Parameter Estimation of Smart Grid using PMU with Kalman Filter and Bayesian Prediction
Asoumitra Kumar Sarker1, Anand Mohan2

1Soumitra Kumar Sarker, m.Tech Research Scholar, 1Department of Electrical Engineering Alakh Prakash Goyal Shimla University, Mehli Shoghi Bypass Road, Shimla-171009, (Himachal Pradesh) India.
2Anand Mohan, professor, Department of Electrical Engineering Alakh Prakash Goyal Shimla University, Mehli Shoghi Bypass Road, Shimla-171009, (Himachal Pradesh) India.

Manuscript received on 30 June 2019 | Revised Manuscript received on 05 July 2019 | Manuscript published on 30 July 2019 | PP: 1604-1612 | Volume-8 Issue-9, July 2019 | Retrieval Number: I8153078919/19©BEIESP | DOI: 10.35940/ijitee.I8153.078919
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Load flow is the main issue which occurs in power grid systems. To improve the performance, reduce the cost and enhance the reliability in power systems, smart grids have been proposed. In electricity distribution system, smart devices like smart meters are used for effective performance. The real concern in these devices is to protect the data from unauthorized parties and noise occurring in data. Smart device reader acts as the bridge which connects the smart grid devices with smart grid clouds. In many of the instances of circuit-based analysis, the network parts are restricted to the regarded value of impedances with voltage and current resource. But the issue of load flow is usually diverse in the sense that rather of impedances, the known amounts are active and reactive powers in most of the network buses, since the performance of most of the load in a great deal of instances are as continuous power loads, presuming that voltages used on them stay within suitable ranges. There are various methods which are used to solve these problems. Kalman filters are proposed to achieve the optimal performance on the smart grid devices. This filter identifies the device failures, unusual disturbance, and malicious data attacks. Kalman Filter is a dynamic state estimation method which is mainly used in this paper for noise variation estimation. The use of dynamic state estimation methods such as the Kalman filter provides an optimal solution to the process of real-time data prediction and reduces the problem based on non-linearity. The analysis of real-time data depends on Phasor Measuring Units (PMU) which plays a significant role in power transmission and distribution processes due to their ability to monitor the power flow within a network. The process of PMU-based monitoring improves the quality of the smart grid. Simultaneously, the implementation of PMU increases the dynamics of noise variance which further inflates the uncertainty in noise-based distribution. This paper presents a method to reduce the amount of uncertainty in noise by using a linear quadratic estimation method (LQE), usually known as Kalman filter along with Taylor expansion series but this process is time-consuming and is vulnerable to a large number of errors at the time of testing. The main reason behind this approach is the high complexity of the system which makes it very hard to derive the process. The proposed studies adopts a technique to work on covariance earlier based estimation using Bayesian method together with the estimation of dynamic polynomial prior by using Particle Swarm Optimization (PSO). The experimental evaluation compares the outcomes received from the primary Kalman filter, PSO optimized Kalman filter out and Kalman filter Covariance Bayesian method. Finally, the effects received from the analysis highlights the truth that the PSO optimized Kalman clear out to be more effective than the Kalman filter out with Covariance Bayesian approach
Keywords: PMU, Bus, Filters, Kalman, PSO, Taylor Expansion, Voltage.

Scope of the Article: Predictive Analysis