Normalized Performance Index for Feedback Control Setpoint Alarming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for calculating performance indices of feedback controllers in HVAC systems are cumbersome, as they depend on system-specific variables and are computationally expensive, making it difficult to compare performance across different systems or controllers.
Innovation Solution
A feedback control system that uses exponentially-weighted moving average (EWMA) statistics to generate normalized performance indices, allowing for real-time calculation without requiring batch data storage, and sets alarms based on control loop-agnostic thresholds derived from standard deviation and time constant ratios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional performance measures are used to quantify feedback controller performance, then the performance can be evaluated, but the measures depend on system-specific variables and cannot be directly compared across different systems or controllers
Solution Approach 1:
The patent transforms traditional performance measures by normalizing them using system-specific parameters (standard deviation of error signal, time constant of EWMA values) to create control loop agnostic indices. This parameter transformation enables direct comparison across different HVAC control loops while eliminating dependency on system-specific variables through normalization.
2Productivity
If traditional methods for calculating normalized performance indices are used, then normalized indices can be generated, but the calculation is computationally expensive and limited to offline batch analysis
Solution Approach 1:
The patent pre-calculates and stores the standard deviation of the error signal and the time constant of the EWMA values during normal operation. These pre-computed parameters are then reused for real-time calculation of normalized performance indices, avoiding repeated computationally expensive calculations and enabling real-time monitoring with reduced computational burden.
Solution Approach 2:
The patent uses exponentially-weighted moving average (EWMA) values as a simplified representation of the error signal characteristics. Instead of performing complex batch analyses of the entire error signal history, the system copies the essential statistical properties (standard deviation and time constant) into EWMA values that can be updated efficiently in real-time.
3Ease of operation
If setpoint error monitoring is implemented without normalization, then alarms can be generated, but the alarm thresholds must be system-specific and require knowledge of system parameters
Solution Approach 1:
The patent transforms alarm threshold setting from a system-specific process to a universal one by normalizing the setpoint error using the standard deviation and time constant. The normalized performance index allows the use of universal alarm thresholds (e.g., I1 < 0.5) that do not require system parameter knowledge, simplifying alarm configuration and operation across different HVAC control loops.
Data Source
AI summary
A plant includes equipment that operate to affect a variable state or condition of the plant in response to a control signal. A feedback controller monitors a process variable received via a feedback signal from the plant and uses an error signal representing a setpoint error between the process variable and a setpoint to generate the control signal for the plant. An exponentially-weighted moving average (EWMA) calculator generates EWMA values based on the setpoint error. A normalized index generator uses the EWMA values to calculate a normalized performance index. A threshold generator generates a threshold based on a first alarm parameter κ representing a normalized magnitude of the setpoint error and a second alarm parameter η representing a normalized duration that the setpoint error has persisted. An alarm manager generates an alarm in response to the normalized performance index crossing the threshold.