CN-121980451-A - Fault diagnosis method and system for self-adaptive sand supply device

CN121980451ACN 121980451 ACN121980451 ACN 121980451ACN-121980451-A

Abstract

The invention relates to the technical field of chips of the Internet of things, and discloses a fault diagnosis method and system of a self-adaptive sand supply device. The method comprises the steps of obtaining multichannel original data of vibration, temperature and pressure of a sand supply device, recombining a signal sequence after time synchronization and frequency division processing, obtaining a smooth data set through multi-scale decomposition denoising, extracting time domain and frequency domain features to construct a feature vector set, comparing the feature vector set with historical feature data, analyzing a long-term trend, locating an abnormal point and calculating coupling influence intensity and abnormal fluctuation probability when the feature vector set is abnormal, deducing a fault evolution path and defining a signal change range, calculating a fault probability increment dynamic index, comparing risk threshold value grading evaluation, fusing an evaluation result and the index change trend, and generating an accurate fault early warning signal. The method can realize accurate fault identification and early warning of the sand supply device.

Inventors

ZHOU XIAOMOU
GU FEI
SHEN YONGBO
MA YANSONG
SHI JUNZE
CHEN SHUO

Assignees

江苏和瑞鑫智能科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260115

Claims (10)

1. The fault diagnosis method of the self-adaptive sand supply device is characterized by comprising the following steps of: Acquiring multichannel original data for the operation of a sand supply device, wherein the multichannel original data comprise vibration data, temperature data and pressure data; Performing time synchronization and frequency division processing on the multichannel original data, intercepting a time domain signal segment exceeding a preset reference energy threshold, and recombining the time domain signal segment to obtain a recombined signal sequence; performing multi-scale decomposition and denoising treatment on the recombined signal sequence to obtain a smooth data set; extracting time domain features and frequency domain features from the smoothed data set, and constructing a multi-dimensional feature vector set; Comparing the feature vector set with the pre-acquired historical feature data in a time dimension, and analyzing the long-term change trend of the feature; if the long-term change trend shows an abnormal fluctuation phenomenon, positioning potential abnormal points and analyzing the coupling influence intensity among multiple variables, and converting the coupling influence intensity into abnormal fluctuation probability; Deducing a fault evolution path according to the abnormal fluctuation probability, defining a signal change range, and calculating a dynamic index with the ascending fault probability by combining the abnormal fluctuation probability, the signal change range and the fluctuation amplitude of the long-term change trend; And comparing the dynamic index with a preset risk judgment threshold value, carrying out grading evaluation on the fault risk according to the amplitude of the dynamic index exceeding the risk judgment threshold value, and fusing a grading evaluation result and the change trend of the dynamic index to generate an accurate fault early warning signal.
2. The method for diagnosing a fault in an adaptive sand supply apparatus according to claim 1, wherein acquiring multi-channel raw data for operation of the sand supply apparatus, the multi-channel raw data including vibration data, temperature data, and pressure data, comprises: and deploying a sensor array, and collecting vibration data, temperature data and pressure data when the sand supply device operates to obtain multichannel original data.
3. The method for diagnosing faults of the adaptive sand supply device according to claim 1, wherein the steps of performing time synchronization and frequency division processing on the multi-channel original data, intercepting a time domain signal segment exceeding a preset reference energy threshold, recombining the time domain signal segment to obtain a recombined signal sequence, and include: Performing time stamp alignment on the vibration data, the temperature data and the pressure data to generate a time-synchronized data stream; Carrying out frequency spectrum analysis on the time synchronization data stream, and splitting to obtain a multi-band sub-signal set; If the instantaneous energy value of any frequency band in the sub-signal set exceeds a preset reference energy threshold, locking a corresponding time point and intercepting a time domain signal segment; And vectorizing and recombining the time domain signal segments according to the time sequence to obtain a recombined signal sequence.
4. The fault diagnosis method of the adaptive sand supply device according to claim 1, wherein the performing multi-scale decomposition and denoising on the recombined signal sequence to obtain a smoothed data set includes: Performing multi-scale decomposition on the recombined signal sequence to obtain an approximation coefficient and a detail coefficient; performing signal denoising and reconstruction by combining the approximation coefficient and the detail coefficient to generate a denoised signal sequence; and calculating the first-order differential variance of the denoising signal sequence, and screening continuous sequence fragments with minimum variance to form a smooth data set.
5. The method for fault diagnosis of an adaptive sand supply apparatus according to claim 1, wherein the extracting time domain features and frequency domain features from the smoothed data set, constructing a multi-dimensional feature vector set, includes: calculating the mean value, variance, peak value and kurtosis of the smooth data set to form a time domain feature set; performing frequency spectrum conversion on the smooth data set, extracting main frequency, power spectrum peak value and frequency band energy, and forming a frequency domain feature set; And splicing the time domain feature set and the frequency domain feature set according to dimensions to construct a multi-dimensional feature vector set.
6. The method for diagnosing a fault in an adaptive sand supply device according to claim 1, wherein the comparing the feature vector set with the pre-acquired historical feature data in a time dimension, and analyzing a long-term change trend of a feature, comprises: Constructing a historical data window matrix according to the pre-acquired historical characteristic data; calculating the difference value of the corresponding dimension in the feature vector set and the historical data window matrix to generate a dimension difference sequence; smoothing the dimension difference sequence, and filtering short-term random fluctuation to obtain a stable difference sequence; Generating a trend change curve according to the stationary difference sequence fitting; And calculating the fluctuation amplitude of the trend change curve, and judging that the long-term change trend of the characteristic presents an abnormal fluctuation phenomenon if the fluctuation amplitude exceeds a preset trend fluctuation threshold.
7. The method according to claim 6, wherein if the long-term variation trend shows an abnormal fluctuation phenomenon, locating potential abnormal points and analyzing coupling influence intensities among multiple variables, converting the coupling influence intensities into abnormal fluctuation probabilities, comprises: performing second-order differential operation on the trend change curve to generate a trend change rate sequence; identifying time coordinate points of symbol overturn in the trend change rate sequence, and marking the time coordinate points as a candidate abnormal point set; Constructing a multi-variable state matrix based on the candidate abnormal point set, calculating a cross covariance value among variables in the multi-variable state matrix, and quantifying coupling influence intensity; if the coupling influence intensity exceeds a preset coupling intensity threshold value, mapping the coupling influence intensity into a probability value through probability modeling to obtain abnormal fluctuation probability.
8. The method for diagnosing faults of an adaptive sand supply device according to claim 1, wherein the step of calculating a dynamic index of increasing fault probability by combining the abnormal fluctuation probability, the signal variation range and the fluctuation amplitude of the long-term variation trend includes the steps of: integrating equipment operation parameters according to the abnormal fluctuation probability, and constructing a fault evolution initial state vector; Simulating a plurality of groups of future state transition tracks of the initial state vector, analyzing the statistical distribution of the tracks, calculating a confidence interval according to the analysis result and defining a signal change range; dividing risk intervals based on the signal change range, and calculating the weight of each risk interval by combining the fluctuation amplitude of the long-term change trend; And according to the abnormal fluctuation probability, the risk interval weight and the distribution density of the track statistical distribution, weighting calculation is carried out to obtain a dynamic index with increased fault probability.
9. The fault diagnosis method of the adaptive sand supply device according to claim 1, wherein the comparing the dynamic index with a preset risk determination threshold value, performing a hierarchical assessment on a fault risk according to an amplitude of the dynamic index exceeding the risk determination threshold value, and fusing a hierarchical assessment result and a variation trend of the dynamic index to generate an accurate fault early warning signal, includes: comparing the dynamic index with a preset risk judgment threshold value, and dividing risk grades according to the exceeding amplitude to obtain a grading evaluation result; Calculating the time change slope of the dynamic index, and determining the fault probability increment rate; Fusing the grading evaluation result and the fault probability increment rate to construct a comprehensive early warning vector; Analyzing the comprehensive early warning vector, determining fault associated components and development trends, and generating accurate fault early warning signals.
10. An adaptive sand supply device fault diagnosis system, comprising: the data acquisition module is used for acquiring multichannel original data of the sand supply device, wherein the multichannel original data comprise vibration data, temperature data and pressure data; The signal processing module is used for carrying out time synchronization and frequency division processing on the multichannel original data, intercepting a time domain signal segment exceeding a preset reference energy threshold value, and recombining the time domain signal segment to obtain a recombined signal sequence; the denoising module is used for performing multi-scale decomposition and denoising processing on the recombined signal sequence to obtain a smooth data set; the feature extraction module is used for extracting time domain features and frequency domain features from the smooth data set and constructing a multi-dimensional feature vector set; The trend judging module is used for comparing the feature vector set with the pre-acquired historical feature data in time dimension and analyzing the long-term change trend of the features; the anomaly analysis module is used for locating potential anomaly points and analyzing the coupling influence intensity among multiple variables if the long-term change trend shows an anomaly fluctuation phenomenon, and converting the coupling influence intensity into anomaly fluctuation probability; the index calculation module is used for deducing a fault evolution path according to the abnormal fluctuation probability, defining a signal change range and calculating a dynamic index with the ascending fault probability by combining the abnormal fluctuation probability, the signal change range and the fluctuation amplitude of the long-term change trend; The early warning generation module is used for comparing the dynamic index with a preset risk judgment threshold value, carrying out grading evaluation on the fault risk according to the amplitude of the dynamic index exceeding the risk judgment threshold value, and fusing a grading evaluation result and the change trend of the dynamic index to generate an accurate fault early warning signal.

Description

Fault diagnosis method and system for self-adaptive sand supply device Technical Field The invention relates to the technical field of chips of the Internet of things, in particular to a fault diagnosis method and system of a self-adaptive sand supply device. Background At present, the rapid development of the chip technology of the Internet of things provides a core support for intelligent monitoring of industrial equipment, and the characteristics of low power consumption and high integration level can realize efficient acquisition and real-time processing of multi-dimensional data. The sand supply device is used as key equipment in industrial production, the operation stability of the sand supply device directly influences production continuity and economic benefit, and an accurate and efficient fault diagnosis system constructed based on an Internet of things chip becomes a need to be solved in industry. In the prior art, the fault diagnosis of the sand supply device mostly adopts a traditional fixed rule judgment or single threshold monitoring mode, and the fault diagnosis is judged by directly comparing the single dimension data such as vibration, temperature or pressure and the like acquired by an independent sensor with a preset threshold value. The method does not combine the technical advantages of the internet of things chip, cannot realize synchronous integration and deep analysis of multichannel data, and can only passively respond to dominant faults. In sum, the prior art is difficult to distinguish the effective normal fluctuation of the sand supply device from the real fault signal, and the requirement of industrial production on intelligent operation and maintenance of the sand supply device under the energization of the chip of the Internet of things cannot be met. Disclosure of Invention The invention provides a fault diagnosis method and a fault diagnosis system for a self-adaptive sand supply device, which are used for distinguishing effective normal fluctuation and real fault signals of the sand supply device and meeting the requirement of industrial production on intelligent operation and maintenance of the sand supply device under the energization of an Internet of things chip. In order to solve the above technical problems, the present invention provides a fault diagnosis method and system for an adaptive sand supply device, including: Acquiring multichannel original data for the operation of a sand supply device, wherein the multichannel original data comprise vibration data, temperature data and pressure data; Performing time synchronization and frequency division processing on the multichannel original data, intercepting a time domain signal segment exceeding a preset reference energy threshold, and recombining the time domain signal segment to obtain a recombined signal sequence; performing multi-scale decomposition and denoising treatment on the recombined signal sequence to obtain a smooth data set; extracting time domain features and frequency domain features from the smoothed data set, and constructing a multi-dimensional feature vector set; Comparing the feature vector set with the pre-acquired historical feature data in a time dimension, and analyzing the long-term change trend of the feature; if the long-term change trend shows an abnormal fluctuation phenomenon, positioning potential abnormal points and analyzing the coupling influence intensity among multiple variables, and converting the coupling influence intensity into abnormal fluctuation probability; Deducing a fault evolution path according to the abnormal fluctuation probability, defining a signal change range, and calculating a dynamic index with the ascending fault probability by combining the abnormal fluctuation probability, the signal change range and the fluctuation amplitude of the long-term change trend; And comparing the dynamic index with a preset risk judgment threshold value, carrying out grading evaluation on the fault risk according to the amplitude of the dynamic index exceeding the risk judgment threshold value, and fusing a grading evaluation result and the change trend of the dynamic index to generate an accurate fault early warning signal. In a second aspect, the present invention provides an adaptive sand supply device fault diagnosis system, comprising: the data acquisition module is used for acquiring multichannel original data of the sand supply device, wherein the multichannel original data comprise vibration data, temperature data and pressure data; The signal processing module is used for carrying out time synchronization and frequency division processing on the multichannel original data, intercepting a time domain signal segment exceeding a preset reference energy threshold value, and recombining the time domain signal segment to obtain a recombined signal sequence; the denoising module is used for performing multi-scale decomposition and denoising processing on the recombined signal s