Neural Network Adaptive Controller for Exhaust EGR Error Minimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control devices for plants, such as exhaust purification systems, face errors in estimating physical quantities like EGR amount and EGR rate due to aging degradation and variability in solids, leading to inadequate control performance.
Innovation Solution
A control device that calculates estimated values using neural networks and adaptive inputs to minimize deviations between detected and estimated values, effectively addressing errors caused by aging and variability, and adjusts control variables accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a map or arithmetic expression established in advance is used to estimate physical quantities, then the control device can calculate estimated values based on detected physical quantities, but error occurs between the estimated value and actual value when aging degradation and variability in solids arise
Solution Approach 1:
The patent implements feedback by detecting the second physical quantity (e.g., exhaust air-fuel ratio) using a sensor and comparing the detected value with the estimated value calculated by the neural network. The deviation between these values is used to correct the adaptive input, creating a closed-loop system that continuously adjusts to maintain accuracy despite aging degradation and variability in solids.
Solution Approach 2:
The patent changes the parameter representation by using adaptive inputs that can dynamically adjust based on system conditions. Instead of fixed map values, the neural network processes variable adaptive inputs that are corrected based on feedback from actual sensor measurements, allowing the system to adapt to aging and variability.
2Ease of manufacture
If fixed algorithms are used for estimation, then the control system is simple to implement, but the estimated values become inaccurate due to system aging and variability
Solution Approach 1:
The patent transforms the static estimation system into a dynamic one by implementing adaptive inputs that change based on feedback. The neural network continuously adjusts its inputs based on the deviation between estimated and detected values, making the system adaptive to aging and variability while maintaining a relatively simple overall structure.
Solution Approach 2:
The system performs self-correction by using the deviation between estimated and detected values to automatically adjust its adaptive inputs. This self-service mechanism allows the system to maintain accuracy without requiring external recalibration or complex manual adjustments.
Data Source
AI summary
A controller for a plant that controls a controlled variable for the plant in accordance with estimated values, allowing to reduce any error in the estimated values that is caused by solid variation or aging of the plant. A controller for an exhaust emission control system has an estimated Inert-EGR value calculation section (711) to calculate the estimated value IEGRHAT for the Inert-EGR amount on the basis of an input vector U through a neural network, an estimated LAF sensor output value calculation section (712) to calculate the estimated value ΦHAT for an exhaust air-fuel ratio correlating with the Inert-EGR amount on the basis of the input vector U through the neural network, an LAF sensor (34) to detect the exhaust air-fuel ratio, and a nonlinear adaptive corrector (713) to calculate the adaptive input UVNS such that the estimated error EHAT between the detected value ΦACT from the LAF sensor (34) and the estimated output value ΦHAT of the LAF sensor (34) is minimized.