ML Classifiers for Vehicle Subsystem Fault Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern vehicles face challenges in diagnosing malfunctions, such as rough idle or misfiring, as the cause is often unclear and can involve multiple potential subsystems or components, making it difficult to identify specific contributors.
Innovation Solution
A system utilizing machine learning classifiers, including one-class, two-class, and multi-class classifiers, receives measurement data from sensing devices to determine if it belongs to a healthy or faulty class, identifying which vehicle subsystems contribute to malfunctions by plotting data in feature spaces and comparing outputs to a knowledge base.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional diagnostic methods are used to identify vehicle malfunctions, then the system can detect problems, but the ability to pinpoint specific contributing subsystems is poor
Solution Approach 1:
The patent divides the vehicle system into multiple subsystems (air intake, fuel injection, ignition, exhaust) and creates separate one-class classifiers for each subsystem. Each classifier independently analyzes measurement data from its associated sensors to determine whether that specific subsystem is contributing to the malfunction, enabling precise fault localization without requiring a single complex diagnostic system.
2Reliability
If multiple sensing devices are used to monitor vehicle subsystems, then measurement coverage is improved, but the difficulty of analyzing and interpreting the data increases
Solution Approach 1:
The one-class classifiers are trained exclusively on measurement data from healthy subsystem operation, enabling them to automatically learn and store the characteristics of normal operation. During diagnostics, the classifiers independently evaluate whether current measurement data falls within the healthy class boundaries, eliminating the need for complex manual analysis of multi-sensor data while maintaining high monitoring reliability.
3Loss of information
If comprehensive measurement data from all subsystems is analyzed together, then overall system health can be assessed, but the ability to isolate specific fault sources is reduced
Solution Approach 1:
Instead of analyzing all measurement data together, the patent segments the data analysis process by creating dedicated one-class classifiers for each subsystem. Each classifier receives and processes only the measurement data relevant to its specific subsystem, preserving fault source information by maintaining separate analysis channels while avoiding the complexity of integrating and analyzing all subsystem data in a single classifier.
Data Source
AI summary
A system for monitoring operation of a vehicle includes a processing device including an interface configured to receive measurement data from sensing devices configured to measure parameters of a vehicle system. The processing device is configured to receive measurement data from each of the plurality of sensing devices, and in response to detection of a malfunction in the vehicle, input at least a subset of the measurement data to a machine learning classifier associated with a vehicle subsystem, the classifier configured to define a class associated with normal operation of the vehicle subsystem. The processing device is also configured to determine whether the subset of the measurement data belongs to the class, and based on at least a selected amount of the subset of the measurement data being outside of the class, output a fault indication, the fault indication identifying the vehicle subsystem as having a contribution to the malfunction.