17 May 2019 in IEEE Transactions on Smart Grid
This paper proposes a hybrid fault location method for smart distribution systems by using available multi-source data, including data from protective devices, fault indicators, smart meters, voltage and current meters, and computer simulations. A fault diagnosis model based on Fuzzy Petri net (FPN) technique is developed that employs discrete evidences such as status of protective devices and fault indicators (FIs) to estimate the faulted section. To eliminate the possible false fault section estimations and pinpoint the actual location of the fault a mis-match index is defined. It quantifies the similarity between recorded measurements such as voltage and current signals during faults and the corresponding values obtained by performing short circuit analysis. The fault scenario with minimum mis-match degree is determined as the actual location of the fault. The proposed method can precisely determine the fault location in distribution systems with many laterals and sub-laterals with heterogenous line parameters in the presence of distributed generators (DGs). The accuracy and effectiveness of the proposed method are validated by applying the proposed method to the simulated IEEE 33 node feeder distribution system.
6 June 2017 in IEEE Transactions on Smart Grid
In smart grids energy storage units are installed at different locations of the network. This paper proposes a method based on model predictive control (MPC) for enhancing the transient stability of power systems by controlling the charging and discharging state of superconducting magnetic energy storage (SMES) systems installed throughout power systems. To enhance the performance of the controller in presence of exogenous disturbances and modelling uncertainty a controller based on tube-based MPC is developed. The proposed method determines output active and reactive powers of SMES units at each time instance to improve the selected transient stability index defined based on the distance of the rotor angles of generators and the center of inertia of the power system. The simulation results demonstrate that the proposed method is able to enhance the transient stability of the system effectively.
6 December 2018 in IEEE Transactions on Industrial Informatics
This paper proposes a two-stage fault location identification method for multi-area power systems. It utilizes available heterogeneous types of data consisting of status data (i.e. discrete data) and analog data (i.e. continues data). A distributed casual model-based diagnosis method using P-invariant Petr nets is proposed that utilizes status data to find faulted sections. Each subsystem/area of multi-area power systems has its own diagnostic model and determines the diagnosis solution based on its local net, the local manifestations, and limited information exchange with the neighboring subsystems/areas. Once a short list of possible fault sections is determined, the second stage further improves the estimated fault location using analog data. The actual fault location is estimated by comparing the measured analog data with their associated calculated values in computer programs using short circuit analysis algorithms. Simulation results demonstrate that the proposed distributed fault location method can diagnose faults in multi-area power systems accurately.
7 June 2018 in IEEE Transactions on Energy Conversion
Conventional loss of excitation (LOE) detection methods for synchronous generators may misoperate in the case of power swings. In this paper, a novel method based on measured slip frequency at the terminal of synchronous generators is proposed to differentiate power swing conditions from LOE condition. To enhance the performance of the proposed method for the noisy measured signals, a modified Prony method based on combination of generalized total least-squares (GTLS) method and conventional Prony method is implemented to estimate the slip frequency and an energy sorting technique is applied to distinguish the dominant frequency component of the measured signals. The performance of the proposed method is evaluated under various operating conditions of a test system that is simulated in a co-simulation platform by linking CAPE and PSS/E software. Case study results illustrate that the proposed method can accurately differentiate the power swing condition from LOE condition.
1 October 2020 in IEEE Systems Journal
In this article, a decentralized controller for improving the transient stability of multimachine power systems is proposed. A coordinated control of excitation and steam turbine of synchronous generators is developed based on model predictive control (MPC) strategy that explicitly takes into account physical constraints of the actuators. A tube-based MPC is developed to ensure the robustness and convergence of the control signal in the presence of exogenous perturbations and modeling uncertainties. To attenuate the noise level in measurement signals that are used in the controller, an improved tracking differentiator is utilized. The proposed controller is implemented locally at individual synchronous generator stations which makes it suitable for practical applications. The performance of the proposed control scheme is studied using 39-bus New England test system. The simulation results show that the proposed robust approach enhances the transient stability of the power systems.