Join 60,000+ competitive exam aspirants
or with email
Your Cart
Subtotal₹0
Total₹0
Examoogle • User • info@examoogle.com • EE-2024-8821
Chapter 1 of 12 • Page 1 of 248🔒 Protected PDF • Watermarked
Join 60,000+ competitive exam aspirants
Identification of bad data is done by
Calculate the elements for error in the measurement vector
Calculate difference of error in the measurement vector
Calculate the error in measurement vector
None of above
Calculate the elements for error in the measurement vector
Quick Summary: Identification of bad data in power system state estimation is performed by evaluating the residuals of the measurement vector. Specifically, the Weighted Least Squares (WLS) approach involves calculating the elements of the normalized residual vector to detect and eliminate spurious measurements that deviate significantly from the expected physical model.
Identification of bad data in power system state estimation is performed by evaluating the residuals of the measurement vector. Specifically, the Weighted Least Squares (WLS) approach involves calculating the elements of the normalized residual vector to detect and eliminate spurious measurements that deviate significantly from the expected physical model.
r=z−h(x) — Measurement residual vector
rN=σr — Normalized residual vector for bad data detection
State estimation relies on minimizing the objective function J(x)=[z−h(x)]TR−1[z−h(x)]. Once the state vector x is estimated, the residuals r=z−h(x) are calculated. Large values in the normalized residual vector rN=σr indicate 'bad data' caused by sensor malfunction or communication errors, which must be identified and removed to maintain accurate grid monitoring.
State estimation maps raw measurements to the grid state using redundant data.
Bad data detection is a crucial part of the 'Gross Error Detection' phase in Energy Management Systems (EMS).
Measurements with high residuals are flagged as outliers and typically discarded before re-running the estimator.
The chi-square test (J(x)<χp,m−n2) is commonly used for global bad data detection.
Ensures high reliability of the energy management system
Corrects erroneous readings from faulty telemetry
Computationally expensive for large-scale power systems
Requires significant data redundancy
Power System SCADA systems
Automatic Generation Control (AGC)
Contingency Analysis in electrical grids
Measurements can be classified as 'Good', 'Suspect', or 'Bad' based on the magnitude of the normalized residual.
Option B and C are incorrect because simply calculating the difference or the error is not sufficient; one must normalize or weight these errors to statistically identify bad data.
A is correct — Bad data is identified by calculating the normalized residual elements of the measurement vector.
In competitive exams, always remember that 'Normalized Residuals' are the key to bad data detection. If the residual is greater than 3, the data is typically classified as bad.