Intelligent Electronic Device Fault Detection Using Bayesian Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting faults in intelligent electronic devices (IEDs) are cumbersome and time-consuming, relying on manual decoding of error codes, leading to high Mean Time to Repair (MTTR) due to inefficient detection and diagnosis processes.
Innovation Solution
A method utilizing a supervision and diagnostic module with a fault tree model and Bayesian network to detect failure events, identify root causes, and initiate corrective actions such as restarting components, filtering input data, or providing alerts through a Human Machine Interface (HMI), enabling quick and efficient fault detection and rectification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual decoding of error codes is used for fault detection, then technical support can analyze the error, but the Mean Time to Repair (MTTR) becomes high due to cumbersome detection process
Solution Approach 1:
The IED performs self-diagnosis by automatically analyzing error codes and monitoring component status without requiring manual technical support intervention. The system uses built-in diagnostic algorithms to identify root causes and suggest corrective actions, enabling the device to serve its own diagnostic needs and significantly reducing repair time.
Solution Approach 2:
The patent replaces the manual mechanical process of technicians physically examining and decoding error codes with an automated electronic diagnostic system. The system uses software-based error code analysis, component status monitoring, and automated root cause identification to substitute the manual diagnostic process, thereby eliminating the time delay associated with human intervention.
2Productivity
If automated fault detection system is implemented, then MTTR is reduced, but the device complexity increases due to additional supervision and diagnostic modules
Solution Approach 1:
The supervision and diagnostic module is designed to perform multiple functions: error code decoding, component status monitoring, root cause analysis, and corrective action recommendation. By consolidating these diverse diagnostic functions into a single multi-functional module, the system achieves automated fault detection without proportionally increasing overall device complexity.
Solution Approach 2:
The patent introduces an intermediary supervision and diagnostic module that acts as a mediator between the IED components and the user/technical support. This module collects data from various components, processes error codes, and presents simplified diagnostic information, thereby automating the diagnostic process while managing complexity through a dedicated intermediary layer.
3Loss of information
If basic error code display is provided in HMI, then fault information is available, but actionable diagnosis is not provided requiring technical support intervention
Solution Approach 1:
The system implements feedback by continuously monitoring component status and error codes, automatically analyzing the collected data, and providing actionable diagnostic recommendations back to the user through the HMI. This closed-loop feedback mechanism transforms raw error information into meaningful diagnostic guidance, enabling users to take immediate corrective actions without requiring external technical support.
Solution Approach 2:
The supervision and diagnostic module performs preliminary analysis of error codes and component status before technical support intervention is needed. By pre-processing the diagnostic information and identifying root causes in advance, the system prepares actionable recommendations that guide users through the repair process, eliminating the need for manual error code decoding and making fault diagnosis easier.
Data Source
AI summary
A method for detecting a fault in an intelligent electronic device that includes components uses a Bayesian network. The method includes detecting a failure event in the components, obtaining a first list of cause of failures in the component using a fault tree model, computing probability of the cause of failures to obtain a second list of probable causes of failure by monitoring of information about the elements identified in the first list, identifying a root cause of failure associated with the element comprised in the component using the Bayesian network based on the second list, and initiating a function. The function may be one of restarting the element having the root cause of failure, a filtering operation for input data provided to that element; and providing an alert in the human machine interface associated with the intelligent electronic device.