46 results about "Sequential probability ratio test" patented technology

A system and method for monitoring a condition of a monitored system. Estimates of monitored parameters from a model of the system provide residual values that can be analyzed using a sequential probability ratio test (“SPRT”). The invention employs empirically derived distributions in the SPRT to provide more accurate and sensitive alerts of impending faults, breakdowns and process deviations. The distributions can be generated from piecewise continuous approximation or spline functions based on the actual distribution of residual data to provide improved computational performance. The distributions can be provided before monitoring, or can be updated and determined during monitoring in an adaptive fashion.

An apparatus and method for monitoring a process involve development and application of a statistically qualified neuro-analytic (SQNA) model to accurately and reliably identify process change. The development of the SQNA model is accomplished in two stages: deterministic model adaption and stochastic model modification of the deterministic model adaptation. Deterministic model adaption involves formulating an analytic model of the process representing known process characteristics, augmenting the analytic model with a neural network that captures unknown process characteristics, and training the resulting neuro-analytic model by adjusting the neural network weights according to a unique scaled equation error minimization technique. Stochastic model modification involves qualifying any remaining uncertainty in the trained neuro-analytic model by formulating a likelihood function, given an error propagation equation, for computing the probability that the neuro-analytic model generates measured process output. Preferably, the developed SQNA model is validated using known sequential probability ratio tests and applied to the process as an on-line monitoring system. Illustrative of the method and apparatus, the method is applied to a peristaltic pump system.

In one embodiment, a method and apparatus for high-efficiency time-series archiving for computer server telemetry signals is disclosed. The method includes selecting one or more telemetry signals of a plurality of telemetry signals by a sequential probability ratio test (SPRT) algorithm, the SPRT algorithm identifying the one or more telemetry signals as not consistent with normal behavior of the plurality of telemetry signals, injecting synthetic samples around the selected one or more telemetry signals to create a continuous time series telemetry sample, and analyzing the continuous time series telemetry sample to identify leading indicators of faults in the target. Other embodiments are also disclosed.

One embodiment provides a system that analyzes telemetry data from a monitored system. During operation, the system periodically obtains the telemetry data as a set of telemetry variables from the monitored system and updates a multidimensional real-time distribution of the telemetry data using the obtained telemetry variables. Next, the system analyzes a statistical deviation of the multidimensional real-time distribution from a multidimensional reference distribution for the monitored system using a multivariate sequential probability ratio test (SPRT) and assesses the integrity of the monitored system based on the statistical deviation of the multidimensional real-time distribution. If the assessed integrity falls below a threshold, the system determines a fault in the monitored system corresponding to a source of the statistical deviation.

The invention discloses a method for detecting leakage of a pipeline of a tank farm by combining the statistical hypothesis testing technology and the pattern recognition technology. The method comprises the steps that measuring meter data and actuator state feedback information of a working site are acquired through a tank farm monitoring system, an oil transportation path is determined, and Kalman filtering is carried out on pressure measuring meter data to generate pressure information; whether the pressure of the pipeline exceeds a limit is detected according to the multi-sequential probability ratio test, and if yes, an over-limit pressure sequence and the multidimensional time domain features of the over-limit pressure sequence are extracted; recognition of leakage of the pipeline and adjustment of working conditions are carried out based on an established neural network leakage recognition model. By means of the method, a small change of the pressure of the pipeline can be detected in the least time within a regulated false alarm rate and a regulated alarm failure rate, and then whether leakage of the pipeline occurs can be determined.

One embodiment of the present invention provides a monitoring system that detects anomalies in data gathered from sensors in a computer system. During operation, the monitoring system samples data from a plurality of sensors located at various sampling points throughout the computer system. Next, the monitoring system interpolates the data from the sampling points to produce a real-time digitized surface. The monitoring system then subtracts a reference digitized surface from the real-time digitized surface to produce a residual digitized surface. Finally, the monitoring system applies a multi-dimensional sequential probability ratio test (SPRT) to the residual digitized surface to detect anomalies in the residual digitized surface which indicate an impending failure of the computer system.

The invention relates to a method for cooperatively detecting spectrum holes by a plurality of subprime users in the field of cognitive radio in communication technology. The method is characterized by making a decision on the basis of sequential probability ratio test in a fusion stage and adaptively adjusting sampling points used in a perceptive stage. By the characteristics, the method can ensure the needed detection precision and minimize the detection time. The method comprises the following steps of: a. carrying out parallel sampling and obtaining a perception result according to the sampling points by all subprime users; b. fusing one perception result of the subprime users by a fusion center; c. if the fused perception result is enough, making the decision; or else, returning backto the step b to continue to fuse the perception result of the subprime users; and d. adaptively adjusting the sampling points, if the detection is not stopped, returning back to the step a to carry out next detection.

A method of detecting and diagnosing system faults, includes detecting the noisy status of a monitor during operations and incorporating a quantified monitor uncertainty level to support fault isolation reasoning. A sequential probability ratio test is used to statistically test the noisy status of a monitor and Shannon's entropy theory is used to quantify the uncertainty levels of the monitor to support the use of the monitor values in fault isolation.

The invention provides a method and apparatus for rapid reliability evaluation of semiconductor, including an efficient and reliable reliability reference (BL) management system and a test scheme baseon sequential probability ratio test (SPRT). In order to ensure the credibility of the SPRT test scheme, the Bayesian approach is used to update the BL database continuously. This is the first time that the SPRT has been applied to reliability risk assessment with two unknown parameters. By using the method and the device provided by the invention, the reliability of the semiconductor can be quickly measured, and the sample size and the test time can be reduced.

We describe a system that performs prognostic-surveillance operations based on an inferential model that dynamically adapts to evolving operational characteristics of a monitored asset. During a surveillance mode, the system receives a set of time-series signals gathered from sensors in the monitored asset. Next, the system uses an inferential model to generate estimated values for the set of time-series signals, and then performs a pairwise differencing operation between actual values and the estimated values for the set of time-series signals to produce residuals. Next, the system performs a sequential probability ratio test (SPRT) on the residuals to produce SPRT alarms. When a tripping frequency of the SPRT alarms exceeds a threshold value, which is indicative of an incipient anomaly in the monitored asset, the system triggers an alert. While the prognostic-surveillance system is operating in the surveillance mode, the system incrementally updates the inferential model based on the time-series signals.

The invention discloses a malicious anchor node detection method based on an isolated forest and a sequential probability ratio test. The malicious anchor node detection method combines an isolated forest algorithm and a voting mechanism to obtain reliable information in anchor nodes, and builds a detection model through the reliable information. According to the malicious anchor node detection method, only a single ranging algorithm is needed for ranging, and multiple ranging algorithms do not need to be used for ranging, and meanwhile, the assumption that one of multiple ranging methods is not attacked completely is avoided, so that the malicious anchor node detection method is more suitable for being used in a real field, that is, it does not need to assume that the malicious anchor node detection method is not attacked completely. Normal samples are screened by utilizing an isolated forest, and reference anchor nodes in the normal samples are screened by utilizing a voting mechanism to realize multiple selection, and the reliability of the reference anchor nodes is ensured, so that the subsequent process of obtaining malicious anchor nodes according to the reference anchor nodes is indirectly ensured; and the difference value information is used for sequential probability ratio inspection, so that the detection of malicious anchor nodes is further improved, and the detection accuracy of the anchor nodes is improved, and the final positioning accuracy of subsequent target nodes is also improved.

During a surveillance mode, the system receives present time-series signals gathered from sensors in the power transformer. Next, the system uses an inferential model to generate estimated values for the present time-series signals, and performs a pairwise differencing operation between actual values and the estimated values for the present time-series signals to produce residuals. The system then performs a sequential probability ratio test on the residuals to produce alarms having associated tripping frequencies (TFs). Next, the system uses a logistic-regression model to compute a risk index for the power transformer based on the TFs. If the risk index exceeds a threshold, the system generates a notification that the power transformer needs to be replaced. The system also periodically updates the logistic-regression model based on the results of periodic dissolved gas analyses for the transformer to more accurately compute the index for the power transformer.

The invention discloses a principal component analysis and sequential probability ratio test-based centrifugal pump fault diagnosis method. The method comprises the following steps of building a modelby using a common impeller and a fault impeller, and acquiring an original vibration signal by employing a centrifugal pump vibration signal acquisition system; then, performing noise reduction on the signal by applying wavelet packet transformation, and extracting a characteristic parameter of the signal by utilizing a time domain analysis method; then, performing dimension-reduced processing onthe extracted characteristic parameter by employing a principal component analysis method, and selecting a principal component of which the contribution ratio is maximum as a test sequence; and finally, analyzing the running status of a centrifugal pump by utilizing a sequential probability ratio test algorithm and performing classification on a fault by combining with a root-mean-square algorithm. According to the method disclosed by the invention, principal component analysis and sequential probability ratio test are mainly utilized to perform fault state diagnosis, and a classification criterion is established; and the method is higher in effectiveness and accuracy in the aspects of fault diagnosis and recognition of the centrifugal pump.

The invention discloses a two-stage DDoS (Distributed Denial of Service) attack detection and defense method in a software defined network, which is characterized in that switch flow table data is acquired based on a northbound interface of a controller, direct features and derived features are extracted, a two-stage attack detection algorithm is designed by adopting an SPRT (Sequential Probability Ratio Test) and a Light Gradient Boosting Machine (Lightweight GBM) for attack detection, and the two-stage DDoS attack detection and defense method is applied to the software defined network. The SPRT first-level attack detection algorithm is used for quickly positioning an attack port in the early stage of an attack, the LightGBM second-level attack detection algorithm is used for specifically classifying the attack, attack defense filters attack traffic in real time by issuing a flow table rule, and a coarse-grained rule is used for quickly responding to the attack, so that the safety of a controller is protected; and the fine-grained rule is used for defending against a specific type of attack to prevent filtering of normal communication traffic, so that the security of the SDN network can be effectively protected when the DDoS attack occurs.

The invention discloses a fault detection method for a redundant sensor based on an ICA-SPRT (Independent Component Analysis-Sequential Probability Ratio Test), which comprises an ICA-SPRT fault detection system, wherein the ICA-SPRT fault detection system comprises a signal independent component analysis module, a signal identification and estimation module and a fault state statistical detectionmodule; the signal independent component analysis module separates independent signals by adopting an ICA algorithm, and finds the independent signal corresponding to an original signal for restoration through the signal identification and estimation module; the fault state statistical detection module obtains fault judgment through an SPRT algorithm. The fault detection method provided by the invention for the redundant sensor based on the ICA-SPRT can effectively eliminate influences of system noise and channel noise, and still can realize accurate fault detection under the condition of lowredundancy.

The present invention discloses a gearbox malfunction identification method based on the sequential hypothesis test. The method is that an identification system carries out an adaptive intelligent query to a propagation channel with available data. Firstly, an extracted vibration signal is pretreated with the wavelet packet analysis method. Secondly, a sequence of kurtosis values of the extractedvibration signal is taken as a test object of the sequential probability ratio test. Then effective identification for the mode and gear crack degradation under four states of the gearbox is performedaccording to the sequential probability ratio test algorithm. Finally the gearbox vibration signal is subjected to the three-layer sequential ratio test by combining the sequential probability ratiotest and the root mean square error algorithm. The invention solves the problems of slow identification speed, inaccurate target identification and low identification efficiency in malfunction detection and multi-malfunction identification.

A method and system using the Sequential Probability Ratio Test to enhance the detection of an elevated level of radiation, by determining whether a set of observations are consistent with a specified model within a given bounds of statistical significance. In particular, the SPRT is used in the present invention to maximize the range of detection, by providing processing mechanisms for estimating the dynamic background radiation, adjusting the models to reflect the amount of background knowledge at the current point in time, analyzing the current sample using the models to determine statistical significance, and determining when the sample has returned to the expected background conditions.