Cansin Yaman Evrenosoglu

Solar Power Prediction Using Interval Type-2 TSK Modeling

The random nature of solar energy resources is one of the obstacles to their large-scale proliferation in power systems. The most practical way to predict this renewable source of energy is to use meteorological data. However, such data are usually provided in a qualitative form that cannot be exploited using traditional quantitative methods but which can be modeled using fuzzy logic. This paper proposes type-1 and interval type-2 Takagi-Sugeno-Kang (TSK) fuzzy systems for the modeling and prediction of solar power plants. The paper considers TSK models with type-1 antecedents and crisp consequents, type-1 antecedents and consequents, and type-2 antecedents and crisp consequents. The design methodology for tuning the antecedents and consequents of each model is described. Finally, input-output data sets from a solar plant are used to obtain the three TSK models and their prediction results are compared to results from the literature. The results show that type-2 TSK models with type2 antecedents and crisp consequents provide the best performance based on the solar plant data.

A Fault Classification and Localization Method for Three-Terminal Circuits Using Machine Learning

This paper presents a traveling-wave-based method for fault classification and localization for three-terminal power transmission systems. In the proposed method, the discrete wavelet transform is utilized to extract transient information from the recorded voltages. Support-vector-machine classifiers are then used to classify the fault type and faulty line/half in the transmission networks. Bewley diagrams are observed for the traveling-wave patterns and the wavelet coefficients of the aerial mode voltage are used to locate the fault. Alternate Transients Program software is used for transients simulations. The performance of the method is tested for different fault inception angles, different fault resistances, nonlinear high impedance faults, and nontypical faults with satisfactory results.

Secure communications in the smart grid

This paper focuses on deployment of smart meters in the power distribution systems to enhance the operation infrastructure. An important challenge in establishing a communication paradigm between the utilities and the customers is that customers are susceptible to privacy concerns. In this paper, we present a model to ensure the privacy and integrity of communicating parties within the smart grid by using smart meters as a gateway between intra- and inter-network communications. In particular, we utilize the smart meter as a firewall to manage incoming and outgoing traffic and mediate household devices based on the instructions from the electric utility. Moreover, third parties are introduced in our model such as service providers so that they can monitor and manage the contracted customers by using the existing communication infrastructure.

Hour-ahead wind power prediction for power systems using Hidden Markov Models and Viterbi Algorithm

This paper presents a new stochastic method for very short-term (1 hour) wind prediction in electrical power systems. The method utilizes Hidden Markov Models (HMM) and the Viterbi Algorithm (VA). Past wind farm power production data are required to develop the HMM model. The accuracy of the predictions improves drastically if hourly weather forecast data are used as pseudo-measurements. Computer simulations using Northwestern weather recordings from the Bonneville Power Administration (BPA) website show good correlation between our predictions and the actual data.

A fault classification method in power systems using DWT and SVM classifier

This paper presents a method for fault classification in the power systems using a combination of support vector machine (SVM) classifier and Wavelet Transformation. Measurements from only one bus are utilized. Discrete Wavelet Transform (DWT) is used to extract the transient information of recorded voltages. The normalized wavelet energy of post-fault voltage and normalized energy of the post-fault currents are used as the input to the classifier. The classifier is trained with different fault scenarios in the power system. The transient voltages and phase currents for different types of faults and locations along the power system are obtained through ATP simulations. MATLAB is used to process the simulated transient voltages and apply the proposed method. The performance of the method is evaluated for two different networks; an overhead line combined with an underground cable and a 6-bus distribution network.

A traveling wave based single-ended fault location algorithm using DWT for overhead lines combined with underground cables

This paper presents a single-ended traveling wave based fault location method for power transmission systems where overhead lines are combined with underground cables. Bewley diagrams are used to determine the traveling wave patterns. The proposed method utilizes Discrete Wavelet Transform (DWT) to extract the transient information from the recorded voltage signals. The squares of the wavelet transform coefficients (WTC2) are calculated in order to determine the energy of the signal which is used to identify the faulted section (underground cable or overhead line) and subsequently to locate the fault. The transient voltages for different types of faults and locations along the overhead section as well as the underground cable section are obtained through EMTP simulations. MATLAB Wavelet Toolbox is used to process the simulated transients and apply the proposed method. The performance of the method is tested on various scenarios.

A regression analysis based state transition model for power system dynamic state estimation

In this paper, a new regression analysis based method is proposed to calculate the power system state transition matrix. This matrix is used to predict the system state which is subsequently corrected through extended Kalman filter in classical dynamic state estimation (DSE). State transition matrix is calculated by using regression analysis for a specified time interval and updated once new online measurement data are available. The preliminary tests on IEEE 14-bus system show improvement in the state forecasting accuracy when compared to existing state forecasting methods in dynamic state estimation.

Blind processing: Securing data against system administrators

Multi-owner systems such as power grid need information from all parties to operate efficiently. However, in general, information sharing is limited by market and other constraints. In addition, the emerging problem of demand side management in distribution systems as a part of “smarter grid” efforts, secure communication and execution between the utilities and the customers is required to ensure the privacy. In this paper, we propose blind processing, a novel communication and execution approach for entities that compete with each other but need to cooperate for the overall good of the system. Our goal is to allow information exchange between system components with protection mechanisms against everyone including system administrators. Shielding information will prevent gaining access to the sensitive data while providing a complete picture of the whole system in computations. Such a security mechanism can be provided by employing the functionality of Trusted Computing, a security technology that utilizes hardware and software modules to improve the trustworthiness of a system.

Power system state forecasting using regression analysis

This paper presents a block-diagonal state transition matrix based on regression analysis. The state transition matrix is used to forecast the system state, which is subsequently corrected through extended Kalman filter in classical dynamic state estimation (DSE). The transition matrix is updated when new online measurement data are available. The forecasting accuracy can be traded off according to the frequency of the updates. The tests on IEEE 14- and 30-bus system show improvement in the state forecasting accuracy when compared to the existing state forecasting methods in dynamic state estimation.

A Unified Approach for Power System Predictive Operations Using Viterbi Algorithm

A paradigm shift in the renewable energy proliferation in the U.S. necessitates a paradigm shift in power system operations to accommodate large-scale intermittent power while keeping the grid reliable and secure. Energy management systems (EMS) will benefit from an auxiliary function, which integrates the wind and load forecasting to state estimation and forecasting. This auxiliary function will create a predictive database for the power system states using the historical states as well as wind and load forecasts. The predictive database can be utilized to provide pseudo-measurements to a static state estimator in the case of loss of observability and bad data processing, or it can be used for short-term congestion and ramping predictions. This paper proposes an auxiliary tool for look-ahead power system state forecasting in electrical power systems with high intermittent renewable energy penetration. The method utilizes Markov models (MMs) and the Viterbi algorithm (VA) with a grid of feasible power system states obtained and updated by using the past states. The proposed algorithm is evaluated on the IEEE 14-bus and 118-bus systems using wind and load data available from the Bonneville Power Administration (BPA). The results show good correlation between the predictions and the actual power system states.