Vehicle Knock Signal Processing via Partial Standard Deviation
Solution Overview
Problem
Conventional methods for processing engine knock signals in vehicle control systems are inefficient due to high memory and throughput requirements, leading to inaccurate engine knock determinations and inefficient operation of vehicle subsystems, particularly because they rely on costly full standard deviation computations.
Innovation Solution
A control system with a signal processing module that computes a partial standard deviation by summing and squaring sample points, allowing for efficient differentiation between normal noise and unwanted conditions, thereby generating a control signal based on the sensor signal and the computed standard deviation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full standard deviation computation method is used, then measurement precision of knock signal is improved, but device complexity and loss of energy increase due to large memory and throughput requirements
Solution Approach 1:
The patent segments the standard deviation computation into two separate accumulations: one for the sum of sample points and another for the sum of squared deviations. This segmentation allows incremental updates without requiring buffering of all sample points, thereby reducing memory requirements while maintaining measurement precision.
Solution Approach 2:
The patent performs preliminary accumulation of the sum of sample points and sum of squared deviations as data is received. By pre-computing and storing these intermediate values, the system avoids the need to buffer all raw sample points and perform computationally intensive operations later, reducing both memory and processing throughput requirements.
2Measurement precision
If full standard deviation computation method is used, then measurement precision of knock signal is improved, but loss of energy increases due to high throughput requirements
Solution Approach 1:
The computation is segmented into separate accumulations for sum of points and sum of squared deviations, allowing incremental updates with minimal processing at each step. This reduces the computational burden and energy consumption compared to computing full standard deviation from buffered data.
Solution Approach 2:
By pre-accumulating the necessary statistical parameters (sum of points, sum of squared deviations) as data streams in, the system performs lightweight incremental updates rather than intensive batch processing, significantly reducing processor throughput requirements and associated energy loss.
3Device complexity
If moving averages methods or first order lag filters are used, then device complexity is reduced, but measurement precision of knock signal deteriorates
Solution Approach 1:
The patent changes the computational parameters from simple moving averages or first-order lag filters to a two-moment accumulation method (sum of points and sum of squared deviations). This parameter change enables calculation of standard deviation, which provides superior description of sample distribution while maintaining low device complexity through incremental computation.
Data Source
AI summary
A control system for a vehicle is provided. The control system includes a signal processing module that receives a sensor signal and extracts a plurality of sample points from the sensor signal. A computation module computes a summation of the sample points, computes a summation of squares of the sample points, and computes a standard deviation based on the summation of the sample points and the summation of the squares of the sample points. A control module generates a control signal based on the sensor signal and the standard deviation.