Measurement Data Preprocessing for Accurate State Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for determining the state of measurement targets, such as industrial plants, face challenges in accurately processing data from field devices due to data loss and noise, which affects the accuracy of state determination using learning models.
Innovation Solution
A determination apparatus and method that includes data preprocessing units for interpolation and waveform compression, acquiring learning data, determining preprocessing algorithms, and learning models to preprocess and output information for specifying these algorithms, enabling accurate state determination of measurement targets despite data loss and noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data preprocessing (interpolation and waveform compression) is applied to handle data loss and noise, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent applies preliminary action by performing interpolation processing to fill missing data points and waveform compression to reduce data volume before the learning model processes the data. This preprocessing prepares the data in advance, ensuring quality input for state determination while managing complexity through standardized processing steps.
Solution Approach 2:
The patent introduces an intermediary preprocessing unit that acts as a mediator between raw data acquisition and the learning model. This unit applies interpolation for missing values and compression for waveform reduction, serving as a buffer that improves measurement precision while isolating the complexity of data handling from the core determination system.
2Measurement precision
If complete data sets are waited for before processing, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent applies partial action by processing available data immediately rather than waiting for complete data sets. The interpolation processing fills in missing portions, allowing the system to perform state determination with partial data while maintaining acceptable precision, thus reducing processing delay significantly.
Solution Approach 2:
The system performs preliminary processing on available data using interpolation to estimate missing values and compression to reduce volume. This allows early state determination to be made without waiting for complete data acquisition, improving response time while maintaining measurement precision through the preprocessing algorithms.
3Device complexity
If raw data is processed without preprocessing, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent introduces a lightweight intermediary preprocessing unit that applies simple interpolation and compression algorithms. This mediator improves measurement precision by handling data quality issues while maintaining relatively simple device complexity through the use of straightforward processing techniques rather than complex systems.
Solution Approach 2:
The system applies preliminary interpolation and compression processing to prepare data before analysis. This preliminary action improves measurement precision by correcting data quality issues in advance, while the simplicity of the preprocessing algorithms keeps device complexity manageable.
4Measurement precision
If interpolation processing is applied to fill missing data, then measurement precision is improved, but loss of time for processing increases
Solution Approach 1:
The patent applies partial interpolation processing that fills missing data points efficiently without overly complex calculations. By using straightforward interpolation methods rather than exhaustive processing, the system improves measurement precision while keeping the additional processing time minimal and acceptable.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Applying a preprocessing algorithm according to the learning model used for determining the state of the measurement target is preferred. Provided is a determination apparatus including: a determination data acquisition unit for acquiring determination data, which are time series data obtained by measuring a measurement target; a determination data preprocessing unit for preprocessing the determination data by a preprocessing algorithm used for learning a learning model that outputs state of the measurement target; and a determination unit for determining state of the measurement target, based on the preprocessed determination data, using the learning model.