Building management system with system identification using multi-step ahead error prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing HVAC control systems face challenges in accurately modeling complex, nonlinear building environments, leading to inefficiencies in heating and cooling processes due to the lack of precise system identification and predictive control.
Innovation Solution
A building management system that includes sensors, a controller, and a model predictive control algorithm, which generates training data, identifies system parameters through a prediction error minimization process, and adjusts control inputs to optimize temperature regulation by using a multi-step ahead prediction error method.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional HVAC control systems are used, then the system structure is simple, but the modeling accuracy of complex nonlinear building environments is poor
Solution Approach 1:
The patent replaces traditional mechanical control approaches with data-driven system identification and model predictive control algorithms. By using multi-step ahead prediction error methods and iterative optimization processes, the system achieves accurate modeling of nonlinear building environments without relying on complex physical models, thus improving modeling accuracy while managing system complexity through computational methods.
2Manufacturing precision
If system identification is not performed, then the control system is simple to implement, but the control precision for temperature regulation is insufficient
Solution Approach 1:
The patent performs system identification and model parameter estimation as preliminary actions before implementing control. By pre-determining the system model parameters through offline system identification using measured input-output data, the complex modeling work is completed in advance, allowing the actual control implementation to use the pre-established model for precise temperature regulation without requiring complex real-time computations.
Solution Approach 2:
The patent employs iterative feedback optimization in the system identification process, where the prediction error is continuously minimized by adjusting model parameters based on the difference between predicted and actual system outputs. This feedback mechanism ensures high control precision by refining the system model to accurately represent the building's thermal dynamics.
3Reliability
If single-step prediction is used, then the computational complexity is low, but the predictive control performance for dynamic building conditions is insufficient
Solution Approach 1:
The patent extends the prediction horizon from single-step to multi-step ahead prediction, adding a temporal dimension to the control approach. By predicting system outputs over multiple future time steps simultaneously using recursive prediction formulas, the system captures dynamic building conditions and thermal inertial effects that single-step prediction misses, thereby improving predictive control performance while managing computational complexity through efficient recursive calculations.
Data Source
AI summary
A building management system includes a controller configured to control building equipment by providing a control input to the building equipment for each of the plurality of time steps and generate a set of training data for a system model for the building. The training data includes input training data and output training data for each of the plurality of time steps. The controller is further configured to perform a system identification process to identify parameters of the system model. The system identification process includes predicting, for each time step, a predicted value for one or more of the output variables for each of a plurality of subsequent time steps, generating a prediction error function by comparing the output training data to the predicted values, and optimizing the prediction error function to determine values for the parameters of the system model that minimize the prediction error function.