Load modeling has significant impact on power system studies. This paper presents a review on load modeling and identification techniques. Load models can be classified into two broad categories: 1) static and 2) dynamic models, while there are two types of approaches to identify model parameters: 1) measurement-based and 2) component-based. Load modeling has received more attention in recent years because of the renewable integration, demand-side management, and smart metering devices. However, the commonly used load models are outdated, and cannot represent emerging loads. There is a need to systematically review existing load modeling techniques and suggest future research directions to meet the increasing interests from industry and academia. In this paper, we provide a thorough survey on the academic research progress and industry practices, and highlight existing issues and new trends in load modeling.

Load Modeling—A Review

Power system model quality is key to safe and reliable electricity delivery. A novel method using disturbance data recorded by phasor measurement units (PMUs) is proposed to improve the integrity of dynamic models, which consists of both model validation and model parameter identification. In the model validation step, event playback is applied to simulate the model's response for a rigorous comparison with measured response captured by PMUs. A large mismatch indicates model deficiency. In the parameter identification step, an automatic calibration method using the extended Kalman filter (EKF) techniques and trajectory sensitivity analysis is formulated to save significant effort and time from manual tuning. Case studies are presented to demonstrate that the proposed method has adequate performance for model validation and parameter calibration in terms of accuracy, good convergence speed and robustness with respect to measurement noises. The proposed method is cost-effective and complements traditional equipment testing for improving dynamic model quality.

Generator dynamic model validation and parameter calibration using phasor measurements at the point of connection

Power system emergency control is generally regarded as the last safety net for grid security and resiliency. Existing emergency control schemes are usually designed off-line based on either the conceived "worst" case scenario or a few typical operation scenarios. These schemes are facing significant adaptiveness and robustness issues as increasing uncertainties and variations occur in modern electrical grids. To address these challenges, for the first time, this paper developed novel adaptive emergency control schemes using deep reinforcement learning (DRL), by leveraging the high-dimensional feature extraction and non-linear generalization capabilities of DRL for complex power systems. Furthermore, an open-source platform named RLGC has been designed for the first time to assist the development and benchmarking of DRL algorithms for power system control. Details of the platform and DRL-based emergency control schemes for generator dynamic braking and under-voltage load shedding are presented. Extensive case studies performed in both two-area four-machine system and IEEE 39-Bus system have demonstrated the excellent performance and robustness of the proposed schemes.

Adaptive Power System Emergency Control using Deep Reinforcement Learning.

Short term voltage stability poses a significant threat to system stability and reliability. This paper applies dynamic VAr injection to ensure short term voltage stability following a large disturbance in a power system with high concentration of induction motor loads. Decelerating and stalling of induction motor loads is considered to be the major cause of fault induced delayed voltage recovery (FIDVR) and short term voltage stability. If system dynamics are not taken into account properly, the proposed control solution may be an expensive over design or an under design that is not capable of eliminating FIDVR problems completely. In this work, the optimal amount and locations for installing dynamic reactive resources are found by control vector parameterization (CVP), a dynamic optimization approach. The efficiency and effectiveness of this approach is improved by utilizing results from trajectory sensitivity analysis, singular value decomposition and linear programming optimization. Dynamic optimization based on CVP approach is tested in an IEEE 162-bus system and a realistic large scale utility power system.

Dynamic Optimization Based Reactive Power Planning to Mitigate Slow Voltage Recovery and Short Term Voltage Instability

This paper develops a novel online fast load shedding strategy aimed at shedding the most effective load to mitigate fault-induced delayed voltage recovery (FIDVR). Induction motor kinetic energy deviation has been shown to be identical to the integral of power imbalance. Instead of using voltage as an indicator, the proposed strategy applies a centralized scheme making use of equivalent motor kinetic energy to identify the most effective loads to shed. In order to derive kinetic energy, the equivalent inertia and the speed of the induction motor are needed. A methodology is proposed to obtain these variables through online measurements. Simulations on a sample power system illustrate the novel load shedding strategy and its effectiveness. The proposed strategy is compared with conventional four-stage under-voltage load shedding (UVLS) scheme. The results have validated that the proposed strategy can effectively mitigate fault-induced delayed voltage recovery and require much less load to be shed.

A Novel Online Load Shedding Strategy for Mitigating Fault-Induced Delayed Voltage Recovery

Parameter identification is the key technology in measurement-based load modeling A hybrid learning algorithm is proposed to identify parameters for the aggregate load model (ZIP augmented with induction motor) The hybrid learning algorithm combines the genetic algorithm (GA) and the nonlinear Levenberg-Marquardt (L-M) algorithm It takes advantages of the global search ability of GA and the local search ability of L-M algorithm, which is a more powerful search technique The proposed algorithm is tested for load parameter identifications using both simulation data and field measurement data Numerical results illustrate that the hybrid learning algorithm can improve the accuracy and reduce the computation time for load model parameter identifications

A Novel Parameter Identification Approach via Hybrid Learning for Aggregate Load Modeling

Load components have strong effects on the power system's behavior and should be modeled accurately in system studies. With more and more disturbance measurement equipments have been installed in transmission systems, it opens up an opportunity to use the measurement data to derive load model parameters. This method is termed as the measurement- based approach. The theoretical foundation of the measurement- based approach is system identification. In this paper, we propose to apply an improved Genetic Algorithm (GA) to derive the parameters of load models using the measured disturbance data. The improved genetic algorithm is based on following aspects: (i) the strategy of keeping the best individual (ii) the adaptive rates of mutation and crossover (iii) the strategy of immigration (iv) the optimal search direction. The improved GA method is compared with the Levenberg-Marquardt method using a 23-bus test system.

Derivation of load model parameters using improved Genetic Algorithm

Relay margin can be applied as a direct tool to evaluate relay status during system disturbances. Based on the practical dynamic security assessment criteria for relay margin, we propose the relay margin ratio and the relay staying time ratio which provide more valuable information than relay margin for relay status evaluation. Motivated by the need of considering relay tripping for contingency evaluations, the most possible tripping severity which incorporates the possibility of relay tripping to evaluate contingency severity is proposed. The relay margin ratio and relay staying time ratio are applied to identify the most critical relays for each contingency based on which the most possible tripping severity is obtained to identify the most critical contingencies in transmission systems. These two inherently related procedures together compose a framework to assess the transmission system vulnerability. Numerical test results are presented using a 23-bus test system.

Transmission system vulnerability assessment based on practical identification of critical relays and contingencies

In this paper relay margin is derived in terms of bus voltage. With the variation of bus voltage during system disturbance the relay margin will also consequently vary to form a trajectory which is proposed as the relay margin trajectory. The sensitivity of the relay margin trajectory is investigated to quantify the rate of the relay margin variation. Based on the relay margin we propose a vulnerable relay indicator which samples the smallest relay margin over the interested time span. Based on the vulnerable relay indicator, another two indicators--severe contingency indicator and system security indicator are proposed. These three inherently related indicators together compose a framework for system security analysis. Vulnerable relay indicator is applied to identify the most vulnerable relays based on which the severe contingency indicator is obtained. System security indicator is sequentially gained based on those two indicators to identify the transmission vulnerability for the power system security analysis. Numerical test result on a test system is presented.

Hua Bai

Papers

A Novel Parameter Identification Approach via Hybrid Learning for Aggregate Load Modeling

A Novel Online Load Shedding Strategy for Mitigating Fault-Induced Delayed Voltage Recovery

Derivation of load model parameters using improved Genetic Algorithm

Transmission system vulnerability assessment based on practical identification of critical relays and contingencies

Relay margin trajectory based identification of transmission vulnerability for power system security assessment