Batch Data Alignment Using DTI and Sliding-Window Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Batch processes in the process industry, particularly in pharmaceutical and biochemical sectors, face challenges with inconsistent batch lengths, leading to difficulties in data alignment, and existing methods like Dynamic Time Warping (DTW) are not mature enough for industrial applications, lacking systematic solutions for both offline and online batch alignments.
Innovation Solution
A comprehensive method using Dynamic Time Interpolation (DTI) with a Golden Section Search (GSS) algorithm for batch data alignment, which performs tasks such as loading historical data, outlier exclusion, reference batch selection, variable weighting, and alignment in both offline and online modes, addressing the limitations of DTW by formulating batch alignment as a continuous optimization problem and using a sliding window for real-time adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Dynamic Time Warping (DTW) algorithm is used for batch data alignment, then batch alignment can be performed, but the alignment results are undesirable and the method is not mature enough for industrial practice
Solution Approach 1:
The patent transforms the discrete DTW alignment approach into a continuous optimization problem by introducing continuous time variables and objective functions. This allows for smooth alignment trajectories and enables the use of continuous optimization methods (Golden Section Search) to find optimal alignment parameters, thereby improving both reliability and precision of batch alignment results
Solution Approach 2:
The patent introduces dynamic adjustment mechanisms where variable weightings are adjusted based on current batch phase identification. The system dynamically adapts the alignment process by modifying objective function parameters during optimization, allowing the alignment to account for different process stages and improve overall alignment quality
2Adaptability or versatility
If a comprehensive batch alignment system is implemented for both offline and online modes, then systematic solutions are provided, but the system complexity increases
Solution Approach 1:
The patent divides the batch alignment system into distinct functional modules: offline alignment module, online alignment module, batch phase identification module, and variable weighting adjustment module. Each module handles specific tasks independently, making the complex system more manageable and easier to implement while maintaining high adaptability across different operating modes
Solution Approach 2:
The patent creates a unified optimization framework that serves both offline and online alignment modes through the same continuous optimization methodology. The core algorithm structure remains consistent across different modes, reducing overall system complexity despite the comprehensive functionality provided
3Measurement precision
If variable weightings are adjusted based on batch phase for optimization, then alignment accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent performs batch phase identification and initial variable weighting setup before the main alignment optimization process. By pre-characterizing the batch process into phases and establishing initial weighting schemes, the system reduces the computational burden during the actual optimization while still achieving high alignment accuracy through the phased approach
Data Source
AI summary
Embodiments include a computer-implemented method (and system) for performing automated batch data alignment for modeling, monitoring, and control of an industrial batch process. The method (and system) loads, scales, and screens plant historian batch data for an industrial batch process. The method (and system) selects a reference batch as basis of the batch alignment, defines and adds or modifies one or more batch phases, and selects one or more batch variables based on one or more profiles and corresponding curvatures of the batch data. The method (and system) estimates one or more weightings, adjust one or more tuning parameters and uses a sliding time window combined with DTW, DTI and GSS algorithms, performs the batch alignment in offline mode or online mode.