Parametric MBR Modeling for Real-Time Vehicle Fault Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Integrated Vehicle Health Management (IVHM) systems face challenges in efficiently diagnosing equipment failures in complex systems due to high false alarm rates and the costly effort required for maintaining models across the life-cycle of vehicles, especially with inconsistent BIT data from different vendors.
Innovation Solution
An on-board diagnostic system utilizing a Model-Based Reasoner (MBR) engine with a parameterized model represented in XML format, incorporating sensor data, BIT data, and a graphical user interface for model development and testing, which reduces false alarms and maintenance costs by providing real-time diagnostic analysis and actionable maintenance tasks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If rules based approach with BIT data is used for engineering diagnostics, then diagnostic coverage is provided, but false alarm rates are high and engineering effort is very costly
Solution Approach 1:
The patent transforms discrete BIT data into continuous parametric models with defined ranges and tolerances. By changing the data representation from binary pass/fail states to parameterized physical quantities with acceptable ranges, the system reduces false alarms while maintaining diagnostic coverage. The parametric approach allows for natural tolerance handling and trend analysis that rules-based systems cannot provide.
Solution Approach 2:
The patent introduces a model-based reasoning layer as an intermediary between raw BIT data and diagnostic conclusions. This intermediary processes BIT data through physical models that incorporate tolerances, uncertainties, and system behavior, thereby filtering out false alarms while preserving true fault detections. The model acts as a mediator that translates inconsistent vendor BIT data into unified diagnostic insights.
2Measurement precision
If comprehensive models are developed for entire vehicle systems, then diagnostic accuracy is improved, but time and resources required to develop and update models increase significantly
Solution Approach 1:
The patent segments the comprehensive vehicle model into modular subsystem models, each with its own parametric definitions and relationships. This segmentation allows independent development, verification, and updating of individual subsystem models without affecting the entire vehicle model. Teams can work on separate modules simultaneously, reducing overall development time while maintaining system-level diagnostic accuracy.
Solution Approach 2:
The patent creates a universal parametric modeling framework that can be applied across different vehicle systems and subsystems. The same modeling approach, XML schema, and reasoning engine serve multiple domains (powertrain, aerodynamics, structures, etc.), reducing the need to develop separate modeling methodologies for each system. This universality accelerates model development and facilitates knowledge transfer across projects.
3Adaptability or versatility
If models are updated in response to hardware and software upgrades, then system currency is maintained, but costs of maintaining systems become immense
Solution Approach 1:
The patent implements dynamic model updating where parametric models automatically adapt to hardware and software changes through version control and configuration management. When upgrades occur, the system dynamically adjusts model parameters, tolerances, and relationships without requiring complete model redevelopment. This dynamic adaptation maintains model currency while minimizing maintenance effort and costs.
Solution Approach 2:
The patent performs preliminary model development and verification activities during the design phase, creating a foundation that anticipates future hardware and software variations. By pre-defining parametric relationships and update mechanisms early in the system lifecycle, the patent reduces the complexity and cost of subsequent model maintenance during operations and support phases.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods and apparatus are provided for an integrated vehicle health management system that uses a model based reasoner to develop real-time diagnostic and maintenance information about a vehicle. A model having nodes is used to represent sensors, components, functions and other elements of the system. Parameterized input sensors provide flexible definitions of each node. Once developed, the model is loaded into an operational flight program to provide real¬ time failure analysis of the vehicle.