Sensor Error Detection in Vehicle Exhaust Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying sensor faults in a motor vehicle's exhaust-gas system require active actuation of the internal combustion engine, leading to adverse effects on driving comfort and emission generation.
Innovation Solution
A method using an artificial neural network to determine a setpoint sensor signal and compare it with actual sensor signals, identifying faults without active engine actuation by analyzing deviations through parameterizable fault models, allowing for qualitative and quantitative fault determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If active actuation of the internal combustion engine is used to identify sensor faults, then sensor fault detection capability is improved, but driving comfort deteriorates and undesired emissions are generated
Solution Approach 1:
The patent creates a virtual copy of the sensor signal through a mathematical model that simulates the expected sensor behavior under various operating conditions. This model-generated setpoint signal is then compared with the actual sensor signal to detect faults, eliminating the need for active engine actuation while maintaining fault detection capability.
Solution Approach 2:
The patent replaces the mechanical approach of actively actuating the engine with a computational approach using mathematical models and signal processing. The physical engine actuation is substituted by comparing actual sensor signals with model-predicted signals, achieving fault detection without mechanical intervention.
2Object-affected harmful factors
If no active engine actuation is used to identify sensor faults, then driving comfort is maintained and emissions are reduced, but sensor fault detection accuracy may deteriorate
Solution Approach 1:
The patent performs preliminary modeling of the sensor system behavior under various operating conditions before actual fault detection. The mathematical model is pre-trained and validated to accurately predict sensor signals, enabling precise fault detection during normal operation without requiring additional active testing.
Solution Approach 2:
The patent implements a feedback mechanism where the model-generated setpoint signal is continuously compared with the actual sensor signal. This feedback loop enables real-time fault detection by identifying deviations between expected and actual sensor behavior, maintaining high detection accuracy during normal vehicle operation.
Data Source
AI summary
A method determines a sensor error of a sensor in an exhaust gas system of a motor vehicle. One step of the method involves determining at least one actual sensor signal of the sensor. Another step of the method involves determining at least one target sensor signal of the sensor by means of a model. A further step of the method involves determining the sensor error of the sensor according to a deviation between the actual sensor signal of the sensor and the target sensor signal of the sensor.

