Search

CN-121982861-A - Attendance gate inhibition machine fault prediction method and system

CN121982861ACN 121982861 ACN121982861 ACN 121982861ACN-121982861-A

Abstract

The invention relates to the technical field of fault prediction and discloses a method and a system for predicting faults of an attendance gate inhibition machine, wherein the method comprises the steps of obtaining operation time stamps and response delay data and constructing an original time sequence; the method comprises the steps of segmenting to obtain a continuous window sequence, calculating window statistical characteristics, marking a window to be detected and recording first window marking information if the window statistical characteristics meet window statistical conditions, generating a delay offset sequence and a delay increment sequence, calculating a homodromous deviation accumulated value, generating a primary early warning signal if the homodromous deviation accumulated value reaches a judging limit, carrying out association operation to obtain second window marking information, calculating to obtain a current performance degradation score, obtaining an adjusted score threshold according to historical performance degradation scores, calculating a continuous overrun count according to the current performance degradation score and the adjusted score threshold, and outputting a predictive maintenance signal if the continuous overrun count reaches the judging times. The method can accurately identify the potential faults of the equipment.

Inventors

  • WU CHAOQIANG

Assignees

  • 广州睿泰智能设备科技股份有限公司

Dates

Publication Date
20260505
Application Date
20260126

Claims (9)

  1. 1. The method for predicting the fault of the attendance gate inhibition machine is characterized by comprising the following steps of: acquiring an operation time stamp and response delay data of an attendance gate inhibition machine, and constructing an original time sequence according to the operation time stamp and the response delay data; segmenting the original time sequence according to a preset time window to obtain a continuous window sequence, calculating window statistical characteristics based on the continuous window sequence, marking a window corresponding to the window statistical characteristics as a window to be detected if the window statistical characteristics meet preset window statistical conditions, and recording window information of the window to be detected as first window marking information; Generating a delay offset sequence according to window response delay data in the window to be detected, and performing first-order difference calculation on the delay offset sequence to obtain a delay increment sequence; Calculating a homodromous deviation accumulated value according to the delay increment sequence, if the homodromous deviation accumulated value is larger than or equal to a preset judgment limit, generating a primary early warning signal, and carrying out association operation on the primary early warning signal and the first window marking information to obtain second window marking information with an early warning state; according to the second window marking information and the homodromous deviation accumulated value, carrying out exponential weighted moving average calculation to obtain a current performance degradation score of a current scoring period; And dynamically adjusting a preset scoring threshold according to the historical performance degradation score of the historical scoring period to obtain an adjusted scoring threshold, calculating a continuous overrun count according to the current performance degradation score and the adjusted scoring threshold, and if the continuous overrun count is greater than or equal to a preset judgment number of times, confirming that the attendance gate inhibition machine enters a performance decay period and outputting a predictive maintenance signal.
  2. 2. The method for predicting the failure of the attendance gate inhibition machine according to claim 1, wherein the steps of obtaining the operation time stamp and the response delay data of the attendance gate inhibition machine, and constructing an original time sequence according to the operation time stamp and the response delay data comprise the following steps: when the card swiping trigger signal is detected, synchronously recording an operation time stamp and response delay data; and removing data points smaller than a preset hardware debounce threshold according to the operation time stamp and the response delay data to obtain an original time sequence.
  3. 3. The method for predicting the failure of the attendance gate inhibition machine according to claim 1, wherein the segmenting the original time sequence according to the preset time window to obtain the continuous window sequence comprises the following steps: calculating subinterval indexes of the original time sequence according to a preset time window, wherein the subinterval indexes comprise a starting index and an ending index of each subinterval in the sliding process; And intercepting the original time sequence according to the subinterval index, and performing null filling to obtain a continuous window sequence.
  4. 4. The attendance gate inhibition machine fault prediction method according to claim 1, wherein calculating window statistics features based on the continuous window sequences, if the window statistics features meet preset window statistics conditions, marking a window corresponding to the window statistics features as a window to be detected, and recording window information of the window to be detected as first window marking information includes: Calculating window mean and window standard deviation of a current window according to the continuous window sequence, and calculating reference mean and reference standard deviation of a previous window according to the continuous window sequence, wherein the window statistical characteristics comprise the window mean and the window standard deviation of the current window, and the reference mean and the reference standard deviation of the previous window; calculating a mean deviation ratio according to the window mean value and the reference mean value; Calculating standard deviation ratio according to the window standard deviation and the reference standard deviation; If the mean deviation ratio is greater than a preset mean deviation threshold and the standard deviation ratio is greater than a preset standard deviation threshold, generating a window mark to be detected, and obtaining window to be detected and corresponding first window mark information, wherein the preset window statistical condition comprises a preset mean deviation threshold and a preset standard deviation threshold.
  5. 5. The method for predicting the failure of the access control device according to claim 1, wherein the generating a delay offset sequence according to the window response delay data in the window to be detected, and performing first-order differential calculation on the delay offset sequence to obtain a delay increment sequence comprises: Extracting window response delay data and window operation time stamps in the window to be detected, determining a historical reference mean value according to a historical window which is not marked as the window to be detected, and performing difference calculation on the window response delay data and the historical reference mean value to obtain a delay offset value; serializing the delay offset values based on the window operation time stamp to construct a delay offset sequence; performing first-order differential operation on the delay offset sequence, and calculating the difference value of adjacent values in the delay offset sequence to obtain delay fluctuation quantity; And collecting the delay fluctuation quantity to obtain a delay increment sequence.
  6. 6. The method for predicting the failure of the attendance gate inhibition machine according to claim 1, wherein the calculating the cumulative value of the homodromous deviation according to the delay increment sequence, if the cumulative value of the homodromous deviation is greater than or equal to a preset judgment limit, generating a primary early warning signal, performing an association operation on the primary early warning signal and the first window marking information to obtain second window marking information with an early warning state, comprises: Extracting continuous delay increment values according to the delay increment sequence, and checking the symbol characteristics of the delay increment values if the number of the delay increment values is greater than or equal to a preset minimum continuous number threshold; If the symbol characteristics accord with consistency verification, accumulating the absolute value of the delay increment value to obtain a homodromous deviation accumulated value; If the homodromous deviation accumulated value is larger than or equal to a preset judgment limit, generating a primary early warning signal, wherein the primary early warning signal comprises a signal operation time stamp; And retrieving a signal window mark corresponding to the first window mark information according to the signal operation time stamp, and correlating the primary early warning signal with the signal window mark to obtain second window mark information with an early warning state.
  7. 7. The method for predicting the failure of the access control device according to claim 6, wherein the performing an exponentially weighted moving average calculation according to the second window marking information and the homodromous deviation accumulated value to obtain a current performance degradation score of a current scoring period includes: integrating the primary early warning signal and the homodromous deviation accumulated value according to the signal operation time stamp to obtain a current period deviation data set; Through statistical analysis on the historical window data set, weight is distributed to the second window marking information, and weight attributes are obtained; Performing numerical mapping calculation on the current period deviation data set according to the weight attribute to obtain an instantaneous performance numerical value of the current scoring period; performing attenuation treatment on the performance degradation score of the previous scoring period according to a preset weighting coefficient to obtain a historical scoring component; And carrying out exponential weighted moving average calculation on the historical scoring component and the instantaneous performance numerical value to obtain a current performance degradation score of a current scoring period.
  8. 8. The method for predicting the failure of the access control device according to claim 1, wherein dynamically adjusting the preset scoring threshold according to the historical performance degradation score of the historical scoring period to obtain an adjusted scoring threshold, and calculating a continuous overrun count according to the current performance degradation score and the adjusted scoring threshold comprises: Calculating to obtain a fluctuation rate characteristic and a trend slope according to the historical performance degradation score of the historical scoring period; constructing an adaptive adjustment factor according to the fluctuation rate characteristics and the trend slope, and acting the adaptive adjustment factor on a preset scoring threshold to obtain an adjusted scoring threshold; and comparing the current performance degradation score with the adjusted score threshold, and accumulating a continuous overrun count if the current performance degradation score is larger than the adjusted score threshold.
  9. 9. An attendance gate inhibition machine fault prediction system, which is characterized by comprising: the data acquisition module is used for acquiring an operation time stamp and response delay data of the attendance gate inhibition machine and constructing an original time sequence according to the operation time stamp and the response delay data; The window marking module is used for segmenting the original time sequence according to a preset time window to obtain a continuous window sequence, calculating window statistical characteristics based on the continuous window sequence, marking a window corresponding to the window statistical characteristics as a window to be detected if the window statistical characteristics meet preset window statistical conditions, and recording window information of the window to be detected as first window marking information; The sequence processing module is used for generating a delay offset sequence according to the window response delay data in the window to be detected, and performing first-order difference calculation on the delay offset sequence to obtain a delay increment sequence; The device early warning module is used for calculating a homodromous deviation accumulated value according to the delay increment sequence, generating a primary early warning signal if the homodromous deviation accumulated value is larger than or equal to a preset judgment limit, and carrying out association operation on the primary early warning signal and the first window marking information to obtain second window marking information with an early warning state; The scoring calculation module is used for carrying out index weighted moving average calculation according to the second window marking information and the homodromous deviation accumulated value to obtain a current performance degradation score of a current scoring period; And the result output module is used for dynamically adjusting a preset scoring threshold according to the historical performance degradation score of the historical scoring period to obtain an adjusted scoring threshold, calculating a continuous overrun count according to the current performance degradation score and the adjusted scoring threshold, and if the continuous overrun count is greater than or equal to the preset judgment times, confirming that the attendance gate inhibition machine enters a performance decay period and outputting a predictive maintenance signal.

Description

Attendance gate inhibition machine fault prediction method and system Technical Field The invention relates to the technical field of fault prediction, in particular to an attendance gate inhibition machine fault prediction method and system. Background In the long-term and high-frequency operation process of the attendance gate inhibition machine, the internal hardware modules and related circuit components of the attendance gate inhibition machine are inevitably influenced by environmental factors and ageing of devices. In order to ensure the stable operation of the system, the operation data of related equipment are gradually incorporated into an industrial data analysis system, and the system has become an application trend in the industry field. In the prior art, the operation state of the attendance gate inhibition machine is monitored mainly by adopting a manual inspection, periodic detection or alarming mode based on a threshold value. For example, by detecting operation parameters such as card swiping response time and card reading success rate, and comparing the detection result with a preset threshold, when the detection value exceeds the threshold, the abnormality of the equipment is determined. The method is simple to implement and can identify obvious faults. In the actual running process, the performance change of the attendance gate inhibition machine usually has a progressive characteristic, the response speed or stability of the attendance gate inhibition machine often slowly changes in a longer time, and the single detection result can still be in a normal range. If the judgment is carried out only according to single or small amount of detection data, the real change trend of the equipment performance is difficult to accurately reflect, and potential abnormal signals are easy to ignore. In summary, it is difficult in the prior art to identify in time the progressive performance degradation of the device during long-term operation, resulting in an inability to accurately identify potential faults of the device. Disclosure of Invention The invention provides an attendance gate inhibition machine fault prediction method and system, which are used for accurately identifying potential faults of equipment. In order to solve the technical problems, the invention provides a method for predicting the faults of an attendance gate inhibition machine, which comprises the following steps: acquiring an operation time stamp and response delay data of an attendance gate inhibition machine, and constructing an original time sequence according to the operation time stamp and the response delay data; segmenting the original time sequence according to a preset time window to obtain a continuous window sequence, calculating window statistical characteristics based on the continuous window sequence, marking a window corresponding to the window statistical characteristics as a window to be detected if the window statistical characteristics meet preset window statistical conditions, and recording window information of the window to be detected as first window marking information; Generating a delay offset sequence according to window response delay data in the window to be detected, and performing first-order difference calculation on the delay offset sequence to obtain a delay increment sequence; Calculating a homodromous deviation accumulated value according to the delay increment sequence, if the homodromous deviation accumulated value is larger than or equal to a preset judgment limit, generating a primary early warning signal, and carrying out association operation on the primary early warning signal and the first window marking information to obtain second window marking information with an early warning state; according to the second window marking information and the homodromous deviation accumulated value, carrying out exponential weighted moving average calculation to obtain a current performance degradation score of a current scoring period; And dynamically adjusting a preset scoring threshold according to the historical performance degradation score of the historical scoring period to obtain an adjusted scoring threshold, calculating a continuous overrun count according to the current performance degradation score and the adjusted scoring threshold, and if the continuous overrun count is greater than or equal to a preset judgment number of times, confirming that the attendance gate inhibition machine enters a performance decay period and outputting a predictive maintenance signal. In a second aspect, the present invention provides an attendance gate inhibition machine fault prediction system, including: the data acquisition module is used for acquiring an operation time stamp and response delay data of the attendance gate inhibition machine and constructing an original time sequence according to the operation time stamp and the response delay data; The window marking module is used for segmenting th