The emergence of smart grids has posed great challenges to traditional power system control given the multitude of new risk factors. This paper proposes an online supplementary learning controller (OSLC) design method to compensate the traditional power system controllers for coping with the dynamic power grid. The proposed OSLC is a supplementary controller based on approximate dynamic programming, which works alongside an existing power system controller. By introducing an action-dependent cost function as the optimization objective, the proposed OSLC is a nonidentifier-based method to provide an online optimal control adaptively as measurement data become available. The online learning of the OSLC enjoys the policy-search efficiency during policy iteration and the data efficiency of the least squares method. For the proposed OSLC, the stability of the controlled system during learning, the monotonic nature of the performance measure of the iterative supplementary controller, and the convergence of the iterative supplementary controller are proved. Furthermore, the efficacy of the proposed OSLC is demonstrated in a challenging power system frequency control problem in the presence of high penetration of wind generation.

This paper proposes an adaptive critic design (ACD) based control method of a generator in a single machine infinite bus (SMIB) power system, by using a biologically-inspired artificial neural network (BIANN). A heuristic dynamic programming (HDP) based optimal controller applying three BIANNs is designed for the control of the turbo generator in the SMIB system. By developing a numerical derivative calculation algorithm, the difficulties in back-propagating errors through a BIANN are overcome. The HDP-BIANN controller for the turbo generator in the SMIB system is trained and run online for different operating conditions. The results show that the HDP-BIANN controller improves the system damping as well as dynamic transient stability more effectively than the conventional controllers (CONVC) for not only small step changes, but also large disturbances such as a three phase short circuit. This is the first time that a BIANN is used for closed-loop optimal control of a power system.

This paper presents a generalized space vector modulation (SVM) method for any modular multilevel converter (MMC). The proposed SVM method produces the maximum level number (i.e., 2n+1, where n is the number of submodules in the upper or lower arm of each phase) of the output phase voltages and a higher equivalent switching frequency than other modulation methods, which consequently leads to reduced harmonics in the output voltages and currents. Compared with earlier modulation methods for MMCs, the proposed SVM method provides two more degrees of freedom, i.e., the redundant switching sequences and the adjustable duty cycles, thus offering significant flexibility for optimizing the circulating current suppression and capacitor voltage balancing. This SVM method is a useful tool for further studies of MMCs, as it can be conveniently extended for any control objectives. The demonstrated results validate the analysis.

Dynamic reactive power control of doubly fed induction generators (DFIGs) plays a crucially important role in maintaining transient stability of power systems with high penetration of DFIG based wind generation. Based on approximate dynamic programming (ADP), this paper proposes an optimal adaptive supplementary reactive power controller for DFIGs. By augmenting a corrective regulation signal to the reactive power command of rotor-side converter (RSC) of a DFIG, the supplementary controller is designed to reduce voltage sag at the point of common connection (PCC) during a fault, and to mitigate output active power oscillation of the wind farm after a fault. As a result, the transient stability of both DFIG and the power grid is enhanced. An action dependent cost function is introduced to provide real-time online ADP learning control. Furthermore, a policy iteration algorithm using high-efficiency least square method is employed to train the supplementary controller in an online model-free manner. By using such techniques, the supplementary reactive power controller is endowed with capability of online optimization and adaptation. Simulations carried out on a benchmark power system integrating a large DFIG wind farm show that the ADP based supplementary reactive power controller can significantly improve the transient system stability in changing operation conditions.

Non-intrusive load monitoring (NEM) is a load monitoring technique proposed to be used in today's residential energy auditor. It is expected to automatically provide the information of the type, energy consumption, and operation status of the electric loads without getting access to the loads. However, there still not exists any commercialized product so far, mainly because of the extraordinary large load sets comparing with the limited learning data. The fast emerging of new types of loads further aggravates the problem. This paper proposes an adaptive non-intrusive load identification model to address this problem. The proposed model is not dedicated to identify all the loads around the world, but it will grasp knowledge from samples that are not identified in the real application, and gradually form a new learning procedure so as to identify more and more new samples correctly. Random forest algorithm is introduced here to realize the objective and a case study is carried out to verify the effectiveness of the model.

This paper presents a methods of utilizing building-level demand response in microgrid frequency regulation. Based on an in-depth study of building loads, a bi-level framework of building demand response control is presented to determine the controllability and demand response participation timescale of different loads. Based on this framework, the internal control scheme of buildings is presented in analogy to negative generators. The contribution of building-level demand response to microgrid frequency regulation is discussed in details. The large power loss on transmission lines in microgrids is considered, and the heterogeneity among buildings in their contribution to system frequency deviation is discussed. A test case is studied to prove the efficacy of building-level demand response in microgrid frequency regulation, and the advantage of building-level demand response over single-load demand response.

In this paper, the authors propose a particle swarm optimization (PSO) for a discrete-time inverse optimal control scheme of a doubly fed induction generator (DFIG). For the inverse optimal scheme, a control Lyapunov function (CLF) is proposed to obtain an inverse optimal control law in order to achieve trajectory tracking. A posteriori, it is established that this control law minimizes a meaningful cost function. The CLFs depend on matrix selection in order to achieve the control objectives; this matrix is determined by two mechanisms: initially, fixed parameters are proposed for this matrix by a trial-and-error method and then by using the PSO algorithm. The inverse optimal control scheme is illustrated via simulations for the DFIG, including the comparison between both mechanisms.

The optimal shunt capacitor bank placement problem determines the optimal number, location, sizing, control type and setting of capacitor banks in order to maximize the benefits against the cost, energy loss, and voltage distortion. Many factors need to be taken into consideration, which makes the problem complex. This paper reviews and compares three major category of existing methods: analytical methods, optimization methods, and computational intelligent methods. Based on this survey, some recommended future work are proposed.

Miscellaneous electric loads (MELs) currently consume more electricity than any other single major category of electric appliances. MELs provide valuable energy consumption and performance information which can be utilized to meet the raising needs and opportunities of energy saving, demand response, peak shaving, and building management. A reliable intelligent method to identify different MELs is a prerequisite of all purposes. A support-vector-machine (SVM) based hybrid identification method of MELs is proposed in this paper. Studies on applying only SVM as well as a combination of SVM and supervised Self-Organizing Map (SSOM) are presented. SSOM first cluster a large number of MELs into several classes. MELs with similar feature values fall into the same class. SVM is then utilized to identify similar MELs. The proposed method shows satisfactory accuracy in tests using real-world data.

Plug-in Hybrid Electric Vehicles (PHEVs) are becoming increasingly popular all over the world. Therefore, some associated public infrastructures will have to be established, such as smart parking zones, charging stations, and charging piles, etc. This paper only focuses on the smart charging station. A comprehensive analysis upon the impacts of electricity price on the return of investment (ROI) of the charging station with a coordinating charging is presented. First, a coordinated charging and discharging model is presented which considers the profits of all three partners, namely the utility, the charging station and the customers. Second, the bidirectional money flow model of the station with both utility and customer are presented as well. Third, the hierarchical optimization model is introduced to solve the Mixed Integer Linear Programming problems. Fourth, a simulation based on a real database of a typical city is run and the Bass' diffusion model is applied to forecast the amount of PHEVs arriving at the charging station per day. Lastly, it is proved that an optimal coordinating charging strategy will reduce the years of ROI significantly and a figure presenting the relationship between price and ROI years is given.

Inspired by living neuron networks (LNNs) in the brain, artificial neural networks (ANNs) have been broadly used in various applications as a computational intelligence tool. However, due to many fundamental differences between ANNs and LNNs, despite the mature training mechanisms for ANNs, it is often challenging to use LNNs as a computational intelligence tool. To bridge the gap between ANNs and LNNs, a novel type of artificial neural network, i.e. biologically-inspired artificial neural network (BIANN) is proposed in this paper. The BIANN, which is based on spiking neuron models of LNNs, processes information in a more “brain-like” fashion than conventional ANNs. A reservoir-computing-based training approach is also proposed for BIANNs to serve as a novel modeling and control tool for practical applications. The feasibility of the proposed BIANN is illustrated for the prediction of a synchronous generator's speed and terminal voltage signals in a single machine infinite bus electric power system setup. The proposed BIANN model is able to provide an accurate prediction for online monitoring of a generator.

A large nonlinear dynamic system usually has complex dynamic modes corresponding to the system's eigenvalues. These eigenvalues govern the system's local behavior and thus are critical information for designing system operation and control strategies. Without the availability of the system's analytical model, which is often the case for large nonlinear systems, the system's eigenvalues need to be estimated. A linear echo state network (ESN) based method for extracting observable eigenvalues of a dynamic system together with the participation factors of these eigenvalues in the accessible system states is presented in this paper. A linear ESN is first trained to track the dynamic system's local responses under injected small perturbation signals. The dynamic system's eigenvalues are then extracted from the ESN's weight matrices. Given the merit of fast training of ESNs, the ESN can be quickly retrained once the system operating point changes, and the system eigenvalues can be reestimated. Application of the proposed eigenvalue extraction method in the power system small-signal analysis is presented to demonstrate the effectiveness of the proposed method.

A shunt active filter is a power electronic device used in a power system to decrease ldquoharmonic current pollutionrdquo caused by nonlinear loads. The Echo State Network (ESN) has been widely used as an effective system identifier with much faster training speed than the other Recurrent Neural Networks (RNNs). However, only a few attempts have been made to use an ESN as a system controller. As the first attempt to use an ESN in indirect neurocontrol, this paper proposes an indirect adaptive neurocontrol scheme using two ESNs to control a shunt active filter in a multiple-reference frame. As the first step in the proposed neurocontrol scheme, an online system identifier using an ESN is implemented in the Innovative Integration M67 card consisting of the TMS320C6701 processor to identify the load harmonics in a typical electric ship power system. The shunt active filter and the ship power system are simulated using a Real-Time Digital Simulator (RTDS) system. The required computational effort and the system identification accuracy of an ESN with different dynamic reservoir size are discussed, which can provide useful information for similar applications in the future. The testing results in the real-time implementation show that the ESN is capable of providing fast and accurate system identification for the indirect neurocontrol of a shunt active filter.

Dramatic reductions in sensor, computing and communications costs, coupled with significant performance enhancements has increased the possibility of realizing widely and massively distributed power line sensor networks (PLSNs) to monitor utility asset status for enhancing line reliability and utilization. One of the important applications of such a PLSN is to evaluate the overhead power line dynamic current capacity down to dasiaper spanpsila level of granularity. Due to the inherent non-linearity of overhead conductor thermal behavior, it is usually quite complex to directly calculate the conductor temperature. Therefore the prediction for the conductor dynamic thermal behavior becomes difficult. In this work, an echo state network (ESN) is proposed to identify the overhead conductor thermal dynamics in real-time. The well trained ESN model is used to predict the dynamic thermal behavior, and thus to evaluate the dynamic current capacity of the line under current ambient weather conditions. This paper addresses the design and implementation issues for such an ESN for this specific application. Simulation results reveal that the ESN model is very effective to predict the conductor temperature and to identify the conductor thermal dynamics subject to wide variations in line current and ambient weather conditions.

An efficient multi-objective design method with Particle Swarm Optimization (PSO) is developed for Surface Mount Permanent Magnet machines to reduce the complexity in the PMSM machine design. First an analytical model of the PMSM machine’s geometry is developed and results are verified by finite element analysis. With proper design specification and assumption, the design input variables in this model can be reduced to as low as two, which significantly simplifies the optimization process. PSO is then applied to this analytical model. Compared to the traditional machine design methods, this proposed algorithm finds the optimized solution with fast computation and high convergence.

Power system networks are complex systems that are highly nonlinear and non-stationary, and therefore, their performance is difficult to optimize using traditional optimization techniques. This paper presents an enhanced particle swarm optimizer for solving constrained optimization problems for power system applications, in particular, the optimal allocation of multiple STATCOM units. The study focuses on the capability of the algorithm to find feasible solutions in a highly restricted hyperspace. The performance of the enhanced particle swarm optimizer is compared with the classical particle swarm optimization (PSO) algorithm, genetic algorithm (GA) and bacterial foraging algorithm (BFA). Results show that the enhanced PSO is able to find feasible solutions faster and converge to feasible regions more often as compared with other algorithms. Additionally, the enhanced PSO is capable of finding the global optimum without getting trapped in local minima.

Many areas in power systems require solving one or more nonlinear optimization problems. While analytical methods might suffer from slow convergence and the curse of dimensionality, heuristics-based swarm intelligence can be an efficient alternative. Particle swarm optimization (PSO), part of the swarm intelligence family, is known to effectively solve large-scale nonlinear optimization problems. This paper presents a detailed overview of the basic concepts of PSO and its variants. Also, it provides a comprehensive survey on the power system applications that have benefited from the powerful nature of PSO as an optimization technique. For each application, technical details that are required for applying PSO, such as its type, particle formulation (solution representation), and the most efficient fitness functions are also discussed.

A near-optimal neurofuzzy external controller is designed in this paper for a static compensator (STATCOM) in a multimachine power system. The controller provides an auxiliary reference signal for the STATCOM in such a way that it improves the damping of the rotor speed deviations of its neighboring generators. A zero-order Takagi-Sugeno fuzzy rule base constitutes the core of the controller. A heuristic dynamic programming (HDP) based approach is used to further train the controller and enable it to provide nonlinear near-optimal control at different operating conditions of the power system. Based on the connectionist systems theory, the parameters of the neurofuzzy controller, including the membership functions, undergo training. Simulation results are provided that compare the performance of the neurofuzzy controller with and without updating the fuzzy set parameters. Simulation results indicate that updating the membership functions can noticeably improve the performance of the controller and reduce the size of the STATCOM, which leads to lower capital investment.

Identification and control of nonlinear systems depend on the availability and quality of sensor measurements. Measurements can be corrupted or interrupted due to sensor failure, broken or bad connections, bad communication, or malfunction of some hardware or software (referred to as missing sensor measurements in this paper). This paper proposes a novel fault-tolerant indirect adaptive neurocontroller (FTIANC) for controlling a static synchronous series compensator (SSSC), which is connected to a power network. The FTIANC consists of a sensor evaluation and (missing sensor) restoration scheme (SERS), a radial basis function neuroidentifier (RBFNI), and a radial basis function neurocontroller (RBFNC). The SERS provides a set of fault-tolerant measurements to the RBFNI and RBFNC. The resulting FTIANC is able to provide fault-tolerant effective control to the SSSC when some crucial time-varying sensor measurements are not available. Simulation studies are carried out on a single machine infinite bus (SMIB) as well as on the IEEE 10-machine 39-bus power system, for the SSSC equipped with conventional PI controllers (CONVC) and the FTIANC without any missing sensors, as well as for the FTIANC with multiple missing sensors. Results show that the transient performances of the proposed FTIANC with and without missing sensors are both superior to the CONVC used by the SSSC (without any missing sensors) over a wide range of system operating conditions. The proposed fault-tolerant control is readily applicable to other plant models in power systems.

Nowadays power distribution systems typically operate with nonsinusoidal voltages and currents. Harmonic currents from nonlinear loads propagate through the system and cause harmonic pollution. The premise of IEEE 519 is that there exists a shared responsibility between utilities and customers regarding harmonic control. Maintaining reasonable levels of harmonic voltage distortion depends upon customers limiting their harmonic current injections and utilities controlling the system impedance characteristics. Measurements of current taken at the point of common coupling (PCC) to a customer are expected to determine whether the customer is in compliance with IEEE 519. These measurements yield the combination of nonlinear load harmonics and nonlinear current due to supply voltage harmonics and typically the customer is required to take corrective actions to compensate the harmonics. This paper presents a neural network scheme whereby, it is possible to do data modeling of the customer's impedance and predict the resulting voltage distortion at the PCC if the customer were to take corrective actions. Experimental results from field measurements are provided. The proposed scheme is applicable to single as well as three phase systems

Wide-area coordinating control is becoming an important issue and a challenging problem in the power industry. This paper proposes a novel optimal wide-area monitor and wide-area coordinating neurocontroller (WACNC), based on wide-area measurements, for a power system with power system stabilizers, a large wind farm, and multiple flexible ac transmission system (FACTS) devices. The wide-area monitor is a radial basis function neural network (RBFNN) that identifies the input-output dynamics of the nonlinear power system. Its parameters are optimized through a particle swarm optimization (PSO) based method. The WACNC is designed by using the dual heuristic programming (DHP) method and RBFNNs. It operates at a global level to coordinate the actions of local power system controllers. Each local controller communicates with the WACNC, receives remote control signals from the WACNC to enhance its dynamic performance, and therefore helps improve system-wide dynamic and transient performance.

This paper describes a single phase switched rectifier for current control using model reference adaptive control (MRAC) with a continually on-line trained artificial neural network (COT-ANN). The results obtained with the proposed scheme are similar to the ones obtained in a previous work but using a simpler control structure. Simulations are used to test the validity of the proposed algorithm and the results are later verified by a practical implementation of this system.

This paper investigates the application of a new kind of recurrent neural network called echo state networks (ESNs) for the problem of measuring the actual amount of harmonic current injected into a power network by a nonlinear load. The interaction between loads connected to a point of common coupling (PCC) is a highly dynamic process. The determination of true harmonic current injection by individual loads is further complicated by the fact that the supply voltage waveform at the PCC is distorted by other loads at the PCC or further upstream and is therefore rarely a pure sinusoid. Harmonics in a power system are classified as either load harmonics or as supply harmonics. The principles of ESN are based on the use of a recurrent neural network (RNN) as a dynamic reservoir. In order to compute the desired output dynamics, only the weights of connections from the reservoir to the output units are calculated. This is simply a linear regression problem. Experimental results presented in this paper confirm that attempting to predict the total harmonic distortion (THD) of a load by simply measuring the load's current may not be accurate. The main advantage of this new method is that only waveforms of voltages and currents at the PCC have to be measured. This method is applicable for both single and three phase loads.

This paper presents a novel fuzzy logic based controller for a Static Compensator (STATCOM) connected to a power system. Type-II fuzzy systems are selected that enable the controller to deal with design uncertainties and the noise associated with the measurements in the power system. Interval type-II fuzzy is computationally more effective than the ordinary type-II fuzzy systems and is more suitable for the power network with fast changing dynamics. Using a proportional-integrator approach the proposed controller is capable of dealing with actual rather than deviation signals. The STATCOM is connected to a multimachine power system in order to provide extra voltage support and improve the system dynamic performance. Simulation results are provided to show that the proposed controller outperforms a conventional PI controller during large scale faults as well as small disturbances. The type-II fuzzy membership functions provide a robust performance for the controller and eliminate the need for a model based adaptive control scheme.

When subjected to transient disturbances in the power grid, the variable frequency converter (VFC) is the most sensitive part in the variable-speed wind turbine generator system (WTGS) equipped with a doubly fed induction generator (DFIG). The VFC is normally controlled by a set of PI controllers. Tuning these PI controllers is a tedious work and it is difficult to tune the PI gains optimally due to the nonlinearity and the high complexity of the system. This paper presents an approach to use the particle swarm optimization algorithm to design the optimal PI controllers for the rotor-side converter of the DFIG. A new time-domain fitness function is defined to measure the performance of the controllers. Simulation results show that the proposed design approach is efficient to find the optimal parameters of the PI controllers and therefore improves the transient performance of the WTGS over a wide range of operating conditions.

This paper proposes a neural network solution methodology for the problem of measuring the actual amount of harmonic current injected into a power network by a non-linear load. The determination of harmonic currents is complicated by the fact that the supply voltage waveform is distorted by other loads and is rarely a pure sinusoid. Harmonics may therefore be classified as contributions from the load on the one hand and contributions from the power system or supply harmonics on the other hand. A recurrent neural network architecture based method is used to find a way of distinguishing between the load contributed harmonics and supply harmonics, without disconnecting the load from the network. The main advantage of this method is that only waveforms of voltages and currents have to be measured. This method is applicable for both single and three phase loads. This could be fabricated into a commercial instrument that could be installed in substations of large customer loads, or used as a hand-held clip on instrument.

A novel robust artificial neural network identifier (RANNI) model is proposed in this paper. This RANNI can continuously track the dynamics of the plant model on-line when some sensor measurements are unavailable. A static synchronous series compensator (SSSC) connected to a small power system is used as a test system to examine the validity of the proposed model. In the simulation, one sensor is assumed to be missing; simulation results show that the proposed RANNI tracks the plant dynamics with good precision during the steady state, the small disturbance, and the transient state after a large disturbance. The proposed RANNI is readily applicable to other plant models in power systems.

Backpropagation algorithm is the most commonly used algorithm for training artificial neural networks. While being a straightforward procedure, it suffers from extensive computations, relatively slow convergence speed and possible divergence for certain conditions. The efficiency of this method as the training algorithm of a radial basis function neural network (RBFN) is compared with that of particle swarm optimization, for neural network based identification of a small power system with a static compensator. The comparison of the two methods is based on the convergence speed and robustness of each method.

A neural network based identifier is designed for effective control of a small power system. The power network in this work is considered from an external point of view, i.e., from a. supervisory level. Such a neuroidentifier can serve as a general model of such a plant, and then used for different neural network based control schemes.

This paper compares two indirect adaptive neurocontrollers, namely a multilayer perceptron neurocontroller (MLPNC) and a radial basis function neurocontroller (RBFNC) to control a synchronous generator. The different damping and transient performances of two neurocontrollers are compared with those of conventional linear controllers, and analyzed based on the Lyapunov direct method.

A novel method for nonlinear identification of a static compensator connected to a power system using continually online trained (COT) artificial neural networks (ANNs) is presented in this paper. The identifier is successfully trained online to track the dynamics of the power network without any need for offline data and can be used in designing an adaptive neurocontroller for a static compensator connected to such system.

In this paper, the dual heuristic programming (DHP) optimization algorithm is used for the design of a nonlinear optimal neurocontroller that replaces the proportional-integral (PI) based conventional linear controller (CONVC) in the internal control of a power electronic converter based series compensator in the electric power transmission system. The performance of the proposed DHP based neurocontroller is compared with that of the CONVC with respect to damping low frequency oscillations. Simulation results using the PSCAD/EMTDC software package are presented.

The design of nonlinear optimal neurocontrollers based on the Adaptive Critic Designs (ACDs) family of algorithms has recently attracted interest. This paper presents a summary of these algorithms, and compares their performance when implemented on two different types of artificial neural networks, namely the multilayer perceptron neural network (MLPNN) and the radial basis function neural network (RBFNN). As an example for the application of the ACDs, the control of synchronous generator on an electric power grid is considered and results are presented to compare the different ACD family members and their implementations on different neural network architectures.

This paper presents the design and practical hardware implementation of optimal neurocontrollers that replace the conventional automatic voltage regulator (AVR) and the turbine governor of turbogenerators on multimachine power systems. The neurocontroller design uses a powerful technique of the adaptive critic design (ACD) family called dual heuristic programming (DHP). The DHP neurocontrollers' training and testing are implemented on the Innovative Integration M67 card consisting of the TMS320C6701 processor. The measured results show that the DHP neurocontrollers are robust and their performance does not degrade unlike the conventional controllers even when a power system stabilizer (PSS) is included, for changes in system operating conditions and configurations. This paper also shows that it is possible to design and implement optimal neurocontrollers for multiple turbogenerators in real time, without having to do continually online training of the neural networks, thus avoiding risks of instability.

This paper presents the design of an optimal neurocontroller that replaces the conventional automatic voltage regulator (AVR) and the turbine governor for a turbogenerator connected to the power grid. The neurocontroller design uses a novel technique based on the adaptive critic designs (ACDs), specifically on heuristic dynamic programming (HDP) and dual heuristic programming (DHP). Results show that both neurocontrollers are robust, but that DHP outperforms HDP or conventional controllers, especially when the system conditions and configuration change. This paper also shows how to design optimal neurocontrollers for nonlinear systems, such as turbogenerators, without having to do continually online training of the neural networks, thus avoiding risks of instability.

This paper deals with modern control systems technology that is frequently applied to power conversion systems. The discussion goes far beyond the basic level of switch control in switching regulators. System-level control issues are important in expanding the market base of power electronics. Improvement in system performance involves not only the use of advanced control techniques, but also the integration of several converters or converter systems into a larger system. A fully controlled rectifier and an induction motor drive system are presented as typical applications to illustrate the use of many techniques for controlling power converters. The paper then points out how a wide variety of control techniques can be applied to enhance the controller performance and bandwidth in power electronics. These techniques include observers and adaptive control, nonlinear control, sliding and dead-beat control, and intelligent control. Finally, the paper points out how power electronic system performance can be enhanced through the use of condition monitoring and diagnostics.

This paper presents the design of a neurocontroller for a turbogenerator that augments/replaces the conventional automatic voltage regulator and the turbine governor. The neurocontroller uses a novel technique based on the adaptive critic designs with emphasis on heuristic dynamic programming (HDP) and dual heuristic programming (DHP). Results are presented to show that the DHP based neurocontroller is robust and performs better than the HDP based neurocontroller, as well as the conventional controller, especially when the system conditions and configuration changes.

This paper presents the design of a continually online trained (COT) artificial neural network (ANN) controller for a laboratory turbogenerator system connected to the infinite bus through a transmission line in real time. Two COT ANNs are used for the implementation: one ANN to identify the complex nonlinear dynamics of the power system, and the other ANN to control the turbogenerator. Practical results are presented to show that COT ANN controllers can control turbogenerators under steady state as well as transient conditions in the laboratory environment.