Neural Network Gas Estimation for Coal Boiler Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control devices for thermal power plants struggle to maintain optimal combustion efficiency and reduce CO and NOx emissions as plant characteristics change over time, due to mismatched model and actual plant characteristics, and errors in measurement data used for neural network modeling.
Innovation Solution
A control device that calculates manipulation signals using a combination of learning signals and correction signals, based on the deviation between measured and target values, to adjust for changes in plant characteristics and reduce estimation errors in neural network models, thereby maintaining control characteristics and improving estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a model is used to simulate plant characteristics for control, then control characteristics can be maintained, but estimation errors increase when plant characteristics change over time
Solution Approach 1:
The patent applies dynamics by making the model adaptable to changing plant characteristics. The learning section continuously updates the model parameters based on actual plant behavior, transforming a static model into a dynamic one that evolves with the plant. This resolves the contradiction by allowing the model to maintain control characteristics while adapting to prevent estimation errors when plant characteristics change.
Solution Approach 2:
The patent implements feedback through the learning section that compares model outputs with actual plant measurements and uses this information to correct model parameters. This feedback mechanism ensures that the model remains accurate over time, resolving the contradiction between maintaining control characteristics and preventing estimation errors by continuously aligning the model with actual plant behavior.
2Adaptability or versatility
If measurement data is used for neural network modeling, then control adaptation is enabled, but errors in measurement data reduce estimation accuracy
Solution Approach 1:
The patent introduces an intermediary correction mechanism between the measurement data and the neural network model. The learning section acts as a mediator that processes measurement data, identifies errors, and applies corrections to both the model parameters and the manipulation signals. This intermediary process enables control adaptation while compensating for measurement errors, thus resolving the contradiction.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting model parameters based on the quality and characteristics of measurement data. When measurement errors are detected, the system modifies the parameters used in estimation and control, allowing the system to adapt to changing conditions while maintaining estimation accuracy despite the presence of erroneous measurement data.
3Adaptability or versatility
If learning algorithms are used to adapt control to plant changes, then control flexibility improves, but model-plant mismatch increases estimation errors
Solution Approach 1:
The patent uses feedback through the learning section that continuously monitors the difference between model predictions and actual plant behavior. This feedback loop enables the system to adapt control flexibly while simultaneously correcting estimation errors by adjusting model parameters based on observed discrepancies, thus resolving the contradiction between adaptability and estimation accuracy.
Solution Approach 2:
The patent replaces the traditional mechanical approach of fixed model parameters with an intelligent system that uses learning algorithms to dynamically adjust parameters. This substitution allows the system to achieve both control flexibility and maintained estimation accuracy by using software-based adaptation rather than fixed mechanical parameters, resolving the contradiction between flexibility and accuracy.
Data Source
AI summary
A gas concentration estimation device of a coal-burning boiler adapted to estimate the concentration of the gas component included in an exhaust gas emitted from a coal-burning boiler using a neural network, including: a process database section adapted to store process data of a coal-burning boiler; a filtering processing section adapted to perform filtering processing for extracting data suitable for learning of a neural network from the process data stored in the process database section; a neural-network learning processing section adapted to perform learning processing of the neural network based on the data extracted by the filtering processing section and suitable for learning of the neural network; and a neural-network estimation processing section adapted to perform estimation processing of the CO concentration or the NOx concentration in the exhaust gas emitted from the coal-burning boiler based on the learning processing of the neural-network learning processing section.