AI Forecasting of Industrial Transient States for Automatic Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automation systems in industrial processes struggle to effectively monitor and control complex processes due to the large volume of operational data, often requiring human intervention, which can lead to inappropriate actions that exacerbate deviations or cause new deviations.
Innovation Solution
A system and method for predicting parameter values, plant states, and alarms using AI-based parameter and plant state forecast models trained on historical data, allowing for the prediction of steady and transient states, and enabling automatic adjustments to mitigate unplanned transient states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human operators monitor and control industrial processes, then appropriate control actions can be taken when deviations occur, but inexperienced operators may take inappropriate actions that worsen deviations or cause new deviations
Solution Approach 1:
The system enables self-service by implementing automated monitoring and control that performs functions independently of human operators. The automated system continuously monitors process parameters, detects deviations, and executes control actions without requiring human intervention, thereby eliminating the reliability issues associated with operator experience while maintaining continuous process control
Solution Approach 2:
The system replaces the mechanical human decision-making process with an automated computational system. Instead of relying on human operators to analyze process data and make control decisions, the patent implements an automated system that uses algorithms to monitor parameters, detect deviations, and execute control actions, substituting human cognitive functions with mechanical/computational processes
2Productivity
If current automation systems monitor industrial processes, then continuous monitoring is achieved, but the systems are not able to take all necessary actions when deviations occur and human interaction is needed
Solution Approach 1:
The system performs preliminary action by continuously monitoring process parameters and detecting potential deviations before they escalate into significant problems. The automated system is prepared in advance with predefined control strategies, enabling it to immediately execute appropriate control actions when deviations are detected, eliminating the delay associated with human response
Solution Approach 2:
The system implements feedback by continuously monitoring process parameters and using this information to automatically adjust control actions. The automated system creates a closed-loop control mechanism where process measurements are fed back to the control algorithm, which then modifies control outputs to maintain process stability, achieving both continuous monitoring and automatic control
3Reliability
If experienced operators manage industrial processes, then appropriate control actions are taken timely, but human intervention is required which reduces productivity
Solution Approach 1:
The system replaces human operators with an automated computational system that processes process data and executes control actions. This substitution eliminates the time required for human perception, analysis, and decision-making, enabling instantaneous response to process deviations while maintaining or improving control accuracy through consistent algorithmic application
Solution Approach 2:
The system ensures continuity of useful action by implementing uninterrupted automated monitoring and control. Unlike human operators who require breaks and are subject to fatigue, the automated system operates continuously without interruption, maintaining constant surveillance of process parameters and executing control actions at any time, thereby improving both reliability and productivity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A plurality of tags each identify a corresponding process parameter of an industrial process and historical values for the process parameter. A parameter forecast model is trained for each of the parameters that are identified by the plurality of tags, wherein each of the parameter forecast models is trained based at least in part on the received historical values for at least some of the parameters that are identified by the plurality of tags. A forecasted parameter value is generated for each of the parameters identified by the plurality of tags based at least in part on the corresponding parameter forecast model that corresponds to the respective parameter. A plurality of forecasted steady state periods and a plurality of forecasted transient state periods are predicted for each of one or more of the identified parameters.