Plant Equipment Data Sorting for Faster Accurate Monitoring Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for creating models of plant equipment are limited by the number of parameters that can be set, requiring extensive experience and time, and are inefficient for large and complex equipment, leading to longer computational processing times and a need for highly accurate data to achieve model accuracy.
Innovation Solution
A data sorting device and method that sorts operating data from plant equipment by identifying physically and engineeringly associated data, creating a model that excludes abnormal values, and using this model to form a normal-value database for monitoring and diagnosis, allowing for efficient model creation and improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all operating data from plant equipment is used to create a model, then the model accuracy improves, but the computational processing time increases and the model becomes more complex
Solution Approach 1:
The patent extracts only the necessary operating data that is highly associated physically and engineeringly with the target equipment, separating it from unrelated data. This extraction process reduces the data volume used for model creation while maintaining model accuracy by focusing on relevant parameters only.
Solution Approach 2:
The patent segments operating data into different categories based on physical and engineering associations. By dividing the data into relevant and irrelevant segments, the system processes only the necessary portions for model creation, reducing computational load while preserving model accuracy.
2Adaptability or versatility
If multiple diversified monitoring items are consolidated into a single monitoring device, then the device versatility improves, but the model and device size increase leading to longer processing time
Solution Approach 1:
The patent creates a universal model structure that can handle multiple diversified monitoring items through a single unified framework. This universal model uses common processing logic and data structures that work across different equipment types and monitoring parameters, maintaining versatility without proportionally increasing complexity.
Solution Approach 2:
The patent changes parameters by selectively including only highly associated operating data for each monitoring item rather than processing all possible parameters. This parameter selection approach allows the single monitoring device to efficiently handle multiple items by adjusting which specific parameters are processed for each case.
3Ease of manufacture
If equipment information parameters are limited when creating an initial model, then the model creation process simplifies, but the model accuracy deteriorates requiring extensive experience and operating time to improve
Solution Approach 1:
The patent enables the model creation process to automatically identify and select highly associated operating data without requiring extensive manual expertise. The system self-services by autonomously determining which parameters are relevant based on physical and engineering associations, reducing the need for experienced operators while maintaining model accuracy.
Data Source
AI summary
The present invention provides a data sorting device and method and a monitoring and diagnosis device, which are able to create a model conveniently. A data sorting device and a monitoring and diagnosis device of the present invention include an operating data database which stores operating data of the plant equipment in a time-series manner. The devices input the operating data that are highly associated physically and engineeringly among the operating data stored in the operating data database, inputting processing attributes relevant to the operating data that are highly associated physically and engineeringly, creating a model simulating characteristics of the plant equipment, and performing data sorting, monitoring, and diagnosis through use of the model.