Systems for constructing hierarchical training data sets for use with machine-learning and related methods therefor

1. A system comprising:

a processor; and

a tangible, non-transitory memory configured to communicate with the processor,

the tangible, non-transitory memory having instructions stored thereon that, in response to execution by the processor, cause the processor to perform operations comprising:

receiving a first training data set comprising first training data having been generated by a first controlled device;

determining that the first training data set is inadequate to fully form a hierarchical training data set such that the first training data set does not include a first optimal control input value of a first control input for each potential combination of controlled device inputs and controlled device outputs;

receiving a second training data set comprising second training data having been generated by the first controlled device and a second controlled device;

receiving first free inputs of the first controlled device and second free inputs of the second controlled device, wherein the first free inputs and the second free inputs are respective measurable inputs of the first controlled device and the second controlled device that are not subject to the control of an operator of the first controlled device and the second controlled device;

supplementing the first training data set with the second training data set to fill in missing combinations of the controlled device inputs, the controlled device outputs, the first free inputs and the second free inputs in the first training data set and to form the hierarchical training data set;

supplementing the hierarchical training data set with simulated optimal values of the controlled device inputs, the controlled device outputs, the first free inputs and the second free inputs, in response to the first training data set with the second training data set being inadequate;

training a neural network using the hierarchical training data set to determine the first optimal control input value of the first control input for the first controlled device of multiple controlled devices;

wherein the multiple controlled devices comprise the first controlled device and the second controlled device;

using the hierarchical training data set with the trained neural network, the controlled device inputs, the first free inputs and the second free inputs to determine the controlled device outputs, and

implementing a perturbation system configured to cause a control system of the first controlled device to operate the first controlled device with values of the first control input other than the first optimal control input value of the first control input.

2. The system of claim 1 , wherein the multiple controlled devices further comprise a third controlled device and a fourth controlled device.

3. The system of claim 1 , wherein the first controlled device is associated with a first controlled device classification.

4. The system of claim 1 , wherein the second controlled device is associated with a second controlled device classification.

5. The system of claim 1 , wherein the multiple controlled devices further comprise a third controlled device, and wherein the third controlled device is not associated with a first controlled device classification.

6. The system of claim 1 , wherein the multiple controlled devices further comprise a fourth controlled device, and wherein the fourth controlled device is associated with a first controlled device classification but not associated with a second controlled device classification.

7. The system of claim 1 , further comprising determining, by using array programming, that the first training data set includes the first optimal control input value of a first control input.

8. The system of claim 1 , wherein the multiple controlled devices further comprise a third controlled device and a fourth controlled device, and further comprising:

determining that the first training data set and the second training data set are inadequate to fully form the hierarchical training data set such that the first training data set and the second training data set do not include the first optimal control input value of the first control input for each potential combination of controlled device inputs and controlled device outputs;

receiving a third training data set comprising third training data having been generated by all of the multiple controlled devices; and

supplementing the first training data set and the second training data set with the third training data set to form the hierarchical training data set.

9. The system of claim 1 , wherein the controlled device outputs are different for each of the multiple controlled devices.

10. The system of claim 9 , wherein a first of the controlled device outputs of the first controlled device comprises a quantity of water generated by the first water generating unit, and wherein the first optimal control input value corresponds to a maximum quantity of water generated by the first water generating unit.

11. The system of claim 9 , wherein each of the multiple water generating units comprises at least one sensor configured to detect at least one of: an ambient temperature at the first water generating unit, an ambient humidity at the first water generating unit, a solar insolation flux at the first water generating unit, a quantity of water generated by the first water generating unit, a rate of movement of a process fluid, a rate of movement of a regeneration fluid, and a rate of exposure of a regeneration fluid to a desiccant.

12. The system of claim 9 , wherein the controlled device outputs is a function of the first free inputs and the second free inputs comprising one or more of: an ambient temperature at the first water generating unit, an ambient relative humidity at the first water generating unit, and a solar insolation flux at the first water generating unit.

13. The system of claim 1 , wherein the first optimal control input value of a first control input varies between the multiple controlled devices due to design variables in the multiple controlled devices.

14. The system of claim 9 , wherein the first controlled device is associated with a first controlled device classification and the second controlled device is associated with a second controlled device classification different than the first controlled device classification;

wherein the first controlled device classification and the second controlled device classification are associated with one or more design variables associated with the first water generating unit including one or more of: a hygroscopic material and a system geometry;

wherein the first optimal control input value of the first control input corresponds to a maximum quantity of water generated by the first water generating unit, and a second optimal control input value of a second control input corresponds to a maximum quantity of water generated by a second water generating unit; and,

wherein the first optimal control input value of the first control input varies from the second optimal control input value of the second control input.

15. The system of claim 1 , wherein training coefficients correspond to the first optimal control input value of the first control input, and wherein training coefficients are configured to train a control system of the first control device to learn the first optimal control input value of the first control input of the first control device.

16. The system of claim 15 , wherein the communication system is configured to perform an operation comprising at least one of:

transmit a value of the first control input;

transmit a value of a first controlled output;

transmit a value of a free input of the multiple controlled devices;

transmit one or more training coefficients for storage at the central database, wherein the one or more training coefficients correspond to the first optimal control input value of the first control input.

17. The system of claim 1 , wherein an optimal value of the controlled device outputs comprise at least one of a minimum value, a maximum value, or a desired value.

18. A method being implemented via execution of computer instructions configured to run on one or more processors and configured to be stored at on one or more non-transitory memory storage devices, the method comprising:

receiving a first training data set comprising first training data having been generated by a first controlled device;

determining that the first training data set is inadequate to fully form a hierarchical training data set such that the first training data set does not include a first optimal control input value of a first control input for each potential combination of controlled device inputs and controlled device outputs;

receiving a second training data set comprising second training data having been generated by the first controlled device and a second controlled device;

receiving first free inputs of the first controlled device and second free inputs of the second controlled device, wherein the first free inputs and the second free inputs are respective measurable inputs of the first controlled device and the second controlled device that are not subject to the control of an operator of the first controlled device and the second controlled device;

supplementing the first training data set with the second training data set to fill in missing combinations of the controlled device inputs, the controlled device outputs, the first free inputs and the second free inputs in the first training data set and to form the hierarchical training data set;

supplementing the hierarchical training data set with simulated optimal values of the controlled device inputs, the controlled device outputs, the first free inputs and the second free inputs, in response to the first training data set with the second training data set being inadequate;

training a neural network using the hierarchical training data set to determine the first optimal control input value of the first control input for the first controlled device of multiple controlled devices;

wherein the multiple controlled devices comprise the first controlled device and the second controlled device;

using the hierarchical training data set with the trained neural network, the controlled device inputs, the first free inputs and the second free inputs to determine the controlled device outputs; and

causing, using a perturbation system, a control system of the first controlled device to operate the first controlled device with values of the first control input other than the first optimal control input value of the first control input.