An automatic inspection method for intelligent manufacturing

Abstract

The present invention discloses an automatic inspection method for intelligent manufacturing, which relates to the technical field of equipment inspection and solves the problem that when data is inspected and recorded in a top-down manner and some nodes of some equipment fail to generate data in time, the system will be left waiting, thereby prolonging the inspection time. The present invention prioritizes confirming the inspection record collection node, and then analyzes whether the corresponding node is a normal point or an abnormal point, and then analyzes the point data of the abnormal point to confirm the interval time difference between the past time. Subsequently, the virtual time point corresponding to the abnormal point is obtained, and then the specific time point of the virtual time point is used to construct a sorting table of the abnormal point, and then the overall inspection record collection table is generated by combining with the normal point. The inspection system collects the inspection data according to the inspection record collection table, changes the original data inspection collection method, improves the automatic inspection efficiency, and fully reduces the automatic inspection time.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of equipment inspection, in particular to an automatic inspection method for intelligent manufacturing. BACKGROUND

[0002] Intelligent manufacturing is derived from the research of artificial intelligence. Generally, intelligence is the sum of knowledge and intelligence, the former is the basis of intelligence, and the latter refers to the ability to acquire and use knowledge to solve problems.

[0003] The application with the patent publication No. CN109120071A discloses an automatic inspection system and method for pumped storage power station equipment. A plurality of high-definition panoramic cameras are arranged in each production area of the pumped storage power station. Each camera in the same production area is connected with a regional inspection data processing terminal of the production area. Each regional inspection data processing terminal is connected with a power station equipment inspection master control workstation. After the equipment inspection is started, the equipment is automatically inspected according to the pre-set parameters. The images of the inspection process are automatically recorded and compared with the image sample library for identification and analysis. It is automatically judged whether the equipment is abnormal. After the inspection is completed, the inspection image video file is automatically generated. The completion of this inspection and the abnormal situation are displayed in the form of short message. If there is an abnormal situation, an alarm prompt is sent and the abnormal image and video are pushed to remind the operation and maintenance personnel to check the equipment state. The application can repair the abnormal equipment in time, has strong timeliness, and has wide coverage in one inspection.

[0004] During the operation of the intelligent manufacturing equipment, a corresponding inspection system is used to record the data of each manufacturing equipment. However, there are still the following deficiencies in the specific inspection recording process, which need to be improved:

[0005] 1. The data is recorded in a top-down manner. When some devices and nodes do not generate data in time, the system will wait, thereby prolonging the inspection time.

[0006] 2. The inspection recorded data is not bundled with the associated data, so that the corresponding associated data cannot be found in time when the operator reviews it. SUMMARY

[0007] In view of the deficiencies of the prior art, the present application provides an automatic inspection method for intelligent manufacturing, which solves the problem that the data is recorded in a top-down manner, and when some devices and nodes do not generate data in time, the system will wait, thereby prolonging the inspection time.

[0008] To achieve the above purpose, the present application realizes the following technical scheme: an automatic inspection method for intelligent manufacturing, comprising the following steps:

[0009] S1, according to the inspection route formulated by the operator, the inspection system runs adaptive inspection on the intelligent manufacturing equipment belonging to the internal inspection route;

[0010] S2, during the inspection record process of different intelligent manufacturing equipment, according to the corresponding time point of the parameter data generated by the corresponding node of the intelligent manufacturing equipment, the inspection record collection table of the corresponding intelligent manufacturing equipment is constructed, and the specific mode is:

[0011] S21, confirm the collection object and the current time, and confirm a plurality of to-be-inspected collection nodes of the collection object, and then mark the point with data display as a normal point and the point without data display as an abnormal point from the plurality of to-be-inspected collection nodes;

[0012] S22, confirm the specific time point of the past parameters of the corresponding abnormal point of the collection object from the cloud, and mark different specific time points as SJ i-k , wherein i represents different abnormal points, k represents different time points, and CZ q = SJ i-k -SJ i-(k-1) is adopted, and the interval period difference CZ q between a plurality of different time points is obtained, wherein q=k-1, and k≥2;

[0013] S23, the maximum value and the minimum value existing in the plurality of interval period difference CZ q belonging to the same abnormal point are removed, and then the remaining plurality of interval period difference CZ q is subjected to mean value processing to obtain a limited mean value XJ, and the confirmed limited mean value XJ is taken as the standard value of the abnormal point;

[0014] S24, the time point of the last data display on the abnormal point is confirmed again, which is calibrated as an initial time point, and the initial time point+standard value=virtual time point is adopted, and then the plurality of abnormal points are sorted according to the specific value of the virtual time point to obtain an abnormal point sorting table;

[0015] S25, the normal points are sorted in a top-down manner to generate a normal point sorting table, the normal point sorting table and the abnormal point sorting table are combined, and the normal point sorting table is sorted in front, and the inspection record collection table is generated by combination, and subsequently, the inspection system inspects the records of different nodes in the equipment according to the inspection record collection table, and the recorded inspection data is subjected to next step processing;

[0016] S3, data capacity reduction processing is performed on the recorded inspection data, repeated data appearing in the inspection data is confirmed, and the repeated data is replaced using a corresponding repeated mark, the processed inspection data is marked as to-be-inducted data, and the next step is processed, and the specific mode is:

[0017] S31, the past data of the corresponding equipment in the inspection data is combined and analyzed to confirm the repeated data, and the repeated data appearing in the inspection data is marked, after the marking processing is completed, the different repeated data is replaced using the repeated mark;

[0018] S32, after the repeated data in the inspection data is completely replaced, the corresponding repeated mark and repeated data are bundled to generate a repeated data matching table and stored, after the repeated data is replaced, the corresponding inspection data is marked as to-be-inducted data;

[0019] S4, when the inspection data of different nodes in the to-be-inducted data is inducted, the corresponding data with the correlation degree of the current inspection data is analyzed and confirmed from the to-be-inducted data, and the similar data is sorted according to the similarity between the data and placed at the back end of the current inspection data, after the processing is completed, the corresponding device inspection data package is generated, and the specific mode is:

[0020] S41, a group of inspection data is randomly selected from the to-be-inducted data, which is marked as the current inspection data, the other data in the to-be-inducted data is analyzed with the current inspection data, and the corresponding coincidence degree parameter is recorded, and the recorded coincidence degree parameter is marked as SH t , wherein t represents different other data;

[0021] S42, according to the data trend table, whether the other data belongs to the upstream data or downstream data of the current inspection data is analyzed and confirmed, if it belongs to, then: t = SH t × C1+ JS obtains the correlation value GL t of the corresponding other data, wherein C1 is a preset fixed coefficient factor, and JS is a preset basic base;

[0022] If not, then: t = SH t × C1 obtains the correlation value GL t of the corresponding other data, wherein C1 is a preset fixed coefficient factor;

[0023] S43, according to the confirmed correlation value GL t , the other data corresponding to the numerical value is sorted to confirm the data sorting table;

[0024] S44, confirm the specific storage location of other data in the confirmed data ranking table, and directly generate a portal, and place the generated several portals at the rear end position of the current inspection data according to the ranking mode of the ranking table, to complete the correlation degree processing corresponding to the current inspection data;

[0025] S45, the other inspection data is processed in the same way as steps S41-S44, if there is no corresponding coincidence degree parameter, no processing is performed.

[0026] Also includes the steps of:

[0027] S5, the inspection data in the device inspection data packet is compared with the preset data interval, when the inspection data is in the preset data interval, no processing is performed, otherwise, a warning signal is generated and directly displayed.

[0028] Beneficial effects

[0029] The application provides an automatic inspection method for intelligent manufacturing.

[0030] Beneficial effects:

[0031] The application preferentially confirms the inspection record collection node, analyzes whether the corresponding node is a normal point or an abnormal point, analyzes the point data of the abnormal point, confirms the interval difference value between the past time, subsequently, obtains the virtual time point corresponding to the abnormal point, constructs the ranking table of the abnormal point according to the specific time point of the virtual time point, and generates the overall inspection record collection table by combining the normal point, so that the inspection system collects the inspection data according to the inspection record collection table, changes the original data inspection collection mode, improves the automatic inspection efficiency, and sufficiently reduces the automatic inspection time.

[0032] When storing data subsequently, the correlation degree between the data is analyzed, and the data having mutual correlation is generated into a corresponding portal and placed at the rear end of the corresponding data, so that the operation personnel can find the corresponding associated data in time when reviewing the inspection data in the later stage, thereby improving the overall review effect of the operation personnel and facilitating the operation of the operation personnel. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 The figure is a method flowchart of the application;

[0034] Figure 2 The figure is an inspection record collection table confirmation diagram of the application. DETAILED DESCRIPTION

[0035] 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 the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0036] Example 1

[0037] See also Figure 1 , this application provides an automatic inspection method for intelligent manufacturing, comprising the following steps:

[0038] S1. Based on the inspection route planned by the operator, the inspection system performs adaptive inspections on the intelligent manufacturing equipment within this inspection route. Specifically, the inspection route is planned in advance by the operator and stored in the inspection system. The inspection system records the equipment data by identifying the equipment identifier, and the inspection route is set according to the manufacturing process of the corresponding product.

[0039] S2. Combination Figure 2 During the inspection and recording process of different intelligent manufacturing equipment, an inspection record collection table of the corresponding intelligent manufacturing equipment is constructed according to the corresponding time point when the corresponding node of the intelligent manufacturing equipment generates parameter data. The specific method of constructing this inspection record collection table is as follows:

[0040] S21. Confirm the collection object and the current time, and confirm several collection nodes to be inspected for this collection object. Then, from the several collection nodes to be inspected, mark the points where data is displayed as normal points, and mark the points where no data is displayed as abnormal points. Specifically, during inspection and collection, not all collection nodes to be inspected can be used for data collection. Some nodes may not have corresponding data. If inspection and collection are performed in a top-down manner, the time it takes for some nodes to process data will affect the overall inspection and collection time, resulting in an excessively long inspection and collection time.

[0041] S22. Confirm the specific time point at which the abnormal point corresponding to the collection object generates the past parameters from the cloud, and mark the different specific time points as SJ. i-k , where i represents different abnormal points, k represents different time points, and CZ q =SJ i-k -SJ i-(k-1) , get the interval difference CZ between several different time points q , where q = k-1, and k ≥ 2. Since there are k time points, there is an interval difference between each time point. If there are 5 time points, there are 4 groups of interval differences, so q = k-1;

[0042] S23, several interval period difference values CZ belonging to the same abnormal point are confirmed q The maximum value and the minimum value existing in the internal are eliminated, and the remaining several interval period difference values CZ are obtained q The mean value is processed to obtain the limited mean value XJ, and the confirmed limited mean value XJ is taken as the standard value of the abnormal point;

[0043] S24, the time point of the last data display of the abnormal point is confirmed again, which is marked as the initial time point, and the initial time point + standard value = virtual time point is used, and then the specific numerical value of the virtual time point is used to sort several abnormal points to obtain an abnormal point sorting table;

[0044] S25, the normal points are sorted in a top-down manner to generate a normal point sorting table, the normal point sorting table and the abnormal point sorting table are combined, and the normal point sorting table is sorted in front, a patrol record collection table is generated by combination, and then the patrol system records the patrol data of different nodes in the device according to the patrol record collection table, and the recorded patrol data is processed in the next step;

[0045] S3, the recorded patrol data is processed to reduce the data capacity, the repeated data appearing in the patrol data is confirmed, and the corresponding repeated mark is used to replace the repeated data, and the processed patrol data is marked as the data to be induced, and the next step is processed, wherein the specific way of replacement is:

[0046] S31, the past data of the corresponding device in the patrol data is combined and analyzed to confirm the repeated data, and the repeated data appearing in the patrol data is marked, and after the marking is completed, the repeated mark is used to replace the different repeated data;

[0047] S32, after the repeated data in the patrol data is completely replaced, the corresponding repeated mark and repeated data are bundled to generate a repeated data matching table and stored, and after the repeated data is replaced, the corresponding patrol data is marked as the data to be induced.

[0048] Specifically, when the same intelligent manufacturing device is data patrolled, if the device is in normal operation state, a large amount of repeated data will be generated, if each group of repeated data is stored, the storage space in the system will be wasted, so the repeated data needs to be processed, and the repeated mark is used to replace the repeated data to reduce the overall capacity of the corresponding patrol data.

[0049] S4, when the data to be induced is induced, the corresponding data associated with the current inspection data is analyzed and confirmed from the data to be induced, and the similar data is sorted according to the similarity between the data, and placed at the back end of the current inspection data, and after processing, the corresponding device inspection data package is generated, wherein the specific way of induction is:

[0050] S41, a group of inspection data is randomly selected from the data to be induced, and is marked as the current inspection data, and the other data in the data to be induced is analyzed with the current inspection data, and the corresponding coincidence parameter is recorded, and the recorded coincidence parameter is marked as SH t , wherein t represents different other data;

[0051] S42, according to the data trend table, wherein the data trend table is a preset table, and the specific setting method is set by the operator according to experience, and the data trend table is the data trend of the corresponding intelligent manufacturing equipment, which can be understood as only when the upstream data appears, the downstream data exists, for example: when the motor starts, only when the start data exists, the rotating speed data of the rotating shaft exists, analyze whether the other data belongs to the upstream data or the downstream data of the current inspection data, if yes, use: GL t = SH t ×C1+JS to obtain the correlation value GL t of the corresponding other data, wherein C1 is a preset fixed coefficient factor, and JS is a preset basic base, and JS generally takes 20;

[0052] If not, use: GL t = SH t ×C1 to obtain the correlation value GL t of the corresponding other data, wherein C1 is a preset fixed coefficient factor;

[0053] S43, according to the confirmed correlation value GL t , the other data corresponding to the numerical value is sorted, and the data sorting table is confirmed;

[0054] S44, the specific storage location of the other data in the data sorting table is confirmed, and the transfer door is directly generated, and the generated several transfer doors are placed at the back end of the current inspection data according to the sorting method of the sorting table, and the correlation degree processing of the current inspection data is completed;

[0055] S45, the other inspection data is processed in the same way as steps S41-S44, if there is no corresponding coincidence parameter, no processing is performed;

[0056] Specifically, in the induction process of the inspection data, the general induction method adopted is direct storage, and subsequent operators directly extract the corresponding data from the specific storage location and analyze it, but this method cannot timely confirm the existence of mutually related data. If the above method is used to confirm the existence of mutually related data, and then sort according to the specific value of the similarity, the data with higher similarity is sorted first, and the data with lower similarity is sorted last, if the corresponding data is stored again in the corresponding position, it will cause excessive waste of system capacity. Therefore, by generating a portal, the interface is directly transmitted, which not only allows the similar data to be reviewed in the first time, but also does not occupy the overall capacity of the system.

[0057] Embodiment two

[0058] Based on the above embodiments, the present embodiment further includes the following steps in the specific implementation process:

[0059] S5, comparing the inspection data in the device inspection data packet with the preset data interval, when the inspection data is in the preset data interval, no processing is performed, otherwise, a warning signal is generated and directly displayed for external personnel to view and timely maintain and repair the specified manufacturing equipment.

[0060] Embodiment three

[0061] The specific difference of the present embodiment in the specific implementation process is:

[0062] In the step S42, the value of JS is 30;

[0063] Embodiment four

[0064] The present embodiment in the specific implementation process contains all the implementation processes of the above three groups of embodiments.

[0065] Some data in the above formula are dimensionless for numerical calculation, and the contents not described in detail in the specification all belong to the prior art known to those skilled in the art.

[0066] The above embodiments are only used to illustrate the technical method of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. An automatic inspection method for intelligent manufacturing, characterized in that: The following steps are involved: S1. Based on the inspection route planned by the operator, the inspection system performs adaptive inspection on the intelligent manufacturing equipment within this inspection route; S2. During the inspection and recording process of different intelligent manufacturing equipment, an inspection record collection table for the corresponding intelligent manufacturing equipment is constructed according to the corresponding time point at which the parameter data of the corresponding node of the intelligent manufacturing equipment is generated. The specific method is as follows: S21, confirming the collection object and the current time, and confirming several collection nodes to be inspected for this collection object, and then marking the points where data is displayed as normal points from the several collection nodes to be inspected, and marking the points where no data is displayed as abnormal points; S22. Confirm the specific time point at which the abnormal point corresponding to the collection object generates the past parameters from the cloud, and mark the different specific time points as SJ. i-k , where i represents different abnormal points, k represents different time points, and CZ q =SJ i-k -SJ i-（k-1） , get the interval difference CZ between several different time points q , where q=k-1 and k≥2; S23, the difference CZ of several intervals belonging to the same abnormal point q The maximum and minimum values ​​existing inside are eliminated, and the remaining interval differences CZ are calculated. q Perform mean processing to obtain the limited mean XJ, and use the confirmed limited mean XJ as the standard value of this abnormal point; S24. Confirm the time point when the data of the abnormal point was last displayed, and mark it as the initial time point. Use the formula: initial time point + standard value = virtual time point. Then, sort the abnormal points according to the specific value of the virtual time point to obtain an abnormal point sorting table. S3. Perform data capacity reduction processing on the recorded inspection data, identify duplicate data in the inspection data, replace the duplicate data with corresponding duplicate markers, mark the processed inspection data as data to be summarized, and proceed to the next step of processing; S4. When summarizing the inspection data of different nodes within the data to be summarized, the corresponding data that is associated with the inspection data is analyzed and confirmed from the data to be summarized, and the similar data is sorted according to the similarity between the data, and placed at the back end of the inspection data. After processing, the corresponding equipment inspection data packet is generated.

2. The automatic inspection method for intelligent manufacturing according to claim 1, characterized in that: The step S2 further includes: S25. Sort the normal points from top to bottom to generate a normal point sorting table, combine the normal point sorting table with the abnormal point sorting table, and sort the normal point sorting table in front, and generate an inspection record collection table by combining them. Subsequently, the inspection system records the inspections of different nodes inside the device according to the inspection record collection table, and processes the recorded inspection data for the next step.

3. The automatic inspection method for intelligent manufacturing according to claim 2, characterized in that: In step S3, the specific method of replacing duplicate data with duplicate markers is as follows: S31, combining and analyzing the past data of the corresponding equipment in the inspection data, identifying duplicate data, and marking the duplicate data in the inspection data. After the marking process is completed, the duplicate data is replaced with the duplicate marker; S32. After the duplicate data in the inspection data is completely replaced, the corresponding duplicate marks and duplicate data are bundled to generate a duplicate data matching table and store it. After the duplicate data is replaced, the corresponding inspection data is marked as data to be summarized.

4. The automatic inspection method for intelligent manufacturing according to claim 1, characterized in that: In step S4, the specific method of summarizing the inspection data of different nodes within the data to be summarized is: S41. Randomly select a group of inspection data from the data to be summarized and mark it as the current inspection data. Perform coincidence analysis on the other data in the data to be summarized and the current inspection data, and record the corresponding coincidence parameters. The recorded coincidence parameters are marked as SH. t , where t represents different other data; S42. According to the data trend table, which is a preset table, analyze and confirm whether other data belongs to the upstream data or downstream data of the inspection data. If so, use: GL t =SH t ×C1+JS gets the associated value GL corresponding to other data t , where C1 is the preset fixed coefficient factor, and JS is the preset base base; If not, use: GL t =SH t ×C1 obtains the associated value GL corresponding to other data t , where C1 is the preset fixed coefficient factor; S43, according to the confirmed correlation value GL t , sort the other data corresponding to the value and confirm the data sorting table; S44. Confirm the specific storage location of other data in the confirmed data sorting table, and directly generate a transfer gate. Place the generated several transfer gates at the back end position of the inspection data according to the sorting method of the sorting table to complete the correlation processing corresponding to the inspection data.

5. The automatic inspection method for intelligent manufacturing according to claim 4, characterized in that: The step S4 further includes: S45. Process the remaining inspection data in the same manner as steps S41-S44. If there is no corresponding coincidence parameter, no processing is performed.

6. The automatic inspection method for intelligent manufacturing according to claim 1, characterized in that: Also includes the steps: S5. Compare the inspection data in the equipment inspection data packet with the preset data interval. When the inspection data∈the preset data interval, no processing is performed. Otherwise, an early warning signal is generated and displayed directly.

Images