Victor Kurbatsky

Optimal Training of Artificial Neural Networks to Forecast Power System State Variables

The problem of forecasting state variables of electric power system is studied. The paper suggests data-driven adaptive approach based on hybrid-genetic algorithm which combines the advantages of genetic algorithm and simulated annealing algorithm. The proposed method has two stages. At the first stage the input signal is decomposed into orthogonal basis functions based on the Hilbert-Huang transform. The genetic algorithm and simulated annealing algorithm are applied to optimal training of the artificial neural network and support vector machine at the second stage. The results of applying the developed approach for the short-term forecasts of active power flows in the electric networks are presented. The best efficiency of proposed approach is demonstrated on real retrospective data of active power flow forecast using the hybrid-genetic support vector machine algorithm.

A random forest-based approach for voltage security monitoring in a power system

Voltage collapse is a critical problem that impacts power system operational security. Timely and accurate assessment of voltage security is necessary to detect alarm states in order to prevent a large-scale blackout. This paper presents an on-line voltage security assessment scheme using periodically updated random forest-based decision trees. We demonstrated the proposed method on the modified 53-bus IEEE power system. Results are presented and discussed.

Application of two stages adaptive neural network approach for short-term forecast of electric power systems

The paper presents the two-stapes adaptive approach for short-term forecast of parameters of expected operating conditions. The first stage involves decomposition of the time series into intrinsic modal functions and subsequent application of the Hilbert transform. During the second stage the computed modal functions and amplitudes are employed as input functions for artificial neural networks. Their optimal combinations is constructed using methods of simulated annealing and neural-genetic input selection approach. The efficiency of developed approach is demonstrated on real time the problem of forecasting power flow and voltage level.

Preventive and emergency control of intelligent power systems

The adaptive emergency control concept is based on realization of a tradeoff between preventive and emergency control by combining preventive and emergency actions. The paper proposes to complement the existing control strategies by predictive control strategy - adaptive emergency control, which implies identification of the emergency, until the emergency has occurred. In this paper presented the smart predictive voltage stability assessment (VSA) for the adaptive emergency control techniques on the basis of the artificial neural network methodology. The main idea of the smart predictive VSA here is a creation of a neural network model, based on a self-organized Kohonen map SOM that will be able to perform monitoring and prediction of emergency condition. Simulation results are obtained by the proposed scheme for different power system networks to assess the security level of the network.

Smart Technologies in Emergency Control of Russia's Unified Energy System

The objective trends in electric power systems (EPSs) call for prompter and more adequate response of control systems. New smart measurement, communication and control tools, information and computer technologies can be used to improve EPS controllability. The distinctive features of the Unified Energy System (UES) of Russia are discussed and the current emergency control system is presented in the paper. A modern approach to monitoring, forecasting and control is suggested. Some artificial intelligence applications for development of emergency control in the UES of Russia are presented.

Electricity prices neural networks forecast using the Hilbert-Huang transform

The problem of forecasting of electicity prices is addressed in terms of joint approach employing the general regression artificial neural network and empirical mode decomposition approaches (EMD) which is part of Hilbert-Huang transform. The application of developed approach to day-ahead hourly time series has demonstrated the whole accuracy increase as well as peaks prediction.

Intelligent monitoring in electric power system

The paper presents new approaches for effective organization of the system of IPS operating condition monitoring The presence of efficient system for wide-scale monitoring and forecasting of electric power system (EPS) is one of the key conditions for reliable work of systems intended for EPS operation and emergency control.

Development of software for modelling decentralized intelligent systems for security monitoring and control in power systems

With rapidly increasing complexity of power grids in Europe, North America and Asia, liberalization of electricity markets and increasing penetration of renewable energy, the risk of large-scale emergencies and blackouts increases. This paper proposes a novel approach for development of software for modelling of decentralized intelligent systems for security monitoring and control in power systems. The innovation here is to joint use the modern computing environments - MATLAB, R and Java Agent Development Framework platform. The proposed intelligent system was tested on the modified 53-bus IEEE power system.

Hybrid genetic algorithms for forecasting power systems state variables

A problem of forecasting state variables of electric power system is studied. The paper suggests data-driven adaptive approach based on hybrid-genetic algorithm which combines the advantages of genetic algorithm and simulated annealing algorithm. The input signal is decomposed into orthogonal basis functions using the Hilbert-Huang transform. The hybrid-genetic algorithm is applied to optimal training of the support vector machine and artificial neural network. The results of applying the developed approach for the short-term forecasts of active power flows in the electric networks are presented. The best efficiency of proposed approach is demonstrated on real retrospective data of active power flow forecast using the hybrid-genetic support vector machine algorithm empowered with the Hilbert-Huang transform.

The hybrid model based on Hilbert-Huang Transform and neural networks for forecasting of short-term operation conditions of power system

The paper addresses the conventional approaches to the short-term forecasting of nonstationary processes in complex power systems using the methodology of artificial neural networks (ANNs). In many practical cases the application of different ANNs can provide a satisfactory forecast. But data preprocessing and analysis can significantly improve the forecast. In this paper the Hilbert-Huang Transform (HHT) is used as one of the most promising tools in this area. Here we focus on HHT since this transform underlies the proposed two-stage intelligent approach to short-term forecasting of nonstationary processes.