Plant Behavior Analysis Using Structure and Variation Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing analysis methods for plant behavior are limited by the assumption of constant device characteristics, failing to accurately account for variations over time, which can lead to inaccurate predictions and simulations.

Innovation Solution

An analysis apparatus and method that utilize a combination of structure models and variation models, where structure models represent the initial plant configuration and variation models simulate changes in device characteristics over time, allowing for more accurate analysis of plant operations by considering actual operating conditions and historical data from multiple plants.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If initial device characteristics are used for analysis, then analysis simplicity is maintained, but analysis accuracy deteriorates due to device deterioration over time

Engineering Contradiction:
Improveanalysis accuracyVSAvoidanalysis complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The analysis system segments device characteristics into two distinct components: structure models representing initial device characteristics and variation models representing deterioration over time. This segmentation allows the system to maintain separate, manageable model types while achieving comprehensive and accurate plant behavior analysis by combining both components.

Inventive Principle:
Principle #1Segmentation

2Reliability

If variation models from multiple plants are combined, then predictive capability is improved, but data processing complexity increases

Engineering Contradiction:
Improvepredictive capabilityVSAvoiddata processing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The variation model serves multiple functions: it represents device deterioration for a single plant, captures operational variations across different plants, and enables predictive analysis for future time points. This multi-functionality allows the same model structure to handle diverse data sources and analysis objectives without requiring separate specialized models for each function.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system merges structure models and variation models into a unified analysis framework. By combining the initial characteristics from structure models with the time-dependent variations from variation models, the system achieves comprehensive plant behavior prediction that accounts for both design specifications and actual operational deterioration.

Inventive Principle:
Principle #5Merging (Combining)

3Adaptability or versatility

If constant device characteristics are assumed, then analysis simplicity is maintained, but operational realism deteriorates

Engineering Contradiction:
Improveoperational realismVSAvoidmodel complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system transitions from static device characteristic assumptions to dynamic modeling by introducing variation models that capture how device characteristics change over time and with operational conditions. This dynamic approach allows the analysis to adapt to actual plant behavior while maintaining a structured modeling framework through the separation of structure and variation components.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4057095B1Analysis apparatus, analysis method and program
Publication Date: 2023.11.08 YOKOGAWA ELECTRIC CORP
  • EP4057095B1 patent drawingFigure 1
  • EP4057095B1 patent drawingFigure 2
  • EP4057095B1 patent drawingFigure 3

AI summary

[Problem] To accurately analyze a behavior of a plant. Provided is an analysis apparatus comprising: a variation model storage unit configured to store a plurality of variation models indicating variation in characteristics of a plant corresponding to an operating condition of the plant; a model extraction unit configured to acquire structure information indicating a structure model of an analysis target plant and to extract the variation model corresponding to the structure model; and an analysis unit configured to analyze the analysis target plant, based on the structure model of the analysis target plant and the variation model extracted by the model extraction unit.