Smart Device Real-Time Positioning Using Accelerometer Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing real-time positioning methods for smart devices suffer from low accuracy due to the limitations of MEMS sensors and the complexity of human motion, leading to significant cumulative errors and errors introduced by incorrect device orientation.
Innovation Solution
A method that utilizes tri-axial acceleration data from accelerometers and magnetic field sensors, combined with Principal Component Analysis and band-pass filtering, to determine the moving direction and displacement of a smart device, allowing for improved accuracy and real-time positioning without requiring a specific device orientation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If Dead Reckoning and Zero Velocity Update (ZUPT) are utilized to estimate position by integrating acceleration data, then positioning can be achieved without external references, but significant cumulative error occurs due to low MEMS sensor accuracy
Solution Approach 1:
The patent applies periodic action by using Zero Velocity Update (ZUPT) at specific moments (when the device is stationary or during heel-strike phases of walking) to reset cumulative errors. The acceleration data is processed in periodic segments, with error correction applied at regular intervals rather than continuously integrating raw data, thereby reducing cumulative drift while maintaining autonomous positioning capability
Solution Approach 2:
The patent changes parameters by transitioning from direct integration of raw acceleration data to a hybrid approach that combines ZUPT-corrected velocity estimates with periodic GPS or other external reference corrections. This parameter change in the processing methodology (from pure Dead Reckoning to corrected Dead Reckoning) significantly reduces cumulative error while preserving the ability to operate without continuous external references
2Ease of operation
If the moving direction is determined by the moving orientation of the smart device requiring a specific axis to point to the moving direction, then positioning can be calculated by combining gravitational acceleration and magnetic field orientation, but considerable error is introduced when the user does not hold the device in a pre-defined orientation
Solution Approach 1:
The patent applies dynamics by transitioning from a static device orientation requirement to a dynamic adaptation approach. The system dynamically identifies the user's actual holding orientation through accelerometer and gyroscope data, then dynamically adjusts the coordinate transformation and moving direction calculation accordingly. This allows the positioning system to work accurately regardless of whether the user holds the device in portrait or landscape mode, or at various angles, eliminating the need for strict pre-defined orientation requirements
3Measurement precision
If users' gait is analyzed according to waveforms of acceleration values to determine moving direction by detecting instant acceleration in the deceleration phase, then cumulative error can be avoided, but the method is not universally applicable due to the randomness and complexity of human motions
Solution Approach 1:
The patent applies universality by developing a multi-functional approach that combines multiple gait analysis methods. Instead of relying on a single waveform detection method that may fail for certain users or motion patterns, the system integrates multiple indicators including acceleration waveform analysis, orientation changes, step detection algorithms, and temporal patterns. This universal approach adapts to different user characteristics, walking styles, and motion complexities, making the system broadly applicable across diverse populations and scenarios
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 method enhances the accuracy of real-time positioning by filtering interference and incorporating a step threshold to accurately count steps and determine the moving direction, effectively reducing cumulative errors and improving the stability of the positioning system.
Implementation Method 1
getting tri-axial acceleration in the device coordinate by calculating real-time data acquired by an accelerometer
Implementation Method 2
getting tri-axial gravitational acceleration in the device coordinate acquired by the gravity accelerometer
Implementation Method 3
getting tri-axial magnetic flux in the device coordinate acquired by the magnetic field sensor
Data Source
AI summary
The present invention discloses a method for real-time positioning of a smart device, comprising: getting tri-axial acceleration sequences truncated in turn by using a preset time period in the world coordinate individually; acquiring a dominate frequency fstep by applying fast Fourier transform (FFT) on the truncated Z-axis acceleration sequence, using a band-pass filter with a pass band [fstep−0.5 Hz, fstep+0.5 Hz] to filter the truncated tri axial acceleration sequences individually; comparing the time domain waveform form from the filtered Z-axis acceleration sequence with a preset step threshold, and counting each peak above the threshold as a step to acquire a user's step number in the present time period; determining the moving direction of the smart device in the time period by calculating the filtered X-axis and Y-axis acceleration sequences; determining the current position of the smart device by calculating the user's step number and the moving direction of the smart device.