A method and apparatus are provided for sensor fault prediction. A time-sequence of output values is received from a sensor. A plurality of features are extracted from the received values. A model is applied to the features to obtain a health score for the sensor. The trend of the health score over time is calculated to detect degrading performance of the sensor, and a time at which the sensor will become faulty is predicted.

A monitoring device includes a process data acquisition unit configured to acquire process data indicating an operation condition of a machine having a rotating shaft, a shaft vibration value acquisition unit configured to acquire a measurement value of a shaft vibration value of the rotating shaft under the operation condition indicated by the process data, a determination model configured to determine a normal value of the shaft vibration value according to the operation condition created on the basis of the shaft vibration value measured during an operation of the machine and the shaft vibration value calculated on the basis of a predetermined shaft vibration calculation model, and a monitoring unit configured to evaluate the measurement value of the shaft vibration value on the basis of the process data, the measurement value of the shaft vibration value, and the determination model.

A system for controlling an operation of a machine for performing a task is disclosed. The system submits a sequence of control inputs to the machine and receives a feedback signal. The system further determines, at each control step, a current control input for controlling the machine based on the feedback signal including a current measurement of a current state of the system by applying a control policy transforming the current measurement into the current control input based on current values of control parameters in a set of control parameters of a feedback controller. Furthermore, the system may iteratively update a state of the feedback controller defined by the control parameters using a prediction model predicting values of the control parameters and a measurement model updating the predicted values to produce the current values of the control parameters that explain the sequence of measurements according to a performance objective.

Examples described herein provide a computer-implemented method that includes receiving training data indicative of incipient compressor surge for cabin air compressors. The method further includes generating, using the training data, a training spectrogram. The method further includes training, by a processing system, a machine learning model to detect incipient compressor surge events for the cabin air compressors using the spectrogram. The method further includes receiving, at a microcontroller associated with a cabin air compressor, operating data associated with the cabin air compressor. The method further includes generating, at the microcontroller and using the operating data, an operating spectrogram. The method further includes detecting, by the microcontroller associated with the cabin air compressor, an incipient compressor surge event by applying the machine learning model to the operating spectrogram. The method further includes implementing a corrective action to correct the incipient compressor surge event.

A method for learning and detecting an abnormal part of a device through artificial intelligence comprises: an information collection step for collecting a current waveform of a current value that changes over time in a driving state of at least one device and collecting information about a faulty part of the device, together with current waveform information before a fault occurs in the device; a model setting step for learning, by a control unit, information collected in the information collection step and setting a reference model of a current waveform for each faulty part of the device; and a detection step for, when an abnormal symptom of the device is detected in a real-time driving state, comparing, by the control unit, a real-time current waveform of the device and the reference model, and detecting and providing an abnormal part regarding the abnormal symptom of the device.

A gas delivery apparatus is provided, comprising a system controller configured to collect valve position information and sensor information from at least a plurality of the sensors and valves, store the valve position information and sensor information into the monitored gas delivery process data, and execute the discriminative classifier including a first artificial intelligence (AI) model configured to extract features in a first input image of the monitored gas delivery process; a second AI model configured to extract features in a second input image of a golden gas delivery process; and a contrastive loss function configured to calculate a similarity between the first input image and the second input image based on outputs of the first AI model and the second AI model, and output a repeatability confidence value based on a similarity index between the first input image and the second input image.

By accurately predicting industrial aging processes (IAP), such as the slow deactivation of a catalyst in a chemical plant, it is possible to schedule maintenance events further in advance, thereby ensuring a cost-efficient and reliable operation of the plant. So far, these degradation processes were usually described by mechanistic models or simple empirical prediction models. In order to accurately predict IAP, data-driven models are proposed, comparing some traditional stateless models (linear and kernel ridge regression, as well as feed-forward neural networks) to more complex stateful recurrent neural networks (echo state networks and long short-term memory networks). Additionally, variations of the stateful models are discussed. In particular, stateful models using mechanistical pre-knowledge about the degradation dynamics (hybrid models). Stateful models and their variations may be more suitable for generating near perfect predictions when they are trained on a large enough dataset, while hybrid models may be more suitable for generalizing better given smaller datasets with changing conditions.

Systems and methods for operating an energy plant are disclosed herein. A time series of performance variable associated with a device in the energy plant is obtained. An auto-correlation function data of the device is obtained based on the time series of the performance variable associated with the device. An electronic model of the device is generated based on the auto-correlation function data. Time, at which a future event of the device is predicted to occur, is predicted based on the electronic model. A report indicating the future event of the device and the predicted time may be generated. The device may be automatically configured, according to the future event and the predicted time.

An abnormal irregularity cause identifying device includes a process data acquisition unit that reads process data output by sensors included in a production facility performing a batch stage and a continuous stage, a preprocessing unit that associates a range of a complete timing of the batch stage with an output timing of process data of the process data in the continuous stage based on a residence time of the processing target in the production facility, an abnormality determination unit that calculates an abnormality degree by using process data in the batch stage and process data in the continuous stage associated with each other by the preprocessing unit, and a cause diagnosis unit that determines, for each of the process data output by the corresponding one of the plurality of sensors, whether the abnormality degree calculated by the abnormality determination unit satisfies a predetermined criterion.

An information processing apparatus includes an n-th parameter adjuster and an (n+1)-th parameter adjuster. The n-th parameter adjuster adjusts an n-th parameter set so that an n-th evaluation value set based on the n-th parameter set approaches an n-th target value set. The (n+1)-th parameter adjuster adjusts an (n+1)-th parameter set so that an (n+1)-th evaluation value set based on the (n+1)-th parameter set approaches an (n+1)-th target value set. In addition, the n-th parameter adjuster acquires, based on initial value set or search value set of the n-th parameter set, an n-th actual measured value set or an n-th predicted value set, acquires an (n+1)-th target value set based on the initial value set or the search value set of the n-th parameter set, and searches for the n-th parameter set that optimizes the (n+1)-th target value set under a restriction that the n-th evaluation value set approaches the n-th target value set using the acquired n-th actual measured value set or the n-th predicted value set and the acquired (n+1)-th target value set.