Plant Operation Control Using Simulated Dangerous Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current plant control systems in industrial settings face challenges in ensuring safety and efficiency, often requiring human intervention that can lead to errors and increased costs, while AI systems may not adequately secure plant safety without direct operator intervention.
Innovation Solution
A plant control device and method that uses machine learning, specifically a support vector machine, to model and predict dangerous operation conditions, determining optimal operation parameters to prevent safety hazards and automate instruction outputs, thereby reducing human error and operational costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human operators (board operators and field operators) are used to monitor and control plant operations, then operational flexibility and safety checks are improved, but human error and operational costs increase
Solution Approach 1:
The plant control system performs self-monitoring and self-control through automated acquisition of operation conditions, automatic determination of operation parameters, and automated output of operation instructions. The system serves itself by replacing human operators with autonomous computational processes that continuously optimize plant operations without human intervention.
Solution Approach 2:
The patent replaces the mechanical human operator system with an automated computational system. Instead of human perception and decision-making, the system uses automatic acquisition means to sense plant conditions, determination means to compute optimal parameters, and output means to execute control instructions, thereby eliminating human error and reducing operational costs.
2Productivity
If AI systems are used to determine operation parameters, then operational costs are reduced, but plant safety may be compromised due to inadequate safety checks
Solution Approach 1:
The automated system continuously acquires current operation conditions, compares them against learned operation models, and adjusts operation parameters in real-time based on feedback from the plant's actual state. This closed-loop control ensures safety by constantly monitoring and correcting deviations from safe operating parameters.
Solution Approach 2:
The system performs preliminary determination of operation parameters before actual plant operations proceed. By pre-calculating safe and optimal operation parameters based on learned models, the system prevents unsafe conditions from occurring rather than reacting to them after the fact.
3Extent of automation
If manual operation instructions are output by AI systems, then automation benefits are achieved, but direct operator intervention is required which reduces efficiency
Solution Approach 1:
The system completes the automation loop by having the output means directly implement operation instructions based on determination results, without requiring operator confirmation or manual input. The system serves itself entirely, from sensing conditions to executing control actions, maximizing both automation extent and operational efficiency.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
A plant control device of the present invention includes a register configured to register a simulated dangerous condition which is a simulated representation of an operation condition under which a plant is dangerous, a first acquirer configured to acquire an operation condition of the plant, a learner configured to learn the operation condition acquired and the simulated dangerous condition registered and produce an operation model of the plant, a determiner configured to determine an operation parameter of the plant on the basis of the operation condition acquired and the operation model produced, and an instructor configured to instruct an operation of the plant on the basis of the operation parameter determined.