Accuracy-optimal control decisions for systems

• US 9,291,358 B2
• Filed: 12/12/2014
• Issued: 03/22/2016
• Est. Priority Date: 08/20/2010
• Status: Active Grant

• Current Assignee: Vigilent Corporation
• Original Assignee: Vigilent Corporation
• Inventors: Federspiel, Clifford C., Chin, Jerry
• Primary Examiner(s): Jarrett, Ryan
• Application Number:
  US14/568,965
• Publication Number:
  US 20150100165A1
• Time in Patent Office: 466 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: F24F 11/30 for purposes related to the operation of the system, e.g. for safety or monitoring
• CPC Class Codes: F24F 11/46 Improving electric energy efficiency or saving
• CPC Class Codes: F24F 11/58 using Internet communication
• CPC Class Codes: F24F 11/62 characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
• CPC Class Codes: F24F 11/63 Electronic processing
• CPC Class Codes: F24F 11/77 by controlling the speed of ventilators
• CPC Class Codes: G05B 13/026 using a predictor
• CPC Class Codes: G05D 23/1934 each space being provided with one sensor acting on one or more control means
• CPC Class Codes: H05K 7/1498 Resource management, Optimisation arrangements, e.g. configuration, identification, tracking, physical location thermal management H05K7/20836
• CPC Class Codes: H05K 7/20836 Thermal management, e.g. server temperature control

Abstract

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).

Claims

1. A method of controlling an environmental maintenance system that includes a plurality of actuators and a plurality of sensors, each sensor measuring a value of a physical condition of an environment, the method comprising:

receiving sensor values S corresponding to the sensors, the sensor values S corresponding to operational levels U₁ of the actuators;

receiving new operational levels U₂ of the actuators, where U₂−

U₁=dU;

predicting a change dS in the sensor values S, the change dS corresponding to the change dU, wherein the prediction is performed in accordance with a transfer model;

operating the actuators with the new operational levels U₂;

subsequently, measuring new sensor values S_new;

repeating the receiving, prediction, operation, and subsequent measurement at a plurality of times;

determining a metric of an accuracy of the transfer model by comparing the predicted sensor values to the measured sensor values;

receiving a proposed change to the operation levels of the actuators, the proposed change being based on the transfer model;

based on the accuracy metric, adjusting the proposed change to the operation levels of the actuators; and

setting the operation levels of the actuators based on the adjustment to the proposed change.

2. The method of claim 1, wherein determining the accuracy metric includes:

storing the predicted sensor value and the measured sensor values;

for each sensor;

calculating a difference between the predicted sensor value and the measured sensor value;

aggregating the differences; and

determining the metric based on a magnitude of the aggregated differences.

3. The method of claim 2, wherein the predicted sensor values are stored as differences in predicted sensor values, and wherein the measured sensor values are stored as differences in measured sensor values.

4. The method of claim 2, wherein calculating a difference between a predicted sensor value and a measured sensor value includes scaling and/or offsetting at least one of the values before calculating the difference.

5. The method of claim 2, wherein the accuracy metric is determined using the Pearson correlation coefficient.

6. The method of claim 1, wherein the proposed change is determined from an optimization of a penalty function that includes:

a first contribution that includes a difference between predicted sensor values and setpoint values; and

a second contribution that includes a cost associated with operation levels resulting from the proposed change.

7. The method of claim 1, wherein measuring new sensor values S_new is performed at a predetermined time after starting to operate the actuators with the new operational levels U₂.

8. The method of claim 1, wherein predicting a change dS in the sensor values S is accomplished by multiplying dU by the transfer model.

9. The method of claim 1, wherein receiving new operational levels U₂ of the actuators is accomplished by receiving dU and U₁.

10. The method of claim 1, wherein adjusting the proposed change to the operation levels of the actuators includes:

specifying a range of operation of the actuators; and

adjusting the proposed change based on the range.

11. The method of claim 10, wherein the range limits the operation to a smaller range than full operability, wherein the proposed change is a percentage of operation, and wherein the adjusting modifies the percentage based on the limited range.

12. The method of claim 1, wherein the transfer model is a matrix of numerical values.

13. The method of claim 12, wherein the proposed change is determined by inverting the transfer matrix and multiplying the inverted transfer matrix by a difference of sensor values and setpoint values.

14. A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to control an environmental maintenance system that includes a plurality of actuators and a plurality of sensors, each sensor measuring a value of a physical condition of an environment, the instructions comprising:

receiving sensor values S corresponding to the sensors, the sensor values S corresponding to operational levels U₁ of the actuators;

receiving new operational levels U₂ of the actuators, where U₂−

U₁=dU;

predicting a change dS in the sensor values S, the change dS corresponding to the change dU, wherein the prediction is performed in accordance with a transfer model;

operating the actuators with the new operational levels U₂;

subsequently, measuring new sensor values S_new;

repeating the receiving, prediction, operation, and subsequent measurement at a plurality of times;

determining a metric of an accuracy of the transfer model by comparing the predicted sensor values to the measured sensor values;

receiving a proposed change to the operation levels of the actuators, the proposed change being based on the transfer model;

based on the accuracy metric, adjusting the proposed change to the operation levels of the actuators; and

setting the operation levels of the actuators based on the adjustment to the proposed change.

15. The computer product of claim 14, wherein determining the accuracy metric includes:

storing the predicted sensor value and the measured sensor values;

for each sensor;

calculating a difference between the predicted sensor value and the measured sensor value;

aggregating the differences; and

determining the metric based on a magnitude of the aggregated differences.

16. The computer product of claim 14, wherein the proposed change is determined from an optimization of a penalty function that includes:

a first contribution that includes a difference between predicted sensor values and setpoint values; and

a second contribution that includes a cost associated with operation levels resulting from the proposed change.

17. The computer product of claim 14, wherein adjusting the proposed change to the operation levels of the actuators includes:

specifying a range of operation of the actuators; and

adjusting the proposed change based on the range.

18. An environmental maintenance system comprising:

a plurality of actuators;

a plurality of sensors, each sensor measuring a value of a physical condition of an environment; and

one or more processors configured to;

receive sensor values S corresponding to the sensors, the sensor values S corresponding to operational levels U₁ of the actuators;

receive new operational levels U₂ of the actuators, where U₂−

U₁=dU;

predict a change dS in the sensor values S, the change dS corresponding to the change dU, wherein the prediction is performed in accordance with a transfer model;

operate the actuators with the new operational levels U₂;

subsequently, measure new sensor values S_new;

repeat the receiving, prediction, operation, and subsequent measurement at a plurality of times;

determine a metric of an accuracy of the transfer model by comparing the predicted sensor values to the measured sensor values;

receive a proposed change to the operation levels of the actuators, the proposed change being based on the transfer model;

based on the accuracy metric, adjust the proposed change to the operation levels of the actuators; and

set the operation levels of the actuators based on the adjustment to the proposed change.

19. The environmental maintenance system of claim 18, wherein the one or more processors configured to determine the accuracy metric by:

storing the predicted sensor value and the measured sensor values;

for each sensor;

calculating a difference between the predicted sensor value and the measured sensor value;

aggregating the differences; and

determining the metric based on a magnitude of the aggregated differences.

20. The environmental maintenance system of claim 18, wherein the proposed change is determined from an optimization of a penalty function that includes:

a first contribution that includes a difference between predicted sensor values and setpoint values; and

a second contribution that includes a cost associated with operation levels resulting from the proposed change, and

wherein the one or more processors configured to adjust the proposed change to the operation levels of the actuators by;

specifying a range of operation of the actuators; and

adjusting the proposed change based on the range.