Fourier-Based Misfire Detection Using DFT Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing misfire detection systems in internal combustion engines face challenges in accurately distinguishing misfire conditions from normal combustion, especially across wide ranges of engine speed and load, and in higher cylinder engines, due to variations in crankshaft speed and engine resonance effects.
Innovation Solution
A method involving the production of a crankshaft angular speed signal, conversion into a frequency-domain misfire detection metric using Discrete Fourier Transform (DFT), normalization with a non-misfire metric to remove inherent variability, and detection of misfire by exceeding a predetermined threshold, allowing for accurate identification of the affected cylinder(s).
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional misfire detection methods are used, then detection can be performed on four and six cylinder engines, but detectability is compromised on higher cylinder engines and at low load high speed conditions
Solution Approach 1:
The patent transforms the crankshaft speed signal from time domain to frequency domain using Fourier Transform, changing the parameter domain to enable detection of misfire-related spectral components that are not apparent in the time domain. This allows detection across wider operating regions including low load high speed conditions where time-domain methods fail.
Solution Approach 2:
The patent adds a frequency dimension to the analysis by applying Fourier Transform to the crankshaft speed signal. This dimensional transformation enables separation of misfire-related frequency components from normal combustion variations, improving detectability on higher cylinder engines and across broader operating conditions.
2Measurement precision
If frequency domain analysis with adaptive thresholds is used, then misfire detection can be performed, but normal variations in crankshaft speed due to tooth error and engine effects mask misfire conditions
Solution Approach 1:
The patent segments the frequency spectrum into distinct components, isolating misfire-related spectral components from normal combustion variations through frequency domain analysis. This segmentation allows selective measurement of misfire indicators without interference from tooth error and engine effect variations.
Solution Approach 2:
The patent introduces a reference-based normalization mechanism that acts as an intermediary to remove the effects of normal crankshaft speed variations. By comparing against reference measurements taken under known good conditions, the system eliminates masking effects while preserving genuine misfire signals.
3Reliability
If crankshaft speed variations are monitored, then misfire can be detected, but variations due to tooth error and engine resonance complicate detection
Solution Approach 1:
The patent replaces mechanical/time-domain analysis with frequency-domain signal processing. By using Fourier Transform to convert the crankshaft speed signal into the frequency domain, the system substitutes complex time-varying mechanical analysis with spectral analysis, simplifying the detection process while improving reliability.
Data Source
AI summary
A computer implemented system for engine misfire detection employs a crankshaft-originated speed signal that includes normal variability due to target wheel tooth error and the like. A Discrete Fourier Transform (DFT) is performed on the speed signal to convert it into a frequency-domain raw misfire detection metric. The system is configured to normalize the raw misfire detection metric using a non-misfire metric that is obtained by and corresponds to non-misfire operation of an internal combustion engine. A resulting normalized misfire detection metric has such normal variations removed leaving variations attributable to misfire. The system detects a misfire when the normalized misfire detection metric exceeds a predetermined threshold and is bounded by a predetermined phase angle region. The system detects both continuous misfire as well as intermittent misfire. The system also detects single cylinder misfire and multiple cylinder misfire.