Segmental Process Tracing for Partial Interlock Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional process modeling techniques require all variables to be known for all components before the model can converge, making it inefficient and unreliable for monitoring systems and predicting changes in industrial processes.
Innovation Solution
A segmental analysis method that allows analyzing a portion of the system by sending results from one element to adjacent elements connected by segments, enabling predictions of overall system outcomes without needing complete system data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional process modeling techniques are used to analyze the entire system, then complete system coverage is achieved, but computational resources are excessively consumed and analysis time increases
Solution Approach 1:
The patent divides the process control system into multiple segments, where each segment represents a portion of the overall system. The segmentation allows the system to analyze only the relevant portion affected by a change event, rather than requiring analysis of the entire system. This is achieved by identifying a first segment containing the changed element and a second segment containing the element to be actuated, and analyzing the path between these segments without requiring analysis of all other system components.
2Measurement precision
If all variables for all components are required before model convergence, then complete system accuracy is achieved, but the system cannot provide timely predictions for partial system changes
Solution Approach 1:
The system performs preliminary identification of the analysis path between the changed element and the target element before actual analysis begins. By pre-identifying which segments and elements are relevant to the specific change event, the system prepares the minimal necessary dataset in advance, avoiding the need to collect and process all system variables before analysis can commence.
3Reliability
If the entire system is analyzed simultaneously, then comprehensive system understanding is achieved, but computational complexity and resource requirements increase significantly
Solution Approach 1:
The patent extracts and isolates only the relevant segments and elements that are directly affected by the change event from the overall system. By taking out the specific path between the changed element and the target element for focused analysis, the system eliminates the computational burden of analyzing unrelated system components while maintaining accuracy for the affected portions.
4Measurement precision
If complete system data is required for analysis, then comprehensive predictions are achieved, but the system cannot provide predictions when sensor data is missing from certain components
Solution Approach 1:
The system performs analysis on a partial basis, requiring only the data necessary for the specific segment being analyzed rather than complete system data. This partial action approach allows the system to provide predictions even when sensor data is missing from components outside the analyzed segment, as long as the necessary data for the affected path is available.
Data Source
AI summary
The disclosure is directed to a novel system and methods for modeling process components using segmental analysis. As opposed to analyzing the entire system at once, each element representing a process component is assigned parameter inputs that enable the system to determine a current state. Changes in the element's parameter inputs are fed in a bi-directional manner to other elements connected by segments where the other elements use the parameter inputs to determine a predictive or current process condition that are used as new parameter inputs for the next element. In some embodiments, the system allows for the prevention of component state change with only a portion of the entire process being solved, as abnormal conditions can be detected without solving for the entire system.