Energy-optimal control decisions for systems

Claim 3

The method of claim 1 , wherein setting the operation levels of the actuators based on the optimal change includes: adjusting the optimal change; and adding the adjusted optimal change to U.

Claim 4

The method of claim 2 , wherein adjusting the optimal change includes at least one of normalization, randomization, and scaling.

Claim 5

The method of claim 2 , further comprising: determining an accuracy of the transfer model by comparing predicted sensor value with measured sensor values, wherein the optimal change is adjusted based on an accuracy of the transfer model.

Claim 6

The method of claim 1 , wherein setting the operation levels of the actuators based on the optimal change includes: receiving new settings for the actuators; and adjusting the new settings based on the optimal change.

Claim 7

The method of claim 5 , further comprising: calculating the new settings using a PID object, wherein the PID object provides new settings based on: a comparison of a midpoint of the sensor values to a midpoint reference, or a comparison of an extremum of the sensor values to an extremum reference.

Claim 8

The method of claim 1 , wherein a maximum number of potential changes to the operation levels U are analyzed to determine the optimal change, and wherein the optimal change is selected from the plurality of potential changes if at least one of the corresponding potential changes has a value for the penalty function that is less than the penalty function with no change to U.

Claim 9

The method of claim 1 , wherein the cost of operating an actuator with a new operation level includes at least one of an energy cost and a maintenance cost.

Claim 10

The method of claim 8 , wherein the cost of operating the actuator includes: receiving a cost per rate of energy usage; and inputting the cost per rate of energy usage and the new operation level into a function that outputs the cost of operating the corresponding actuator.

Claim 11

The method of claim 1 , wherein using S New to determine the first contribution includes: for each sensor: inputting S New into a respective function that has at least one setpoint of the sensor as a parameter to obtain an output of a partial first contribution; and summing the partial first contributions from the sensors.

Claim 12

The method of claim 10 , wherein a respective function includes: taking a difference between the corresponding new sensor value S New and a setpoint; and multiplying the difference by a gain factor to obtain the partial contribution for the sensor.

Claim 13

The method of claim 1 , wherein determining a potential change dU to operation levels U of the actuators is accomplished by receiving the new operation levels U new , where U new =U+dU, and wherein applying a transfer model to dU to determine a predicted change dS in the sensor values S is accomplished by applying the transfer model to the potential new value for U.

Claim 14

The method of claim 1 , wherein determining a potential change dU to operation levels U of the actuators uses an optimization algorithm that incorporates past potential changes to determine the current potential change.

Claim 15

The method of claim 1 , wherein the setpoint values include a range of sensor values.

Claim 16

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

Claim 17

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; determining a potential change dU to operation levels U of the actuators; calculating a first contribution to a penalty function by: applying a transfer model to dU to determine a predicted change dS in the sensor values S, the transfer model providing a relation between changing an operation level of an actuator and resulting changes in sensor values; determining new sensor values S New from the predicted change dS; and using S New to determine the first contribution to the penalty function, the first contribution being based on a relationship of each new sensor value relative to one or more respective setpoint values; calculating a second contribution to the penalty function by: determining new operation levels U new corresponding to the potential change dU to the operation levels U; for each new operation level, determining a cost of operating the corresponding actuator with the new operation level; and aggregating the costs to obtain the second contribution; determining a first value of the penalty function based on the first and second contributions for the potential change dU; determining an optimal change to the operation levels U of the actuators based on values of the penalty functions for a plurality of potential changes to the operation levels U; and setting the operation levels of the actuators based on the optimal change.

Claim 18

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; determining a potential change dU to operation levels U of the actuators; calculating a first contribution to a penalty function by: applying a transfer model to dU to determine a predicted change dS in the sensor values S, the transfer model providing a relation between changing an operation level of an actuator and resulting changes in sensor values; determining new sensor values S New from the predicted change dS; and using S New to determine the first contribution to the penalty function, the first contribution being based on a relationship of each new sensor value relative to one or more respective setpoint values; calculating a second contribution to the penalty function by: determining new operation levels U new corresponding to the potential change dU to the operation levels U; for each new operation level, determining a cost of operating the corresponding actuator with the new operation level; and aggregating the costs to obtain the second contribution; determining a first value of the penalty function based on the first and second contributions for the potential change dU; determining an optimal change to the operation levels U of the actuators based on values of the penalty functions for a plurality of potential changes to the operation levels U; and setting the operation levels of the actuators based on the optimal change.

Claim 19

The computer product of claim 16 , wherein setting the operation levels of the actuators based on the optimal change includes: adjusting the optimal change; and adding the adjusted optimal change to U.

Claim 20

The computer product of claim 17 , wherein the instructions further comprise: determining an accuracy of the transfer model by comparing predicted sensor value with measured sensor values, wherein the optimal change is adjusted based on an accuracy of the transfer model.

Claim 21

The computer product of claim 16 , wherein setting the operation levels of the actuators based on the optimal change includes: receiving new settings for the actuators; adjusting the new settings based on the optimal change; and calculating the new settings using a PID object, wherein the PID object provides new settings based on: a comparison of a midpoint of the sensor values to a midpoint reference, or a comparison of an extremum of the sensor values to an extremum reference.

Claim 22

The computer product of claim 16 , wherein a maximum number of potential changes to the operation levels U are analyzed to determine the optimal change, and wherein the optimal change is selected from the plurality of potential changes if at least one of the corresponding potential changes has a value for the penalty function that is less than the penalty function with no change to U.

Claim 23

The computer product of claim 16 , wherein using S New to determine the first contribution includes: for each sensor: inputting S New into a respective function that has at least one setpoint of the sensor as a parameter to obtain an output of a partial first contribution; and summing the partial first contributions from the sensors.

Claim 24

The computer product of claim 16 , wherein determining a potential change dU to operation levels U of the actuators is accomplished by receiving the new operation levels U new , where U new =U+dU, and wherein applying a transfer model to dU to determine a predicted change dS in the sensor values S is accomplished by applying the transfer model to the potential new value for U.

Claim 25

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; determine a potential change dU to operation levels U of the actuators; calculate a first contribution to a penalty function by: applying a transfer model to dU to determine a predicted change dS in the sensor values S, the transfer model providing a relation between changing an operation level of an actuator and resulting changes in sensor values; determining new sensor values S New from the predicted change dS; using S New to determine the first contribution to the penalty function, the first contribution being based on a relationship of each new sensor value relative to one or more respective setpoint values; calculate a second contribution to the penalty function by: determining new operation levels U new , corresponding to the potential change dU to the operation levels U; for each new operation level, determining a cost of operating the corresponding actuator with the new operation level; aggregating the costs to obtain the second contribution; determine a first value of the penalty function based on the first and second contributions for the potential change dU; determine an optimal change to the operation levels U of the actuators based on values of the penalty functions for a plurality of potential changes to the operation levels U; and set the operation levels of the actuators based on the optimal change.

Claim 26

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; determine a potential change dU to operation levels U of the actuators; calculate a first contribution to a penalty function by: applying a transfer model to dU to determine a predicted change dS in the sensor values S, the transfer model providing a relation between changing an operation level of an actuator and resulting changes in sensor values; determining new sensor values S New from the predicted change dS; using S New to determine the first contribution to the penalty function, the first contribution being based on a relationship of each new sensor value relative to one or more respective setpoint values; calculate a second contribution to the penalty function by: determining new operation levels U new , corresponding to the potential change dU to the operation levels U; for each new operation level, determining a cost of operating the corresponding actuator with the new operation level; aggregating the costs to obtain the second contribution; determine a first value of the penalty function based on the first and second contributions for the potential change dU; determine an optimal change to the operation levels U of the actuators based on values of the penalty functions for a plurality of potential changes to the operation levels U; and set the operation levels of the actuators based on the optimal change.

Claim 27

The system of claim 23 , wherein setting the operation levels of the actuators based on the optimal change includes: adjusting the optimal change; and adding the adjusted optimal change to U.

Claim 28

The system of claim 24 , wherein the one or more processors are further configured to: determine an accuracy of the transfer model by comparing predicted sensor value with measured sensor values, wherein the optimal change is adjusted based on an accuracy of the transfer model.

Claim 29

The system of claim 23 , wherein setting the operation levels of the actuators based on the optimal change includes: receiving new settings for the actuators; adjusting the new settings based on the optimal change; and calculating the new settings using a PID object, wherein the PID object provides new settings based on: a comparison of a midpoint of the sensor values to a midpoint reference, or a comparison of an extremum of the sensor values to an extremum reference.

Claim 30

The system of claim 23 , wherein a maximum number of potential changes to the operation levels U are analyzed to determine the optimal change, and wherein the optimal change is selected from the plurality of potential changes if at least one of the corresponding potential changes has a value for the penalty function that is less than the penalty function with no change to U.

Claim 31

The system of claim 23 , wherein using S New to determine the first contribution includes: for each sensor: inputting S New into a respective function that has at least one setpoint of the sensor as a parameter to obtain an output of a partial first contribution; and summing the partial first contributions from the sensors.

Claim 32

The system of claim 23 , wherein determining a potential change dU to operation levels U of the actuators is accomplished by receiving the new operation levels U new , where U new =U+dU, and wherein applying a transfer model to dU to determine a predicted change dS in the sensor values S is accomplished by applying the transfer model to the potential new value for U.