Gait Cadence Control Using Frequency Domain Signal Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining cadence using sensors in mobile and wearable devices are prone to errors, especially at low walking velocities, due to reliance on thresholding in the time domain and are inaccurate in noisy or weak motion signals, which is problematic for healthcare applications.
Innovation Solution
The use of a combination of Fourier transformation, wavelet transformation, autocorrelation, and peak counting to analyze motion sensor signals, with techniques like sensor fusion and adaptive filtering to improve accuracy and handle abrupt changes in cadence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If thresholding in the time domain is used to detect steps, then the method is simple to implement, but accuracy deteriorates especially at low walking velocities and with noisy signals
Solution Approach 1:
The patent replaces the mechanical thresholding approach in the time domain with signal processing techniques in the frequency domain (FFT, wavelet transformation). This substitution allows for more accurate cadence determination by analyzing the spectral characteristics of the acceleration signal rather than relying on simple amplitude thresholds, thereby resolving the contradiction between implementation simplicity and measurement accuracy.
Solution Approach 2:
The patent transforms the analysis from the time domain to the frequency domain by changing the parameter space in which the signal is processed. By applying FFT and wavelet transformation, the system analyzes frequency components and temporal-frequency representations of the signal, which provides robustness against noise and low-velocity conditions while maintaining computational feasibility.
2Measurement precision
If frequency analysis (FFT) is used to obtain fundamental frequency, then cadence determination improves, but computational complexity increases
Solution Approach 1:
The patent segments the signal analysis process into multiple stages: first applying FFT to obtain frequency spectrum, then using wavelet transformation for time-frequency analysis, and finally applying peak detection algorithms. This segmentation of the computational process allows each step to be optimized independently and facilitates implementation on mobile devices with limited processing power while maintaining high measurement precision.
Solution Approach 2:
The patent performs preliminary signal processing by applying bandpass filtering and detrending before the main frequency analysis. This preliminary action removes irrelevant components and prepares the signal for more efficient FFT and wavelet transformation, reducing the computational burden while preserving the accuracy of cadence determination.
3Measurement precision
If adaptive thresholds are used to improve accuracy, then measurement precision improves, but the system becomes more complex and still relies on time domain analysis
Solution Approach 1:
The patent moves the analysis from the one-dimensional time domain to the two-dimensional time-frequency domain using wavelet transformation. This dimensional change allows the system to capture both temporal and spectral characteristics of the gait signal simultaneously, providing accurate cadence determination without relying on adaptive thresholding in the time domain, thus achieving high precision with a more elegant solution.
4Device complexity
If traditional sensor methods are used, then the system is simple, but reliability deteriorates with weak or noisy motion signals typical of walking
Solution Approach 1:
The patent introduces wavelet transformation as an intermediary between the raw acceleration signal and the final cadence measurement. This intermediary technique acts as a powerful signal processing tool that enhances the reliability of detection by providing time-frequency localization, allowing the system to reliably detect gait patterns even in the presence of noise and weak signals typical of walking activities.
Data Source
AI summary
Methods for controlling an aspect of an application in a mobile or wearable device and a mobile or wearable device user's representation in real time are described, where the user is performing a gait activity with a gait cadence, and the gait cadence is used for control. Additional user's mobility characteristics leveraged for control may include velocity and stride length, and the sensors utilized to obtain any contextual information may be accelerometers.