Time filter

Source Type

Xiang P.,Urban Power Supply Company | Xiang X.,Shanghai Electric Power Company
ICHVE 2016 - 2016 IEEE International Conference on High Voltage Engineering and Application | Year: 2016

According to the electro-acoustic pulse method for space charge measurement system of high voltage pulse power supply, the paper presents the pulse forming line and the oscillating circuit principle to realize the ultra narrow pulse high voltage. In this paper, based on the detailed discussion and analysis of the influence of the parameters of the pulse generating principle and the forming line, the ultra narrow high voltage pulse power supply is fabricated and tested. The results of the test can be seen that the high-voltage pulse generator based on the principle of pulse forming line can form a stable ultra narrow high voltage pulse (ns class). With the pulse width of the pulse forming line length increases, and it can be applied to space charge measurement system of pulse electro acoustic method. © 2016 IEEE.

Menhas M.I.,Shanghai University | Fei M.,Shanghai University | Wang L.,Shanghai University | Qian L.,Shanghai Electric Power Company
Energy Conversion and Management | Year: 2012

In this paper, multivariable PID controller design based on cooperative and coevolving multiple swarms is demonstrated. A simplified multi-variable MIMO process model of a ball mill pulverizing system with steady state decoupler is considered. In order to formulate computational models of cooperative and coevolving multiple swarms three different algorithms like real coded PSO, discrete binary PSO (DBPSO) and probability based discrete binary PSO (PBPSO) are employed. Simulations are carried out on three composite functions simultaneously considering multiple objectives. The cooperative and coevolving multiple swarms based results are compared with the results obtained through single swarm based methods like real coded particle swarm optimization (PSO), discrete binary PSO (DBPSO), and probability based discrete binary PSO (PBPSO) algorithms. The cooperative and coevolving swarms based techniques outperform the real coded PSO, PBPSO, and the standard discrete binary PSO (DBPSO) algorithm in escaping from local optima. Furthermore, statistical analysis of the simulation results is performed to calculate the comparative reliability of various techniques. All of the techniques employed are suitable for controller tuning, however, the multiple cooperative and coevolving swarms based results are considerably better in terms of mean fitness, variance of fitness, and success rate in finding a feasible solution in comparison to those obtained using single swarm based methods. © 2011 Elsevier Ltd. All rights reserved.

Cai D.,Shanghai Electric Power Company | Regulski P.,University of Manchester | Osborne M.,UK National Grid Corporation | Terzija V.,University of Manchester
IEEE Transactions on Smart Grid | Year: 2013

This paper presents the results of the development of Smart Grid transmission network applications in the Great Britain (GB) power system. A new Wide Area Monitoring System (WAMS) application for monitoring inter-area oscillations is developed. The core of this novel application is a fast nonlinear algorithm for the real-time estimation of the dominant inter-area oscillation mode, which processes GPS synchronized information obtained from Phasor Measurement Units (PMUs) installed in the power system. It is based on the Newton-Type Algorithm (NTA), an efficient parameter estimator. The paper focuses on the practical application of the new WAMS application: two data sets were tested, one based on computer simulations and the other based on real-life data records. The computer simulated oscillatory signals were obtained through dynamic simulations of the full GB power system model consisting of over 200 generators. The real-life data records used information collected by the FlexNet Wide Area Monitoring System (FlexNET-WAMS) installed in the GB network. Based on these data records, the features of inter-area oscillations in the GB network are drawn. © 2010-2012 IEEE.

Liu L.,Shanghai JiaoTong University | Wang H.,Shanghai Electric Power Company | Cheng H.,Shanghai JiaoTong University | Liu J.,State Power Economic Research Institute
Dianli Xitong Zidonghua/Automation of Electric Power Systems | Year: 2012

In order to overcome the neglect of medium-and long-term costs while underestimating short-term investment in current power systems in economic evaluation, a 3-dimensional life cycle cost (LCC) model for the power system as a whole is developed in the perspective of the component dimension, the cost dimension and the time dimension. The cost dimension is studied by structurally analyzing the device layer, the system layer and the cost of surroundings. A series of strategies for economic evaluation based on LCC is presented. The conversion methods are studied for devices with different life cycles. Multistage calculation is applied for devices with different runtimes. The traditional economic evaluation indicators are improved and the efficiency indicators based on LCC are proposed. Case studies are made on an actual 500 kV substation and 110 kV distribution network, respectively. It is shown that the economic evaluation results are much more accurate and valid, providing a reference for further deepening asset management of power systems.

Zhang W.-S.,Dalian University of Technology | Sun G.,Dalian University of Technology | Guo X.,Dalian University of Technology | Shan P.,Shanghai Electric Power Company
Gongcheng Lixue/Engineering Mechanics | Year: 2013

A novel numerical approach for the simultaneous optimization of structural topology and the layout of embedding structural components is proposed. The distinctive feature of this approach is that a level set function is employed to describe the arbitrary irregular shape of embedding structural components implicitly. With the use of this description, the non-overlap constraints for embedding structural components can be dealt with in a convenient way. Numerical examples demonstrated that the comparison with the existing methods in the literature, the simultaneous optimization of the structural topology and components layout can be achieved via the present approach with much less computational effort.

Yang Y.,Shanghai University of Electric Power | Wei G.,Shanghai University of Electric Power | Zhou B.,Shanghai University of Electric Power | Zhang X.,Shanghai Electric Power Company
Diangong Jishu Xuebao/Transactions of China Electrotechnical Society | Year: 2011

In traditional distribution network planning, uncertain planning parameters is difficult to deal with. To study further this problem, load forecast values is taken into account and the credibility theory and uncertain programming is introduced into it. Then a distribution network planning model is presented, which is based on fuzzy expected value model. The optimal solution is found using genetic algorithm. In this model, trapezoidal fuzzy variable is used to represent load predictive value. Moreover, the objective function and constraints of the model possess definite mathematical meaning, and the solution can be found by the strict mathematical approach. This model overcomes the defects of traditional distribution network planning model, such as, neglects the fuzziness of the load forecasting results or constructed fuzzy optimal planning model has no obvious mathematical meaning. An example illustrates that comparing to the traditional distribution network planning, the grid planning result obtained by this method is more adaptable to the uncertainty of the load in the future.

Lei Z.,Shanghai University of Electric Power | Wei G.,Shanghai University of Electric Power | Cai Y.,Shanghai University of Electric Power | Zhang X.,Shanghai Electric Power Company
Dianli Xitong Zidonghua/Automation of Electric Power Systems | Year: 2011

With a large scale of distributed generations (DGs) being installed and operated in distribution network, variety of power sources and complexity of their operation will greatly effects the reliability of distribution network. Different DGs'models and various operation situations have been studied, in which a network model called zone-nodes including DGs is introduced. The calculation model of reliability is build based on different failure types affected by zone-nodes, and the principle of island operations is definite. Then, on the above basis, a Monte Carlo simulation method is applied to solve this model. An example demonstrates the validity and practicability of the model and simulation method. The result of test example indicates that DGs can improve the reliability of distribution network, but also have some adverse impacts on it. The traditional reliability indices can not perfectly evaluate advantages or disadvantages of DGs'impact, it is recommended to standardize the evaluation system of the impacts of DGs on distributed system. © 2010 State Electric Power Research Institute Press.

Yin J.,North China Electrical Power University | Zhu Y.,North China Electrical Power University | Yu G.,Shanghai Electric Power Company | Shao Y.,Shanghai Electric Power Company | Guan H.,Shanghai Electric Power Company
Diangong Jishu Xuebao/Transactions of China Electrotechnical Society | Year: 2013

Gaussian process classifier based on Laplace approximation method (LGPC) is constructed. It can optimize the hyper parameters of the LGPC automatically, output classification results in probability, and be convenient to analyze problems' uncertainty. Therefore, LGPC can overcome the inherent limitations of SVM whose regularization factors and kernel function parameters are difficult to determine. In this paper, performance of LGPC is analyzed and validated by typical classification datasets, and transformer fault diagnosing method based on LGPC is presented and described in details. Experimental results show that the diagnosing correctness ratios are higher when mean function adopts a constant function, covariance function adapts a full square exponential function and likelihood function adopts an error function. Compared with methods based on SVM, the proposed method has higher classification accuracy, which proves it is effective.

Yin J.-L.,North China Electrical Power University | Zhu Y.-L.,North China Electrical Power University | Yu G.-Q.,Shanghai Electric Power Company
Dianli Xitong Baohu yu Kongzhi/Power System Protection and Control | Year: 2013

The transformer fault diagnosis is naturally a multi-classification problem with few sample data and a lot of uncertainties. Among the existing transformer fault diagnosis methods, a large number of sample data and amount of computation are needed for Bayesian Network (BN), and the adjustment of the coefficient is difficult for support vector machine (SVM). So a new method of transformer fault diagnosis based on multi-class relevance vector machine (M-RVM) is proposed. The method takes ratios of feature gases as inputs and Fast Type-II ML and expectation maximization (EM) are adopted. Diagnostic outputs are probability for each fault category and fault type with the highest probability is taken as diagnosis result. Experimental results show that the diagnosis speed is sufficient for project needs and M-RVM shows higher diagnosis accuracy compared with BN and SVM.

Yin J.,North China Electrical Power University | Zhu Y.,North China Electrical Power University | Yu G.,Shanghai Electric Power Company
Dianli Zidonghua Shebei/Electric Power Automation Equipment | Year: 2012

Analysis and typical data classification examples validate the classification performance of RVM (Relevance Vector Machine) is better than that of SVM (Support Vector Machine). A transformer fault diagnosis method based on RVM is put forward, which takes the normalized contents of transformer feature gases as inputs and adopts the binary tree classification means. Experimental results show that, compared with the fault diagnosis method based on SVM, it gets comparable or better diagnostic accuracy with less vector amount and faster diagnosis speed.

Loading Shanghai Electric Power Company collaborators
Loading Shanghai Electric Power Company collaborators