Distribution Network State Estimation with Bad Data Elimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distribution network state estimation methods fail to accurately identify and eliminate erroneous measurement values, particularly in non-redundantly metered systems, leading to complex data errors due to biases, drifts, or wrong connections.
Innovation Solution
A system with a bad data detection and elimination module is introduced, comprising a distribution system state estimation unit and a supervisory control and data acquisition unit, which uses load group scaling factors and normalized residuals to identify and discard obviously erroneous measurements, and applies additional steps to detect and eliminate bad data in non-redundant network parts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional measurement steps are taken to ensure all measurements are correct, then measurement accuracy is improved, but system complexity and processing time increase
Solution Approach 1:
The patent segments the bad data detection process into distinct phases: preparation phase detection using normalized residuals, area estimation phase detection, and post-estimation verification. This segmentation allows comprehensive measurement verification without requiring a single complex system, thereby improving measurement accuracy while managing system complexity through modular processing steps.
2Productivity
If bad data detection is performed only in preparation phase or after area estimation, then processing speed is improved, but detection reliability deteriorates
Solution Approach 1:
The patent applies preliminary action by performing bad data detection during the preparation phase using normalized residuals and load group scaling factor analysis. This preliminary detection identifies and eliminates obvious bad data before the main state estimation process, ensuring that subsequent processing operates on cleaner data and improving overall detection reliability without significantly impacting processing speed.
Solution Approach 2:
The patent implements feedback mechanisms where measurement data is continuously evaluated against expected ranges and patterns. When measurements deviate from expected behavior, the system provides feedback to identify potential bad data sources, allowing for iterative refinement of data quality assessment and improving detection reliability through multiple verification passes.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention inter alia relates to a method for state estimation of a distribution network based on real time measurement values, the method comprising the steps of: calculating load scaling factors at least for non-redundantly measurable parts of the distribution network, determining whether each of the load scaling factors related to a non-redundantly measurable part of the distribution network is within a given range (RLSF), discarding the measurement values related to the corresponding scaling factor for all load scaling factors outside the given range (RLSF), and estimating the network state of the distribution network based on the remaining measurement values.