Neural Network Time Series Alignment for Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for aligning asynchronous intercorrelated time series datasets from different sensors, such as ECG and BCG, face challenges due to clock drift, noise, and morphological differences, limiting their adaptability and precision, especially when dealing with high-frequency signals and non-stationary data.
Innovation Solution
A computer-implemented method using two neural networks to filter and transform morphologically distinct time series into a common space, with iterative training and multi-model fitting algorithms to determine segment shifts and align datasets, allowing for flexible adaptation to various sensor inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional DTW or CCLA methods are used for time series alignment, then alignment can be performed, but the methods collapse time series points together or cannot handle large time series noise and morphologically strongly different time series datasets
Solution Approach 1:
The patent transforms time series data by changing parameters through neural network-based transformations. The method applies learned transformations to normalize and align time series from different sensors, enabling precise alignment while maintaining adaptability to various sensor inputs through the universal transformation model.
Solution Approach 2:
The patent replaces traditional mechanical alignment methods (DTW, CCLA) with a neural network-based transformation system. This substitution allows the system to handle noise and morphological differences that traditional methods cannot process, achieving both precision and versatility.
2Reliability
If manual mask creation and filtering are applied to time series, then noise can be reduced, but the process becomes strongly dependent on the nature of input time series and requires sensor-specific masks and filters
Solution Approach 1:
The patent creates a universal filtering and alignment system using neural networks that can process multiple types of time series from different sensors without requiring sensor-specific masks or filters. The learned transformations serve multiple functions: noise filtering, normalization, and alignment, eliminating the need for complex sensor-specific processing configurations.
3Measurement precision
If high-resolution alignment is performed on entire time series, then alignment precision can be improved, but computational time and complexity increase significantly
Solution Approach 1:
The patent segments time series data into smaller segments before processing. By dividing the entire time series into manageable segments, the method reduces computational complexity while maintaining alignment precision. The neural networks process these segments independently, then the results are combined to produce the final aligned time series.
Data Source
AI summary
A computer-implemented method for aligning intercorrelated asynchronous time series datasets includes the steps of:(a) retrieving a first time series dataset (x) and a second time series dataset (y), the first and second time series dataset being intercorrelated,(b) segmenting each of the first and second time series dataset (x, y) into a plurality of consecutive smaller segments (xi, yi), all segments of the first and second time series dataset (x, y) having the same length,(c) determining pairs of corresponding segments by associating successive segments of the first time series dataset (x) with corresponding segments of the second time series dataset (y),(d) optimizing, for each pair of corresponding segments, a correlation function to obtain an approximation of a first times series transformation function (f1) and of a second time series transformation function (f2),(e) using the first and second time series transformation functions (f1,f2) to determine a vector of segment shifts (s) whose components contain approximations of the shifts between the first and the second segment in a pair of corresponding segments (xi, yi),(f) applying a multi-model fitting algorithm to the segment shift vector (s), said multi-model algorithm outputting a shift function (fopt) for aligning segments of each pair of corresponding segments (xi, yi), and(g) aligning the first time series dataset (x) with the second series dataset by applying said shift function (fopt) to all pairs of corresponding segments (xi, yi),wherein the first time series transformation function (f1) is parametrized by weights (w1) of a first neural network (N1) and outputs a first highly correlated time series dataset, and wherein the second time series transformation function (f2) is parametrized by weights (w2) of a second neural network (N2) and outputs a second highly correlated time series dataset.