function (RBF) networks to estimate the voltage stability level of the system under contingency state. Maximum L-index of the load buses in the system is taken as the indicator of voltage stability. Pre-contingency state power flows are taken as the input to the neural network. The key feature of the proposed method is the use of dimensionality reduction techniques to improve the performance of the developed network. Mutual information based technique for feature selection is proposed to enhance overall design of neural network. The effectiveness of the proposed approach is demonstrated through voltage security assessment in IEEE 30-bus system and Indian practical 76 bus system under various operating conditions considering single and double line contingencies and is found to predict voltage stability index more accurate than feedforward neural networks trained by back propagation algorithm and AC load flow. Experimental results show that the proposed method reduces the training time and improves the generalization capability of the network than the multilayer perceptron networks.     

A New Optimal Under-frequency Load-shedding Method Using Hybrid Culture–Particle Swarm Optimization–Co-evolutionary Algorithm and Artificial Neural Networks

Optimum under-frequency load shedding during contingency situations is one of the most important issues in power system security analysis; if carried out online fast enough, it will prevent the system from going to a complete blackout. This article presents a new fast load-shedding method in which the amounts of active and reactive power to be shed are optimized with a dynamic priority list by using a hybrid culture–particle swarm optimization–co-evolutionary algorithm and artificial neural network method. The proposed method uses a five-step load-shedding scenario and is able to determine the necessary active and reactive load-shedding amounts in all steps simultaneously on a real-time basis. An artificial neural network database is established by using offline N – K (K = 1, 2, and 3) contingency analysis of the IEEE 118-bus test system. The Levenberg–Marquardt back-propagation training algorithm is used for the artificial neural network, and the training process is optimized by using a genetic algorithm. The artificial neural network database is updated based on new contingency events that occur in the system. The simulation results show that the proposed algorithm will give optimal load shedding for different N – K contingency scenarios in comparison with other available under-frequency load-shedding methods.     

基于XGBoost-EE的电力系统暂态稳定评估方法

深度学习在暂态稳定评估中发挥着越来越重要的作用,然而电网规模逐渐扩大导致数据出现维数灾难,这对模型的性能提出了更高的要求.目前,暂态稳定特征构建需要依靠人工经验,具有主观性;深度学习的模型在设计和训练上耗时、耗力.针对以上两点,结合极限梯度提升(XGBoost)算法和实体嵌入(EE)网络,提出了一种基于XGBoost-...     

A feed-forward artificial neural network with enhanced feature selection for power system transient stability assessment

This paper describes an approach where an artificial neural network is used to predict the stability status of the power system. This efficient and robust approach combines the advantages of the time–domain integration schemes and artificial neural network for on-line transient stability assessment of the power system. The transient stability index has been obtained by the extended equal area criterion method and is used as an output of the neural network. Two feature selection techniques have been used to identify the input variables best suitable for training. The proposed technique predicts the transient stability index correctly, without any false alarm. In addition, the transient stability index as an output of the neural network helps to implement possible control actions. The results obtained demonstrate the potential for neural network to be a part of any on-line dynamic security assessment tool.     

Robust decentralized neural networks based LFC in a deregulated power system

In this paper, a decentralized radial basis function neural network (RBFNN) based controller for load frequency control (LFC) in a deregulated power system is presented using the generalized model for LFC scheme according to the possible contracts. To achieve decentralization, the connections between each control area with the rest of system and effects of possible contracted scenarios are treated as a set of input disturbance signals. The idea of mixed H₂/H_∞ control technique is used for the training of the proposed controller. The motivation for using this control strategy for training the RBFNN based controller is to take large modeling uncertainties into account, cover physical constraints on control action and minimize the effects of area load disturbances. This newly developed design strategy combines the advantage of the neural networks and mixed H₂/H_∞ control techniques to provide robust performance and leads to a flexible controller with simple structure that is easy to implement. The effectiveness of the proposed method is demonstrated on a three-area restructured power system. The results of the proposed controllers are compared with the mixed H₂/H_∞ controllers for three scenarios of the possible contracts under large load demands and disturbances. The resulting controller is shown to minimize the effects of area load disturbances and maintain robust performance in the presence of plant parameter changes and system nonlinearities.     

电力系统暂态稳定性实时预测与控制

利用外部实时同步测量的思想和能量原理,提出一种电力系统暂态稳定性实时预测方法,并在此基础上提出了暂态稳定紧急控制的方法。该预测和控制方法简单．概念清楚,无需要预先知道系统的网络结构和参数,计算速度快,能准确反映系统的真实运行情况,适于在线应用。     

An online power system static security assessment module using multi-layer perceptron and radial basis function network

Efficient contingency screening and ranking method has gained importance in modern power systems for its secure operation. This paper proposes two artificial neural networks namely multi-layer feed forward neural network (MFNN) and radial basis function network (RBFN) to realize the online power system static security assessment (PSSSA) module. To assess the severity of the system, two indices have been used, namely active power performance index and voltage performance index, which are computed using Newton–Raphson load flow (NRLF) analysis for variable loading conditions under N − 1 line outage contingencies. The proposed MFNN and RBFN models based PSSSA module, are fed with power system operating states, load conditions and N − 1 line outage contingencies as input features to train the neural network models, to predict the performance indices for unseen network conditions and rank them in descending order based on performance indices for security assessment. The proposed approaches are tested on standard IEEE 30-bus test system, where the simulation results prove its performance and robustness for power system static security assessment. The comparison of severity obtained by the neural network models and the NRLF analysis in terms of time and accuracy, signifies that the proposed model is quick, accurate and robust for power system static security evaluation for unseen network conditions. Thus, the proposed PSSSA module implemented using MFNN and RBFN models are found to be feasible for online implementation.     

基于卷积神经网络的N-2线路开断潮流快速计算

交流潮流(AC)算法需迭代求解,难以满足实际电力系统在线安全校核的需求。文中基于卷积神经网络,提出一种电力系统线路开断潮流的快速计算方法。离线训练阶段,从线路开断前后工况与拓扑的变化中提取特征作为输入信号(原始特征图),经大量算例训练后,卷积神经网络构建了原始特征图与线路开断后潮流结果的非线性映射关系。在线应用时,直接生成原始特征图,并基于离线训练的卷积神经网络计算测试集的潮流结果。经4个IEEE典型系统的N-2潮流仿真验证,文中方法具有良好的泛化能力。相比传统交流算法,文中方法将速度提高了接近80倍;相比传统人工神经网络模型,文中方法将精度提高近了1个数量级。     

基于生成对抗网络的暂态稳定预防控制

针对电力系统暂态稳定预防控制在线计算的复杂性,提出一种基于生成对抗网络的暂态稳定预防控制方法。通过将暂态稳定预防控制建模为样本空间映射问题,该方法利用数据驱动方法训练生成模型,建立从暂态失稳运行空间到暂态稳定运行空间的映射。模型通过调整电网中发电机的有功出力,提高电网的暂态稳定裕度,使电网运行点满足暂态稳定校核的要求。与传统优化建模方法相比,所提方法通过神经网络的前馈推断求解控制策略,无需迭代求解,极大地提高了求解效率。基于新英格兰39节点系统的测试结果验证了所提方法的可行性和有效性。     

Modal based stability analysis of power systems using energy functions

In this paper a physically based decomposition technique is exploited to perform direct stability analysis of power systems using the energy function method. The slow-coherency analysis decomposes the system into r areas associated with (r- 1) slow modes of the linearized system. The centres of inertia (COIs) of these areas from the slow subsystem and the rest of the fast modes are associated with different areas. The system transient energy function E is decomposed into E_slow associated with the slow variables of the slow subsystem, E_fast associated with the fast variables of the r-areas and E_fast-slow which is a function of both fast and slow variables. Depending on the fault location and strength of connection between areas, the sum of E_slow, E_fast-slow and E_fast of the disturbed arease constitutes a good approximation to the system energy function. Using this partial E-function and the potential energy boundary surface (PEBS) method, both E_cr and t_cr are computed. Numerical results on a 10-machine 39-bus system are presented in support of the technique.     

Power system transient stability enhancement by STATCOM with nonlinear H∞ stabilizer

A robust control design is presented for a single machine infinite bus system (SMIB) with a STATCOM. The controller for the nonlinear system is designed using the recently developed nonlinear H_∞ theory. The approach combines state feedback exact linearization with linear H_∞ principle, which avoids the difficulty of solving Hamilton–Jacoby–Issacs inequality. Simulation results with a number of disturbances like torque pulses and three-phase faults on the generator show that the proposed robust controller can ensure transient stability of the power system over a wide range of operating points     

Performance comparison of neural networks for intelligent management of distributed generators in a distribution system

The Multilayer Perceptron (MLP) neural network has been proven to be a very successful type of neural network in many applications. The MLP activation function is one of the important elements to be considered in neural network training in which proper selection of the activation function will give a huge impact on the network performance. This paper presents a comparative study of the four most commonly used activation functions in the neural network which include the sigmoid, hyperbolic tangent and linear functions used in the MLP neural network and the Gaussian function used in the Radial Basis Function (RBF) network for managing active and reactive power of distributed generation (DG) units in distribution systems. Simulation results show that the sigmoid activation functions give better performance in predicting the optimal power reference of the DG units. However, the RBF neural network gives the fastest conversion time compared to the MLP neural network.     

自组织映射神经网络用于动态电压稳定分析的新方法

介绍了一种利用人工神经网络(ANN)进行动态电压稳定分析的新方法。这种多层自组织网络(SHNN)综合利用了自组织映射网络(文中使用Kohonen网络)和多层感知机网络（MLP）。Kohonen网络把输入样本按运行条件的相似性进行聚类,从而使MLP网络的性能得到提高。使用2个SHNN模型,一个用于判定电力系统是否处于动态稳定,另一个预测动态稳定系统的PQ节点的电压幅值。通过动态模拟得到训练样本。最后对WSCC 9节点系统和New England 39节点系统进行数字仿真,证明了该方法的有效性。