Logic Circuit Fault Location Estimation for Multiple Faults
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional fault location estimation systems for logic circuits face challenges such as long processing times for large-scale circuits, reduced accuracy due to neglecting error propagating paths, inability to handle multiple faults effectively, and difficulty in prioritizing physical analysis from output results.
Innovation Solution
The system classifies fault candidates into groups based on error-observation nodes, calculates combinations considering inclusion relations, and performs multiple-fault simulation to accurately estimate fault locations, reducing diagnostic time and improving accuracy by accounting for interacting error paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional fault location estimation systems use fan-in cone classification and single-fault simulation for each internal node, then fault location can be estimated, but processing time becomes excessively long for large-scale circuits
Solution Approach 1:
The patent segments the fault diagnosis process into two distinct phases: (1) single-fault simulation phase where each internal node is simulated independently to generate fault candidates and error propagating paths, and (2) multiple-fault combination phase where pre-computed fault candidate groups are combined based on error-observation node relationships. This segmentation allows the computationally intensive single-fault simulations to be performed once and reused, avoiding redundant calculations in the multiple-fault analysis phase.
Solution Approach 2:
The patent performs preliminary single-fault simulations for all internal nodes before the actual multiple-fault diagnosis. During this preliminary phase, fault candidates and error propagating paths are pre-computed and stored. In the subsequent multiple-fault analysis phase, these pre-computed results are reused to generate fault candidate groups and determine combinations, eliminating the need to re-simulate single-fault conditions and significantly reducing overall processing time.
2Productivity
If conventional systems neglect error propagating paths in fault simulation, then processing is simplified, but accuracy of fault location estimation deteriorates
Solution Approach 1:
The patent performs preliminary error propagating path analysis during the single-fault simulation phase, tracing how errors propagate from each internal node to error-observation nodes. These pre-computed error propagating paths are stored and reused in the multiple-fault combination phase to determine which fault candidates should be grouped together, ensuring accurate fault location estimation without requiring complex real-time path analysis.
3Measurement precision
If conventional systems perform single-fault simulation for each internal node without grouping, then fault candidates can be identified, but the ability to handle multiple faults effectively is reduced
Solution Approach 1:
The patent merges multiple single-fault simulation results by grouping fault candidates based on their error propagating paths to common error-observation nodes. Fault candidates that propagate errors to the same error-observation nodes are merged into a single fault candidate group. This merging process enables the system to handle multiple faults effectively by considering combinations of fault candidates within groups, thereby improving the system's adaptability to multiple fault conditions.
Solution Approach 2:
The patent creates a universal framework that handles both single-fault and multiple-fault conditions through the same error-observation node based grouping mechanism. The fault candidate groups and their combinations can explain various error output patterns, making the system universally applicable to different fault scenarios including single faults, multiple faults, and complex interacting fault conditions.
4Quantity of substance
If conventional systems output fault candidates without prioritization, then complete fault locations can be listed, but difficulty in prioritizing physical analysis increases
Solution Approach 1:
The patent applies local quality by differentiating fault candidate groups based on their specific error propagating paths and error-observation node relationships. Each fault candidate group is characterized by the specific error-observation nodes it affects and the error propagating paths it follows. This local differentiation enables prioritization of fault candidates based on their specific characteristics and impact, making physical analysis more efficient by focusing on the most likely fault locations first.
Data Source
AI summary
A fault location estimation system includes single-fault-assumed diagnostic unit nodes; error-observation node basis candidate classification unit; inclusion fault candidate group selection unit; inter-pattern overlapping unit; and multiple-fault simulation checking unit.