J. Fitch

Wide-Area Monitoring, Protection, and Control of Future Electric Power Networks

Wide-area monitoring, protection, and control (WAMPAC) involves the use of system-wide information and the communication of selected local information to a remote location to counteract the propagation of large disturbances. Synchronized measurement technology (SMT) is an important element and enabler of WAMPAC. It is expected that WAMPAC systems will in the future reduce the number of catastrophic blackouts and generally improve the reliability and security of energy production, transmission, and distribution, particularly in power networks with a high level of operational uncertainties. In this paper, the technological and application issues are addressed. Several key monitoring, protection, and control applications are described and discussed. A strategy for developing a WAMPAC system in the United Kingdom is given as well.

Optimal coordinated voltage control for power system voltage stability

An optimal coordinated voltage controller (OCVC) is developed based on the spirit of model predictive control (MPC) method. The OCVC consists of three components, namely a predictor, a control candidate pool, and a selector. It has been used in secondary voltage control (SVC) to coordinate dissimilar control actions at different geographical locations in order to maintain desired voltage profiles in a global sense in emergencies. A single-stage Euler state predictor (SESP) is utilized, based on the system model, to predict voltage performance under selected control actions; the selection of the optimum control action from the pool is a complex optimization problem that is achieved by a pseudogradient evolutionary programming (PGEP) technique. Simulation results on a six-bus benchmark system and the New England 10-generator-39-bus system are given to show the potential of this method for online usage.

Agent-based substation automation

Agent technology is one of the most interesting developments in the field of distributed artificial intelligence. It has a wide range of applications, with information management, intelligent user interfaces, personal assistants, and Internet commerce among the most popular. This article defines a generic architecture that applies the multiagent systems methodology to the field of substation automation, describes the design of a system to be implemented based on this architecture, and proposes several possible applications. Compared with SCADA or client-server substation automation solutions, an agent-based system offers a number of advantages. Each function or task of the system, such as the management of a single IED, can be encapsulated within a separate agent, making the system highly modular. Agents are loosely coupled, typically communicating via messaging rather than by procedure calls (remote or local), and, using directory services, new functions can easily be added to an agent-based system by creating a new agent, which will then make its capabilities available to others. The inherently distributed power system architecture is suited ideally to a multiagent system, which provides greater autonomy to each of the constituent parts than a traditional system.

A multi-agent based intelligent monitoring system for power transformers in distributed substations

Power Transformers are distributed across a large number of substations, and many measures have been adopted for their onsite monitoring. Current approaches to transformer monitoring focus primarily on the sampling and display of vital parameters such as voltage and current, and that have been proved insufficient for transformer maintenance. This paper presents a new approach, based on a multi-agent system and an transformer equivalent heat circuit thermal model, which improves the quality and efficiency of distributed onsite transformer condition monitoring in power grids. The thermal model provides transformer temperature predictions and evaluation suggestions, while multi-agents playing the main roles of SCADA system maintenance, transformer autonomous diagnosis and information management.

Implementation of a power transformer temperature monitoring system

This paper describes the implementation of a power transformer temperature monitoring system that contains a recently developed equivalent heat circuit based thermal model. In this system, four fundamental components cooperate to perform onsite condition monitoring and intelligent analysis, including data acquisition, thermal model prediction and analysis, data storage, and HTTP services. The offline test shows that this thermal model based transformer temperature monitoring system provides useful information for transformer monitoring and maintenance.

Enhanced state estimation with real-time updated network parameters using SMT

This paper presents a brief explanation how Synchronized Measurement Technology (SMT) and WAMPAC architecture can support State Estimation and it discusses the existing correlation between each other. Additionally, it is presented a practical implementation of Phasor Measurement Units for updating network parameter data bases to improve the performance of state estimators. The methodology presented can be implemented only when full observability of the network parameters is available. The performance of the enhanced state estimator is assessed for different types of measurements and parameter errors. It is shown how the proposed solution delivers a more reliable estimation, especially when parameter errors are difficult to detect.

Condition Assessment of Power System Apparatuses Using Ontology Systems

Ontology based annotation provides clear and rigorous vocabularies for applications of knowledge management. It is explicitly defined and independent of the applications implementation. This paper proposes an ontology system for the condition assessment of power substation apparatus. The proposed methodology demonstrates the capability of annotating facts and their relationships in power system operation. The promising results obtained for power system apparatus condition assessment are achieved from laboratory experiments

A particle swarm optimiser with passive congregation approach to thermal modelling for power transformers

This paper employs an intelligent learning technique based on a particle swarm optimiser with passive congregation (PSOPC) algorithm to identify the thermal parameters of a simplified thermoelectric analogous thermal model (STEATM) for transformers, based upon only a few onsite measurements instead of experimental methods. The model outputs deliver good agreements with the onsite data based upon a single set of parameters obtained from the PSOPC learning with a fast convergence rate. The simulation results are compared with that obtained using an artificial neural network (ANN) approach.

An agent brokering-based scheme for anti-islanding protection of distributed generation

Anti-islanding protection of distributed generation (DG) is a significant technical barrier for the DG industry. This paper presents an innovative agent brokering-based anti-islanding scheme developed in response to the challenge. The scheme enables relay operational signals to be transferred to downstream DG units for coordinating their connection status using the multi-agent system (MAS) technology. A DG unit is considered as islanded from the upstream system if a trip signal is received at the DG unit. A number of software agents have been developed based on a generic structure for device control, data transfer and communication and the proposed scheme has been simulated using an established distribution system model integrated with four DG units. Two scenarios are designed to evaluate the performance of the scheme in handling multiple faults and change of system topology issues respectively. The simulation results are promising and the flexibility and scalability of the scheme are discussed.

An adaptive distance relaying algorithm with a morphological fault detector embedded

This paper presents an adaptive distance relaying algorithm (ADRA) for transmission line protection. In ADRA, a fault detector designed based on mathematical morphology (MM) is used to determine the occurrence of a fault. The Euclidean norm of the detector output is then calculated for fault phase selection and fault type classification. With respect to a specific type of fault scenario, an instantaneous circuit model applicable to a transient fault process is constructed to determine the position of the fault. The distance between the fault position and the relay is calculated by a differential equation of the instantaneous circuit model which is resolved in a recursive manner within each sampling interval. Due to the feature of recursive calculation, the protection zone of the relay varies from a small length to large, which increases as an augment in the sample window length. ADRA is evaluated on a transmission model based on PSCAD/EMDTC, under a variety of different fault distances, fault types, fault resistances and loading angles, respectively. The simulation results show that in comparison with conventional DFT-based protection methods, by which the fault distance is calculated using phasor measurements of voltage and current signals in a fixed-length window, ADRA requires much fewer samples to achieve a same degree of the accuracy of fault distance calculation, which enables much faster tripping, and its protection zone can be extended as more samples are used.