Lockheed Martin Corporation

Richard M Hall

Sreerupa Das

Gregory Anthony Harrison

Michael Allen Bodkin

Eric W Worden

Stefan Herzog, Jr

Michelle L Harris

Deriving a clean timing signal from a waveform is disclosed. A sensor-of-interest (SOI) sample set and a waveform sample set that correspond to the SOI sample set in time is collected. The waveform sample set is partitioned into a plurality of waveform sample subsets, and the SOI sample set is partitioned into a plurality of SOI sample subsets, each SOI sample subset corresponding to one of the plurality of waveform sample subsets. A plurality of waveform sample subset angular speeds is determined, wherein each waveform sample subset angular speed corresponds to a different waveform sample subset. An aggregate mean angular speed based on the plurality of waveform sample subset angular speeds is determined. Each SOI sample subset is resampled to the aggregate mean angular speed based on the corresponding waveform sample subset angular speed to generate a plurality of resampled SOI subsets.

Mechanisms for deriving an accurate timing signal from a noisy waveform

Quantification of a condition of a selected item using a neural network is disclosed. A device defines a good hypertube in a neural state space based on good item state points obtained from one or more items that exhibit desired operating characteristics, and a bad hypertube in the neural state space based on bad item state points obtained from one or more items that exhibit undesirable operating characteristics. A current item state hyperpoint is determined in the neural state space based on a current item state point of the selected item. A condition of the selected item is quantified as a function of a location of the current item state hyperpoint with respect to at least a portion of the good hypertube and with respect to at least a portion of the bad hypertube.

Quantifying a condition with a neural network

Systems and methods are disclosed for more accurately estimating a remaining useful life (RUL) of an item. The RUL of the item is estimated by determining probability values throughout a time period. Each probability value quantifies a probability that a failure event of the item will occur by a time in a time segment of the time period. Based on the probability values, the particular time segment in which the failure event is most likely to occur is determined. Once the particular time segment is determined, the presentation of a visual representation of at least a portion of the time period can be effected. The visual representation indicates that the failure event of the item is most likely to occur during the particular time segment.

Systems and methods for estimating a remaining useful life of an item

Estimating a remaining useful life (RUL) of an apparatus is disclosed. A computer device may obtain a priori RUL data of an apparatus. The a priori RUL data identifies a priori RULs values of the apparatus as a function of time. Buckets are then defined in the a priori RUL data, wherein each of the buckets corresponds to a different set of the a priori RUL values in the a priori RUL data. An operational event indicator may then be obtained for the apparatus that indicates a current operational event of the apparatus. The RUL of an apparatus is estimated by determining probability values throughout a time period. Probability values are then determined based on the operational event indicator where each probability value quantifies a probability that a current RUL value of the first apparatus is within one of the buckets.

Adaptive learning network for fusion of remaining useful life with machinery condition

Mechanisms for predicting a remaining useful life of a cutter head of a milling machine that includes a plurality of flutes are disclosed. Features are extracted from reference data associated with a plurality of reference cutter heads. The reference data includes reference vibration data and reference wear data. At least two Neural Network predictive models are trained in parallel for predicting the remaining life of a new cutter head based upon the extracted features. Operational data associated with the new cutter head is obtained. The operational data includes operational vibration data and operational wear data. Features extracted from the operational data are input into an optimal predictive model of the at least two Neural Network predictive models. A remaining useful life of the new cutter head is estimated by the optimal predictive model.

System and method for estimating remaining useful life

Real-time condition-based analysis is performed on a machine for providing diagnostic and prognostic outputs indicative of machine status includes a signal processor for receiving signals from sensors adapted for measuring machine performance parameters. The signal processor conditions and shapes at least some of the received signals into an input form for a neural network. A fuzzy adaptive resonance theory neural network receives at least some of the conditioned and shaped signals, and detects and classifies a state of the machine based upon the received conditioned and shaped signals, and upon a predetermined ontology of machine states, diagnostics, and prognostics. The neural network can also determine from the machine state a health status thereof, which can comprise an anomaly, and output a signal representative of the determined health status. A Bayesian intelligence network receives the machine state from the neural network and determines a fault probability at a future time.

Growing community of inventors

Orlando, FL, United States of America

Richard M Hall