Wavelet Analysis for DC Distribution Fault Location
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for locating phase-to-ground faults in ungrounded or high-resistance grounded DC distribution systems are inefficient, requiring additional signal sources and sensors, and are prone to human error and noise interference, especially in systems with DC/DC converters.
Innovation Solution
A method utilizing wavelet analysis, specifically Multi-Resolution Analysis (MRA), to process voltage signals and identify characteristic noise patterns for fault localization, eliminating the need for additional signal sources and sensors by leveraging high-frequency noise generated by power electronic converters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional AC signal injection method is used for fault location, then fault can be detected, but additional signal sources and sensors are required increasing system complexity
Solution Approach 1:
The patent utilizes the inherent high-frequency switching noise already present in the DC distribution system from power electronic converters as the detection signal source. The system serves itself by using its own operational noise rather than requiring external AC signal generators and additional sensors, thereby eliminating the need for extra hardware while maintaining fault detection capability
Solution Approach 2:
The patent converts the harmful high-frequency switching noise generated by power electronic converters into a useful detection signal for fault location. Instead of treating the noise as interference to be eliminated, the invention leverages this noise as the primary carrier for detecting and locating ground faults, transforming a detrimental effect into a beneficial resource
2Measurement precision
If handheld AC current sensors are used for fault location, then fault can be traced, but the process is time consuming and subject to human error
Solution Approach 1:
The patent replaces the mechanical handheld sensor approach with an automated electronic signal processing system. By using wavelet analysis to automatically process voltage signals and identify fault locations, the system eliminates manual intervention, reducing both detection time and human error while maintaining or improving location accuracy
Solution Approach 2:
The patent implements continuous monitoring of voltage signals with pre-configured wavelet analysis algorithms ready to immediately detect and locate faults. The system is prepared in advance with the detection methodology embedded, allowing instant fault identification without the need for manual sensor deployment or step-by-step tracing procedures
3Extent of automation
If fixed mounted sensors are used for automated fault detection, then automation is achieved, but sophisticated data communication links are required and reliability may be reduced
Solution Approach 1:
The patent makes the existing voltage measurement infrastructure serve multiple functions: both normal system operation monitoring and fault detection. By processing the same voltage signals through wavelet analysis, the system achieves automated fault location without requiring separate dedicated sensors or additional communication infrastructure, thereby maintaining automation while reducing complexity
Solution Approach 2:
The patent merges the fault detection function with the existing voltage measurement and control system. Instead of creating a separate automated detection system with its own sensors and communication links, the invention combines fault detection capabilities into the existing operational framework, eliminating redundant components and simplifying the overall system architecture
4Measurement precision
If AC signal generator is applied to locate ground faults, then fault location can be achieved, but the method is not tolerant of system noise and random variables
Solution Approach 1:
The patent changes the frequency domain parameters by focusing analysis on high-frequency components above the converter switching frequencies. By transforming the detection approach to analyze specific frequency ranges using wavelet transforms, the system achieves noise immunity by targeting the unique spectral signature of fault-related high-frequency transients rather than being affected by lower-frequency operational noise and random variables
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables rapid and accurate detection and location of phase-to-ground faults without additional hardware, improving reliability and reducing the risk of secondary faults by correlating noise patterns with specific fault locations within the system.
Implementation Method 1
A method utilizing wavelet analysis, specifically Multi-Resolution_analysis (MRA), to process voltage signals and identify characteristic noise patterns for fault localization
Data Source
AI summary
A method for locating phase to ground faults in DC distribution systems. The method includes utilizing wavelet analysis using Multi-Resolution Analysis (MRA) as a signal processing tool for recognition of characteristic features in the voltage signal. The voltage signal contains characteristic information in the high frequency range above the switching frequencies of the PE converters which allows for localization of the fault.