Control sequence generation system and methods
Various embodiments described herein relate to a method, device, and non-transitory machine-readable storage medium for controlling a system having a plurality of devices including one or more of the following: calculating a demand curve for a controlled space representing at least one target amount of state for delivery to the controlled space; creating a cost function that compares the demand curve to simulated state curves produced by a model of the controlled system based on respective sets of control actions issuable to the plurality of controlled devices; performing an optimization method that tunes a candidate set of control actions to reduce a cost output by the cost function based on the candidate set of control actions; and issuing control actions of the candidate set of control actions to respective ones of the plurality of controlled devices to cause the controlled system to deliver state to the controlled space.
1. A method for controlling a controlled system having a plurality of controlled devices, the method comprising:
calculating a demand value representing at least one target outcome from the controlled system using a first neural network, the first neural network having a plurality of nodes, wherein at least two nodes represent at least two corresponding controlled devices in the controlled system;
creating a cost function that compares the demand value to simulated outcome values produced by the first neural network, and which takes as input respective sets of control actions issuable to the plurality of controlled devices;
performing an optimization method that tunes a candidate set of control actions to reduce a cost output by the cost function based on the candidate set of control actions; and
issuing control actions of the candidate set of control actions to the respective ones of the plurality of controlled devices to cause the controlled system to adjust an environmental state of an area associated with the controlled system.
2. The method of claim 1 , wherein performing the optimization method comprises:
computing partial derivatives using the first neural network; and
using the partial derivatives to perform gradient descent to tune the candidate set of control actions.
3. The method of claim 2 , wherein performing gradient descent comprises:
randomly generating the candidate set of control actions; and
iteratively tuning the candidate set of control actions according to the partial derivatives to reduce a cost indicated by the cost function.
4. The method of claim 1 , wherein the demand value is a demand curve representing variance in the at least one target outcome over a period of time.
5. The method of claim 1 , wherein:
calculating a demand value comprises calculating a plurality of demand values corresponding to a plurality of desired outcomes, respectively; and
in comparing the demand value to simulated outcome values produced by the first neural network, the cost function compares the plurality of demand values to a plurality of simulated outcome values produced by the first neural network, respectively.
6. The method of claim 1 , wherein the optimization method that tunes a candidate set of control actions to reduce a cost output by the cost function based on the candidate set of control actions comprises a method that runs a neural network, the neural network comprising nodes that correlate to controlled devices.
7. A device for controlling a controlled system having a plurality of controlled devices, comprising:
a communication interface configured to communicate with the plurality of controlled devices;
a memory comprising a model of the controlled system; and
a processor configured to:
calculate a demand curve for a controlled space representing at least one target amount of state for delivery to the controlled space using a first neural network, the first neural network having a plurality of nodes, wherein at least two nodes represent at least two corresponding controlled devices,
create a cost function that compares the demand curve to simulated state curves produced by the first neural network, and which takes as input respective sets of control actions issuable to the plurality of controlled devices,
perform an optimization method that tunes a candidate set of control actions to reduce a cost output by the cost function based on the candidate set of control actions, and
issue control actions of the candidate set of control actions to respective ones of the plurality of controlled devices to cause the controlled system to deliver an environmental state to the controlled space.
8. The device of claim 7 , wherein, in performing the optimization method, the processor is configured to:
compute partial derivatives using the model; and
use the partial derivatives to perform gradient descent to tune the candidate set of control actions.
9. The device of claim 8 , wherein, in performing gradient descent, the processor is configured to:
randomly generate the candidate set of control actions; and
iteratively tune the candidate set of control actions according to the partial derivatives to reduce a cost indicated by the cost function.
10. The device of claim 7 , wherein, in calculating a demand curve, the processor is configured to calculate the demand curve as at least one target amount of state for delivery to the controlled space to achieve a comfort curve representing a desired state for the controlled space.
11. The device of claim 7 , wherein:
the controlled space comprises a plurality of zones;
in calculating a demand curve, the processor is configured to calculate a plurality of demand curves corresponding to the plurality of zones, respectively; and
in comparing the demand curve to simulated state curves produced by the model, the cost function compares the plurality of demand curves to a plurality of simulated state curves produced by the model corresponding to the plurality of zones, respectively.
12. A non-transitory computer-readable storage medium configured with instructions executable by a processor for controlling a controlled system having a plurality of controlled devices, the non-transitory computer-readable storage medium comprising:
instructions for calculating a demand value representing at least one target outcome from the controlled system using a first neural network, the first neural network having a plurality of nodes, wherein at least two nodes represent at least two corresponding controlled devices;
instructions for creating a cost function that compares the demand value to simulated outcome values produced by the first neural network, which takes as input respective sets of control actions issuable to the plurality of controlled devices;
instructions for performing an optimization method that tunes a candidate set of control actions to reduce a cost output by the cost function based on the candidate set of control actions; and
instructions for issuing control actions of the candidate set of control actions to respective ones of the plurality of controlled devices to cause the controlled system to adjust an environmental state of an area associated with the controlled system.
13. The non-transitory computer-readable storage medium of claim 12 , wherein performing the optimization method comprises:
instructions for computing partial derivatives using the first neural network; and
instructions for using the partial derivatives to perform gradient descent to tune the candidate set of control actions.
14. The non-transitory computer-readable storage medium of claim 13 , wherein performing gradient descent comprises:
instructions for randomly generating the candidate set of control actions; and
instructions for iteratively tuning the candidate set of control actions according to the partial derivatives to reduce a cost indicated by the cost function.
15. The non-transitory computer-readable storage medium of claim 12 , wherein the demand value is a demand curve representing variance in the at least one target outcome over a period of time.
16. The non-transitory computer-readable storage medium of claim 12 , wherein:
instructions for calculating a demand value comprises instructions for calculating a plurality of demand values corresponding to a plurality of desired outcomes, respectively; and
in comparing the demand value to simulated outcome values produced by the first neural network, the cost function compares the plurality of demand values to a plurality of simulated outcome values produced by the first neural network, respectively.
17. The non-transitory computer-readable storage medium of claim 12 , wherein the optimization method that tunes a candidate set of control actions to reduce a cost output by the cost function based on the candidate set of control actions comprises a method that runs a neural network.
18. The non-transitory computer-readable storage medium of claim 17 , wherein the neural network comprises nodes that represent the plurality of controlled devices such that each device is represented by a node.
19. The non-transitory computer-readable storage medium of claim 18 , wherein at least two of the nodes are arranged in the neural network corresponding to their equivalent controlled device.
20. The non-transitory computer-readable storage medium of claim 19 , wherein at least two nodes have different activation functions, and wherein at least one activation function has multiple variables.