R.W. Dunn

A novel fault classification technique for double-circuit lines based on a combined unsupervised/supervised neural network

Summary form only as given. The work described in this paper addresses the problems encountered by conventional techniques in fault type classification in double-circuit transmission lines; these arise principally due to the mutual coupling between the two circuits under fault conditions, and this mutual coupling is highly variable in nature. It is shown that a neural network based on a combined unsupervised/supervised training methodology provides the ability to accurately classify the fault type by identifying different patterns of the associated voltages and currents. The technique is compared with that based solely on a supervised training algorithm (i.e., bad-propagation network classifier). It is then tested under differed fault types, location resistance and inception angle; different source capacities and load angles are also considered. All the test results show that the proposed fault classifier is very well suited for classifying fault types in double-circuit lines.

Design and implementation of an adapative single pole autoreclosure technique for transmission lines using artificial neural networks

Adaptive single pole autoreclosure (SPAR) offers many advantages over conventional techniques. In the case of transient faults, the secondary arc extinction time can be accurately determined and in the case of a permanent fault, breaker reclosure can be avoided. This paper describes, in some detail, the design and implementation of a SPAR technique using artificial neural networks (ANNs). The design described includes special methods for extracting features from post-circuit breaker opening fault data, which is a prerequisite for setting up training data sets. The technique is then implemented in hardware based on a high performance T800 transputer system and some results obtained from laboratory tests of this equipment are presented.

One-hour-ahead wind speed prediction using a Bayesian methodology

The contribution of wind power in market-driven power systems together with the uncertain nature of the wind resource have led to many research efforts on methodologies to predict future wind speed/power production. Applications such as the operational balancing market in the UK would benefit from accurate one-hour-ahead forecasts of the available power from all generators, wind being no exception. This paper focuses on one-hour-ahead wind speed prediction using a Bayesian approach to characterise the wind resource. To test the approach, two years of wind speed data from a weather station were modelled as an autoregressive process. In this paper, the methodology used is described together with the model employed and prediction results are presented and compared to the persistence method. The results obtained indicate that Bayesian inferencing can be a useful tool in wind speed/power prediction, particularly due to the flexibility inherent to the methodology

Spatially Correlated Wind Speed Modelling for Generation Adequacy Studies in the UK

Generation adequacy assessment is an essential tool in the analysis of the security and planning aspects of power systems. This, associated to the growth in wind power plants worldwide, has driven the need to better characterise the resource to support the integration of this technology into the power system. This paper introduces a time-series-based model of the wind speed for adequacy studies using the GB system (UK). The modelling methodology is described and simulation results presented using the Loss of Load Probability (LOLP) as the adequacy index shown against differing plant margin backgrounds.

Stability constrained optimal power flow for the balancing market using genetic algorithms

Angle stability (both transient and oscillatory) is an important constraint in power system operation. The work presented in this paper describes a genetic algorithm (GA) based approach for solving the problem of angle stability constrained optimal power flow. The control parameter modeled in the chromosome of the GA is generation power of the units. The application presented here is the UK balancing market using balancing mechanism units (BMUs). The BMUs put into the GA list depend on either their bid/offer price or their impact of generation change on system stability. Sensitivity factors, obtained by doing perturbations, are used to represent a BMUs impact on system stability. A novel mapping method is employed to maintain power balance. Stability constraints are dealt with as penalty cost, and their contribution to the fitness of the objective function is evaluated independently, so that the search for the optimal solution concentrates on feasible solutions. Tests on a reduced UK system show that the proposed GA is able to cope with the highly nonlinear optimization problem. Numerical simulation results of the test system are presented.

Modelling of operator heuristics in dispatch for security enhancement

This paper addresses the problem of dispatching active and reactive power on a large interconnected power system to maintain security. It models operator heuristics in the decision making process by means of fuzzy sets. Sensitivity analysis is then used to derive the necessary movements in control settings within a practical expert system that seeks a low cost, low number-of-controllers solution. Results are presented for a 20 machine, 100 bus reduced model of the UK national grid. Comparisons are shown with a linear programming approach and a conventional production rule based expert system implementation. The flexibility of the new approach is demonstrated whereby it can be used to model any number of operational criteria.

Bayesian Inferencing for Wind Resource Characterisation

The growing role of wind power in power systems has motivated R&D on methodologies to characterise the wind resource at sites for which no wind speed data is available. Applications such as feasibility assessment of prospective installations and system integration analysis of future scenarios, amongst others, can greatly benefit from such methodologies. This paper focuses on the inference of wind speeds for such potential sites using a Bayesian approach to characterise the spatial distribution of the resource. To test the approach, one year of wind speed data from four weather stations was modelled and used to derive inferences for a fifth site. The methodology used is described together with the model employed and simulation results are presented and compared to the data available for the fifth site. The results obtained indicate that Bayesian inference can be a useful tool in spatial characterisation of wind

Controlled island ng scheme for power systems

System islanding is often considered as the final stage of power system defense plans. The goal is to preserve stable areas of the faulted power systems. The islanding scheme plays an important role in the power system restoration phase as it can make the power system restoration less complex and reduce the overall restoration time. The basis for islanding is not standard but rather depends upon the nature of the utility. Even though the formation of islands is dominated by geographical proximity of the synchronous generators to maintain generation-load balance, there are some factors which can assist in designing a better islanding scheme. These factors are the type and location of the fault and the dynamic performance of every island on the system against the fault. This paper presents an optimization technique to obtain the optimal formation of islands taking into consideration the geographical distribution of the synchronous generators and the dynamic performance of every island in the system against the extreme and credible faults that lead to full system breakdown. In order to show the validity of this algorithm, the Algorithm is applied to IEEE 118 Bus System and a comparison between the proposed islanding scheme and an islanding scheme based on the geographical distribution of the synchronous generators is presented. The results presented in this paper show that taking into the account the type and location of the extreme and credible faults helps to preserve more stable area than that of the traditional islanding scheme.

A neural network based protection technique for combined 275 kV/400 kV double circuit transmission lines

Abstract The work described in this paper addresses the problems encountered by conventional distance relays when protecting double-circuit transmission lines comprising different voltage levels. The problems arise principally as a result of the mutual coupling between the two circuits under different fault conditions; this mutual coupling is highly nonlinear in nature. An adaptive protection scheme is proposed for such lines based on employing a neural network (NN). A NN has the ability to classify the nonlinear relationship between measured signals and the faulted zone by identifying different patterns of the associated voltages and currents. One of the key points of this paper is data preprocessing of the measured signals essentially to extract the most significant features from the signals. The adaptive protection scheme is tested under a specific fault type, but varying fault location, fault resistance, fault inception angle and different source capacities. All the test results clearly show that the proposed adaptive protection technique is well suited for double circuits with different voltage levels.

Backbone-network reconfiguration for power system restoration using genetic algorithm and expert system

During the last few years many blackouts have been experienced throughout the world. It seems that modern power systems are more exposed to major blackouts. This raises the necessity of having an obvious restoration plan to rebuild the power system as soon as possible. This problem is characterized by a large solution space which can be constrained with expert knowledge. This paper describes a new power system restoration algorithm jointly using Genetic Algorithms (GA) and Expert systems (ES). GAs are used to obtain optimized Skeleton Networks for power systems, while ES acts as an effective system operator to constrain the solution space for the GA. Also ES allows the GA to be more informed about the overall power system physical performance. This includes, for example, Frequency response to sudden load pick up, Reactive power balance, load-generation balance, Stability limits, high and low voltage levels limits, MW and MVAR reserve requirement and line transfer capability, etc. In order to show the advantages of combining the GA and ES to this problem, this paper presents a comparative result between the hybrid algorithm and pure ES method. The case study presented in this study is 39 IEEE bus systems. The results presented in this paper show that the application of ES can be significantly enhanced by the stated combination.