# US 2019/0171967 A1 · Systems for constructing hierarchical training data sets for use with machine-learning and related methods therefor > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2019/0171967 A1 - Document ID: US20190171967A1 - Status: Published application - Kind: A1 - Application number: 16/211,896 - Filed: 2018-12-06 - Published: 2019-06-06 - Earliest claimed: 2017-12-06 - Assignees: Zero Mass Water, Inc. - Applicants: Zero Mass Water, Inc. - Inventors: Cody Alden Friesen; Paul Bryan Johnson; Heath Lorzel; Kamil Salloum; Jonathan Edward Goldberg; Grant Harrison Friesen; Jason Douglas Horwitz - Web page: https://patentgrep.com/patent/US20190171967A1 ## CPC - G06N 20/00: Machine learning - G05B 13/0265: {the criterion being a learning criterion} - G05B 17/02: electric - G06N 3/0499: Feedforward networks - G06N 3/08: Learning methods - G06N 3/09: Supervised learning - G06N 3/045: Combinations of networks - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G) ## Abstract Some embodiments include a system operable to construct hierarchical training data sets for use with machine-learning for multiple controlled devices. Other embodiments of related systems and methods are also provided. ## Claims ### Claim 1 1) A system comprising: one or more processors; and one or more non-transitory memory storage devices storing computer instructions configured to run on the one or more processors and perform: constructing a hierarchical training data set; and applying machine-learning to the hierarchical training data set to determine at least one first optimal control input value of a first control input for a first controlled device of multiple controlled devices; wherein: constructing the hierarchical training data set comprises: receiving a first training data set, the first training data set consisting of first training data having been generated by the first controlled device; determining that the first training data set is inadequate to fully form the hierarchical training data set; receiving a second training data set, the second training data set consisting of second training data having been generated by at least two controlled devices of the multiple controlled devices, and the at least two controlled devices comprising the first controlled device; and supplementing the first training data set with the second training data set to form the hierarchical training data set. ### Claim 2 (depends on 1) 2) The system of claim 1 wherein: the multiple controlled devices consist of the at least two controlled devices. ### Claim 3 (depends on 2) 3) The system of claim 2 wherein: constructing the hierarchical training data set further comprises: determining that the first training data set and the second training data set are inadequate to fully form the hierarchical training data set; receiving a third training data set, the third training data set consisting of third training data having been simulated for 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. ### Claim 4 (depends on 1) 4) The system of claim 1 wherein: the multiple controlled devices comprise the at least two controlled devices and at least one other controlled device. ### Claim 5 (depends on 4) 5) The system of claim 4 wherein: the at least two controlled devices are associated with a first controlled device classification; and the at least one other controlled device is not associated with the first controlled device classification. ### Claim 6 (depends on 1) 6) The system of claim 1 wherein: the multiple controlled devices comprise the at least two controlled devices and at least one other controlled device; and constructing the hierarchical training data set further comprises: determining that the first training data set and the second training data set are inadequate to fully form the hierarchical training data set; receiving a third training data set, the third training data set consisting of 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. ### Claim 7 (depends on 6) 7) The system of claim 6 wherein: the multiple controlled devices further comprises at least one further controlled device; the at least two controlled devices are associated with a first controlled device classification and a second controlled device classification; the at least one other controlled device is not associated with the first controlled device classification; the at least one further controlled device is associated with the first controlled device classification but not associated with the second controlled device classification; and the second controlled device classification is a sub-classification of the first controlled device classification. ### Claim 8 (depends on 7) 8) The system of claim 7 wherein: the first controlled device classification is a model of the multiple controlled devices; and the second controlled device classification is a version of the model of the multiple controlled devices. ### Claim 9 (depends on 6) 9) The system of claim 6 wherein: constructing the hierarchical training data set further comprises: determining that the first training data set, the second training data set, and the third training data set are inadequate to fully form the hierarchical training data set; receiving a fourth training data set, the fourth training data set consisting of fourth training data having been simulated for the multiple controlled devices; and supplementing the first training data set, the second training data set, and the third training data set with the fourth training data set to form the hierarchical training data set. ### Claim 10 (depends on 9) 10) The system of claim 9 wherein: the multiple controlled devices further comprises at least one further controlled device; the at least two controlled devices are associated with a first controlled device classification and a second controlled device classification; the at least one other controlled device is not associated with the first controlled device classification; the at least one further controlled device is associated with the first controlled device classification but not associated with the second controlled device classification; and the second controlled device classification is a sub-classification of the first controlled device classification. ### Claim 11 (depends on 10) 11) The system of claim 10 wherein: the first controlled device classification is a model of the multiple controlled devices; and the second controlled device classification is a version of the model of the multiple controlled devices. ### Claim 12 (depends on 1) 12) The system of claim 1 wherein: the multiple controlled devices are multiple water generating units; and the first controlled device is a first water generating unit. ### Claim 13 (depends on 12) 13) The system of claim 12 wherein: the first control input is one of an actuator speed or a circulator speed. ### Claim 14 (depends on 1) 14) The system of claim 1 wherein: the computer instructions further are configured to perform: applying machine-learning to the hierarchical training data set to determine at least one second optimal control input value of a second control input for the first controlled device of the multiple controlled devices. ### Claim 15 (depends on 14) 15) The system of claim 14 wherein: the multiple controlled devices are multiple water generating units; the first controlled device is a first water generating unit; the first control input is an actuator speed; and the second control input is a circulator speed. ### Claim 16 16) A method being implemented via execution of computer instructions configured to run at one or more processors and configured to be stored at one or more non-transitory memory storage devices, the method comprising: constructing a hierarchical training data set; and applying machine-learning to the hierarchical training data set to determine at least one first optimal control input value of a first control input for a first controlled device of multiple controlled devices; wherein: constructing the hierarchical training data set comprises: receiving a first training data set, the first training data set consisting of first training data having been generated by the first controlled device; determining that the first training data set is inadequate to fully form the hierarchical training data set; receiving a second training data set, the second training data set consisting of second training data having been generated by at least two controlled devices of the multiple controlled devices, and the at least two controlled devices comprising the first controlled device; and supplementing the first training data set with the second training data set to form the hierarchical training data set. ### Claim 17 (depends on 16) 17) The method of claim 16 wherein: the multiple controlled devices comprise the at least two controlled devices and at least one other controlled device; and constructing the hierarchical training data set further comprises: determining that the first training data set and the second training data set are inadequate to fully form the hierarchical training data set; receiving a third training data set, the third training data set consisting of 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. ### Claim 18 (depends on 17) 18) The method of claim 17 wherein: constructing the hierarchical training data set further comprises: determining that the first training data set, the second training data set, and the third training data set are inadequate to fully form the hierarchical training data set; receiving a fourth training data set, the fourth training data set consisting of fourth training data having been simulated for the multiple controlled devices; and supplementing the first training data set, the second training data set, and the third training data set with the fourth training data set to form the hierarchical training data set. ### Claim 19 (depends on 16) 19) The method of claim 16 wherein: the multiple controlled devices are multiple water generating units; and the first controlled device is a first water generating unit. ### Claim 20 (depends on 19) 20) The method of claim 19 wherein: the first control input is one of an actuator speed or a circulator speed. ### Claim 21 (depends on 16) 21) The method of claim 16 further comprising: applying machine-learning to the hierarchical training data set to determine at least one second optimal control input value of a second control input for the first controlled device of the multiple controlled devices. ### Claim 22 (depends on 21) 22) The method of claim 21 wherein: the multiple controlled devices are multiple water generating units; the first controlled device is a first water generating unit; the first control input is an actuator speed; and the second control input is a circulator speed. ### Claim 23 23) A system comprising: one or more processors; and one or more non-transitory memory storage devices storing computer instructions configured to run on the one or more processors and perform: constructing a hierarchical training data set; and applying machine-learning to the hierarchical training data set to determine at least one first optimal control input value of a first control input for a first controlled device of multiple controlled devices; and applying machine-learning to the hierarchical training data set to determine at least one second optimal control input value of a second control input for the first controlled device of the multiple controlled devices; wherein: the multiple controlled devices comprise the at least two controlled devices and at least one other controlled device; and constructing the hierarchical training data set comprises: receiving a first training data set, the first training data set consisting of first training data having been generated by the first controlled device; determining that the first training data set is inadequate to fully form the hierarchical training data set; receiving a second training data set, the second training data set consisting of second training data having been generated by at least two controlled devices of the multiple controlled devices, and the at least two controlled devices comprising the first controlled device; supplementing the first training data set with the second training data set to form the hierarchical training data set; determining that the first training data set and the second training data set are inadequate to fully form the hierarchical training data set; receiving a third training data set, the third training data set consisting of third training data having been generated by all of the multiple controlled devices; determining that the first training data set, the second training data set, and the third training data set are inadequate to fully form the hierarchical training data set; receiving a fourth training data set, the fourth training data set consisting of fourth training data having been simulated for the multiple controlled devices; and supplementing the first training data set, the second training data set, and the third training data set with the fourth training data set to form the hierarchical training data set. ### Claim 24 (depends on 23) 24) The system of claim 23 wherein: the multiple controlled devices further comprises at least one further controlled device; the at least two controlled devices are associated with a first controlled device classification and a second controlled device classification; the at least one other controlled device is not associated with the first controlled device classification; the at least one further controlled device is associated with the first controlled device classification but not associated with the second controlled device classification; the second controlled device classification is a sub-classification of the first controlled device classification; the first controlled device classification is a model of the multiple controlled devices; and the second controlled device classification is a version of the model of the multiple controlled devices. ### Claim 25 (depends on 24) 25) The system of claim 24 wherein: the multiple controlled devices are multiple water generating units; and the first controlled device is a first water generating unit.