Plant Facility Diagnosis Using Process and Production State Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional facility diagnosis devices struggle to accurately diagnose the state of facilities due to fluctuations in load conditions and operation environments such as temperature and humidity, leading to unreliable measurement results.
Innovation Solution
A facility diagnosis device that acquires process and production data, generates state information, and uses a determination model to estimate and determine the operation state of facilities, incorporating facility data from sensors and maintenance data to improve diagnostic accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If sensor measurement is used for facility diagnosis, then real-time monitoring capability is improved, but measurement accuracy deteriorates due to environmental fluctuations
Solution Approach 1:
The patent introduces determination models as intermediary components that mediate between raw sensor measurements and final diagnosis results. These models process the measurement data while considering environmental context, acting as a buffer that transforms inaccurate raw measurements into accurate diagnostic conclusions without sacrificing real-time monitoring capability
Solution Approach 2:
The patent changes the parameter representation by transforming raw sensor measurements into determination values through multiple determination models. Each model evaluates different aspects (normal state, abnormal state, cause analysis) and the final diagnosis is based on comparing these transformed parameters rather than raw measurements, thereby improving accuracy while maintaining real-time response
2Measurement precision
If multiple determination models are used to improve diagnosis accuracy, then determination precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the diagnosis function into multiple specialized determination models, each responsible for a specific aspect (normal state determination, abnormal state determination, cause analysis). This segmentation allows each model to be relatively simple and focused, while their combination achieves high overall accuracy without creating an overly complex monolithic system
Solution Approach 2:
The determination models serve multiple functions: they process sensor data, compare it with historical information, analyze causes, and generate diagnoses. This multi-functionality reduces the need for separate dedicated systems for each task, thereby improving determination accuracy while controlling overall system complexity
Data Source
AI summary
A facility diagnosis device includes: a process data acquirer that acquires process data indicating a state of a process executed in a plant; a production data acquirer that acquires production data indicating a production state in the plant; a state information generator that generates state information for estimating an operation state of a facility operating in the process based on the acquired process data or the acquired production data, or the process data and the production data; and a determiner that determines the operation state of the facility based on the acquired process data or the generated state information and the acquired production data.