Plant Control Segmentation for Faster Parameter Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing plant control systems face challenges in optimizing operation states due to a broad learning range and large number of parameters, leading to unrealistically long times for acquiring optimal values or non-convergence during learning.
Innovation Solution
A plant control supporting apparatus that selects segments for learning, defines reward functions, extracts and refines parameters, and uses a relationship model to perform learning, optimizing parameter values for each segment and the whole plant, thereby reducing the complexity of parameter adjustment and facilitating faster convergence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If learning is performed for all parameters in the entire plant, then comprehensive optimization is achieved, but learning time becomes unrealistically long or convergence fails
Solution Approach 1:
The plant is divided into multiple segments based on process flow and functional relationships. Learning is performed separately for each segment rather than for the entire plant at once. This segmentation reduces the complexity of the learning problem for each individual segment, enabling faster convergence while still achieving comprehensive optimization across the whole plant through iterative processing of all segments.
Solution Approach 2:
The system performs preliminary classification of parameters into four categories (variable parameters inside segment, monitor parameters inside segment, variable parameters outside segment, fixed parameters outside segment) before executing learning. This preliminary organization of parameters establishes a structured approach that guides the learning process, allowing the system to focus computational resources efficiently and achieve convergence in reasonable time.
2Loss of time
If the number of parameters to be learned is reduced, then learning time decreases, but optimization completeness may be compromised
Solution Approach 1:
By dividing the plant into segments and performing learning segment by segment, the system reduces the number of parameters learned simultaneously while ensuring that all parameters are eventually optimized. Each segment's learning focuses only on its local parameters, but the iterative process across all segments achieves global optimization completeness.
Solution Approach 2:
The system performs partial learning actions on individual segments rather than attempting to learn all parameters at once. This partial action approach allows the system to make progress on optimization without being overwhelmed by the full complexity, and through repeated iterations across segments, achieves complete optimization over time.
Data Source
AI summary
A plant control supporting apparatus includes a segment selector configured to select, from among a plurality of segments defined in a plant, a segment for which learning for acquiring an optimal value of at least one parameter representing an operation state is executed, a reward function definer configured to define a reward function used for the learning, a parameter extractor configured to extract at least one parameter that is a target for the learning in the selected segment on the basis of input and output information of a device used in the plant and segment information representing a configuration of a device included in the selected segment, and a learner configured to perform the learning for acquiring the optimal value for each segment on the basis of the reward function and the at least one parameter.