Search

Energy management engineers are focusing their interest in tapping maximum profit for their system from substation automation (SSA)/distribution automation (DA). Volt/Var control through fixed/switched capacitors, transformer taps and voltage set points are at different levels of research and implementation. A neural network based solution for voltage-VAR control is proposed with the aim to reduce the real power loss flowing in a power system and subsequently improve the voltage profile. The module consists of two networks. The first network determines the control parameters i.e., generator voltage, transformer taps and shunt capacitance for minimal power loss when the loads at the load buses are specified as inputs. With the obtained parameters, a load flow program is run and power loss is noted and the system is checked for voltage violations. In case of voltage violations, the voltages are fed to the second network, which gives dQ at different buses for voltage violation minimization. These modules are successfully tested for different load patterns on a six-bus system.

Distribution system has been analyzed in this paper using object-oriented approach. The important contribution of this paper is the development of software objects for various distribution system components in such a way that they can be reused in most of the distribution system analysis programs. The design, proposed in this paper, is used for developing load flow analysis program. Object-oriented design replicates the physical system structure exactly in the software. The extensibility of the object-oriented design is exploited to extend the radial load flow analysis module for performing the load flow analysis of weakly meshed system by deriving few specialized objects from the fundamental objects used for radial load flow. Modeled objects are implemented in C++, an object-oriented programming language, and tested with various test systems. The results obtained for 69-bus radial system and 33-bus weakly meshed system are provided. This design is being extended for developing load flow analysis module for 3-phase unbalanced distribution system and distribution system with dispersed generations.

Power system computation software needs revision of its design due to modernization of existing network and automation of the power system operation. Object-oriented design has gained wide spread importance and acceptance in software development due to its advantages over other methodologies. Large-scale power system computation involves sparse matrices. External system equivalents are widely used to reduce computation complexity. This paper proposes an object-oriented design of sparse matrices and field proven Paul Dime's REI equivalent (radial equivalent independent of other nodes). 'Dynamic binding' and 'polymorphism' of object-oriented design have been exploited in this work to develop the software for REI equivalent. The object-oriented design is implemented using C/sup ++/ programming language and tested with 6-bus, IEEE 14-bus and IEEE 30-bus systems. Results for the IEEE 14 bus system in the form of bus admittance matrix using single load REI equivalent, single generator REI equivalent and single generator single load REI equivalent are illustrated.

The controlling functions in the SCADA/EMS system are increasing with the expansion of the power system network as the demand for power increasing. It is difficult to control the functions with the centralized system and rely on a single vendor. Current technologies employ distributed systems that are open (i.e. vendor independent solution) to control the SCADA/EMS functions. With open systems utilities, it tends to incrementally upgrade rather than completely replace the EMS systems. Various open distributed architectures have been proposed to shoulder the SCADA/EMS functions. This paper introduces the concept of open system and examines one of the distributed hardware configurations for open EMS. The proposed work modeled the above contingency analysis procedure for a case study of 6 bus, 11 line network, which was done by a dedicated application server on the pre-specified port.

A new formulation of the power system state estimation problem and a new solution technique is presented. The formulation allows for inequality constraints such as generator VAr limits and transformer tap ratio limits. The solution technique uses a logarithmic barrier function to treat inequality constraints. Non-quadratic criteria for power system state estimation have to facilitate bad data detection and identification. The weighted least absolute value criteria have the desired bad data rejection property. This paper shows that the logarithmic barrier function can accommodate inequality constraints, and then applying the Newton method to solve the KKT equation leads to a problem structure which is identical to that of the non-linear least square problem. The number of iterations needed for convergence does not increase linearly with problem size, unlike the LP algorithm.

This paper presents an efficient and reliable evolutionary based approach to solve the economic load dispatch (ELD) with lineflows and voltage constraints The proposed approach employ a particle swarm optimization (PSO) algorithm for ELD. The incorporation of a type 1 PSO, as a derivative-free optimization technique in solving ELD with voltages and lineflow constraints, significantly relieves the assumptions imposed on the optimized objective function. The proposed approach has been tested on three representative systems, i.e. IEEE 14 bus, IEEE 30 bus and IEEE 57 bus systems respectively. The feasibility of the proposed method is demonstrated and the results are compared with linear programming, quadratic programming and genetic algorithms respectively. The developed algorithms are computationally faster than the other methods.

This paper presents an efficient and reliable evolutionary based approach to solve the economic load dispatch (ELD) with security constraints. The proposed approach employ particle swarm optimization (PSO) algorithm for ELD. Incorporation of type 1 PSO as a derivative-free optimization technique in solving ELD with voltages and lineflow constraints significantly relieves the assumptions imposed on the optimized objective function. The proposed approach has been tested on three representative systems, i.e. IEEE 14 bus, IEEE 30 bus and IEEE 57 bus systems respectively. The feasibility of the proposed method is demonstrated and the results are compared with linear programming, quadratic programming and genetic algorithm respectively. The developed algorithms are computationally faster (no. of load flows) than the other methods.

A new method for the solution of the problems of unit commitment (UC) and economic dispatch (ED) using Hopfield neural network (HNN) is proposed in this paper. The difficulty in combining these problems is that while the first one requires a discrete neuron model, the latter requires a continuous neuron model. The combined solution of these problems using HNN requires the interconnection of discrete and continuous neural network models and the formulation of a unified energy function, which is quite complicated. The important contribution of this work is the proposal of a new architecture for the discrete HNN for UC and the output of the UC module is used as input to the continuous HNN for ED. The associated advantage of using HNN for the combined solution of UC and ED is the decoupling of their interdependency i.e:, both the UC and ED are iteratively solved using respective HNN for the particular period. The implementation of the proposed method causes a considerable reduction in the HNN size and hence complexity and computation requirements, compared to earlier attempts. The method was successfully tested for different cases (3,10,11 and 26 generator units), with varying load pattern of different durations (24 and 168 hours).

An integrated Artificial Neural Network (ANN) approach to Short-Term Load Forecasting (STLF) is proposed in this paper. Four modules consisting of the Basic ANN, Peak and Valley ANN, Averager and Forecaster and Adaptive Combiner form the integrated method for load forecasting. The Basic ANN uses the historical data of load and temperature to predict the next 24h load, while the Peak and Valley ANN uses the past peak and valley data of load and temperatures, respectively. The Averager captures the average variation of the load from the previous load behaviour, while the adaptive combiner uses the weighted combination of outputs from the Basic ANN and the Forecaster, to forecast the final load. The regression based and time series methods are conceptually incorporated into the ANN to obtain an integrated load forecasting approach. ? 2004 Elsevier B.V. All rights reserved.

This paper presents the genetic algorithm approach to adaptive optimal economic dispatch of electrical power systems. Genetic algorithms, also termed as the machine learning approach to artificial intelligence, are powerful stochastic optimization techniques with potential features of random search, hill climbing, statistical sampling and competition. Genetic algorithmic approach to power system optimization, as reported here for a case of economic power dispatch, consists essentially of minimizing the objective function while gradually satisfying the constraint relations. The unique problem solving strategy of the genetic algorithm and their suitability for power system optimization is described. The advantages of the genetic algorithmic approach in terms of problem reduction, flexibility and solution methodology are also discussed. The suitability of the proposed approach is described for the case of a 15 generator power system. >

Fault detection and diagnosis in real-time are areas of research interest in knowledge-based expert systems. Rule-based and model-based approaches have been successfully applied to some domains, but are too slow to be effectively applied in a real-time environment. This paper explores the suitability of using artificial neural networks for fault detection and diagnosis of power converter systems. The paper describes a neural network design and simulation environment for real-time fault diagnosis of thyristor converters used in HVDC power transmission system.

Solution methodology of unit commitment (UC) using genetic algorithms (GA) is presented. Problem formulation of the unit commitment takes into consideration the minimum up and down time constraints, start up cost and spinning reserve, which is defined as minimization of the total objective function while satisfying the associated constraints. Problem specific operators are proposed for the satisfaction of time dependent constraints. Problem formulation, representation and the simulation results for a 10 generator-scheduling problem are presented.

The demand for electricity is known to vary by the time of the day, week, month, temperature, and usage habits of the consumers. Though usage habit is not directly observable, it may be implied in the patterns of usage that have occurred in the past. A short-term load-forecasting (STLF) program that uses an integrated artificial neural network (ANN) approach is capable of predicting load for basic generation scheduling functions, assessing power system security, and providing timely dispatcher information. How well training data is chosen in an ANN is the defining factor in how well the network's output will match the event being modeled.

Large interconnected power systems with dispersed and geographically isolated generators and load constitute a majority of the power network. Present-day power systems are dynamic in nature, where the network topology frequently changes with load demand. With increase in load, the power system network is loaded to its limits, making it susceptible to collapse even under minor disturbances. In order to operate the power system economically, the current operating state of the system must be identified as either secure or insecure. An artificial neural network (ANN) aided method for security assessment is proposed and illustrated for a model six-bus power system. The work demonstrates the feasibility of classification of load patterns for power system static security assessment using a Kohonen self-organizing feature map. The most important aspect of this network is its generalization property. Using 15 different line-loading patterns for training, the network successfully classifies the unknown loading patterns. This powerful and versatile feature is especially useful for power system operation. Research is in progress to include contingency analysis in the security assessment program.

This paper presents a new multiagent-based differential evolution (MADE) approach to the reactive power dispatch problem. Reactive power dispatch problem can be formulated as a mixed-integer non-linear optimisation problem. The proposed method integrates multiagent systems (MASs) and differential evolution (DE) algorithm. An agent in MADE represents an individual to DE and a candidate solution to the optimisation problem. All agents live in a lattice like environment, with each agent fixed on a lattice point. In order to obtain optimal solution quickly, each agent competes and cooperates with its neighbours and it can also use knowledge. Making use of these agent-agent interaction and DE mechanism, MADE realises the purpose of minimising the value of objective function. In this study, the proposed MADE, DE and PSO are applied to minimise the transmission losses on standard IEEE 14, 30 and 57 bus test systems. Simulations results are compared with DE, PSO and existing methods reported in the literature. The results show that the proposed method converges to better solutions much faster than the earlier methods. Copyright � 2013 Inderscience Enterprises Ltd.

Solution methodology of unit commitment (UC) using genetic algorithms (GA) is presented. Problem formulation of the unit commitment takes into consideration the minimum up and down time constraints, start up cost and spinning reserve, which is defined as minimization of the total objective function while satisfying the associated constraints. Problem specific operators are proposed for the satisfaction of time dependent constraints. Problem formulation, representation and the simulation results for a 10 generator-scheduling problem are presented.

Pattern classification is a task common to the areas of artificial intelligence, pattern recognition, neural networks and machine learning. Advanced research in these areas and simillarities in problem solving has led to further investigation into the area of statistcal pattern recognition. Pattern classification approach to diagnosis is a fundamental area of research for problems involving real-time application. Most of the research in pattern classification has evolved from statistical pattern recognition ( bayesian classifiers, linear discriminants ) and neural networks ( back-propagation training algorithm ). An integration of these approaches, termed as “neural network pattern classification” is proposed for efficient classificatory problem solving. Comparitive performance of this approach to conventional classifiers is presented. The suitability of this classifier is tested for two case studies of power systems for discrete and contiouous data training sets.

Real time fault detection and diagnosis (FDD) is an important area of research interest in knowledge based expert systems. Neurocomputing is one of fastest growing areas of research in the fields of artificial intelligence and pattern recognition. The authors explore the suitability of pattern classification approach of neural networks for fault detection and diagnosis. The suitability of using neural networks as pattern classifiers for power system fault diagnosis is described in detail. A neural network design and simulation environment for real-time FDD is presented. An analysis of the learning, recall and generalization characteristic of the neural network diagnostic system is presented and discussed in detail. >

A differential evolution approach to solve optimal power flow problem with multiple and competing objectives is presented. Two sub-problems of optimal power flow namely active power dispatch and reactive power dispatch are considered. The problem is formulated as a nonlinear constrained true multi-objective optimisation problem with competing objectives. Constrain-domination approach have been used to handle inequality constraints, which eliminates the use of penalty factors. The performance of the proposed approach was tested on standard IEEE 30-bus system and is compared with a conventional method. The result demonstrates the capability of the proposed approach to generate diverse and well-distributed Pareto-optimal solutions.

A dynamic topology processor for a radial distribution system using object-oriented methodology is presented. The potential of object-oriented design patterns is also employed to obtain an extendible design. The decorator design pattern is used to separate the topology-based attributes from the attributes representing physical characteristics of the objects. Using decorator pattern, responsibilities can be dynamically attached to or removed from the objects. The state design pattern is used to model the switching states of the tie and sectionalising switches. The proposed topology-processor algorithm is a local updating technique, hence less time consuming. The proposed algorithm has been implemented in C++ and tested with a modified 69-bus system. It has been found that the average CPU time required to modify the links between the objects following a switching operation is only 0.4 ms.