HVAC Predictive Control for Energy-Aware Data Center Cooling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current HVAC systems in data centers operate inefficiently due to decentralized control methods that fail to account for energy costs and redundant units, leading to unnecessary energy waste and instability, especially in variable load conditions and complex airflow patterns.
Innovation Solution
A system utilizing sensors to measure environmental conditions and energy consumption, employing predictive and heuristic feedback control methods to optimize the operation of HVAC units, including the use of transfer models to minimize energy use while maintaining desired temperature and humidity levels, and incorporating a penalty function to account for costs and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If decentralized stand-alone controls are used for each HVAC unit, then each unit can independently control the temperature and humidity of air entering the unit, but redundant HVAC units must run at all times to ensure adequate cooling, wasting energy
Solution Approach 1:
The patent merges decentralized unit controls with a centralized coordination layer that shares temperature sensor data across all HVAC units. This allows the system to maintain independent unit operation while enabling cooperative decision-making to determine when redundant units can be safely turned off, resolving the contradiction between operational independence and energy efficiency.
Solution Approach 2:
The system implements feedback by having each HVAC unit's temperature sensor data shared with other units' controls. This feedback mechanism enables each unit to make informed decisions about whether to operate or shut down based on real-time temperature measurements from the entire data center, eliminating the need for redundant units to run unnecessarily.
2Loss of energy
If manual intervention is used to turn off redundant HVAC units to save energy, then energy consumption is reduced, but the risk of overheating equipment increases and the system becomes unstable under variable loads
Solution Approach 1:
The control system performs self-service by automatically determining which HVAC units can be safely turned off based on real-time temperature sensor data. The system monitors temperature conditions and autonomously makes shutdown decisions without manual intervention, ensuring energy savings while maintaining reliability through continuous monitoring and adaptive decision-making.
Solution Approach 2:
The system implements dynamic control that adapts to variable loads and changing temperature conditions. Rather than static on/off decisions, the controls continuously adjust unit operation based on real-time sensor feedback, enabling the system to safely reduce redundant unit operation while maintaining adequate cooling under varying load conditions.
3Loss of energy
If complex algorithms are attempted to optimize HVAC coordination, then energy efficiency may improve, but the system becomes unstable
Solution Approach 1:
The patent applies partial action by implementing a simplified coordination algorithm that focuses on the essential decision: whether to shut down a redundant HVAC unit based on temperature sensor data. Rather than attempting complex optimization of all system parameters, the system concentrates on this single critical function, achieving energy efficiency improvements while maintaining stability through straightforward, predictable control logic.
Data Source
AI summary
Methods, systems, and apparatuses are provided for controlling an environmental maintenance system that includes a plurality of sensors and a plurality of actuators. The operation levels of the actuators can be determined by optimizing a penalty function. As part of the penalty function, the sensor values can be compared to reference values. The optimized values of the operation levels can account for energy use of actuators at various operation levels and predicted differences of the sensor values relative to the reference values at various operation levels. The predicted difference can be determined using a transfer model. An accuracy of the transfer model can be determined by comparing predicted values to measured values. This accuracy can be used in determining new operational levels from an output of the transfer model (e.g., attenuating the output of the transfer model based on the accuracy).