Knocking Detection Using Multi-Band Vibration Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining knocking in internal combustion engines often incorrectly discriminate between knocking and noise vibrations due to the removal of noise components, leading to erroneous determinations.
Innovation Solution
A device that detects vibrations in multiple frequency bands, including a broader second frequency band to capture noise components, and compares the waveform with a reference model to determine if knocking has occurred, thereby reducing false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a low-pass filter and a high-pass filter are employed to narrow the bandwidth and remove noise components, then the precision in detecting vibration particular to knocking is improved, but the characteristic portions of noise components (such as occurrence timing of vibration, attenuation rate) are removed from the detected waveform, leading to erroneous determination of knocking
Solution Approach 1:
The patent divides the vibration detection into two separate frequency band analyses: a first frequency band (narrow bandwidth) for detecting knocking vibration magnitude, and a second frequency band (wide bandwidth) for detecting waveform characteristics including noise components. This segmentation allows simultaneous preservation of knocking detection precision and noise characteristic information without mutual interference.
Solution Approach 2:
The patent transitions from single-frequency-band analysis to multi-frequency-band analysis, adding a dimensional aspect to the detection system. By analyzing vibrations in both a narrow first frequency band and a wider second frequency band simultaneously, the system gains additional information dimensions that enable distinction between knocking and noise vibrations.
2Device complexity
If only magnitude of vibration is used for determination, then the detection process is simplified, but vibration due to knocking cannot be discriminated from vibration due to noise
Solution Approach 1:
The determination process is segmented into two independent analyses: magnitude determination based on first frequency band vibration and waveform shape determination based on second frequency band vibration. This segmentation maintains relative simplicity while improving discrimination accuracy by evaluating multiple independent criteria.
Solution Approach 2:
The system uses feedback from both magnitude detection and waveform shape detection to make the final knocking determination. Only when both conditions indicate knocking is present does the system confirm knocking occurrence, providing a feedback mechanism that improves accuracy without excessive complexity.
3Loss of information
If a broader frequency band is used to capture noise components, then the waveform characteristics for discrimination are improved, but the precision in detecting knocking vibration magnitude may be reduced
Solution Approach 1:
The patent segments the frequency band analysis into two distinct purposes: the first frequency band (narrow) is dedicated to precise knocking magnitude detection, while the second frequency band (wide) is dedicated to capturing waveform characteristics including noise. This segmentation ensures each analysis optimizes for its specific purpose without compromising the other.
Solution Approach 2:
Different frequency bands are assigned different quality characteristics appropriate for their function: the first frequency band uses narrow bandwidth filtering to provide high precision for knocking magnitude, while the second frequency band uses wide bandwidth to preserve local waveform characteristics and noise information for discrimination purposes.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for precise determination of knocking by considering both knocking and noise characteristics, reducing erroneous determinations and improving the accuracy of knocking detection.
Implementation Method 1
an operation unit detecting a magnitude of vibration in a first frequency band including a frequency of vibration due to knocking, out of vibration occurring in an internal combustion engine
Implementation Method 2
detecting a magnitude of vibration in a second frequency band including the first frequency band and that is broader than the first frequency band
Data Source
AI summary
An engine ECU executes a program including a step of, when it has temporarily been determined that knocking had occurred because of the presence of an integrated value greater than a product of the reference magnitude and coefficient Y among the integrated values of vibration in fourth frequency band D that includes first to third frequency bands A to C, calculating knock magnitude N using the integrated values in the synthesized waveform of first to third frequency bands A to C and correlation coefficient K calculated from a vibration waveform of fourth frequency band D. Based on a comparison between knock magnitude N and determination value V(KX), whether or not knocking has occurred is determined. If there is no integrated value greater than a product of the reference magnitude and coefficient Y, it is determined that knocking has not occurred.