Method and apparatus for efficiently coordinating data center cooling units

• US 8,634,962 B2
• Filed: 08/20/2010
• Issued: 01/21/2014
• Est. Priority Date: 08/21/2009
• Status: Expired due to Fees

• Current Assignee: Vigilent Corporation
• Original Assignee: Vigilent Corporation
• Inventors: Federspiel, Clifford C., Federspiel, Kurt, Chin, Jerry
• Primary Examiner(s): Schechtman, Sean
• Application Number:
  US12/860,820
• Publication Number:
  US 20110203785A1
• Time in Patent Office: 1,250 Days
• Field of Search: 700276-278, 700299-300
• US Class Current: 700/276
• CPC Class Codes: G05D 22/02 characterised by the use of electric means
• CPC Class Codes: G05D 23/1928 sensing the temperature of one space
• CPC Class Codes: H05K 7/20745 within rooms for removing heat from cabinets, e.g. by air conditioning device
• CPC Class Codes: H05K 7/20836 Thermal management, e.g. server temperature control

Abstract

Systems, apparatus, and methods for controlling environmental maintenance modules (e.g. HVAC units) using sensors are provided. Values measured by the sensors can be used to determine a change in operation levels of the modules to keep the sensor values within a desired range. For example, a stopped module can be increased or started for more cooling when a sensor temperature is too hot. The module predicted to have the greatest effect on the temperature of hot sensor can be started. A transfer matrix, which provides a relation between a change in operation level and resulting sensor changes, can be used to perform the above predictions. As another example, a module can be stopped (or otherwise have an operation level decreased), if the sensor temperatures are within range, and the decrease in operation level is predicted not to cause an out-of-range condition.

Claims

1. A method of controlling an environmental maintenance system including a plurality of environmental maintenance modules and a plurality of sensors that measure a value of a physical condition of an environment, the method comprising:

identifying a sensor S_indx whose sensor value is measured to be outside of a range;

for each of a plurality of the modules;

identifying a change in value of an operation level for the module,

using the change in value of the operation level as an input to a transfer function, and

obtaining a respective predicted change in the sensor value of the sensor S_indx as an output of the transfer function, the respective predicted change in the sensor value corresponding to the identified change in the operation level for the module;

selecting at least one module based on the respective predicted change in the sensor value, wherein the respective predicted change in the sensor value of the selected at least one module satisfies a criterion; and

changing the operation level of the selected at least one module to effectuate a change of the sensor value of sensor S_indx.

2. The method of claim 1, wherein the change of operation level is starting or stopping the module.

3. The method of claim 1, wherein obtaining the predicted change in the sensor value for a first module includes determining a predicted change in the sensor value for a change in operation level of each of a plurality of actuators corresponding to the first module.

4. The method of claim 1, wherein the change of operation level is increasing an operation level of the selected module, wherein the selected module was previously running before the change, and wherein the measured sensor value is outside of a range when the measured sensor value is above a threshold value or below a threshold value.

5. The method of claim 1, wherein the module with the largest predicted change value has its operation level changed.

6. The method of claim 1, wherein selecting a module includes:

identifying the predicted change values that are greater than a change threshold; and

selecting one of the modules whose predicted change value is greater than the change threshold.

7. The method of claim 6, wherein if at least one of the modules has a predicted change value greater than the change threshold and has a variable fan speed, then a module with a variable fan speed is selected for changing the operation level.

8. The method of claim 6, wherein selecting one of the modules is performed randomly.

9. The method of claim 1, wherein the respective predicted change in the sensor value of the sensor S_indx obtained as the output of the transfer function corresponds to Y_post−

Y_pre, which is determined as Y_post−

Y_pre=TM(S_indx,P_indx)*Δ

level_est(P_indx) where Y_post is the estimated sensor value of the identified sensor S_indx after changing the operation level of module P_indx, Y_pre is the sensor value of the identified sensor S_indx before changing the operation level of module P_indx, Δ

level_est(P_indx) is an estimated amount of change in the operation level of the module P_indx to be enacted, and wherein TM is a transfer matrix that provides a relation between changing an operation level of a module and resulting changes in sensor values.

10. The method of claim 9, wherein the change of the operation level of module P_indx is a start, wherein N modules are running prior to module S_indx being started, and wherein the estimated operation level of the module P_indx after starting is an average of the operation levels of the operating modules times N/(N+1).

11. The method of claim 9, further comprising updating the transfer matrix TM by:

measuring first sensor values Y1 of the sensors and determining the operation level of the selected at least one module before changing at least one operation level;

after changing the at least one operation level of the selected at least one module, measuring second sensor values Y2 of the sensors and determining the operation level of the selected at least one module;

calculating

12. The method of claim 11, wherein the updated TM is obtained from TM_NEW and the old TM using the equation

TM(i,j)_updated=g*TM(i,j)_new+(1−

g)*TM(i,j)_old,

where g is between 0 and 1.

13. The method of claim 1, further comprising preventing changing an operation level of another module until a predetermined amount of time elapses after the selected module has its operation level changed or until a quasi-steady state condition is achieved after the selected module has its operation level changed.

14. The method of claim 1, wherein obtaining the predicted change in the sensor value uses a transfer matrix, the method further comprising initializing the transfer matrix by:

receiving a first sensor value Y1 for each of the sensors;

changing an operation level of at least two actuators of the modules at a same time;

for each sensor;

after the change in operation levels, receiving a second sensor value Y2; and

calculating a difference between the first sensor value Y1 and the second sensor value Y2; and

determining at least a portion of matrix elements of the transfer matrix based on the differences, the transfer matrix having dimensions of a number of actuators by a number of sensors, wherein the portion includes matrix elements associated with the at least two actuators, and wherein determining a matrix element for sensor S and actuator P includes;

determining a first ratio having a numerator including the difference for sensor S and having a denominator including the change in operation level for actuator P; and

combining the first ratio with at least one other ratio, wherein the at least one other ratio is determined from a previous change of the operation level of actuator P that resulted in a previous difference in the sensor values for sensor S.

15. The method of claim 14, wherein combining the first ratio with at least one other ratio includes using a recursive least squares.

16. The method of claim 14, wherein the change in operation level of an actuator of a module is starting or stopping the respective module.

17. The method of claim 14, wherein combining the first ratio with at least one other ratio includes:

determining an error of the calculated difference relative to a predicted change in the sensor value for sensor S due to the change in operation level of actuator P; and

adding the error to the existing value for the matrix element for sensor S and actuator P, wherein the existing value was determined from the previous difference and previous change in the operation level of actuator P.

18. The method of claim 17, further comprising:

multiplying the error by a learning factor prior to the adding, wherein the learning factor is determined from a covariance matrix.

19. The method of claim 1, wherein determining the predicted change value uses a transfer matrix, the method further comprising initializing the transfer matrix by:

at least one processor determining the transfer matrix TM by;

for each one of the modules;

receiving a first sensor value Y1 for each of the sensors, each first sensor value being received when the respective module has a first operation level;

stopping or starting the respective module; and

after stopping or starting the respective module, for each sensor;

receiving a second sensor value Y2;

calculating a difference between the first sensor value Y1 and the second sensor value Y2; and

calculating a TM matrix element using the difference,

wherein the transfer matrix TM provides a relation between starting and/or stopping a module and resulting changes in the sensor values for each sensor.

20. The method of claim 19, wherein the matrix element equals (Y2−

Y1) divided by a measure of capacity.

21. The method of claim 20, further comprising:

receiving a return temperature TR_s and a discharge temperature TD_s for each module P when all modules are running, wherein the operation level is a capacity measured by

22. The method of claim 19, further comprising:

calculating a load matrix by;

for each respective module P_indx;

for each other module C_indx, calculating a percent change in capacity (Δ

% Cap) that results from changing the operation level of the respective module P_indx; and

using each percent change to determine a load matrix element LOAD(C_indx,P_indx),

wherein the load matrix provides a measure of the effect of changing the operation level of a module on the capacity of use of the remaining modules that remain on.

23. The method of claim 22, wherein

24. The method of claim 23, where

26. An environmental maintenance system comprising:

a plurality of environmental maintenance modules;

a plurality of sensors that measure a value of a physical condition of an environment;

the computer program product of claim 14; and

at least one processor.

25. A computer program product comprising a non-transitory computer readable medium storing a plurality of instructions for controlling a processor to perform an operation for controlling an environmental maintenance system including a plurality of environmental maintenance modules and a plurality of sensors that measure a value of a physical condition of an environment, the instructions comprising:

identifying a sensor S_indx whose sensor value is measured to be outside of a range;

for each of at least a portion a plurality of the modules;

identifying a change in value of an operation level for the module,

using the change in value of the operation level as an input to a transfer function,

obtaining a respective predicted change in the sensor value of the sensor S_indx as an output of the transfer function the respective predicted change in the sensor value corresponding to the identified change in the operation level for the module;

selecting at least one module based on the respective predicted change in the sensor value; and

changing the operation level of the selected at least one module to effectuate a change of the sensor value of sensor S_indx.