Technical System State Prediction Using Dual Time-Series Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting the state of technical systems struggle to maintain precision in long-term behavior without compromising short-term behavior, often leading to inaccuracies or increased computational resources.
Innovation Solution
A computer-implemented method using complementary high-pass and low-pass filters to decompose time series data into separate components for short-term and long-term predictions, with models trained on filtered data to predict states efficiently, reducing computational requirements through sampling and interpolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single model is used for prediction, then the model complexity is low, but the prediction accuracy for both short-term and long-term behavior cannot be maintained simultaneously
Solution Approach 1:
The patent divides the prediction task into two separate models: a first model for short-term behavior prediction and a second model for long-term behavior prediction. Each model is specialized for its respective time horizon, allowing both to achieve high accuracy without the complexity of a single universal model. The segmentation is further applied by dividing the time series data into short-term and long-term components through filtering operations.
2Measurement precision
If full time series data is used for training, then the prediction accuracy improves, but the computational resources required increase
Solution Approach 1:
The patent extracts only the necessary components from the full time series data for each model's training. The first model receives short-term filtered values while the second model receives long-term filtered values. This extraction approach allows each model to be trained on relevant data subsets, reducing overall computational requirements while maintaining prediction accuracy.
Solution Approach 2:
The patent applies partial action by training each model on a subset of the full time series data appropriate for its prediction horizon. Rather than training both models on complete data, the first model uses short-term components and the second model uses long-term components, reducing total computational effort while achieving accurate predictions.
Data Source
AI summary
A device and computer-implemented method for predicting a state of a technical system. A state of the technical system is detected. A time series is provided which includes values which characterize a course of the detected state of the technical system. Using a first filter, first filtered values for predicting the short-term behavior of the technical system are determined as a function of the values of the time series. Using a second filter, second filtered values for predicting the long-term behavior of the technical system are determined as a function of the values of the time series. A first value for the prediction is determined as a function of the filtered first values. A second value for the prediction is determined as a function of the filtered second values. A value of the prediction is determined as a function of the first and second values for the prediction.