Methods and systems provide for fault diagnosis in a vehicular communication network. The methods and systems utilize a trained neural network model which is downloaded to a local computer associated with the vehicular communication network of a given vehicle and which applies inputs from the given vehicle to output maintenance recommendations for the given vehicle.

A computer implemented method to determine aircraft sensor failure and correct aircraft sensor measurement in an aircraft system is provide. The computer implemented method includes determining, using a physics-based high-fidelity model, a high-fidelity system response over operating conditions during which sensor drift of a sensor of interest can be detected, creating, using an aircraft system controller, a reduced order model (ROM) using the high-fidelity system response, wherein the ROM correlates with the sensor of interest when operating normally, calculating, using the ROM, at least one reduced order sensor value, determining an error value between the reduced order sensor value and a sensor measurement reading from the sensor of interest, and comparing the error value to an error threshold, wherein the sensor of interest has failed when the error value is greater than the error threshold.

A method and system for predicting heat exchanger blockage in an aircraft is provided. The method includes generating a reduced order model (ROM) that predicts a ram air fan (RAF) surge margin that correlates to a heat exchanger blockage parameter, calculating, using the ROM, a predicted RAF surge margin value using a sensor signal received from a sensor connected to a ram air fan (RAF), calculating the heat exchanger blockage parameter using at least the predicted RAF surge margin value, and reporting, to a user, the heat exchanger blockage parameter that indicates when a heat exchanger blockage condition is present.

In one embodiment, a plant monitoring system includes a threshold acquiring module, a judging module, and a warning module. The threshold acquiring module acquires in advance, based on a value obtained by leveling the dispersion in first process values corresponding to output values at a first point in a monitoring object and a value representing the dispersion, first thresholds representing a first range of the dispersion in the first process values and second thresholds representing a second range. The judging module compares the first process values chronologically acquired based on the output values at the first point with the corresponding first thresholds and second thresholds. The warning module creates alert information containing information on at least a time point when the first process value exceeds one of the first thresholds and the second thresholds.

A system includes a controller configured to control an operation of a turbine system, and an analytics system coupled to the controller and configured to receive inputs corresponding to the operation of the turbine system, generate an operational impact factor (OIF) value based at least in part on the inputs, generate a turbine system life prediction model configured to predict an operating life of one or more components of the turbine system based at least in part on the OIF value, and provide the OIF value to the controller to perform an action based thereon.

A method for performing technical diagnostics of machines is carried our by means of a diagnostic system of machines that employs at least two sensors to be placed on the machines, wherein the sensors are selected from the group of vibration sensors, strain sensors, position sensors, and distance sensors, and wherein measured data is evaluated by an evaluation process comprising a step of pairing the measured data and a step of comparing processed data with model states.

Mechanisms for generating an analysis result about a machine are provided. A device generates a first health management (HM) analysis result regarding a machine based on real-time first sensor information received during a first period of time and on a first version HM analytic model. The device provides, to an off-board device, a plurality of sensor information comprising the real-time first sensor information and that is generated during the first period of time. The device receives a second version HM analytic model that is based at least in part on the plurality of sensor information and fault information that identifies actual faults that have occurred on the machine. The device generates a second HM analysis result regarding the machine based on real-time second sensor information received during a second period of time and on the second version HM analytic model.

A learning diagnosis apparatus performs learning from failure data to create a diagnostic model, and stores a model, a failure cause part, and sensor data of the equipment in a rare case data table when the number of cases of the failure cause part of the equipment is less than a predetermined number. Then, based on the diagnostic model created by a learning unit, an estimated probability of causing a failure is calculated for each part of the equipment in which a failure has occurred. Based on the rare case data table, a sensor data match rate between sensor data of the equipment in which the failure has occurred and past sensor data of the model of the equipment is calculated. Then, the calculated sensor data match rate for each part of the equipment in which the failure has occurred is displayed.

Computer implemented methods and systems incorporate physics-based and/or chemistry-based constraints into a model of a chemical, physical, or industrial process. The model is derived from a representative dataset of the subject process. The constrained model provides predictions of process behavior that are guaranteed to be consistent with incorporated constraints such as mass balances, atom balances, and/or energy balances while being less computationally intensive than equivalent first principle models. The constrained model can be constructed by matrix multiplication, namely multiplying the solution of an unconstrained linear model by a matrix that enforces the constraints. Improved process control models result, as well as improved process modeling and simulation models result.

A sensor device is coupled to a mechanical machine. The sensor device detects vibrations of the mechanical machine and transmits the vibration data to a remote processing device. The vibration data may be compressed prior to transmission. The remote processing device receives the data and generates a reconstructed version of the vibration data. The remote processing device includes a machine learning model trained to examine vibration data and to identify a motion pattern associated with an error condition. The machine learning model is applied to the reconstructed vibration data and detects an occurrence of an error condition in the mechanical machine. An alert indicating that an error condition has been detected is transmitted to a human operator. The human operator verifies the status of the mechanical machine and confirms that an error condition has occurred. In response to receipt of the confirmation, the machine learning model is further trained on training data updated to include the vibration data generated by the mechanical machine.