Air handling unit and method using reinforcement learning model that replicates model predictive control simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing building management systems (BMS) face inefficiencies in controlling edge equipment due to limited computing power and memory in edge controllers, leading to suboptimal control of parameters such as damper positions and fan speeds, which affects the achievement of temperature setpoints and energy efficiency.
Innovation Solution
Implementing a reinforcement learning model on edge controllers to generate mixed air temperature values and control damper positions, using inputs like temperature setpoints and weather forecasts, and updating the model based on indoor temperature differences and occupancy, trained to replicate model predictive control algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If advanced control approaches (e.g., model predictive control) are used to improve energy efficiency and temperature control, then energy usage is reduced and occupant comfort is improved, but computing power and memory requirements exceed the capabilities of edge controllers
Solution Approach 1:
The patent creates a simplified version (copy) of the complex model predictive control algorithm that can run on edge controllers with limited resources. This copy replicates the essential functionality of the advanced control approach but is optimized to operate within the computational constraints of edge devices, thereby enabling energy efficiency improvements without requiring excessive computing power
Solution Approach 2:
The control system is divided into two parts: a simplified control algorithm running on the edge controller for real-time decisions, and a more complex model predictive control algorithm running on a remote server for optimization. This segmentation allows the edge controller to function with limited computing power while still benefiting from advanced control strategies
2Manufacturing precision
If sophisticated control algorithms are implemented at the edge to achieve BMS temperature setpoints, then temperature control precision is improved, but device complexity increases beyond edge controller capabilities
Solution Approach 1:
A simplified copy of the sophisticated control algorithm is deployed to the edge controller, maintaining the essential temperature control precision while reducing the computational complexity to match edge controller capabilities
3Loss of energy
If real-time control adjustments are made to optimize damper positions and fan speeds, then energy efficiency is improved, but computational requirements exceed available memory and processing power
Solution Approach 1:
The patent deploys a streamlined version of the energy optimization algorithm that replicates the key functionality for controlling dampers and fans but is designed to operate within the memory constraints of edge controllers, enabling real-time energy efficiency improvements without exceeding available memory
Data Source
AI summary
A method includes generating a mixed air temperature value using a reinforcement learning model running on the edge controller. A temperature setpoint and a weather forecast are inputs to the reinforcement learning model. The method also includes controlling damper positions of an air handling unit to achieve the mixed air temperature value.