Predictive control for domestic heating system

1. A method comprising:

obtaining an outside temperature forecast for an area outside of a structure, wherein the outside temperature forecast includes a period of time following a present time;

determining a sequence of boiler set points for a boiler using a model predictive control (MPC) model for the structure using inverse modeling, wherein the MPC model comprises:

unknown parameters of elements of a dynamic heat loss model,

wherein the unknown parameters are constrained within a predetermined expected range of the elements of the dynamic heat loss model, and

wherein gas consumption data from a gas meter for the boiler is used to estimate the unknown parameters,

a first control algorithm that accepts the outside temperature forecast as an input and generates a sequence of heat load set points over a first prediction horizon as an output, each heat load set point of the sequence of heat load set points representing a target heat load to be delivered by the boiler to a hydronic circuit at a time associated with the respective heat load set point; and

a second control algorithm that accepts the sequence of heat load set points and a return water temperature, the return water temperature determined for heated water circulated via a series of pipes, as inputs and generates the sequence of boiler set points over a second prediction horizon as an output based on a determined efficiency for a respective heat load set point of the sequence of heat load set points, each boiler set point of the sequence of boiler set points representing a target parameter of the boiler at a time associated with the respective boiler set point, wherein the second prediction horizon is shorter than the first prediction horizon, the determined efficiency corresponding to a fuel cost for the respective heat load set point of the sequence of heat load set points determined as a function of the respective heat load set point and the return water temperature; and

controlling, based on the MPC model including the sequence of heat load set points and the sequence of boiler set points, the boiler to heat the structure.

2. The method of claim 1 , wherein the elements comprise wall thermal resistivity and windows thermal resistivity.

3. The method of claim 2 , wherein the elements further comprise wall thermal capacity.

4. The method of claim 1 , further comprising determining the unknown parameters of the MPC model via use of cloud computing resources.

5. The method of claim 1 , wherein controlling the boiler to heat the structure comprises controlling the boiler to be on or off using the MPC model to evaluate an optimal control sequence.

6. The method of claim 5 , further comprising selecting the optimal control sequence by evaluating multiple on/off scenarios.

7. The method of claim 1 , wherein the first control algorithm determines a set of predicted operative temperatures based on a set of air temperatures within the structure and a set of wall temperatures within the structure and generates the sequence of heat load set points based on the set of predicted operative temperatures, a set of predicted water temperatures, and a set of slack variables,

wherein each heat load set point of the sequence of heat load set points corresponds to a respective predicted operative temperature of the set of predicted operative temperatures,

wherein each heat load set point of the sequence of heat load set points corresponds to a respective predicted water temperature of the set of predicted water temperatures, and

wherein each heat load set point of the sequence of heat load set points corresponds to a respective slack variable of the set of slack variables.

8. The method of claim 1 , wherein the second control algorithm generates the sequence of boiler set points based on a load delivery cost function that accepts the sequence of heat load set points as an input.

9. The method of claim 1 , wherein the first control algorithm accepts the outside temperature forecast as the input and also accepts an estimated heat accumulation in walls of the structure as a further input.

10. The method of claim 1 , wherein the second prediction horizon comprises a first interval and a second interval distinct from the first interval, wherein the first interval comprises considering various on/off scenarios for the boiler, and wherein the second interval comprises assuming a continuous modulation of the boiler.

11. A non-transitory computer-readable medium storing instructions that when executed by a processor cause the processor to:

obtain an outside temperature forecast for an area outside of a structure, wherein the outside temperature forecast includes a period of time following a present time;

determine a sequence of boiler set points for a boiler using a model predictive control (MPC) model for the structure using inverse modeling, wherein the MPC model comprises:

unknown parameters of elements of a dynamic heat loss model,

wherein the unknown parameters are constrained within a predetermined expected range of the elements of the dynamic heat loss model, and

wherein gas consumption data from a gas meter for the boiler is used to estimate the unknown parameters;

a first control algorithm that accepts the outside temperature forecast as an input and generates a sequence of heat load set points over a first prediction horizon as an output, each heat load set point of the sequence of heat load set points representing a target heat load to be delivered by the boiler to a hydronic circuit at a time associated with the respective heat load set point; and

a second control algorithm that accepts the sequence of heat load set points and a return water temperature, the return water temperature determined for heated water circulated via a series of pipes, as inputs and generates the sequence of boiler set points over a second prediction horizon as an output based on a determined efficiency for a respective heat load set point of the sequence of heat load set points, each boiler set point of the sequence of boiler set points representing a target parameter of the boiler at a time associated with the respective boiler set point,

wherein the second prediction horizon comprises a first interval and a second interval distinct from the first interval, wherein the first interval comprises a consideration of various on/off scenarios for the boiler, and

wherein the second interval comprises an assumption of a continuous modulation of the boiler,

wherein the second prediction horizon is shorter than the first prediction horizon, the determined efficiency corresponding to a fuel cost for the respective heat load set point of the sequence of heat load set points determined as a function of the respective heat load set point and the return water temperature; and

control, based on the MPC model including the sequence of heat load set points and the sequence of boiler set points, the boiler to heat the structure.

12. The computer-readable medium of claim 11 , wherein the MPC model comprises unknown parameters of elements of a dynamic heat loss model, and wherein the unknown parameters are constrained within a predetermined expected range of the elements of the dynamic heat loss model.

13. The computer-readable medium of claim 12 , wherein the elements comprise at least one of wall thermal resistivity, windows thermal resistivity, or wall thermal capacity.

14. The computer-readable medium of claim 11 , wherein controlling the boiler to heat the structure comprises controlling the boiler to be on or off using the MPC model to evaluate an optimal control sequence.

15. The computer-readable medium of claim 14 , wherein the instructions further cause the processor to select the optimal control sequence by evaluating multiple on/off scenarios.

16. The computer-readable medium of claim 11 , wherein the first control algorithm accepts the outside temperature forecast as the input and also accepts an estimated heat accumulation in walls of the structure as a further input.

17. A device comprising:

a processor; and

a memory electrically connected to the processor, wherein the memory is configured to store a program that includes a model predictive control (MPC) model, wherein the processor is configured to:

obtain an outside temperature forecast for an area outside of a structure, wherein the outside temperature forecast includes a period of time following a present time;

determine a sequence of boiler set points for a boiler using MPC model for the structure using inverse modeling, wherein the MPC model comprises:

unknown parameters of elements of a dynamic heat loss model,

wherein the unknown parameters are constrained within a predetermined expected range of the elements of the dynamic heat loss model, and

wherein gas consumption data from a gas meter for the boiler is used to estimate the unknown parameters,

a first control algorithm that accepts the outside temperature forecast as an input and generates a sequence of heat load set points over a first prediction horizon as an output, each heat load set point of the sequence of heat load set points representing a target heat load to be delivered by the boiler to a hydronic circuit at a time associated with the respective heat load set point; and

a second control algorithm that accepts the sequence of heat load set points and a return water temperature, the return water temperature determined for heated water circulated via a series of pipes, as inputs and generates the sequence of boiler set points over a second prediction horizon as an output based on a determined efficiency for a respective heat load set point of the sequence of heat load set points, each boiler set point of the sequence of boiler set points representing a target parameter of the boiler at a time associated with the respective boiler set point, wherein the second prediction horizon is shorter than the first prediction horizon, the determined efficiency corresponding to a fuel cost for the respective heat load set point of the sequence of heat load set points determined as a function of the respective heat load set point and the return water temperature; and

control, based on the MPC model including the sequence of heat load set points and the sequence of boiler set points, the boiler to heat the structure.

18. The device of claim 17 , wherein the elements comprise at least one of wall thermal resistivity, windows thermal resistivity, or wall thermal capacity.

19. The device of claim 17 , wherein the processor is configured to control the boiler to heat the structure by at least controlling the boiler to be on or off using the MPC model to evaluate an optimal control sequence, and wherein the optimal control sequence is selected by evaluating multiple on/off scenarios.

20. The device of claim 17 , wherein the second prediction horizon comprises a first interval and a second interval distinct from the first interval, wherein the first interval comprises a consideration of various on/off scenarios for the boiler, and wherein the second interval comprises an assumption of a continuous modulation of the boiler.