Operation Of Large-scale Power Systems Research Articles

Real-time multi-risk assessment of large scale power system with high-penetration renewable energy

Abstract To accurately assess the impact of real-time uncertainty of high penetration renewable energy sources on the real-time operation of large-scale power systems, this study proposes an innovative method for real-time multi-risk assessment in power systems. Firstly, by utilizing an improved spectral clustering algorithm and mixture density network to analyze historical operational data, effective clustering of different renewable energy generation curves and accurate fitting of prediction errors were achieved. Next, random sampling was performed from the probability density curves of prediction errors to obtain real-time renewable energy outputs. At the same time, a non-sequential Monte Carlo method was used to incorporate interruptions for each component into the scenario simulation. Subsequently, a state assessment was conducted for each sampled scenario by minimizing load shedding and energy curtailment. On this basis, risk indices that comprehensively consider the real-time flexibility of the system and power supply-demand imbalances were established. Finally, validation in a test system verified the accuracy and speed of the proposed multi-risk assessment method.

Wind Power Forecasting Based on WaveNet and Multitask Learning

Accurately predicting the power output of wind turbines is crucial for ensuring the reliable and efficient operation of large-scale power systems. To address the inherent limitations of physical models, statistical models, and machine learning algorithms, we propose a novel framework for wind turbine power prediction. This framework combines a special type of convolutional neural network, WaveNet, with a multigate mixture-of-experts (MMoE) architecture. The integration aims to overcome the inherent limitations by effectively capturing and utilizing complex patterns and trends in the time series data. First, the maximum information coefficient (MIC) method is applied to handle data features, and the wavelet transform technique is employed to remove noise from the data. Subsequently, WaveNet utilizes its scalable convolutional network to extract representations of wind power data and effectively capture long-range temporal information. These representations are then fed into the MMoE architecture, which treats multistep time series prediction as a set of independent yet interrelated tasks, allowing for information sharing among different tasks to prevent error accumulation and improve prediction accuracy. We conducted predictions for various forecasting horizons and compared the performance of the proposed model against several benchmark models. The experimental results confirm the strong predictive capability of the WaveNet–MMoE framework.

Week-ahead Daily Peak Load Forecasting Using Hybrid Convolutional Neural Network

Peak load forecasting (PLF) is essential for the operation of large-scale power systems, such as security-constrained economic dispatch and rolling unit commitment. PLF is critical for peak load shaving and demand side management in distribution systems, too. Results obtained by PLF can assist market participants to develop their bidding schemes in power markets. This work presents a novel method, based on a hybrid convolutional neural network (CNN) that is cascaded with a fully-connected network, to explore week-ahead daily PLF. The proposed method uses a systematic grid search along with Adaptive Moment Estimation (Adam) optimizer to design the hybrid model. The grid search tunes the network structure and hyperparameters (such as kernel size) of the hybrid CNN while Adam optimizer adjusts the synaptic weights and parameters (e.g., values of a kernel). Realistic daily peak load data and meteorological (temperature) data in Taiwan are investigated. Simulation results obtained by the proposed hybrid deep learning model are better than those obtained by the traditional multi-layer neural network, recurrent neural network, vector auto-regressive moving average model and support vector machine.

Protection of power transformers against the effect of magnetic storms

Abstract In the operation of large-scale power systems for the long-distance transmission of large amounts of electricity, a number of cases have been reported in which anomalies in the Earth’s magnetosphere, referred to as geomagnetic storms, have caused a severe system collapse. Changes in the geomagnetic field cause a semi-saturating phenomenon, in which the high-voltage lines and especially the high-voltage windings of the power transformers of the system are overloaded with current and subsequently also thermally. The present article briefly explains the physical nature of magnetic storms and then describes a new device that either eliminates the possibility of a step-down power transformer accident or significantly reduces its effects on the system. The essence of this device are frequency filters, which are connected in parallel to the high-voltage windings of power transformers. At the beginning of a geomagnetic storm, the frequency filter is automatically connected to the system and is automatically disconnected when it subsides. The operation of frequency filters does not require human intervention, acquisition and operating costs are low and their integration into existing power systems is easy.

Load State Transition Curve Based Unit Commitment for Production Cost Modeling With Wind Power

The production cost modeling (PCM) plays an important role in the mid/long-term power system planning. With the increasing penetration of renewable power, it becomes solving a series of unit commitment (UC) problems for time periods up to one or multiple years. This process is greatly challenged by the UC solution efficiency which hinders its practical use. In this paper, a fast UC method is proposed based on the load state transition curve (LSTC). Firstly, LSTC is proposed and obtained via load clustering. It simplifies the load model as well as maintains the necessary temporal characteristics. Secondly, the LSTC based UC model is developed by accommodating the constraint formulation to LSTC. Compared with the conventional ones, the scale of LSTC-based UC is dramatically reduced. The stochastic and volatile wind power is also taken into consideration. The case studies on the modified IEEE-RTS 1979 have verified the validity and feasibility of the proposed method in which the solution time is reduced from ∼300s to ∼12s with the cost error lower than 0.2%. It provides an efficient tool for the optimal planning and operation of large-scale power systems.

Non-Stationary Stabilized Fast Transversal RLS Filter for Online Power System Modal Estimation

Online monitoring of inter-area oscillations is a vital need for secure operation of large-scale power systems. This paper aims at online estimation of inter-area modes by a novel stabilized fast transversal recursive least squares (SFTRLS) filter running on ambient power system data received from phasor measurement units. The SFTRLS filter is a fast finite-impulse response adaptive filter that is widely used in communication applications, such as noise and echo cancelation. It has the lowest computational complexity among least-squares filters and directly calculates coefficients of the parametric model selected for identifying a system under study. In the proposed method, the SFTRLS filter along with digital band-pass Butterworth filter and down-sampling blocks are implemented in online manner for identification of an autoregressive model that describes the measured power system ambient signal. Applicability and accuracy of the proposed method are evaluated using real measurement data of western systems coordinating council (WSCC) breakup on August 10, 1996, as well as simulated ambient data of the IEEE 16-machine, 5-area test system excited by white noise random load modulations. Its performance is investigated under different conditions of load modulations, measurement noises, sudden change in damping ratio, and occurrence of three-phase faults. The simulation results verify efficient and robust performance of the proposed method for modal estimation of real large-scale power systems.

Multi-Objective Optimal Power Flow Based on Hybrid Firefly-Bat Algorithm and Constraints- Prior Object-Fuzzy Sorting Strategy

In this paper, a novel hybrid firefly-bat algorithm with constraints-prior object-fuzzy sorting strategy (HFBA-COFS) is put forward to solve the strictly-constrained multi-objective optimal power flow (MOOPF) problems. The hybrid firefly-bat algorithm (HFBA) integrates the dimension-based firefly algorithm and the modified bat algorithm to improve the population-diversity and global-exploration ability of original algorithm. To handle the unqualified state variables and overcome the deficiency of traditional penalty function approach (PFA), the constraints-prior Pareto-dominant rule (CPR) which takes constraints-violation and objectives-value into consideration is proposed in this paper. Furthermore, an effective constraints-prior object-fuzzy sorting (COFS) strategy based on CPR rule is presented to seek the well-distributed Pareto optimal set (POS) in solving the MOOPF problems. To validate the great advantages of HFBA-COFS algorithm, ten MOOPF cases optimizing active power loss, total emission and fuel cost are simulated on the IEEE 30-bus, IEEE 57-bus and IEEE 118-bus systems. In addition, the generational distance and SPREAD evaluation indexes powerfully demonstrate that the proposed HFBA-COFS algorithm can achieve high-quality POS, which has great significance to realize the safe and economic operation of large-scale power systems.

A novel tool for transient stability analysis of large-scale power systems: Its application to the KEPCO system

Time-domain simulation is an important tool for power system dynamic analysis. We solve a set of differential and algebraic equations (DAE) in order to study the dynamic behavior of power systems. These power systems include thousands of generators, exciters, turbine-governors, loads, and other devices. The resultant large-scale DAEs are very difficult to handle and solve. Nevertheless, solution techniques are needed to not just guarantee accuracy but have computational efficiency. In this paper, we report on a novel tool that we developed to deal with time-domain simulation for dynamic analysis and operation of large-scale power systems. The tool has several major features related to transient stability analysis, contingency screening, and ranking. We mainly discuss the structure of this tool and the accuracy of the included dynamic models. Also, the paper proposes a new load model to overcome the low-voltage problem. The proposed technique provides a good performance and convergence when the terminal voltage is below some predefined value. Compared to the commercial tools, the developed tool is numerically well conditioned by introducing the ZIP model algorithm. This tool has been used to support and enhance power engineering education at both the undergraduate and graduate levels. In the case study, simulation results were validated through comparative simulations with the Power System Simulator for Engineering (PSS/E) and Transient Security Assessment Tool (TSAT).

Splitting strategies for islanding operation of large-scale power systems using OBDD-based methods

System splitting problem (SS problem) is to determine proper splitting points (or called splitting strategies) to split the entire interconnected transmission network into islands ensuring generation/load balance and satisfaction of transmission capacity constraints when islanding operation of a system is unavoidable. For a large-scale power system, its SS problem is very complicated in general because a combinatorial explosion of strategy space happens. This paper mainly studies how to find proper splitting strategies of large-scale power systems using an ordered binary decision diagrams (OBDD)-based three-phase method. Then, a time-based layered structure of the problem solving process is introduced to make this method more practical. Simulation results on IEEE 30- and 118-bus networks show that by this method, proper splitting strategies can be given quickly. Further analyses indicate that this method is effective for larger-scale power systems.

Optimal operation of large-scale power systems

A method for optimal operation of large-scale power systems is presented that it is similar to the one utilized by the Houston Lighting and Power Company. The main objective is to minimize the system fuel costs while maintaining an acceptable system performance in terms of limits on generator real and reactive power outputs, transformer tap settings, and bus voltage levels. Minimizing the fuel costs of such large-scale systems enhances the performance of optimal real power generator allocation and of optimal power flow that results in an economic dispatch. To handle large-scale systems of this nature, the problem is decomposed into a real and a reactive power optimization problem. The control variables are generator real power outputs for the real power optimization problem and generator reactive power outputs, compensating capacitors, and transformer tap settings for the reactive power optimization. The gradient projection method (GPM) is utilized to solve the optimization problems. It is an iterative procedure for finding an extremum of a function of several constraint variables without using penalty functions or Lagrange multipliers. Mathematical models are developed to represent the sensitivity relationships between dependent and control variables for both real and reactive-power optimization procedures and thus eliminate the use of B-coefficients.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>