A method and system for monitoring the operation and maintenance status of sealing cover stamping and forming equipment

Abstract

The present invention relates to the field of data processing technology, and more specifically, to a method and system for monitoring the operation and maintenance status of a sealing cover stamping and forming device. The method comprises: obtaining a pressure data sequence of the sealing cover stamping and forming device; utilizing an improved EWMA algorithm to obtain an observation value of a data point to be filtered in the pressure data sequence; and performing a weighted summation of the observation value and the filter value of the previous data point of the data point to be filtered based on a preset attenuation weight to obtain a filter value of the data point to be filtered, thereby monitoring the operation and maintenance status of the device based on the filter value of each data point in the pressure data sequence. The present invention can obtain accurate pressure data, thereby enabling precise monitoring of the operation and maintenance status of the device based on the accurate pressure data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more particularly to a method and system for monitoring the operation and maintenance status of a sealing cover stamping and forming device. Background Art

[0002] Sealing cover stamping equipment is a mechanical device specifically designed for producing sealing covers. Its production principle is primarily based on the principle of metal plastic deformation. The stamping equipment and die apply pressure to the material, causing it to deform or separate, thereby achieving the desired sealing cover shape and size. During the stamping process, pressure is one of the core process parameters, and its changes directly reflect the equipment's operating status. By collecting and analyzing pressure data in real time, abnormal equipment operating conditions can be effectively identified, providing early warning of potential failures. This allows for precise maintenance and efficient management, further improving equipment reliability and production efficiency.

[0003] Because pressure data collection is susceptible to noise, directly using this data to monitor equipment operation and maintenance status can result in poor accuracy. Therefore, data cleaning is necessary. The EWMA (Exponentially Weighted Moving Average) algorithm is a commonly used time series smoothing technique that captures the latest data trends by assigning different weights to observations at different time points. Therefore, the EWMA algorithm can be used for data cleaning.

[0004] However, due to the uncertainty of noise in the pressure data collected during the equipment production process, directly applying the EWMA algorithm for data cleaning may result in poor filtering results. Specifically, if the collected observations are noisy data, the filtering results will be biased, which will affect the subsequent filtering effect, resulting in the inability to obtain accurate pressure data and thus unable to accurately monitor the operation and maintenance status of the monitoring equipment. Summary of the Invention

[0005] To address the problem that the EWMA algorithm has a poor filtering effect on pressure data during the equipment production process, resulting in an inability to accurately monitor the equipment's operation and maintenance status based on the filtered data, the present invention provides a method and system for monitoring the operation and maintenance status of sealing cover stamping equipment.

[0006] According to a first aspect of the present invention, a method for monitoring the operation and maintenance status of a sealing cover stamping and forming device is provided, comprising:

[0007] Obtaining a pressure data sequence of a sealing cover stamping and forming device;

[0008] The improved EWMA algorithm is used to obtain the observation value of the data point to be filtered in the pressure data sequence. Based on the preset attenuation weight, the observation value and the filtered value of the previous data point of the data point to be filtered are weighted summed to obtain the filtered value of the data point to be filtered. The operation and maintenance status of the equipment is monitored based on the filtered value of each data point in the pressure data sequence.

[0009] The method for obtaining the observation value of the data point to be filtered includes: calculating the noise level of the data point to be filtered, where the noise level represents the mutation of the data point to be filtered within the neighborhood;

[0010] The noise level is corrected using the correlation coefficient between the data point to be filtered and the corresponding data points in the adjacent working cycles. The correction value is negatively correlated with the correlation coefficient. The correction value of the noise level of each data point in the neighborhood of the data point to be filtered is used as the weight. The weighted average of all data points in the neighborhood is performed to obtain the observation value of the data point to be filtered.

[0011] The pressure data sequence contains data points of multiple working cycles.

[0012] The present invention can reduce the influence of noise data when the EWMA algorithm is used to filter the pressure data in the equipment production process, thereby improving the filtering effect and obtaining accurate pressure data, so that based on the accurate pressure data, accurate monitoring of the operation and maintenance status of the sealing cover stamping and forming equipment can be achieved.

[0013] Preferably, the method of distinguishing the data points of each working cycle in the pressure data sequence includes:

[0014] Filter out data points that are identical to the preceding and following adjacent data points from the pressure data sequence to obtain a stable data point set;

[0015] DBSCAN clustering is performed on the stable data point set, and the time period consisting of the corresponding moments of the first data point and the last data point in each cluster is defined as a pressure holding stage. The time intervals between adjacent pressure holding stages in the time series are used to distinguish each working cycle and obtain the working cycle corresponding to each data point.

[0016] The present invention utilizes the characteristics of the pressure data in the pressure holding stage to distinguish the various working cycles in the pressure data sequence, thereby reducing the difficulty of distinction.

[0017] Preferably, the method for obtaining the correlation coefficient between the data point to be filtered and the corresponding data points in the preceding and following adjacent working cycles includes:

[0018] Calculate the Pearson correlation coefficient between the data points in the neighborhood of the data point to be filtered and the data points in the neighborhood of the corresponding data points in the previous and next adjacent working cycles, and take the average value to obtain the correlation coefficient between the data point to be filtered and the corresponding data points in the previous and next adjacent working cycles.

[0019] When calculating the correlation coefficient, the present invention can take into account that noise data may affect the values ​​of surrounding data, thereby increasing the amount of data of the correlation coefficient between the data point to be filtered and the corresponding data points in the previous and next adjacent working cycles, thereby ensuring the accuracy of the determined correlation coefficient.

[0020] Preferably, the noise level of the data point to be filtered satisfies the following relationship:

[0021] ;

[0022] Where, is the first The noise level of each data point; is the first data points; is the first The first data point in the neighborhood of data points; is the first The number of data points in the neighborhood of a data point; is the absolute value symbol; is the hyperbolic tangent function.

[0023] Preferably, calculating the noise level of the data point to be filtered also includes:

[0024] The variance of the data points in the neighborhood of the data point to be filtered is used as the noise level of the data point to be filtered.

[0025] Preferably, the noise level is corrected using the correlation coefficient between the data point to be filtered and the corresponding data points in the adjacent working cycles to satisfy the following relationship:

[0026] ;

[0027] Where, 、 They are the first The noise level of each data point before and after correction; is the first data points; is the first The average value of the Pearson correlation coefficients of the data points in the neighborhood of the data point and the data points in the neighborhood of the corresponding data points in the previous and next adjacent working cycles.

[0028] The present invention can reduce the interference of normal data changes on the identification of noise data, thereby accurately evaluating the possibility that the data point to be filtered is noise data.

[0029] Preferably, the observation value of the data point to be filtered satisfies the following relationship:

[0030] ;

[0031] Where, is the first The observed value of data points; is the first The first data point in the neighborhood of data points; is the first The number of data points in the neighborhood of a data point; is the natural exponential function.

[0032] The observation values ​​determined by the present invention can reduce the influence of noise data, thereby improving the filtering effect of the EWMA algorithm.

[0033] Preferably, the filter value of the data point to be filtered satisfies the following relationship:

[0034] ;

[0035] Where, is the first The filtered value of the data point; is the first The observed value of data points; is the first The filtered value of the data point; The preset attenuation weight.

[0036] Preferably, the neighborhood range of the data point to be filtered is a length range centered on the data point to be filtered and surrounding a preset number of data points.

[0037] According to a second aspect of the present invention, a system for monitoring the operation and maintenance status of a sealing cover stamping and forming device is provided. The system includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the first aspect of the present invention.

[0038] The present invention has the following effects:

[0039] The present invention corrects the noise level, can eliminate the interference of normal data changes on the recognition of noise data, and ensure the accuracy of the weights of each data point in the neighborhood of the data point to be filtered, thereby reducing the impact of noise data on the observation value of the data point to be filtered, and then can realize accurate filtering of the data point to be filtered based on accurate observation values, so as to realize accurate monitoring of the equipment operation and maintenance status based on the filter value of each data point in the pressure data sequence. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood by reading the following detailed description with reference to the accompanying drawings. In the accompanying drawings, several embodiments of the present invention are shown in an illustrative and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0041] Figure 1 The present invention is a flowchart of the steps of a method for monitoring the operation and maintenance status of a sealing cover stamping and forming device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work shall fall within the scope of protection of the present invention.

[0043] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0044] Reference Figure 1 A method for monitoring the operation and maintenance status of a sealing cover stamping and forming device includes steps S1 and S2, specifically as follows:

[0045] S1: Obtain the pressure data sequence of the sealing cover stamping equipment.

[0046] Specifically, the pressure sensor can be installed at the pressure head position of the sealing cover stamping equipment, so that the pressure data during the stamping process can be monitored in real time according to the set sampling frequency, such as 5Hz, to obtain a pressure data sequence. This embodiment does not specifically limit the sampling frequency.

[0047] S2: Use the improved EWMA algorithm to obtain the observation value when filtering each data point in the pressure data sequence, and based on the preset attenuation weight, perform weighted summation on the observation value of any data point and the filtered value of the previous data point of any data point to obtain the filtered value of any data point, so as to monitor the operation and maintenance status of the equipment based on the filtered value of each data point.

[0048] The pressure data sequence contains data points of multiple working cycles.

[0049] It should be noted that the EWMA (Exponentially Weighted Moving Average) algorithm obtains the filtered value by assigning exponentially decreasing weights to data points and performing weighted averaging. That is, the data points closer to the current moment have larger weights, while the data points farther away from the current moment have smaller weights. This allows the algorithm to more sensitively reflect the changing trends of recent data.

[0050] However, due to the influence of factors such as the working environment, the pressure data collected during the stamping process may contain noise data. If the data point closer to the data point at the current moment is noise data, the filter value determined based on the traditional EWMA algorithm will have a large deviation, and the filtering effect will be poor. In addition, the algorithm performs filtering in a recursive manner, that is, the accuracy of the filter value at the current moment will affect the accuracy of the filter value at the next moment, thereby causing the subsequent filtering effects to deviate from the normal level, affecting the final filtering effect. Therefore, the present invention improves the traditional EWMA algorithm. The specific improvements are as follows: first, the noise level of each data point in the neighborhood range of the data point to be filtered is measured; then the noise level of each data point in the neighborhood range is corrected; finally, the correction value is used as a weight, and the weighted average of the calculated data points in the neighborhood range is used as the observation value of the data point to be filtered, so as to improve the determination of the observation value of the data point to be filtered in the traditional EWMA algorithm.

[0051] The present invention only improves the determination of the observation values ​​of the data points to be filtered in the EWMA algorithm, and does not improve the determination process of other parameters.

[0052] Specifically, the process of obtaining the observation value of the data point to be filtered using the improved EWMA algorithm includes the following steps:

[0053] Step 1: Calculate the noise level of the data point to be filtered. The noise level represents the mutation of the data point to be filtered within the neighborhood.

[0054] It should be noted that noise data usually appears as a prominent point in a local range. Therefore, the present invention uses this feature to measure the mutation of each data point in its neighborhood, thereby evaluating the possibility that each data point is noise data.

[0055] In an exemplary embodiment of the present invention, the neighborhood range of the data point to be filtered is a length range centered on the data point to be filtered and encompassing a preset number of data points. In this embodiment, the preset number is 11, and this embodiment does not specifically limit the size of the preset number.

[0056] Furthermore, after the neighborhood range of the data point to be filtered is determined, the noise level of the data point to be filtered can be calculated by evaluating the mutation of the data point to be filtered within its neighborhood range.

[0057] Specifically, the noise level of the data point to be filtered satisfies the following relationship:

[0058] ;

[0059] Where, is the first The noise level of each data point; is the first data points; is the first The first data point in the neighborhood of data points; is the first The number of data points within the neighborhood of a data point, in this embodiment =11; is the absolute value symbol; is the hyperbolic tangent function, used for normalization.

[0060] Among them, when The larger the The deviation between the overall value of a data point and the data points in the neighborhood of the data point is large, which indicates that the data point is likely to be noise data.

[0061] In another embodiment, the noise level of the data point to be filtered can be determined by the following steps:

[0062] The variance of the data points in the neighborhood of the data point to be filtered is used as the noise level of the data point to be filtered.

[0063] Furthermore, the noise level of each data point in the pressure data sequence can be obtained by any noise level calculation method.

[0064] Step 2: Use the correlation coefficient between the data point to be filtered and the corresponding data points in the adjacent working cycles to correct the noise level. The correction value is negatively correlated with the correlation coefficient.

[0065] It's important to note that during the operation of sealing cap stamping equipment, some normal process phenomena can cause sudden changes in pressure data. For example, the material's transition from elastic to plastic deformation, as well as switching between different stages of the stamping process, can cause fluctuations in pressure data, resulting in a significant level of noise in the normally fluctuating pressure data. Therefore, it's necessary to correct the noise level at each data point to reduce the impact of normal data fluctuations.

[0066] It should be further explained that the sealing cover stamping and forming equipment has periodicity during the working process, that is, each production of a sealing cover is a working cycle, and the pressure data changes in each working cycle are similar. Among them, the pressure change in each working cycle is divided into three stages: the pre-tightening stage, the mold applies initial pressure to make the material begin to elastically deform, and the pressure rises slowly; the working stage, the pressure head continues to apply pressure, the material enters plastic deformation and fits the mold, the pressure rises significantly and reaches a peak value; the pressure holding stage, after the material fits the mold, a certain pressure is maintained to ensure the dimensional accuracy and stability of the sealing cover, and the pressure remains stable. Therefore, the present invention uses the correlation coefficient between the data point to be filtered and the corresponding data points in the two adjacent working cycles to correct the noise level of the data point to be filtered, so as to eliminate the influence of normal data changes.

[0067] In an exemplary embodiment of the present invention, the duty cycle of each data point in the pressure data sequence may be determined by the following steps:

[0068] Data points that are identical to the previous and next adjacent data points are screened out from the pressure data sequence to obtain a stable data point set. The stable data point set is clustered using DBSCAN, and the time period consisting of the corresponding moments of the first and last data points in each cluster in the time series is defined as a pressure holding stage. The working cycles are distinguished by the time intervals between adjacent pressure holding stages in the time series, and the working cycle corresponding to each data point is obtained.

[0069] It should be noted that, since the pressure data of the holding stage in each working cycle remains unchanged, the present invention can obtain the data points of the holding stage in all working cycles by screening the stable data point set, so that when clustering is performed using the density-based clustering algorithm, the data points of the holding stage in each working cycle can be clustered into one category, thereby obtaining the holding stage of each working cycle, and then distinguishing each working cycle to obtain the working cycle corresponding to each data point.

[0070] Optionally, when performing DBSCAN clustering on the stable data point set, the selected clustering radius is 2. This embodiment does not impose any special limitation on the size of the clustering radius.

[0071] Furthermore, after determining the duty cycle of each data point, the correlation coefficient between the data point to be filtered and the corresponding data points in the previous and next duty cycles can be calculated, so as to correct the noise level of the data point to be filtered based on the correlation coefficient.

[0072] In an exemplary embodiment of the present invention, the correlation coefficient of the data points to be filtered can be determined by the following steps:

[0073] Calculate the Pearson correlation coefficient between the data points in the neighborhood of the data point to be filtered and the data points in the neighborhood of the corresponding data points in the previous and next adjacent working cycles, and take the average value to obtain the correlation coefficient between the data point to be filtered and the corresponding data points in the previous and next adjacent working cycles.

[0074] For example, the working period of the data point to be filtered can be recorded as , then the previous adjacent working period of the data point to be filtered is , the adjacent working period of the data point to be filtered is , if the data point to be filtered is The data points, then as well as The data points as the data points corresponding to the data points to be filtered in the corresponding working cycle, so that the data points in the neighborhood of the data points to be filtered can be calculated. as well as The Pearson correlation coefficient of the data points in the neighborhood of the corresponding data point in is calculated, and the average value is taken to obtain the correlation coefficient between the data point to be filtered and the corresponding data points in the adjacent working cycles before and after.

[0075] It should be noted that the method for determining the Pearson correlation coefficient between data sequences is a prior art and will not be described in detail in this embodiment.

[0076] Furthermore, the correlation coefficient between the data point to be filtered and the corresponding data points in the adjacent working cycles is used to correct the noise level of the data point to be filtered to satisfy the following relationship:

[0077] ;

[0078] Where, 、 They are the first The noise level of each data point before and after correction; is the first data points; is the first The average value of the Pearson correlation coefficients of the data points in the neighborhood of the data point and the data points in the neighborhood of the corresponding data points in the previous and next adjacent working cycles.

[0079] It should be noted that, since the value range of the Pearson correlation coefficient is -1 to 1, under normal circumstances, the corresponding data points in different working cycles are positively correlated, that is, The value of is positive; then when When the value of is negative, it means that the data points in the neighborhood of the data point to be filtered may be interfered by the noise data, which further indicates that the data point to be filtered may be noise data, and The closer the value of is to -1, the greater the possibility that the data point to be filtered is noise data, and it is necessary to make a smaller correction to the noise level of the data point to be filtered, or even no correction. The larger it is, the less likely the data point to be filtered is noise. The smaller it is, or even negative, the more likely the data point to be filtered is noise.

[0080] It should be further explained that the present invention Characterize the negative correlation between the correction value and the correlation coefficient, and use Limit its value to 0-1.

[0081] In another embodiment, the Spearman rank correlation coefficient, the Kendall rank correlation coefficient, etc. may also be used to determine the correlation coefficient between the data point to be filtered and the corresponding data points in the preceding and following adjacent working cycles.

[0082] Step 3: Use the corrected value of the noise level of each data point in the neighborhood of the data point to be filtered as the weight, and perform weighted averaging on all data points in the neighborhood to obtain the observation value of the data point to be filtered.

[0083] Specifically, the observation value of the data point to be filtered satisfies the following relationship:

[0084] ;

[0085] Where, is the first The observed value of data points; is the first The first data point in the neighborhood of data points; is the first The number of data points in the neighborhood of a data point; is a natural exponential function, where the natural exponential function refers to a function with a natural constant An exponential function with base .

[0086] Optionally, when the correction value of the noise level of any data point in the neighborhood of the data point to be filtered is large, it means that the credibility of the data point is low. Setting a lower weight for the data point can avoid the interference of noise data, thereby ensuring the accuracy of the observation value of the data point to be filtered.

[0087] Furthermore, after the observation value of the data point to be filtered is determined, the observation value and the filtered value of the previous data point of the data point to be filtered may be weightedly summed based on a preset attenuation weight to obtain the filtered value of the data point to be filtered.

[0088] Specifically, the filter value of the data point to be filtered satisfies the following relationship:

[0089] ;

[0090] Where, is the first The filtered value of the data point; is the first The observed value of data points; is the first The filtered value of the data point; To preset the attenuation weight, in this embodiment =0.7.

[0091] It should be noted that when the improved EWMA algorithm is used to filter the pressure data sequence, the initial EWMA value is the observation value of the first data point in the pressure data sequence.

[0092] Furthermore, the data points in the pressure data sequence can be used as data points to be filtered in sequence, so that the improved EWMA algorithm can be used to clean the acquired pressure data to obtain accurate pressure data. Based on the accurate pressure data, anomalies in the pressure data, such as excessively high or too low pressure, can be discovered in a timely manner, so that the abnormal operation and maintenance status of the equipment can be identified in a timely manner, and corresponding measures can be taken to repair the equipment to extend the service life of the equipment and ensure product quality.

[0093] The present invention also provides an operation and maintenance status monitoring system for a sealing cover stamping and forming device. The system includes a memory and a processor, and a computer program is stored in the memory. The computer program integrates the functions of a method for monitoring the operation and maintenance status of a sealing cover stamping and forming device. When the computer program is executed, accurate pressure data can be obtained through a method for monitoring the operation and maintenance status of a sealing cover stamping and forming device, thereby enabling accurate monitoring of the equipment operation and maintenance status based on the accurate pressure data.

[0094] In the description of this specification, "multiple" and "several" mean at least two, such as two, three or more, etc., unless otherwise clearly defined.

[0095] While several embodiments of the present invention have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Numerous modifications, variations, and alternatives will occur to those skilled in the art without departing from the concept and spirit of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein may be employed in practicing the present invention.

Claims

1. A method for monitoring the operation and maintenance status of a sealing cover stamping and forming device, characterized in that: include: Obtaining a pressure data sequence of a sealing cover stamping and forming device; Using the improved EWMA algorithm, the observation value of the data point to be filtered in the pressure data sequence is obtained, and based on the preset attenuation weight, the weighted sum of the observation value and the filtered value of the previous data point of the data point to be filtered is performed to obtain the filtered value of the data point to be filtered, so as to monitor the operation and maintenance status of the equipment based on the filtered value of each data point in the pressure data sequence; The method for obtaining the observation value of the data point to be filtered includes: calculating the noise level of the data point to be filtered, where the noise level represents the mutation of the data point to be filtered within the neighborhood; The noise level is corrected using the correlation coefficient between the data point to be filtered and the corresponding data points in the adjacent working cycles to satisfy the following relationship: , 、 They are the first The noise level of each data point before and after correction, is the first data points; is the first The average value of the Pearson correlation coefficient between the data points within the neighborhood of the data point and the data points within the neighborhood of the corresponding data points in the previous and next adjacent working cycles; the correction value is negatively correlated with the correlation coefficient; The correction value of the noise level of each data point in the neighborhood of the data point to be filtered is used as the weight, and all data points in the neighborhood are weighted averaged to obtain the observation value of the data point to be filtered, which satisfies the following relationship: , is the first The observed value of data points, is the first The first data point in the neighborhood of data points, is the first The number of data points in the neighborhood of a data point, is the natural exponential function; The pressure data sequence contains data points of multiple working cycles.

2. The method for monitoring the operation and maintenance status of a sealing cover stamping and forming device according to claim 1, characterized in that: Methods for distinguishing data points of each working cycle in the pressure data sequence include: Filtering data points that are identical to preceding and following adjacent data points from the pressure data sequence to obtain a stable data point set; DBSCAN clustering is performed on the stable data point set, and the time period consisting of the corresponding moments of the first data point and the last data point in each cluster cluster in the time series is defined as a pressure holding stage. The working cycles are distinguished by the time intervals between adjacent pressure holding stages in the time series to obtain the working cycle corresponding to each data point.

3. The method for monitoring the operation and maintenance status of a sealing cover stamping and forming device according to claim 1, wherein: The noise level of the data point to be filtered satisfies the following relationship: ； Where, is the first The noise level of each data point; is the first data points; is the first The first data point in the neighborhood of data points; is the first The number of data points in the neighborhood of a data point; is the absolute value symbol; is the hyperbolic tangent function.

4. The method for monitoring the operation and maintenance status of a sealing cover stamping and forming device according to claim 1, wherein: The calculation of the noise level of the data points to be filtered also includes: The variance of the data points in the neighborhood of the data point to be filtered is used as the noise level of the data point to be filtered.

5. The method for monitoring the operation and maintenance status of a sealing cover stamping and forming device according to claim 1, wherein: The filtered value of the data point to be filtered satisfies the following relationship: ； Where, is the first The filtered value of the data point; is the first The observed value of data points; is the first The filtered value of the data point; The preset attenuation weight.

6. The method for monitoring the operation and maintenance status of a sealing cover stamping and forming device according to claim 1, characterized in that: The neighborhood range of the data point to be filtered is a length range centered on the data point to be filtered and surrounding a preset number of data points.

7. An operation and maintenance status monitoring system for a sealing cover stamping and forming equipment, characterized in that: The operation and maintenance status monitoring system of the sealing cover stamping and forming equipment includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the operation and maintenance status monitoring method of the sealing cover stamping and forming equipment as described in any one of claims 1-6.

Images