Vapor Compression Control Using Time-Varying Extremum Seeking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional extremum-seeking controllers for vapor compression systems suffer from slow convergence rates, making real-time optimization of energy efficiency and performance metrics challenging due to reliance on averaging and quasi-steady control methods, which are inadequate for dynamic systems.
Innovation Solution
The proposed solution involves a recursive and concurrent estimation of time-varying gradients to update control signals, eliminating the need for averaging and enabling faster convergence by approximating parameters during transient control, allowing the system to track the true gradient and drive the vapor compression system to optimal operating points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional perturbation-based extremum seeking control is used, then the controller can achieve optimal steady state operating point, but the convergence rate is slow due to quasi-steady control and averaging requirements
Solution Approach 1:
The patent transitions from quasi-steady control to dynamic control by allowing the system to operate during transient states. The controller uses real-time gradient estimation without requiring the system to be in steady state, enabling faster adaptation to changing conditions and significantly reducing convergence time while maintaining optimization accuracy.
Solution Approach 2:
The patent eliminates the need for averaging over time windows by implementing continuous gradient estimation. The controller continuously updates control signals based on real-time gradient information, maintaining uninterrupted optimization action rather than relying on periodic averaging, which accelerates convergence while preserving accuracy.
2Use of energy by moving object
If model-based control methods are used, then energy efficiency can be optimized through mathematical models, but the models require extensive calibration and fail to capture installation-specific characteristics
Solution Approach 1:
The patent implements self-service by enabling the controller to automatically adapt to specific installations through real-time gradient estimation. Instead of requiring extensive manual calibration of mathematical models, the system autonomously learns the optimal control strategy for each specific installation context, eliminating calibration complexity while maintaining energy efficiency optimization.
Solution Approach 2:
The patent changes the approach from using fixed mathematical model parameters to dynamically estimating gradient parameters in real-time. This allows the control system to adapt to installation-specific characteristics without requiring pre-calibration of model parameters, simplifying the system while maintaining optimization performance.
3Measurement precision
If conventional ESC with averaging is used, then gradient estimation can be obtained, but the averaging process slows down the response to changing system conditions
Solution Approach 1:
The patent eliminates discontinuous averaging operations and implements continuous gradient estimation. The controller continuously computes gradient information from real-time measurements without requiring time-window averaging, maintaining accurate gradient estimation while enabling immediate response to changing system conditions, thus improving response speed without sacrificing accuracy.
Data Source
AI summary
A method for controlling a vapor compression system determines a value of a metric of performance of the vapor compression system using a previous value of an estimated parameter and a previous value of a control signal determined for a previous time step of the control. The values of the estimated parameter represent a relationship between values of the control signal and values of the metric of performance. A current value of the estimated parameter is determined based on the previous value of the estimated parameter and an error between the determined value of the metric of performance and a measured value of the metric of performance. Next, a current value of the control signal is determined based on the current value of the estimated parameter.