Adaptive Sensor Bias Estimation via Dynamic Dispersion Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion sensors, such as gyrometers, face challenges in accurately updating their bias without requiring additional sensors, as current methods rely on thresholds that can lead to overestimation or underestimation of bias, affecting measurement precision.
Innovation Solution
A process that iteratively updates the bias of motion sensors by calculating a dispersion indicator within defined time ranges, adjusting the dispersion threshold based on signal dispersion, and subtracting the bias from acquired signals, allowing for frequent and accurate bias updates without additional sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a fixed dispersion threshold is used to detect stationary ranges for bias updating, then the device complexity is reduced, but the measurement precision deteriorates due to overestimation or underestimation of bias
Solution Approach 1:
The patent applies the dynamics principle by making the dispersion threshold adaptive rather than fixed. The threshold evolves over time based on the statistical properties of the sensor signals, allowing it to automatically adjust to changing operational conditions. This dynamic adjustment resolves the contradiction by maintaining measurement precision across varying scenarios without increasing device complexity, as the threshold adapts autonomously based on observed signal characteristics.
Solution Approach 2:
The patent implements parameter changes by modifying the dispersion threshold parameter based on the calculated dispersion indicator. Instead of using a static threshold, the system changes the threshold parameter dynamically according to the actual signal dispersion observed during operation. This allows the system to maintain accurate bias detection across different operational states while keeping the overall device complexity low, as only the threshold parameter needs adjustment rather than the entire detection mechanism.
2Productivity
If the dispersion threshold is set high to enable frequent bias updates, then the productivity is improved, but the measurement precision worsens due to bias overestimation when the sensor is not truly stationary
Solution Approach 1:
The patent applies feedback by using the calculated dispersion indicator to inform subsequent threshold settings and bias update decisions. The system continuously monitors the dispersion indicator and adjusts the threshold accordingly, creating a feedback loop that prevents premature bias updates. This feedback mechanism ensures that bias updates occur only when appropriate conditions are met, maintaining measurement precision while still enabling frequent updates when the sensor is truly stationary.
Solution Approach 2:
The patent implements preliminary action by calculating and monitoring the dispersion indicator before proceeding with bias updates. Instead of updating bias solely based on a fixed threshold, the system first evaluates the dispersion indicator to determine whether the sensor is truly stationary. This preliminary evaluation prevents premature bias updates that would occur with high fixed thresholds, ensuring accuracy while maintaining productivity through timely updates when conditions are appropriate.
3Measurement precision
If the dispersion threshold is set low to ensure accurate bias detection, then the measurement precision is improved, but the productivity decreases because bias updates occur less frequently
Solution Approach 1:
The patent resolves this contradiction through dynamics by making the dispersion threshold adaptive rather than statically low. The threshold dynamically adjusts based on the observed dispersion indicator, allowing it to be effectively low when needed for accurate detection but permitting updates to occur more frequently when conditions warrant it. This dynamic behavior maintains measurement precision while improving productivity compared to a fixed low threshold.
4Measurement precision
If additional motion sensors are used to improve bias estimation accuracy, then the measurement precision is improved, but the device complexity increases due to requiring multiple sensor types and complex computational operations
Solution Approach 1:
The patent applies self-service by enabling the single motion sensor to estimate its own bias using its own output signals. The system processes the sensor's signals through dispersion indicator calculation and adaptive threshold comparison to autonomously detect stationary ranges and update bias without requiring additional sensors or complex computational operations. This self-service approach resolves the contradiction by maintaining measurement precision while minimizing device complexity.
Solution Approach 2:
The patent implements taking out by extracting the bias estimation function from a multi-sensor system and implementing it within the single sensor's processing capability. Instead of requiring multiple sensors to provide complementary information for bias estimation, the system extracts and utilizes the necessary information from the single sensor's own output signals. This eliminates the need for additional sensors and reduces computational complexity while maintaining estimation accuracy.
Data Source
Figure 1A~1C
Figure 2A
Figure 2B
AI summary
The invention is a method for estimating and updating the bias of a sensor. After an initialization phase, the method comprises the following steps: - acquisition of a signal (sk) at a measurement time (tk), with each measurement time corresponding to a bias (mk), determined during a previous iteration or during initialization, and a dispersion threshold (vth,k), determined during a previous iteration or during initialization; - association of an analysis time range (Δtk) with the measurement time (tk), and calculation of a dispersion indicator (vk), representing the dispersion of the signals acquired during the analysis time range; - comparison of the dispersion indicator thus calculated with the dispersion threshold (vth,k) taken into account in step b); - depending on the comparison, maintaining the bias at an unchanged value or updating the bias;- subtraction of the bias (mk) resulting from the previous step from the acquired signal (sk) at the time of measurement; - repetition of the steps listed above, incrementing the time of measurement (tk). When the bias is updated, the dispersion threshold is also updated, based on the dispersion indicator (vk) calculated at the time of measurement.