Modular Plant Measurement Confidence via Runtime Uncertainty Propagation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In flexible modular plants, it is challenging to calculate measurement uncertainties of output values during runtime due to the dynamic setup and combination of modular entities, where input values from multiple entities are combined without known arrangements at the design stage.
Innovation Solution
A method and device that assign uncertainty information to each modular entity, with a computing unit calculating output values and their associated uncertainties through propagation of uncertainty, allowing for automatic calculation and updating of uncertainties based on changes in the arrangement of modular entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If modular entities are dynamically arranged during runtime, then flexibility and adaptability of the plant configuration is improved, but the ability to calculate measurement uncertainties during design stage deteriorates
Solution Approach 1:
The patent assigns uncertainty information to each modular entity during the design phase, preparing the necessary data in advance. This preliminary assignment of uncertainty characteristics to individual entities enables later automatic propagation calculations when the actual runtime configuration is established, resolving the contradiction between dynamic flexibility and uncertainty calculation capability
2Extent of automation
If uncertainty information is assigned to each modular entity, then automatic calculation of output uncertainties during runtime is improved, but the complexity of data management and processing increases
Solution Approach 1:
The patent segments the uncertainty calculation by assigning uncertainty information to each individual modular entity separately. This segmentation allows the overall uncertainty propagation to be broken down into manageable individual contributions that can be automatically processed through the computational graph, reducing the complexity of managing global uncertainty data while enabling automatic calculation
3Adaptability or versatility
If multiple input values from different modular entities are combined, then the functionality and output capability of the plant is improved, but the propagation of uncertainty and error accumulation increases
Solution Approach 1:
The patent implements a feedback mechanism where uncertainty information flows through the same computational graph as the measurement data. By tracking how uncertainty propagates through each transformation and combination operation, the system provides feedback on the reliability of output values, enabling identification of error sources and optimization of modular entity arrangements to minimize error accumulation
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The invention provides a method for providing output values with associated uncertainties for a flexible modular plant or machine comprising an arrangement of modular entities (21-2n). To a plurality of said modular entities (21-2n), uncertainty information associated with an operation of said modular entity (21-2n) is assigned. Input values are provided based on an operation of the modular entities (21-2n). A computing unit (12) calculates an output value based on said input values. The computing unit (12) calculates an input value uncertainty for each input value based on the uncertainty information of the modular entity (21-2n). The computing unit (12) calculates at least one output value uncertainty associated with the output value based on propagation of uncertainty and using the input value uncertainties. The output value and the at least one output value uncertainty is outputted.