# US 12,180,892 B2 · Machine learned aero-thermodynamic engine inlet condition synthesis

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,180,892 B2
- Document ID: US12180892B2
- Status: Granted
- Kind: B2
- Application number: 16/265,319
- Filed: 2019-02-01
- Published: 2020-08-06
- Granted: 2024-12-31
- Earliest claimed: 2019-02-01
- Assignees: Rtx Corporation
- Applicants: RTX Corporation
- Inventors: Timothy J. Crowley; Ramesh Rajagopalan; Sorin Bengea
- Web page: https://patentgrep.com/patent/US12180892B2

## CPC

- F02C 7/057: Control or regulation (conjointly with fuel supply control F02C9/50, with nozzle area control F02K1/16)
- F02C 9/00: Controlling gas-turbine plants; Controlling fuel supply in air- breathing jet-propulsion plants (controlling air intakes F02C7/057; controlling turbines F01D; controlling compressors F04D27/00; controlling in general G05)
- G05B 13/027: {using neural networks only}
- G06N 3/0499: Feedforward networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- F01D 17/08: responsive to condition of working-fluid, e.g. pressure
- F01D 17/085: {to temperature}
- F02C 7/047: Heating to prevent icing
- F02K 1/54: Nozzles having means for reversing jet thrust (reversing jet thrust using fluid jets F02K1/32)
- F05D 2220/321: for a special turbine stage
- F05D 2260/80: Diagnostics
- F05D 2260/81: Modelling or simulation
- F05D 2270/092: in particular blow-out and relight
- F05D 2270/20: to optimize the performance of a machine
- F05D 2270/30: Control parameters, e.g. input parameters
- F05D 2270/709: with neural networks
- F05D 2270/71: synthesized, i.e. parameter computed by a mathematical model
- Y02T 50/60: Efficient propulsion technologies, e.g. for aircraft
- F02C: GAS-TURBINE PLANTS; AIR INTAKES FOR JET-PROPULSION PLANTS; CONTROLLING FUEL SUPPLY IN AIR-BREATHING JET-PROPULSION PLANTS (construction of turbines F01D; jet-propulsion plants F02K; construction of compressors or fans F04; generating combustion products of high pressure or high velocity F23R; using gas turbines in compression refrigeration plants F25B11/00)
- 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)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- F01D: NON-POSITIVE DISPLACEMENT MACHINES OR ENGINES, e.g. STEAM TURBINES (machines or engines for liquids F03; non-positive displacement pumps F04D)
- F02K: JET-PROPULSION PLANTS (features of jet-propulsion plants common to gas-turbine plants, air intakes or fuel supply control of air-breathing jet-propulsion plants F02C7/00, F02C9/00)
- F05D: INDEXING SCHEME FOR ASPECTS RELATING TO NON-POSITIVE-DISPLACEMENT MACHINES OR ENGINES, GAS-TURBINES OR JET-PROPULSION PLANTS
- Y02T: CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION

## Abstract

A system for neural network compensated aero-thermodynamic gas turbine engine parameter/inlet condition synthesis. The system includes an aero-thermodynamic engine model configured to produce a real-time model-based estimate of engine parameters, a machine learning model configured to generate model correction errors indicating the difference between the real-time model-based estimate of engine parameters and sensed values of the engine parameters, and a comparator configured to produce residuals indicating a difference between the real-time model-based estimate of engine parameters and the sensed values of the engine parameters. The system also includes an inlet condition estimator configured to iteratively adjust an estimate of inlet conditions based on the residuals and adaptive control laws configured to produce engine control parameters for control of gas turbine engine actuators based on the inlet conditions.

## Claims

### Claim 1

A system forneural network compensated aero-thermodynamic engine model parameter synthesis, the system comprising:
at least one sensor configured to output sensed values of engine parameters of a gas turbine engine;
a controller in signal communication with the at least one sensor and configured to control gas turbine engine actuators of the gas turbine engine, wherein the controller comprises:
an aero-thermodynamic engine model configured to produce real-time model-based estimates of the engine parameters;
a plurality of machine learning models configured to generate model correction errors based at least on a difference between the real-time model-based estimates of the engine parameters and sensed values of the engine parameters;
a first comparator configured to produce residuals indicating the difference between the real-time model-based estimates of the engine parameters and the sensed values of the engine parameters;
an inlet condition estimator configured to iteratively adjust an estimate of inlet conditions based on the residuals to produce selected values of the estimates of the inlet conditions; and
a second comparator configured to compensate the selected values of the estimates of the inlet conditions based on error corrections corresponding to a first aero-thermodynamic transient model for a first operating regime and a second aero-thermodynamic transient model for a second operating regime;
adaptive control laws configured to produce engine control parameters for control of the gas turbine engine actuators based on the selected values of estimates of the of inlet conditions,
wherein each machine learning model of the plurality of machine learning models is a machine neural network system, and wherein a first machine learning model of the plurality of machine learning models represents a first aero thermodynamic transient model error for a first operating regime, and a second machine learning model of the plurality of machine learning models represents a second aero-thermodynamic transient model error for a second operating regime that is different from the first operating regime, and
wherein each machine learning model of the plurality of machine learning models output their respective model correction errors under conditions selected according to the first operating regime and the second operating regime of the first machine learning model and the second machine learning model, respectively.

### Claim 2 (depends on 1)

The system of claim 1, wherein the controller further includes a fault detection and accommodation system configured to detect faults in inlet condition sensors and provide validated sensed engine parameters to each machine learning model of the plurality of machine learning models.

### Claim 3 (depends on 1)

The system of claim 1, wherein the aero-thermodynamic engine model is configured to produce real-time model-based estimate engine parameters based on a previous iteration of the estimates of inlet conditions, and based on engine control parameters.

### Claim 4 (depends on 1)

The system of claim 1, wherein the aero-thermodynamic engine model receives engine control parameters, and provides updates for a next iteration using the aero-thermodynamic model.

### Claim 5 (depends on 1)

The system of claim 1 wherein each each machine learning model of the plurality of machine learning models is trained to identify and learn a difference between responses generated by the aero-thermodynamic engine model and the real gas turbine engine under consideration for selected conditions associated with the real gas turbine engine's respective operating regime.

### Claim 6 (depends on 1)

The system of claim 1, wherein the controller further includes a selection function configured to perform at least one of: select sensed parameters inlet conditions for use by the adaptive control laws in the event of no fault, select the estimates of inlet conditions for use by the adaptive control laws in the event of an inlet condition sensor fault.

### Claim 7 (depends on 6)

The system of claim 6, wherein at least one of the sensed parameters inlet conditions and the estimated parameters inlet conditions includes compressor inlet temperature and compressor inlet pressure.

### Claim 8 (depends on 6)

The system of claim 6, where the selection function is further configured to select a machine learning model among the plurality of machine learning models based on the estimates of inlet conditions for use by the adaptive control laws in the event of the inlet condition sensor fault in a selected operating regime of the gas turbine engine.

### Claim 9 (depends on 8)

The system of claim 8 wherein each machine learning model of the plurality of machine learning models is configured to produce model correction errors based on a respective operating regime of the gas turbine engine.

### Claim 10 (depends on 8)

The system of claim 8 wherein the selected operating regime of the gas turbine engine includes at least one of air start windmilling, thrust reversing, and anti icing modes of operation.