Real-time modeling of heat distributions

Patent metadata

Field	Value
Publication number	US-9857235-B2
Application number	US-201313791140-A
Country	US
Kind code	B2
Filing date	Mar 8, 2013
Priority date	Mar 8, 2013
Publication date	Jan 2, 2018
Grant date	Jan 2, 2018

Abstract

Techniques for real-time modeling temperature distributions based on streaming sensor data are provided. In one aspect, a method for creating a three-dimensional temperature distribution model for a room having a floor and a ceiling is provided. The method includes the following steps. A ceiling temperature distribution in the room is determined. A floor temperature distribution in the room is determined. An interpolation between the ceiling temperature distribution and the floor temperature distribution is used to obtain the three-dimensional temperature distribution model for the room.

First claim

What is claimed is:

1. A method for creating a three-dimensional temperature distribution model for a room having a floor and a ceiling, the method comprising the steps of:
determining an energy balance in the room based on total heat input into the room and total heat extracted from the room;
calculating a mean temperature proximal to the ceiling using the energy balance to ensure that the mean temperature is consistent with a heat load in the room;
obtaining streaming real-time sensor data from a network of sensors located in the room, wherein the network of sensors comprises real-time power sensors, real-time pressure sensors, real-time air flow sensors, and real-time thermal sensors;
weighting the streaming real-time sensor data obtained from each of the real-time thermal sensors by a distance of each of the real-time thermal sensors from the ceiling to account for the real-time thermal sensors being at different heights in the room;
determining a first two-dimensional temperature distribution in the room proximal to the ceiling using the streaming real-time sensor data from a first set of the network of sensors;
scaling the first two-dimensional temperature distribution such that a mean of the first two-dimensional temperature distribution is equal to the mean temperature calculated using the energy balance thereby preserving the energy balance between the heat load in the room and temperature measured using the network of sensors;
determining a second two-dimensional temperature distribution in the room proximal to the floor using the streaming real-time sensor data from a second set of the network of sensors different from the first set; and
interpolating between the first two-dimensional temperature distribution and the second two-dimensional temperature distribution to obtain the three-dimensional temperature distribution model for the room which conforms to a principle of energy balance, wherein the room is a data center having a sub-floor plenum beneath the floor and wherein the three-dimensional temperature distribution model is used to implement cooling within the data center in a manner that improves energy efficiency.

2. The method of claim 1 , wherein cooled air is provided to the data center through the sub-floor plenum by way of perforated tiles in the floor.

3. The method of claim 2 , further comprising the step of:
obtaining power consumption data from equipment in the data center.

4. The method of claim 3 , wherein the power consumption data comprises data from power distribution units in the data center, the equipment in the data center which receive power through the power distribution units and the equipment in the data center which do not receive power through the power distribution units.

5. The method of claim 4 , wherein the equipment in the data center which receive power through the power distribution units comprise lighting in the data center, air conditioning units in the data center and coolant distribution units in the data center.

6. The method of claim 5 , wherein the air conditioning units in the data center each comprise a blower motor to circulate air through the air conditioning unit and to blow cooled air into the sub-floor plenum, the method further comprising the step of:
obtaining one or more of power data, blower setting data and air flow data for blower motors.

7. The method of claim 5 , wherein the data center further comprises a plurality of equipment racks, each of the equipment racks having an air inlet side and an air outlet side, the method further comprising the steps of:
obtaining real-time inlet air temperature data from one or more of the air inlet sides of the equipment racks;
obtaining real-time discharge air temperature data from one or more of the air conditioning units;
obtaining real-time discharge air flow data from one or more of the air conditioning units; and
obtaining real-time sub-floor plenum pressure data from the sub-floor plenum.

8. The method of claim 7 , wherein the real-time inlet air temperature data from one or more of the air inlet sides of the equipment racks and the real-time discharge air temperature data from one or more of the air conditioning units are obtained using the real-time thermal sensors.

9. The method of claim 7 , wherein the real-time discharge air flow data from one or more of the air conditioning units is obtained using the real-time air flow sensors.

10. The method of claim 7 , wherein the real-time sub-floor plenum pressure data from the sub-floor plenum is obtained using the real-time pressure sensors.

11. The method of claim 7 , further comprising the step of:
obtaining real-time outlet air temperature data from one or more of the air outlet sides of the equipment racks using virtual sensors.

12. The method of claim 7 , wherein the real-time inlet air temperature data from one or more of the air inlet sides of the equipment racks is used to determine the ceiling temperature distribution.

13. The method of claim 7 , further comprising the step of:
calculating a sub-floor plenum pressure distribution based on an inverse weighted distance of the real-time sub-floor plenum pressure data from the sub-floor plenum;
using the sub-floor plenum pressure distribution to determine air flow from each of the perforated tiles in the floor; and
using the air flow from each of the perforated tiles in the floor and an average temperature increase distribution in the data center to determine the floor temperature distribution.

14. The method of claim 13 , further comprising the step of:
applying a normalized Lorentzian field to each of the perforated tiles in the floor.

15. The method of claim 2 , further comprising the step of:
creating a convolution mask for the sub-floor plenum which indicates what portions of the sub-floor plenum comprise open space and what portions of the sub-floor plenum are blocked off.

16. The method of claim 15 , further comprising the step of:
obtaining real-time sub-floor plenum temperature data from the sub-floor plenum;
using the real-time sub-floor plenum temperature data to calculate a sub-floor plenum temperature distribution; and
applying the convolution mask for the sub-floor plenum to the sub-floor plenum temperature distribution to determine a mean sub-floor plenum temperature.

17. The method of claim 16 , wherein the sub-floor plenum temperature distribution is calculated based on the inverse weighted distance of the real-time sub-floor plenum temperature data.

18. The method of claim 16 , further comprising the step of:
using the mean sub-floor plenum temperature to determine a mean ceiling temperature based on a heat load in the data center and a total air flow in the data center attributable to the air conditioning units.

19. The method of claim 1 , further comprising the steps of:
creating a convolution mask for the room which indicates what portions of the room comprise open space and what portions of the room are blocked off; and
creating a convolution mask for the ceiling which indicates what portions of the ceiling comprise open space and what portions of the ceiling are blocked off.

20. An apparatus for creating a three-dimensional temperature distribution model for a room having a floor and a ceiling, the apparatus comprising:
a memory; and
at least one processor device, coupled to the memory, operative to:
determine an energy balance in the room based on total heat input into the room and total heat extracted from the room;
calculate a mean tempe

Assignees

• IBM

Inventors

• Hamann Hendrik F
• Li Hongfei
• Yarlanki Srinivas

Classifications

• G01K2213/00

  Temperature mapping · CPC title

• G01K1/02

  Means for indicating or recording specially adapted for thermometers · CPC title

• G01K7/42 Primary

  Circuits effecting compensation of thermal inertia; Circuits for predicting the stationary value of a temperature · CPC title