Mevludin Glavic

Wide-Area Detection of Voltage Instability From Synchronized Phasor Measurements. Part I: Principle

This two-part paper deals with the early detection of an impending voltage instability from the system states provided by synchronized phasor measurements. Recognizing that voltage instability detection requires assessing a multidimensional system, the method fits a set of algebraic equations to the sampled states, and performs an efficient sensitivity computation in order to identify when a combination of load powers has passed through a maximum. The important effects of overexcitation limiters are accounted for. The approach does not require any load model. This first part of the paper is devoted to theoretical foundations of sensitivity calculation along the system trajectory, derivation of the algebraic model, and illustration on a simple five-bus system involving the long-term dynamics of a load tap changer and a field current limiter.

Power systems stability control: reinforcement learning framework

In this paper, we explore how a computational approach to learning from interactions, called reinforcement learning (RL), can be applied to control power systems. We describe some challenges in power system control and discuss how some of those challenges could be met by using these RL methods. The difficulties associated with their application to control power systems are described and discussed as well as strategies that can be adopted to overcome them. Two reinforcement learning modes are considered: the online mode in which the interaction occurs with the real power system and the offline mode in which the interaction occurs with a simulation model of the real power system. We present two case studies made on a four-machine power system model. The first one concerns the design by means of RL algorithms used in offline mode of a dynamic brake controller. The second concerns RL methods used in online mode when applied to control a thyristor controlled series capacitor (TCSC) aimed to damp power system oscillations.

Wide-Area Detection of Voltage Instability From Synchronized Phasor Measurements. Part II: Simulation Results

This two-part paper deals with the early detection of an impending voltage instability from the system states provided by synchronized phasor measurements. Recognizing that voltage instability detection requires assessing a multidimensional system, the method fits a set of algebraic equations to the sampled states, and performs an efficient sensitivity in order to identify when a combination of load powers has passed through a maximum. This second part of the paper presents simulation results obtained from detailed time-domain simulation of the Nordic32 test system, without and with measurement noise, respectively. Several practical improvements are described such as anticipation of overexcitation limiter activation, and use of a moving average filter. Robustness to load behavior, non-updated topology and unobservability is also shown. Finally a comparison with Thevenin impedance matching criterion is provided.

Contingency Filtering Techniques for Preventive Security-Constrained Optimal Power Flow

This paper focuses on contingency filtering to accelerate the iterative solution of preventive security-constrained optimal power flow (PSCOPF) problems. To this end, we propose two novel filtering techniques relying on the comparison at an intermediate PSCOPF solution of post-contingency constraint violations among postulated contingencies. We assess these techniques by comparing them with severity index-based filtering schemes, on a 60-and a 118-bus system. Our results show that the proposed contingency filtering techniques lead to faster solution of the PSCOPF, while being more robust and meaningful, than severity-index based ones.

Reinforcement Learning Versus Model Predictive Control: A Comparison on a Power System Problem

This paper compares reinforcement learning (RL) with model predictive control (MPC) in a unified framework and reports experimental results of their application to the synthesis of a controller for a nonlinear and deterministic electrical power oscillations damping problem. Both families of methods are based on the formulation of the control problem as a discrete-time optimal control problem. The considered MPC approach exploits an analytical model of the system dynamics and cost function and computes open-loop policies by applying an interior-point solver to a minimization problem in which the system dynamics are represented by equality constraints. The considered RL approach infers in a model-free way closed-loop policies from a set of system trajectories and instantaneous cost values by solving a sequence of batch-mode supervised learning problems. The results obtained provide insight into the pros and cons of the two approaches and show that RL may certainly be competitive with MPC even in contexts where a good deterministic system model is available.

A short survey of methods for voltage instability detection

This paper shortly surveys existing and proposed methods for voltage instability detection. The emphasis is on methods relying on real-time measurements as well as on long-term voltage instability. Methods are classified according to the required measurement configuration: local vs. wide-area, standard SCADA-type vs. synchronized phasor measurements, etc. In the various categories, some of the features are summarized, and what appears to the authors as advantages or limitations is shortly discussed. An important feature is the ability to anticipate instability. Some tracks for further work on the subject are also outlined.

Receding-Horizon Multi-Step Optimization to Correct Nonviable or Unstable Transmission Voltages

In this paper, a receding-horizon multi-step optimization is proposed to correct nonviable transmission voltages and prevent long-term voltage instability. The proposed control scheme is based on real-time control, inspired by model predictive control, and steady state power-flow-based equations. In order to anticipate load behavior and avoid using dynamic equations in the control scheme, explicit formulations are used to model evolution of load with time. The simulation results of the proposed technique are presented on the Nordic32 test system.

Estimation of rotor angles of synchronous machines using artificial neural networks and local PMU-based quantities

This paper investigates a possibility for estimating rotor angles in the time frame of transient (angle) stability of electric power systems, for use in real-time. The proposed dynamic state estimation technique is based on the use of voltage and current phasors obtained from a phasor measurement unit supposed to be installed on the extra-high voltage side of the substation of a power plant, together with a multilayer perceptron trained off-line from simulations. We demonstrate that an intuitive approach to directly map phasor measurement inputs to the neural network to generator rotor angle does not offer satisfactory results. We found out that a good way to approach the angle estimation problem is to use two neural networks in order to estimate the sin(@d) and cos(@d) of the angle and recover the latter from these values by simple post-processing. Simulation results on a part of the Mexican interconnected system show that the approach could yield satisfactory accuracy for real-time monitoring and control of transient instability.

Model Predictive Control to Alleviate Thermal Overloads

An approach inspired by model predictive control is proposed to determine a sequence of control actions aimed at alleviating thermal overloads. The algorithm brings the line currents below their limits in the time interval left by protections while accounting for constraints on control changes at each step. Its closed-loop nature allows to compensate for model inaccuracies.

Reconstructing and tracking network state from a limited number of synchrophasor measurements

A method is proposed to reconstruct and track network state from a limited number of phasor measurement unit (PMU) data. To deal with the resulting unobservability, the state with bus powers and generator voltages closest to previously estimated values is computed. Those values, treated as pseudo-measurements, are obtained from the last reconstructed state, in a recursive manner. The method involves solving an optimization problem with linear constraints. It is scalable insofar as it accommodates from a few PMUs up to configurations ensuring full network observability. Reconstruction of only a region is possible. These and other features are demonstrated on the Nordic32 test system, with synchronized phasors obtained from detailed time simulation of a situation evolving towards instability. Suitable choices of PMU location and pseudo-measurements are also discussed.