Intelligent piecework method for a sewing machine

Abstract

The application discloses an intelligent piecework method of a sewing machine, and comprises the following steps: judging whether a standard piecework template exists in a current piecework algorithm, obtaining data of the standard piecework template if the standard piecework template exists, obtaining at least three pieces of marking data to obtain a temporary piecework template if the standard piecework template does not exist; recording a sewing machine action signal in a sewing process of the sewing machine; constantly updating the temporary piecework template if the temporary piecework template is obtained; matching the sewing machine action signal with the standard piecework template or the latest temporary piecework template to judge whether the sewing of a piece of clothing is completed, counting the number of the clothing sewing pieces if yes, otherwise, continuously recording the sewing machine action signal in the sewing process of the sewing machine; and the application reduces the influence of data marking and improves the piecework accuracy through the standard piecework template and the constant optimization of the temporary piecework template.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mechanical equipment piecework, and in particular to an intelligent piecework method of a sewing machine. BACKGROUND

[0002] At present, the Internet of Things sewing equipment has basically replaced the ordinary sewing machine, and the factory has changed from manual piecework of workers to automatic piecework method using work ticket software, hanging system, piecework algorithm and the like. However, most of the algorithms on the market at present need workers to mark the data generated in the sewing process before piecework (at least the data of the first three pieces of sewing need to be marked) to obtain a template matched with the data, so the three pieces of data marked will have a great influence on the piecework result of the algorithm. Since the three pieces of data form a fixed template, once rework, manual thread cutting and the like occur, it is easy to cause the sewing machine data received to be unmatched with the template, and thus the piecework accuracy greatly decreases. Therefore, how to reduce the influence of data marking by technical means is a problem that needs to be solved at present.

[0003] For example, the "intelligent piecework method, system, memory and sewing machine of a sewing machine" provided in Chinese patent CN107345346A collects and records the start signal of starting to sew each piece of clothing in the sewing process of the sewing machine; collects and records the number of stitches and the number of thread cutting in the sewing process of the sewing machine; determines whether the sewing of each piece of clothing is completed according to the number of stitches and the number of thread cutting collected and the target number of stitches and the target number of thread cutting of each piece of clothing; and accumulates the number of pieces of clothing sewn when it is determined that the sewing of each piece of clothing is completed. In the process of piecework of the clothing, the template data is fixed, so that the piecework accuracy is low. SUMMARY

[0004] The present application mainly solves the problem of low piecework accuracy of the sewing machine in the prior art; an intelligent piecework method of a sewing machine is provided, which reduces the influence of data marking on the piecework algorithm and improves the piecework accuracy.

[0005] The above technical problems of the present application are mainly solved by the following technical scheme: an intelligent piecework method of a sewing machine, comprising the following steps:

[0006] determining whether a standard piecework template exists in the current piecework algorithm, if a standard piecework template exists, obtaining the data of the standard piecework template, and if a standard piecework template does not exist, obtaining at least three pieces of marked data to obtain a temporary piecework template;

[0007] recording the sewing machine action signal in the sewing process of the sewing machine; if the temporary piecework template is obtained, the temporary piecework template is constantly updated;

[0008] The sewing machine action signal is matched with a standard piecework template or a latest temporary piecework template to determine whether the sewing of a piece of garment is completed, if yes, the piece number of the garment is accumulated, otherwise, the sewing machine action signal during the sewing process of the sewing machine is continuously recorded.

[0009] Preferably, the obtaining of the temporary piecework template from the at least three pieces of labeling data comprises: if there is a difference in the obtained labeling data, the next piece of labeling data is continuously obtained, difference calculation is performed, the calculation result is compared with a difference threshold, if the calculation result is less than the difference threshold, the current labeling data is used as the piecework template, otherwise, the next piece of labeling data is continuously obtained and the difference calculation is performed until the calculation result is less than the difference threshold.

[0010] Preferably, if the temporary piecework template is obtained, the temporary piecework template is continuously updated, which comprises: a preset update cycle time is obtained, when the update cycle time is reached, the temporary piecework template is optimized by the sewing machine action signal collected in the update cycle time to obtain a new temporary piecework template.

[0011] Preferably, the temporary piecework template is optimized by the sewing machine action signal collected in the update cycle time to obtain a new temporary piecework template, which comprises: the sewing machine action signal collected in the update cycle time is grouped to obtain new template data, if there is a difference between the new template data and the temporary piecework template, the new template data is used as the new temporary piecework template, otherwise, the original temporary piecework template is still maintained.

[0012] Preferably, if the original temporary piecework template is still maintained, a no-change count is performed, a preset stop number is obtained, and when the accumulated no-change count reaches the stop number, the updating of the temporary piecework template is stopped, and the current temporary piecework template is stored as a standard piecework template.

[0013] Preferably, the obtaining of the standard piecework template comprises: all sewing machine action signals corresponding to a standard number of garments are obtained, the sewing machine action signals are segmented based on the standard number to obtain the standard piecework template.

[0014] Preferably, the labeling data comprises: the sewing machine action signal, the needle number correction signal and the presser foot lifting signal.

[0015] Preferably, the sewing machine action signal comprises: the number of needles and the number of thread cutting during the sewing process of the sewing machine.

[0016] The present application has the advantages that: by setting the standard piecework template and continuously optimizing the temporary piecework template, the influence of data labeling is reduced, the piecework accuracy is improved, no additional hardware device is needed, the cost is effectively saved, the piecework process is automatically performed, no manual operation is needed, and the work efficiency is improved. Attached Figure Description

[0017] Figure 1 This is a flowchart of the intelligent piece counting method according to an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the annotation data in an embodiment of the present invention.

[0019] Figure 3 This is a schematic diagram of template updating according to an embodiment of the present invention.

[0020] Figure 4 This is a schematic diagram of obtaining the piecework template according to an embodiment of the present invention.

[0021] Figure 5 This is a schematic diagram of the piecework template data according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the invention.

[0023] Example:

[0024] A smart piece counting method for sewing machines, such as Figure 1 As shown, it includes the following steps:

[0025] S1: Determine if a standard piecework template exists in the current piecework algorithm. If a standard piecework template exists, obtain the data from the standard piecework template. If no standard piecework template exists, obtain at least three annotation data points to obtain a temporary piecework template.

[0026] The process of obtaining at least three annotation data points to obtain a temporary piecework template includes: if there are differences in the obtained annotation data, then continue to obtain the next annotation data point, perform differential calculation, compare the calculation result with the differential threshold, and if the calculation result is less than the differential threshold, then use the current annotation data point as the piecework template; otherwise, continue to obtain the next annotation data point and perform differential calculation until the calculation result is less than the differential threshold.

[0027] like Figure 2 As shown, a set of labeled data consists of multiple event data. The event data is divided into three parts separated by commas: the time the event occurred, the event type, and the event parameters. Event types 0-10 represent various actions or states of the sewing machine. For example, 2 represents the motor starting, 3 represents the motor stopping, and the number of stitches sewn in this instance is appended after the comma (i.e., the event parameters). 4 and 5 represent raising and lowering the presser foot, respectively, and 6 and 7 represent the sewing machine's thread cutting and back thread cutting actions, respectively.

[0028] Specifically, the gateway device sends the first three pieces of annotation data to the cloud platform in real time, and the algorithm server performs differential calculation based on the thread cutting, needle count, and presser foot lifting count in the three pieces of annotation data. If the algorithm server calculates that the difference degree of the three pieces of annotation data is less than the difference degree threshold, the three pieces of annotation data are used as a temporary piecework template.

[0029] If the difference degree of the three pieces of annotation data is greater than the difference degree threshold, the algorithm server requests a fourth piece of annotation data from the sewing machine and performs difference degree calculation on the fourth piece of annotation data and the first three pieces of annotation data. Similarly, if the difference degree is less than the difference degree threshold, the four pieces of annotation data are used as a template for piecework. Otherwise, a fifth piece of annotation data is requested from the sewing machine. In this way, the algorithm server can improve the accuracy of the piecework template and improve the accuracy of piecework.

[0030] In actual use, when the first two pieces of annotation data are both 2 times of thread cutting (i.e., the number of times of 6,0 appears), and the third piece of annotation data is only 1 time of thread cutting, the piecework algorithm can consider that the standard thread cutting times (i.e., the template of the algorithm) is 2 (taking the mode, and regarding the third time as an anomaly). The next piece of annotation data can also be requested from the sewing machine system to obtain the fourth piece of annotation data. If the fourth piece of annotation data has 2 times of thread cutting, the fifth piece of annotation data is not requested. If it is not 2, the fifth piece of annotation data is requested, and the number of needles is counted, for example, the number of needles of the first piece of annotation data is 33 (21+12), the number of needles of the second piece of annotation data is 39 (18+21), and the number of needles of the third piece of annotation data is 21. Similarly, the algorithm counts the number of needles of the fourth and fifth pieces of annotation data, and matches the similarity through the two dimensions of needle count and thread cutting to obtain the standard thread cutting times.

[0031] The standard piecework template acquisition includes: acquiring all sewing machine action signals corresponding to a standard number of garments, and segmenting the sewing machine action signals based on the standard number to obtain a standard piecework template.

[0032] Specifically, as shown in Figure 4 A worker receives a package of clothes for sewing (assuming a standard number x pieces), and inputs the number x after completing the sewing of the package. As shown in Figure 5 The algorithm server segments the data based on the number x and the number of needles of the sewing machine, and can clearly obtain the data generated by the worker when sewing each garment (ααβ is one piece). The algorithm server uses the x pieces of data as a standard piecework template algorithm, which greatly reduces the influence of less annotation data on the accuracy of the piecework algorithm.

[0033] The probability of error in obtaining the piecework template using three clothes is relatively large, so a package of data is directly used to obtain the piecework template, for example, a worker makes 20 clothes in a package, and the number of cutting lines is 3 times, then it is found from the data that the total number of each cutting line is 60 A (AAA / AAA / AAA, ABC represents the needle number group), or 40 A 20 B (AAB / AAB / AAB or its permutation combination, only the number is concerned), or 20 A 20 B 20 C (ABC / ABC / ABC), then the standard cutting line number can be clearly obtained from the data of this package, that is, the standard piecework template is obtained.

[0034] The sewing machine action signal mentioned in the application includes the number of stitches, the number of cutting lines during the sewing process of the sewing machine, and also includes the number of presser foot lifting and the interval time of the number of stitches and the interval time of the cutting lines. Whether the current sewing is complete is determined by the number of stitches and the number of cutting lines, whether the current sewing is smooth and whether a sewing failure occurs is determined by the interval time of the number of stitches and the interval time of the cutting lines, and the accuracy of piecework of the piecework template is improved by comparing the number of presser foot lifting with the number of stitches.

[0035] The labeled data includes the sewing machine action signal, the stitch number correction signal and the presser foot signal, and the sewing machine action signal includes the number of stitches and the number of cutting lines during the sewing process of the sewing machine.

[0036] S2: record the sewing machine action signal in the sewing process of the sewing machine; if a temporary piecework template is obtained, the temporary piecework template is constantly updated.

[0037] If the temporary piecework template is obtained, the temporary piecework template is constantly updated, including: obtaining a preset update period time, when the update period time arrives, the sewing machine action signal collected in the update period time is used to optimize the temporary piecework template, and a new temporary piecework template is obtained.

[0038] The sewing machine action signal collected in the update period time is used to optimize the temporary piecework template, and a new temporary piecework template is obtained, including: the sewing machine action signal collected in the update period time is grouped to obtain new template data, if the new template data is different from the temporary piecework template, the new template data is used as the new temporary piecework template, otherwise the original temporary piecework template is still maintained.

[0039] If the original temporary piecework template is still maintained, a no-change count is performed once, a preset stop number is obtained, when the cumulative no-change count reaches the stop number, the updating of the temporary piecework template is stopped, and the current temporary piecework template is stored as a standard piecework template.

[0040] Specifically, as Figure 3As shown, the first three pieces of labeled data are used as templates for piece counting. If the number of pieces sewn by the worker is calculated every 1 h, the algorithm server obtains action data of the sewing machine every 1 h, uses the temporary piece counting template to calculate the number of pieces for 1 h of data, and optimizes the temporary piece counting template during the calculation process. For example, if the number of thread trimming times for each of the first three pieces of data is 3, 3, and 4, respectively, the algorithm will have a different number of thread trimming times in the temporary piece counting template at this time, so the algorithm cannot accurately determine how many times of thread trimming are needed to complete each piece of the process. Therefore, the algorithm optimizes the temporary piece counting template based on the remaining data (timestamp, number of presser foot lifts, number of stitches) and other data to assist in the calculation, unifies the number of thread trimming times, and uses the optimized temporary piece counting template to process subsequent data, further improving the accuracy of piece counting. Similarly, if the number of thread trimming times is still unstable within 1 h, the temporary piece counting template is optimized every 1 h until the number of thread trimming times can be unified.

[0041] If the number of thread trimming times does not converge in the first three pieces of labeled data, the algorithm uses the mode (i.e., 2) as the standard thread trimming time. After 1 h of data is calculated, for example, the action sequence, number of stitches, and other data within each thread trimming are grouped, and it is found that the main classes are not 2 (the data within each thread trimming for 1 h of data is counted, and the label is given to the classification result, for example, {“Label 0”:[thread trimming 1 data, thread trimming 4 data, thread trimming 7 data], “Label 1”:[thread trimming 2 data, thread trimming 5 data, thread trimming 8 data], “Label 2”:[thread trimming 3 data, thread trimming 6 data, thread trimming 9 data]}. It can be clearly seen that three main classes are divided by thread trimming, i.e., the actual standard thread trimming time for this process is three times. The algorithm updates the temporary piece counting template, and the better data within this time period is used as the temporary piece counting template for calculation. Similarly, 2 h is also calculated once, and the temporary piece counting template is optimized in real time.

[0042] S3: Match the sewing machine action signal with the standard piece counting template or the latest temporary piece counting template to determine whether a piece of clothing is completed, if yes, count the number of pieces of clothing, otherwise, continue to record the sewing machine action signal during the sewing process of the sewing machine.

[0043] When using the standard piece counting template for piece counting, the number of correction stitches and the number of thread return stitches during the sewing process are obtained, and it is determined whether the obtained number of stitches is greater than or equal to the difference between the template stitch number, the correction stitch number, and the thread return stitch number. If not, continue to record the number of stitches during the sewing process, if yes, determine whether the number of thread trimming times is greater than or equal to the template thread trimming time, if yes, determine that a piece of clothing is completed, otherwise, continue to record the number of stitches and the number of thread trimming times during the sewing process of the sewing machine, and perform piece counting when it is determined that a piece of clothing is completed.

[0044] After one piece is completed, the recorded needle number and thread cutting number during the sewing process of the sewing machine are cleared.

[0045] When piecework is performed by using the temporary piecework template, the needle and thread time interval of each process of one piece of garment is obtained, process matching is performed, if the time interval of the current process is greater than or less than the needle and thread time interval, needle interval time matching is performed, if the needle interval time matching is successful and the needle number is the same, the current process is not matched, target process rotation is performed, until the same process is matched, when the processes of the temporary piecework template are all matched and there is no repetition, it is determined that the sewing of one piece of garment is completed, and one piece is counted.

[0046] The present application reduces the influence of data labeling, improves the piecework accuracy, does not need to increase additional hardware devices, effectively saves the cost, the piecework process is automatically performed, does not need manual operation, and improves the work efficiency by setting the standard piecework template and continuously optimizing the temporary piecework template.

[0047] The above-mentioned embodiment is only a preferred scheme of the present application, and does not limit the present application in any form, and other variants and modifications can be made without exceeding the technical scheme recorded in the claims.

Claims

1. An intelligent piecework method for a sewing machine, characterized in that, The method comprises the following steps: S1, judging whether a standard piecework template exists in a current piecework algorithm, if the standard piecework template exists, obtaining data of the standard piecework template, if the standard piecework template does not exist, obtaining at least three pieces of marking data to obtain a temporary piecework template; S2, recording sewing machine action signals in a sewing process of the sewing machine; if the temporary piecework template is obtained, the temporary piecework template is constantly updated; a preset update period time is obtained, when the update period time arrives, sewing machine action signals collected in the update period time are grouped to obtain new template data, if the new template data is different from the temporary piecework template, the new template data is taken as a new temporary piecework template, otherwise, the original temporary piecework template is maintained; S3, matching the sewing machine action signals with the standard piecework template or the latest temporary piecework template to judge whether sewing of a piece of clothing is completed, if yes, a piece of clothing sewing count is accumulated, otherwise, the sewing machine action signals in the sewing process of the sewing machine are continuously recorded.

2. The intelligent piecework method of a sewing machine according to claim 1, wherein, The obtaining of the temporary piecework template from the at least three pieces of marking data comprises: if the marking data has differences, the next piece of marking data is continuously obtained, difference calculation is performed, the calculation result is compared with a difference threshold value, if the calculation result is less than the difference threshold value, the current marking data is taken as the piecework template; otherwise, the next piece of marking data is continuously obtained and the difference calculation is performed until the calculation result is less than the difference threshold value.

3. The intelligent piecework method of a sewing machine according to claim 1, wherein, If the original temporary piecework template is maintained, a no-change count is performed, a preset stop number is obtained, when the accumulated no-change count reaches the stop number, the updating of the temporary piecework template is stopped, and the current temporary piecework template is stored as the standard piecework template.

4. The intelligent piece counting method for a sewing machine as claimed in claim 1, wherein, The obtaining of the standard piecework template comprises: obtaining all sewing machine action signals corresponding to a standard number of clothes, segmenting the sewing machine action signals based on the standard number to obtain the standard piecework template.

5. The intelligent piecework method of the sewing machine according to claim 1, wherein the marking data comprises: the sewing machine action signals, the needle number correction signals and the presser foot lifting signals.

6. The intelligent piecework method of the sewing machine according to claim 1, wherein the sewing machine action signals comprise: the needle number in the sewing process of the sewing machine and the thread trimming number.

7. The intelligent piecework method of the sewing machine according to claim 6, wherein the sewing machine action signals comprise: the presser foot lifting number, the needle number interval time and the thread trimming interval time. The step S3 comprises: when the standard piecework template is used for piecework, the correction needle number and the return thread needle number in the sewing process are obtained, it is judged whether the obtained needle number is greater than or equal to the difference between the template needle number, the correction needle number and the return thread needle number, if not, the needle number in the sewing process is continuously recorded, if yes, it is judged whether the thread trimming number is greater than or equal to the template thread trimming number, if yes, it is determined that the sewing of a piece of clothing is completed.

8. The intelligent piece counting method for a sewing machine as claimed in claim 1, wherein, ​ 9. The intelligent piecework method of a sewing machine according to claim 1, wherein, The step S3 comprises: when piecework is carried out by using the temporary piecework template, the needle and thread time interval of each working procedure of a piece of garment is obtained; when the working procedure matching is carried out, if the time interval of the current working procedure is greater than or less than the needle and thread time interval, the needle number interval time matching is carried out; if the needle number interval time matching is successful and the needle number is same, the current working procedure is not matched, the target working procedure rotation is carried out until the same working procedure is matched; when the working procedures of the temporary piecework template are all matched and there is no repetition, it is determined that the sewing of a piece of garment is completed.

Images