Drifting Sensor State Association Using Clustered Signal Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sensors often experience output value drift over time, making it difficult to accurately determine states, especially in noisy conditions, and existing methods are limited in handling multiple states with limited memory resources.
Innovation Solution
A method involving signal clustering and compression to dynamically associate states with new values, using techniques like k-means clustering and piecewise constant function approximation, which allows for state determination without assuming a normal state and reduces memory usage by storing only significant changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a simple average or sliding window average is used to determine sensor states, then the method is simple and computationally efficient, but it cannot accurately handle multiple states with important probabilities and assumes a normal state which may not exist
Solution Approach 1:
The patent transforms the sensor signal from raw values to a compressed representation using piecewise constant functions, changing the parameter space. This allows the system to work with fewer, more meaningful parameters (compression points and values) rather than all raw sensor readings, enabling accurate multi-state detection without requiring complex processing of the entire signal history.
Solution Approach 2:
The patent creates a compressed copy of the sensor signal using piecewise constant approximation. Instead of processing the original high-resolution signal, it works with a simplified representation that captures the essential state changes. This copied representation is sufficient for state determination while reducing computational complexity and memory requirements.
2Loss of information
If all sensor signal values are stored for analysis, then complete information is available for state determination, but memory resources are excessively consumed
Solution Approach 1:
The patent extracts only the essential information from the sensor signal by identifying compression points where state changes occur. Instead of storing all signal values, it extracts and stores only the critical points and their corresponding state values. This extraction process retains the necessary information for accurate state determination while dramatically reducing the quantity of data that must be stored in memory.
3Measurement precision
If clustering is performed on raw sensor values, then accurate state identification is possible, but the computational load and memory requirements increase significantly
Solution Approach 1:
The patent performs preliminary compression of the sensor signal using piecewise constant approximation before applying clustering algorithms. By pre-processing the signal to identify and store only the critical compression points and values, the system reduces the dataset size that requires clustering. This preliminary action maintains state identification accuracy while significantly reducing the computational load and memory requirements of the subsequent clustering operation.
Data Source
AI summary
The present invention relates to a method of associating at least one state in a plurality of states to a new value output by a drifting sensor, the method comprising: /a/ receiving a signal from the sensor, said signal comprising a plurality of values; /b/clustering the values of said signal into a number of clusters equal to the number of the plurality of states, each cluster being associated with a respective state in the plurality of states; /c/ for the new value of the signal, associating at least one state in said plurality of states or a probability rating representing the probability to be associated with one state in said plurality of states for said new value of the signal, the associated state or the associated probability rating being determined based on at least distances (dH, dL) of said new value of the signal to respective clusters.