# US 11,111,937 B2 · Fault prediction in hydraulic systems > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,111,937 B2 - Document ID: US11111937B2 - Status: Granted - Kind: B2 - Application number: 16/023,514 - Filed: 2018-06-29 - Published: 2020-01-02 - Granted: 2021-09-07 - Earliest claimed: 2018-06-29 - Assignees: The Boeing Company - Applicants: The Boeing Company - Inventors: Omid B. Nakhajavani; Greg Jackson - Web page: https://patentgrep.com/patent/US11111937B2 ## CPC - B64C 25/28: with indicating or warning devices - F15B 19/005: {Fault detection or monitoring} - F15B 21/02: Servomotor systems with program control derived from a store or timing device; Control devices therefor - B64C 25/22: fluid - F15B 2211/6313: the pressure being a load pressure - F15B 2211/857: Monitoring of fluid pressure systems - F15B 2211/865: Prevention of failures - B64C: AEROPLANES; HELICOPTERS - F15B: SYSTEMS ACTING BY MEANS OF FLUIDS IN GENERAL; FLUID-PRESSURE ACTUATORS, e.g. SERVOMOTORS; DETAILS OF FLUID-PRESSURE SYSTEMS, NOT OTHERWISE PROVIDED FOR ## Abstract Fault prediction improvements are provided herein via pruning a dataset by selecting a subset of the dataset for use as a pruned dataset, wherein the pruned dataset includes pressure data for a hydraulic system correlated in time with command signals for a plurality of deployment events of the hydraulic system, and in response to determining that a given pressure value for a given deployment event in the pruned dataset satisfies a low pressure threshold and that a change in pressure values over time during the given deployment event satisfies a pressure slope threshold, generating a service message for the hydraulic system that indicates that the hydraulic system is in a prefault state. ## Claims ### Claim 1 A method, comprising: pruning a dataset by selecting a subset of the dataset for use as a sanitized dataset based on both command signals for a plurality of deployment events of a hydraulic system and pressure data for the hydraulic system, wherein the pruned dataset includes the pressure data for the hydraulic system correlated in time with the command signals for the plurality of deployment events of the hydraulic system; determining that a first pressure value measured during a first extension or retraction of a piston in the hydraulic system, and included in the pruned dataset, satisfies a prefault pressure threshold, wherein the hydraulic system relates to a landing gear on an aircraft, and wherein the first extension or retraction of the piston relates to deployment of the landing gear; identifying a first pressure slope value relating to a first change in a first plurality of pressure values over time during the first extension or retraction of the piston; identifying a second pressure slope value relating to a second change in a second plurality of pressure values over time during a second extension or retraction of the piston in the hydraulic system, wherein the second extension or retraction of the piston relates to deployment of the landing gear, wherein the second plurality of pressure values are included in the pruned dataset, and wherein the second extension or retraction of the piston occurs earlier in time than the first extension or retraction of the piston; determining that the first pressure slope value satisfies a prefault trend threshold based on comparing the first pressure slope value with the second pressure slope value; and determining, based on both the first pressure value satisfying the prefault pressure threshold and the first pressure slope value satisfying the prefault trend threshold, that the hydraulic system is in a pre-fault state, and in response generating a service message for the hydraulic system. ### Claim 2 (depends on 1) The method of claim 1, wherein pruning the dataset comprises selecting the subset of the dataset based on at least one of: wherein the plurality of deployment events selected last for at least a predetermined amount of time; wherein the plurality of deployment events selected include a deployment condition; and wherein the pressure data during the plurality of deployment events selected satisfy a valid-pressure threshold. ### Claim 3 (depends on 2) The method of claim 2, wherein pruning the dataset further comprises selecting the plurality of deployment events based on a difference between a maximum pressure value in a particular deployment event and a minimum value in the particular deployment event satisfying a range threshold. ### Claim 4 (depends on 2) The method of claim 2, wherein the deployment condition includes at least one of: a given altitude at which the deployment events occur; and a given temperature at which the deployment events occurs. ### Claim 5 (depends on 1) The method of claim 1, wherein the first pressure slope value is correlated to a low pressure spike occurring at a start of a first deployment event comprising the first extension or retraction of the piston. ### Claim 6 (depends on 1) The method of claim 1, wherein at least one of the prefault pressure threshold or the prefault trend threshold are determined using a supervised machine learning model trained using historical operational data and indicated fault states relating to one or more hydraulic systems. ### Claim 7 A non-transitory computer-readable storage device including processor executable instructions that enable a processor to perform operations comprising: pruning a dataset by selecting a subset of the dataset for use as a sanitized dataset based on both command signals for a plurality of deployment events of a hydraulic system and pressure data for the hydraulic system, wherein the pruned dataset includes the pressure data for the hydraulic system correlated in time with the command signals for the plurality of deployment events of the hydraulic system; determining that a first pressure value measured during a first extension or retraction of a piston in the hydraulic system, and included in the pruned dataset, satisfies a prefault pressure threshold, wherein the hydraulic system relates to a landing gear on an aircraft, and wherein the first extension or retraction of the piston relates to deployment of the landing gear; identifying a first pressure slope value relating to a first change in a first plurality of pressure values over time during the first extension or retraction of the piston; identifying a second pressure slope value relating to a second change in a second plurality of pressure values over time during a second extension or retraction of the piston in the hydraulic system, wherein the second extension or retraction of the piston relates to deployment of the landing gear, wherein the second plurality of pressure values are included in the pruned dataset, and wherein the second extension or retraction of the piston occurs earlier in time than the first extension or retraction of the piston; determining that the first pressure slope value satisfies a prefault trend threshold based on comparing the first pressure slope value with the second pressure slope value; and determining, based on both the first pressure value satisfying the prefault pressure threshold and the first pressure slope value satisfying the prefault trend threshold, that the hydraulic system is in a pre-fault state, and in response generating a service message for the hydraulic system. ### Claim 8 (depends on 7) The non-transitory computer-readable storage device of claim 7, wherein pruning the dataset comprises selecting the subset of the dataset based on at least one of: wherein the plurality of deployment events selected last for at least a predetermined amount of time; wherein the plurality of deployment events selected include a deployment condition; and wherein the pressure data during the plurality of deployment events selected satisfy a valid-pressure threshold. ### Claim 9 (depends on 8) The non-transitory computer-readable storage device of claim 8, wherein pruning the dataset further comprises selecting the plurality of deployment events based on a difference between a maximum pressure value in a particular deployment event and a minimum value in the particular deployment event satisfying a range threshold. ### Claim 10 (depends on 8) The non-transitory computer-readable storage device of claim 8, wherein the deployment condition includes at least one of: a given altitude at which the deployment events occur; and a given temperature at which the deployment events occurs. ### Claim 11 (depends on 7) The non-transitory computer-readable storage device of claim 7, wherein the first pressure slope value is correlated to a low pressure spike occurring at a start of a first deployment event comprising the first extension or retraction of the piston. ### Claim 12 A system, comprising: a processor; and a memory, the memory including instructions that, when executed by the processor, enable the processor to perform operations comprising: pruning a dataset by selecting a subset of the dataset for use as a sanitized dataset, wherein the pruned dataset includes pressure data for a hydraulic system correlated in time with command signals for a plurality of deployment events of the hydraulic system; determining that a first pressure value measured during a first extension or retraction of a piston in the hydraulic system, and included in the pruned dataset, satisfies a prefault pressure threshold, wherein the hydraulic system relates to a landing gear on an aircraft, and wherein the first extension or retraction of the piston relates to deployment of the landing gear; identifying a first pressure slope value relating to a first change in a first plurality of pressure values over time during the first extension or retraction of the piston; identifying a second pressure slope value relating to a second change in a second plurality of pressure values over time during a second extension or retraction of the piston in the hydraulic system, wherein the second extension or retraction of the piston relates to deployment of the landing gear, wherein the second plurality of pressure values are included in the pruned dataset, and wherein the second extension or retraction of the piston occurs earlier in time than the first extension or retraction of the piston; determining that the first pressure slope value satisfies a prefault trend threshold based on comparing the first pressure slope value with the second pressure slope value; and determining, based on both the first pressure value satisfying the prefault pressure threshold and the first pressure slope value satisfying the prefault trend threshold, that the hydraulic system is in a pre-fault state, and in response generating a service message for the hydraulic system. ### Claim 13 (depends on 12) The system of claim 12, wherein pruning the dataset comprises selecting the subset of the dataset based on at least one of: wherein the plurality of deployment events selected last for at least a predetermined amount of time; wherein the plurality of deployment events selected include a deployment condition; and wherein the pressure data during the plurality of deployment events selected satisfy a valid-pressure threshold. ### Claim 14 (depends on 13) The system of claim 13, wherein pruning the dataset further comprises selecting the plurality of deployment events based on a difference between a maximum pressure value in a particular deployment event and a minimum value in the particular deployment event satisfying a range threshold. ### Claim 15 (depends on 13) The system of claim 13, wherein the deployment condition includes at least one of: a given altitude at which the deployment events occur; and a given temperature at which the deployment events occurs. ### Claim 16 (depends on 12) The system of claim 12, wherein the first pressure slope value is correlated to a low pressure spike occurring at a start of a first deployment event comprising the first extension or retraction of the piston.