Maximum Time Constant Estimation for Stable Plant Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control algorithms for plants, such as HVAC systems, face challenges in estimating the time constant, especially in non-linear or higher order systems, leading to inadequate control inputs due to underestimation, which can result in instability and inefficient operation.
Innovation Solution
A controller system that calculates a maximum time constant by performing an open loop step test, monitoring the plant's response, and using a time constant estimator to determine a normalized variable and corresponding time constants, allowing for the generation of more effective control inputs based on the maximum time constant.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard time constant estimation methods are used, then the estimation process is simple, but the estimated time constant is inaccurate for non-linear or higher order systems
Solution Approach 1:
The patent segments the time constant estimation process into multiple discrete steps: (1) performing an open-loop step test to collect plant response data, (2) calculating a normalized variable from the response data, (3) computing multiple time constant estimates at different time points, and (4) determining the maximum time constant from these estimates. This segmentation transforms a single complex estimation into a systematic multi-step procedure that improves accuracy for non-linear systems.
Solution Approach 2:
The patent applies preliminary action by first performing an open-loop step test to excite the plant and collect response data before attempting time constant estimation. This preliminary data collection ensures that sufficient information is available to accurately characterize the plant's dynamic behavior, especially for non-linear systems where direct estimation from normal operation would be inaccurate.
2Ease of operation
If a single time constant is used, then the control algorithm is simple to implement, but the control performance deteriorates for higher order systems
Solution Approach 1:
The patent changes the time constant parameter from a single fixed value to a maximum time constant derived from multiple estimates at different time points. By identifying and using the maximum time constant (which corresponds to the slowest response mode), the control algorithm becomes more reliable for higher order systems while maintaining relative simplicity through the use of a single conservative parameter value for controller tuning.
Solution Approach 2:
The patent incorporates feedback by using the calculated maximum time constant to adjust and optimize control inputs. The time constant estimator continuously monitors plant response and feeds back the maximum time constant value to the control algorithm, enabling adaptive tuning that ensures stability while maintaining ease of operation through automated parameter adjustment.
3Speed
If the sampling interval is too short, then the control response is fast, but the energy consumption increases and the system becomes inefficient
Solution Approach 1:
The patent applies dynamics by making the sampling interval adaptive rather than fixed. By calculating the maximum time constant from plant response characteristics and using this to dynamically adjust the sampling interval, the system achieves optimal balance between response speed and energy efficiency. The sampling rate automatically adapts to the plant's actual dynamics, avoiding unnecessarily frequent sampling when the plant responds slowly.
Solution Approach 2:
The patent changes the sampling interval parameter based on the calculated maximum time constant. By setting the sampling interval as a function of the maximum time constant (e.g., a fraction of it), the control system automatically adjusts its operating parameters to match the plant's dynamics, achieving efficient energy usage while maintaining adequate control response speed.
Data Source
AI summary
A controller includes a communications interface configured to provide a control input to and receive feedback from a plant. The feedback is representative of a response of the plant to the control input over a response period. The controller further includes a time constant estimator. The time constant estimator calculates a normalized variable based on the feedback, each value of the normalized variable representative of the response of the plant at a different time during the response period. The time constant estimator calculates a plurality of time constant estimates, based on the plurality of values of the normalized variable. The time constant estimator determines a maximum time constant from the time constant estimates. The controller further includes a control input generator that generates the control input for the plant using the maximum time constant. The control input affects a variable state or condition of the plant.