A scalable wide area monitoring system using cellular neural networks

Balasubramaniam, Karthikeyan; Luitel, Bipul; Venayagamoorthy, G.K.

doi:10.1109/ijcnn.2012.6252647

Cited by 5 publications

(7 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Different forecasting techniques can be applied to estimate F k , g k , and Q k . The Kalman filter in [16], exponential smoothing in [11], and artificial neural networks (ANN) in [17]- [18] have been utilized successfully under this context.…”

Section: Ekf-based Dsementioning

confidence: 99%

See 1 more Smart Citation

Dynamic state estimation for distribution networks with renewable energy integration

Nguyen¹,

Venayagamoorthy²,

Kling³

et al. 2013

SGCE

View full text Add to dashboard Cite

The massive integration of variable and unpredictable Renewable Energy Sources (RES) and new types of load consumptions increases the dynamic and uncertain nature of the electricity grid. Emerging interests have focused on improving the monitoring capabilities of network operators so that they can have accurate insight into a network's status at the right moment and predict its future trends. Though state estimation is crucial for this purpose to trigger control functions, it has been used mainly for steady-state analysis. The need for dynamic state estimation (DSE), however, is increasing for real-time control and operation. This paper addresses the important role of DSE over conventional static-state estimation in this new distribution network context. Computational burden mitigates the state-of-the-art utilizations of DSE in real large-scale networks, although DSE was introduced several decades ago. This paper the unscented Kalman filter (UKF) to alleviate computational burden with DSE. The UKF-based approach does not use a linearization procedure and thus outperforms the conventional Extended Kalman Filter based approach to cope with non-linear models. The performance of the UKF method is investigated with a simulation of an 18-bus distribution network on the real-time digital simulator (RTDS) platform. A distribution network with considerable integration of renewable energy production is used to evaluate the UKF-based DSE approach under different types of events.

show abstract

Section: Ekf-based Dsementioning

confidence: 99%

“…ANN as a typical application of CI has been studied extensively in [17]- [18]. The prediction model can be improved by integrating load forecasting, which was proposed as a forecasting-aided state estimation (FASE) concept in [15].…”

Section:  Step 1 -Parameter Identificationmentioning

confidence: 99%

Dynamic state estimation for distribution networks with renewable energy integration

Nguyen¹,

Venayagamoorthy²,

Kling³

et al. 2013

SGCE

View full text Add to dashboard Cite

show abstract

“…2 and 3. Each block, referred to as a cell in the CNN is a computing unit, which models the behavior of a power system component [13]. For example, each cell in the transient stability margin prediction layer represents a generator.…”

Section: Continuous Stability Limit Prediction Systemmentioning

confidence: 99%

“…Cellular neural networks have been used in literature for predictive modeling of power systems. References [13], [14] use CNNs to predict power system state variables. CNNs provide a generic framework that can take advantage of distributed computing power.…”

Section: Introductionmentioning

confidence: 99%

CNN based power system transient stability margin and voltage stability index prediction

Balasubramaniam

Venayagamoorthy

Watson

2013

2013 IEEE Computational Intelligence Applications in Smart Grid (CIASG)

Self Cite

View full text Add to dashboard Cite

Operators at electric grid control centers are faced with the task of making important decisions in real-time. With the plethora of data available it becomes important to extract information from the available data, based on which knowledge of system condition can be formed. This knowledge can then be used in decision making. Metrics such as transient stability margin (TSM) and voltage stability load index (VSLI) help in assessing the stability of the system. In this study, cellular neural network (CNN) based stability margin prediction system is developed in a distributed computing framework. The developed system not only extracts information from available data but also predicts the same, one step ahead of time. Moreover, the framework employed uses distributed computing and hence could be used on a large scale power system with a linear increase in computation time instead of an exponential increase. A reduced version of New Zealand's South Island power system is used as the test system to demonstrate the feasibility of CNNs for TSM and VSLI prediction.Index Terms-Cellular neural networks, dynamic security assessment, real-time monitoring, smart grid, transient stability margin, voltage stability load index.

show abstract

“…In the online process, these controllers revive a learned policy in the presence of an associated input or generalizes for a new input. Neural Network (NN) as a distinct learning-based function approximator has been effectively implemented as a power system intelligent controller in several works [7][8][9], and their ability to adapt during nonlinear transient conditions have been discussed [7][8][9][10][11][12].These architectures use NNs in the form of supervised learning as an intelligent PSS for damping generator oscillations. However, majority of these works have used the intelligent controller by itself.…”

Section: Introductionmentioning

confidence: 99%

A Lyapunov function based optimal hybrid power system controller for improved transient stability

Yousefian

Kamalasadan

2016

Electric Power Systems Research

View full text Add to dashboard Cite

In this paper, an intelligent power system stabilizer based on a stable and optimal hybrid learning-based adaptive control architecture is proposed which is evolved from approximate dynamic programming technique. The hybrid learning controller is based on two algorithms a) A reinforcement learning neural network controller implemented by Adaptive Critic Design (ACD) and b) A model reference adaptive controller (MRAC). The proposed method uses a novel value priority architecture for integrating these two algorithms. The designed ACD approximates nonlinear functional dynamics of the power system and interacts with the MRAC that adapts based on parametric changes. The value priority scheme developed based on a softmax network generates a hybrid control signal derived from a Lyapunov stability function that evaluates identification, performance, and stability of each controller. The overall hybrid control algorithm ensures convergence to an optimal control solution without using an explicit model of the system and at the same time ensures overall system stability. Theoretical results are validated by simulation studies for electric-generator stabilization on a 2-area 5-generator power system and as a wide-area controller on IEEE 68-bus system.

show abstract

A scalable wide area monitoring system using cellular neural networks

Cited by 5 publications

References 10 publications

Dynamic state estimation for distribution networks with renewable energy integration

Dynamic state estimation for distribution networks with renewable energy integration

CNN based power system transient stability margin and voltage stability index prediction

A Lyapunov function based optimal hybrid power system controller for improved transient stability

Contact Info

Product

Resources

About