Building management system with saturation detection and removal for system identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for saturation detection and removal in HVAC systems are inadequate as they assume nonlinearity only occurs in inputs or outputs, failing to handle inherent dynamic nonlinearity, leading to inefficiencies in system identification and control.
Innovation Solution
A building management system with a saturation detector that identifies non-transient regions based on error calculations and deadband analysis, allowing for the detection and removal of saturation periods, enabling the generation of linear models for system identification and online control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If nonlinear system identification is used to capture model dynamics under saturation, then model accuracy is improved, but computational complexity increases making it too expensive for on-line optimization based control
Solution Approach 1:
The patent segments the operating range into saturated and non-saturated regions, and segments the identification process into offline (nonlinear) and online (linear) phases. This allows capturing nonlinear dynamics during offline identification while using computationally efficient linear models for online control optimization.
Solution Approach 2:
The patent performs saturation detection and removal as a preliminary action before system identification. By pre-processing the data to remove saturated segments, the subsequent linear system identification can proceed without dealing with nonlinear saturation effects, achieving both accuracy and computational efficiency.
2Ease of operation
If conventional saturation detection methods are used, then detection simplicity is improved, but ability to handle inherent dynamic nonlinearity deteriorates
Solution Approach 1:
The patent introduces dynamic thresholds for saturation detection that adapt to changing operating conditions. The thresholds are calculated based on the local characteristics of the data, allowing the detection method to handle inherent dynamic nonlinearity while maintaining operational simplicity through automated adaptation.
Solution Approach 2:
The patent employs feedback mechanisms where the detected saturation information is used to improve subsequent detection accuracy. The system continuously monitors and adjusts detection parameters based on observed system behavior, enabling effective handling of dynamic nonlinearity without increasing operational complexity.
Data Source
AI summary
A building management system includes building equipment, a sensor, and a saturation detector. The building equipment is configured to operate at an operating capacity to drive a variable state or condition of a building zone toward a setpoint. The operating capacity and the setpoint vary over time. The sensor is in the building zone and is configured to provide a zone measurement of the variable state or condition of the building zone. The saturation detector is configured to determine whether the operating capacity is in a non-transient region for a threshold amount of a time period upon determining that an error for the building zone exists for the time period, and, in response to a determination that the operating capacity is in the non-transient region for at least the threshold amount of the time period, indicate the time period as a saturation period.