Self-adaptive adjustment method for thread tension of flatcar sewing machine

Abstract

The invention relates to the technical field of thread tension adjusting control, in particular to a thread tension self-adaptive adjusting method of a flat sewing machine. Current and historical line tension time sequence data are obtained, and data support is provided for adjustment. By comparing the difference between the current tension and the historical preset tension, a tension deviation index is determined, and abnormal fluctuation is found preliminarily. Thirdly, comparing the numerical difference of the tension time sequence data of each line and the change along with time, determining a tension characteristic value, and reflecting the tension change condition; and integrating the current deviation index and the deviation of all the tension characteristic values to obtain a current tension adjustment degree value so as to realize comprehensive evaluation. In order to achieve more accurate adjustment, the current adjustment value is adjusted based on the deviation index and the adjustment degree value of the historical monitoring period to obtain a new value, so that the adjustment is more continuous and stable. Finally, according to the new adjustment value, the current preset tension and the historical preset tension, the self-adaptive tension value of the next period is determined, dynamic and self-adaptive adjustment is achieved, and the accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of thread tension adjustment and control, and particularly relates to a method for adaptively adjusting the thread tension of a flatbed sewing machine. Background Art

[0002] In the field of sewing machines, especially in flatbed sewing machines, thread tension control is one of the key factors affecting sewing quality. Thread tension includes the tension of the top thread (i.e., the needle thread) and the bottom thread (i.e., the bobbin thread). Excessive tension will cause the top thread to break, the fabric to wrinkle, or the bottom thread to be exposed; too loose tension will make the stitch loose and easy to come off the thread. The ideal state is that the tensions of the upper and lower threads are balanced, and the stitch knots are buried in the middle of the fabric, which can ensure that the stitches are uniform and firm during the sewing process, and avoid problems such as thread breakage and skipped stitches.

[0003] When the prior art adjusts and controls the thread tension in a flatbed sewing machine, a fixed threshold is usually set for each thread, and the thread tension is adjusted based on the deviation between the tension during the sewing process and the fixed threshold; however, in the actual sewing process, there is an interaction relationship between the top thread and the bottom thread, and the tension will also constantly fluctuate and change. Adjusting and controlling the tension only based on the relationship between the tension value at a single moment and the fixed threshold will result in a low credibility of the obtained tension adjustment magnitude. Summary of the Invention

[0004] In order to solve the technical problem that in the actual sewing process, there is an interaction relationship between the top thread and the bottom thread, and the tension will also constantly fluctuate and change, and adjusting and controlling the tension only based on the relationship between the tension value at a single moment and the fixed threshold will result in a low credibility of the obtained tension adjustment magnitude, the purpose of the present invention is to provide a method for adaptively adjusting the thread tension of a flatbed sewing machine, and the specific technical solution adopted is as follows:

[0005] Obtain thread tension time series data, including the monitored thread tension time series data of the current monitoring period in the current sewing process and the normal thread tension time series data in the historical sewing process; wherein, the thread includes the top thread and the bottom thread;

[0006] Compare the difference between the current monitoring period and the preset thread tension in the historical sewing process to determine the tension deviation index of the current monitoring period; for each type of thread, compare the numerical differences between the thread tension time series data, and combine the change of the tension value with time in the thread tension time series data to determine the tension characteristic value of each thread tension time series data; comprehensively consider the tension deviation index of the current monitoring period and the deviation between all tension characteristic values for each type of thread to obtain the tension adjustment degree value of the current monitoring period for each type of thread.

[0007] Adjust the tension adjustment degree value of the current monitoring cycle based on the tension deviation index and the tension adjustment degree value of the historical monitoring cycle during the current sewing process to obtain the new tension adjustment degree value for each type of thread; for each type of thread, determine the adaptive tension value for the next monitoring cycle based on the new tension adjustment degree value, the current monitoring cycle, and the preset thread tension during the historical sewing process.

[0008] Further, the method for obtaining the tension deviation index includes:

[0009] For each type of thread, compare the difference between the preset thread tension during the current monitoring cycle and that during the historical process, and determine the first tension deviation factor corresponding to the current monitoring cycle;

[0010] In the current monitoring cycle and each historical sewing process respectively, take the difference between the preset thread tensions of the two types of threads as the difference factor, take the absolute value of the difference between the difference factors during the current monitoring cycle and each historical sewing process as the difference parameter, and take the sum value of all the difference parameters corresponding to the current monitoring cycle as the second tension deviation factor corresponding to the current monitoring cycle;

[0011] Take the value obtained by normalizing the sum of the first tension deviation factor and the second tension deviation factor corresponding to the current monitoring cycle as the tension deviation index of the current monitoring cycle.

[0012] Further, the method for obtaining the first tension deviation factor includes:

[0013] For the same type of thread, take the absolute value of the difference between the preset thread tension of the current monitoring cycle and that of each historical sewing process as the difference factor, and take the sum value of all the difference factors corresponding to the current monitoring cycle as the difference index;

[0014] Take the sum value of the difference indexes of the current monitoring cycle for the two types of threads as the first tension deviation factor corresponding to the current monitoring cycle.

[0015] Further, the method for obtaining the tension characteristic value includes:

[0016] For the same type of thread, randomly select a thread tension time series data as the data to be measured. If the data to be measured is the monitoring thread tension time series data, then all the normal thread tension time series data are used as the comparison time series data. If the data to be measured is the normal thread tension time series data, then the remaining normal thread tension time series data are used as the comparison time series data;

[0017] For the same type of thread, compare the numerical differences between the data to be measured and all the corresponding comparison time series data, and determine the tension fluctuation index of the data to be measured;

[0018] Under the same wire type, analyze the change of the tension value with time in the to-be-tested time-series data, and determine the change disorder index of the to-be-tested time-series data;

[0019] The value obtained by normalizing the product of the tension fluctuation index and the change disorder index of the to-be-tested time-series data is used as the tension characteristic value of the to-be-tested time-series data for each wire type.

[0020] Further, the method for obtaining the tension fluctuation index includes:

[0021] Under the same wire type, calculate the mean value of the tension values of all comparison time-series data of the to-be-tested time-series data at the same moment to obtain the tension mean value, and sort the tension mean values at all moments according to time, so as to obtain the reference time-series data of the to-be-tested time-series data;

[0022] Under the same wire type, in the to-be-tested time-series data and the reference time-series data, take the absolute value of the difference between the tension values at the same moment as the corresponding wire tension difference at the same moment;

[0023] Under the same wire type, compare the tension values at the same moment in the to-be-tested time-series data and the reference time-series data. If the tension values are the same, mark them to obtain the reference points;

[0024] Based on the moments corresponding to the reference points, segment the time series to obtain all time periods. In each time period, the value obtained by normalizing the sum value of the wire tension differences corresponding to all moments is used as the fluctuation degree value of each time period;

[0025] Take the time periods with the fluctuation degree value greater than the preset fluctuation threshold as the fluctuation segments, and the value obtained by standardizing the product of the sum value of the fluctuation degree values of all fluctuation segments and the total number of fluctuation segments is used as the tension fluctuation index of the to-be-tested time-series data for each wire type.

[0026] Further, the method for obtaining the change disorder index includes:

[0027] For each wire type, obtain the value range of the tension values in the to-be-tested time-series data, and perform dimensionality reduction segmentation on the tension values within the value range based on the piecewise aggregate approximation algorithm to obtain each tension segmentation point;

[0028] In the to-be-tested time-series data, segment the to-be-tested time-series data based on the tension segmentation points in the time series, so as to obtain multiple tension data segments;

[0029] Obtain the middle moment of each tension data segment. In the time series, calculate the absolute value of the difference between the middle moments of every two adjacent tension data segments as the moment deviation value, and the value obtained by normalizing the variance of all moment deviation values corresponding to the to-be-tested time-series data is used as the change disorder index of the to-be-tested time-series data for each wire type.

[0030] Further, the method for obtaining the tension adjustment degree value includes:

[0031] Under each type of thread, take the mean value of the tension characteristic values corresponding to all normal tension time series data as the standard value;

[0032] Process the difference between the tension characteristic value corresponding to the monitored thread tension time series data and the standard value to obtain a deviation factor, and the value range of the deviation factor is (-1, 1);

[0033] Normalize the product of the deviation factor and the tension deviation index of the current monitoring period, so as to obtain the tension adjustment degree value of the current monitoring period under each type of thread.

[0034] Further, the method for obtaining the new tension adjustment degree value includes:

[0035] Under each type of thread, in every two adjacent historical monitoring periods during the current sewing process, take the difference between the tension deviation index of the previous historical monitoring period and the tension deviation index of the next historical monitoring period as the first change factor of the next historical monitoring period, and take the difference between the absolute value of the tension adjustment degree value of the previous historical monitoring period and the absolute value of the tension adjustment degree value of the next historical monitoring period as the second change factor of the next historical monitoring period;

[0036] Take the ratio of the second change factor of the last historical monitoring period to the maximum value of the second change factors of all historical detection periods as the weight factor;

[0037] Take the value after negatively correlating and mapping the product of the weight factor and the first change factor of the last historical monitoring period as the adjustment factor of the last historical monitoring period;

[0038] Normalize the product of the adjustment factor and the tension adjustment degree value of the current monitoring period, and take the resulting value as the new tension adjustment degree value.

[0039] Further, the method for obtaining the adaptive tension value includes:

[0040] Under each type of thread, take the mean value of the preset thread tensions in all historical sewing processes as the reference value, and take the difference between the preset thread tension of the current monitoring period and the corresponding reference value as the deviation degree value;

[0041] Take the product of the deviation degree value and the new tension adjustment degree value as the adjustment index, and take the difference between the preset thread tension of the current monitoring period and the adjustment index as the adaptive tension value of each thread in the next monitoring period.

[0042] Further, the value range of the preset fluctuation threshold is (0.5, 1).

[0043] The present invention has the following beneficial effects:

[0044] The present invention analyzes the tension fluctuation within a period of time, so that the influence relationship between the threads and the fluctuation of the thread tension can be taken into consideration. First, the monitoring thread tension time series data of the current monitoring period in the current sewing process and the normal thread tension time series data in the historical sewing process are obtained, which provide sufficient data support for the precise adjustment of the thread tension. Comparing the difference between the preset thread tension of the current monitoring period and that of the historical sewing process to determine the tension deviation index of the current monitoring period helps to initially detect abnormal fluctuations in the thread tension. When the tension of the upper thread or the bobbin thread is too large or too small, it will cause irregular fluctuations in the thread during the sewing process, and there are changes in different fluctuation amplitudes. Therefore, under each type of thread, the numerical differences between the thread tension time series data are compared, and combined with the change of the tension value with time in the thread tension time series data, the tension characteristic value of each thread tension time series data is determined to reflect the change of the thread tension. Then, by comprehensively considering the tension deviation index of the current monitoring period and the deviation between all the tension characteristic values under each type of thread, the tension adjustment degree value of the current monitoring period under each type of thread is obtained, realizing a comprehensive evaluation of the thread tension. In order to achieve more precise thread tension adjustment, the present invention adjusts the tension adjustment degree value of the current monitoring period based on the tension deviation index and the tension adjustment degree value of the historical monitoring period in the current sewing process, obtaining a new tension adjustment degree value under each type of thread. By introducing the historical monitoring period data, the adjustment becomes more continuous and stable. Finally, under each type of thread, based on the new tension adjustment degree value, the preset thread tension between the current monitoring period and the historical process, the adaptive tension value of the next monitoring period is determined, realizing the dynamic, adaptive adjustment and control of the thread tension, improving the accuracy of the adjustment control, and keeping the thread tension in the sewing process within a suitable range all the time. Description of the Drawings

[0045] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0046] Figure 1 It is a method flow chart of a method for self - adaptive adjustment of thread tension of a flatbed sewing machine provided by an embodiment of the present invention;

[0047] Figure 2The flowchart of a method for obtaining a tension deviation index provided by an embodiment of the present invention;

[0048] Figure 3 The flowchart of a method for obtaining a tension characteristic value provided by an embodiment of the present invention. Detailed implementation manners

[0049] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following combines the drawings and preferred embodiments to specifically describe a method for self-adaptive adjustment of thread tension of a flatbed sewing machine according to the present invention, including its specific implementation manners, structures, features and effects in detail. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0051] The following specifically describes the specific solution of a method for self-adaptive adjustment of thread tension of a flatbed sewing machine provided by the present invention with reference to the drawings.

[0052] Please refer to Figure 1 , which shows the flowchart of a method for self-adaptive adjustment of thread tension of a flatbed sewing machine provided by an embodiment of the present invention. The method includes the following steps:

[0053] Step S1: Obtain the thread tension time series data, including the monitored thread tension time series data of the current monitoring period in the current sewing process and the normal thread tension time series data in the historical sewing process; wherein, the thread includes the top thread and the bobbin thread.

[0054] The tension setting of the sewing machine directly affects the stitch quality and fabric adaptability. Excessive tension will cause the top thread to break, the fabric to wrinkle or the bobbin thread to be exposed; too loose tension will make the stitches loose and easy to come off. The ideal state is that the tensions of the top thread and the bobbin thread are balanced, and the stitch knots are buried in the middle of the fabric. In the prior art, when making self-adaptive adjustment of the tension, the mutual influence relationship between the top thread and the bobbin thread and the fluctuating change characteristics of the tension during the sewing process are not considered, and the obtained adjustment magnitude is not accurate enough. To solve this problem, in the embodiment of the present invention, the preset magnitudes of the tensions of the top thread and the bobbin thread in the current sewing can be compared with those under historical normal conditions to roughly obtain the current tension setting deviation magnitude. Furthermore, the adjustment magnitude of the tension is judged by the fluctuation amplitude and change rule of the tension fluctuation curve under different tension settings. And the adjustment value is further corrected in combination with the change situation of the feedback data before and after the adjustment.

[0055] Therefore, first, the line tension time series data can be obtained. At key positions of the flatbed sewing machine, such as the thread tensioning device for the upper thread and the bobbin case for the lower thread, high-precision tension sensors are installed. These tension sensors can monitor the tension changes of the upper thread and the lower thread, record the tension values at a certain sampling frequency, and form the monitored tension time series data for the current monitoring period in the current sewing process. In the embodiment of the present invention, the line includes the upper thread (needle thread) and the lower thread (bobbin thread). Therefore, the monitored tension time series data is divided into the monitored tension time series data of the upper thread and the monitored tension time series data of the lower thread.

[0056] Meanwhile, the normal line tension time series data in the historical sewing process can be obtained. The line tension time series data recorded in the past sewing process (when the fabric has no defects, that is, the normal line tension time series data of the product) can be extracted from the historical database. These data should be consistent with the sewing conditions in the current sewing process (such as the same fabric thickness, the same sewing speed, etc.). Taking them as the normal line tension time series data, similarly, the normal line tension time series data is also divided into the normal line tension time series data of the upper thread and the normal line tension time series data of the lower thread.

[0057] It should be noted that in this embodiment of the present invention, the length of the monitoring period in the current sewing process is set to 2 minutes; the lengths of all the normal line tension time series data should be the same. The number of times of the historical sewing process can be adjusted according to the implementation scenario, but it should include at least 4 times; in this embodiment of the present invention, the data acquisition frequency is set to 10 times per second. The setting of the length of the monitoring period and the data acquisition frequency can be adjusted according to the implementation scenario and will not be limited here.

[0058] Step S2: Compare the difference between the current monitoring period and the preset line tension in the historical sewing process to determine the tension deviation index for the current monitoring period; for each type of thread, compare the numerical differences between the line tension time series data, and combine the change of the tension value with time in the line tension time series data to determine the tension characteristic value for each line tension time series data; based on the tension deviation index for the current monitoring period and the deviation between all the tension characteristic values for each type of thread, obtain the tension adjustment degree value for each type of thread in the current monitoring period.

[0059] To ensure good sewing results and avoid problems such as exposed stitch intersections, skipped stitches, and thread breakage, appropriate settings should be made for the tensions of the upper thread and the bobbin thread. For different sewing fabrics, due to differences in materials, thicknesses, etc., the tension settings are often different. Therefore, the historical setting data of the upper thread and the bobbin thread during sewing with the same fabric can be referred to first to roughly judge the appropriateness of the tension settings of the upper thread and the bobbin thread during the current sewing machine sewing, and a tension deviation index for the current monitoring period can be obtained. Then, when the tension of the upper thread or the bobbin thread is too large, it will cause the thread to fluctuate frequently during sewing. For example, when the upper thread is too tight, the elastic deformation of the thread take-up spring intensifies, and the sensor will record more high-frequency oscillation signals. At the same time, when the thread tension is too small, the thread cannot maintain a uniform tension during transportation, resulting in irregular fluctuations in tension, but the fluctuations in the tension signal changes during normal sewing operations are more regular. When the thread is too tight, the high-frequency oscillation signal of the thread take-up spring shows non-stationary and non-continuous fluctuations, and the amplitude of this tension is usually large. When the tension of each stitch during sewing is significantly lower than the normal value, the overall tension amplitude during sewing will be smaller. Thus, the situation of the thread tension setting can be characterized by the fluctuation regularity and numerical difference of the tension change of the upper thread or the bobbin thread, and a tension characteristic value can be obtained. Then, by comprehensively considering the tension deviation index of the current monitoring period and the deviation situation among all the tension characteristic values for each type of thread, the tension adjustment degree value for each type of thread in the current monitoring period can be determined, which is used to reflect the preliminary adjustment degree of the tension in the next monitoring period. Among them, the preset thread tension of each type of thread in the current monitoring period during the current sewing process can be obtained according to the implementation scenario, and the preset thread tension of each type of thread in the historical sewing process can be obtained according to the historical database.

[0060] First, the difference between the preset thread tension in the current monitoring period and that in the historical sewing process can be compared to determine the tension deviation index for the current monitoring period.

[0061] Preferably, in an embodiment of the present invention, the method for obtaining the tension deviation index includes:

[0062] Please refer to Figure 2 , which shows a flowchart of the method for obtaining the tension deviation index in an embodiment of the present invention. The method includes the following steps:

[0063] Step S201: Under each type of thread, compare the difference between the preset thread tension in the current monitoring period and that in the historical process to determine the first tension deviation factor corresponding to the current monitoring period.

[0064] Given that no quality problems occurred in the fabrics sewn during the historical sewing process, the preset thread tension during the historical sewing process can be used as a reference standard to judge whether the preset thread tension setting in the current monitoring period is appropriate.

[0065] Under the same thread type, the absolute value of the difference between the preset thread tension of the current monitoring cycle and the preset thread tension of each historical sewing process is used as the difference factor. The larger the difference factor, the greater the deviation between the preset thread tension of the current monitoring cycle and the preset thread tension of the normal historical process, indicating that the current monitoring cycle is more likely to have an abnormal tension setting. At this time, under each thread type, there is a difference factor between the current monitoring cycle and each historical sewing process. Therefore, the sum value of all the difference factors corresponding to the current monitoring cycle is used as the difference index. Similarly, the larger the difference index, the higher the possibility of abnormal tension setting in the current monitoring cycle.

[0066] Finally, comprehensively considering the tension difference between the upper thread and the bobbin thread in the current monitoring cycle, more comprehensive tension deviation information is provided: the sum value of the difference indexes of the current monitoring cycle under the two thread types is used as the first tension deviation factor corresponding to the current monitoring cycle. Based on the foregoing analysis, it can be seen that the larger the first tension deviation factor here, the more the tension setting of the current monitoring cycle does not meet the requirements of the current fabric and the less appropriate the setting is.

[0067] Step S202: Compare the deviation of the preset thread tensions of the two threads respectively in the current monitoring cycle and each historical sewing process to obtain the difference factor. Based on the difference characteristics of the difference factors between the current monitoring cycle and the historical sewing processes, determine the second tension deviation factor corresponding to the current monitoring cycle.

[0068] During the sewing process, the upper thread and the bobbin thread also affect each other. Therefore, the deviation characteristics between them can also be determined according to the deviation of the preset thread tensions between the upper thread and the bobbin thread, providing a reference for quantifying the appropriateness of the tension setting in the current monitoring cycle.

[0069] In the current monitoring cycle and each historical sewing process respectively, the difference between the preset thread tensions of the two threads is used as the difference factor. At this time, the current monitoring cycle and each historical sewing process each correspond to a difference factor, and the difference factor is used to reflect the deviation characteristics of the tension setting between the bobbin thread and the upper thread during each sewing. Then, the absolute value of the difference between the difference factors of the current monitoring cycle and each historical sewing process is used as the difference parameter. The larger the difference parameter, the greater the difference between the deviation characteristics of the tension setting between the bobbin thread and the upper thread in the current monitoring cycle and the deviation characteristics under normal historical conditions, indicating that the possibility of unreasonable settings of the bobbin thread and the upper thread in the current monitoring cycle is higher.

[0070] At this time, there is a difference parameter corresponding to the current monitoring cycle and each historical sewing process. Finally, the sum of all the difference parameters corresponding to the current monitoring cycle is used as the second tension deviation factor corresponding to the current monitoring cycle. Based on the above analysis, the larger the second tension deviation factor is, the less the tension setting of the current monitoring cycle meets the requirements of the current fabric, and the more inappropriate the setting is.

[0071] Step S203: The first tension deviation factor and the second tension deviation factor corresponding to the current monitoring period are integrated to obtain a tension deviation index of the current monitoring period.

[0072] Based on the analysis in the previous steps, it can be known that the first tension deviation factor and the second tension deviation factor of the current monitoring cycle are both positively correlated with the inappropriateness of the tension setting of the current monitoring cycle. Therefore, the sum of the first tension deviation factor and the second tension deviation factor corresponding to the current monitoring cycle is normalized as the tension deviation index of the current monitoring cycle. At this time, the larger the tension deviation index is, the more unreasonable the tension setting of the current monitoring cycle is, and the more adjustment is needed. Normalization is a technical means well known to those skilled in the art, and the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0073] After obtaining the tension deviation index of the current monitoring cycle, we can further compare the numerical differences between the tension time series data under each line, and combine the time variation characteristics of the tension value in the line tension time series data to determine the tension characteristic value of each line tension time series data as another indicator reflecting the rationality of the tension setting.

[0074] Preferably, in one embodiment of the present invention, the method for obtaining the tension characteristic value includes:

[0075] See also Figure 3 , which shows a method flow chart of a method for obtaining a tension characteristic value in one embodiment of the present invention, the method comprising the following steps:

[0076] Step S211: Under the same type of wire, select any one of all wire tension time series data as the time series data to be tested and determine the comparison time series data corresponding to the time series data to be tested.

[0077] To facilitate subsequent explanation and illustration, under the same line, any line tension timing data can be selected as the timing data to be tested. If the timing data to be tested is the monitoring line tension timing data, all normal line tension timing data will be used as comparison timing data. If the timing data to be tested is the normal line tension timing data, the remaining normal line tension timing data will be used as comparison timing data.

[0078] Step S212: Under the same thread, compare the numerical differences between the to-be-tested timing data and all corresponding comparison timing data, and determine the tension fluctuation index of the to-be-tested timing data.

[0079] When the thread tension is set too tight or too loose, it will cause the wire to fail to maintain a uniform tension during sewing, and then the tension timing data will fluctuate frequently. Therefore, the numerical differences between the to-be-tested timing data and all corresponding comparison timing data can be compared to determine the tension fluctuation index, which is used to reflect the fluctuation characteristics of the wire during sewing.

[0080] First, under the same thread, calculate the mean value of the tension values of all comparison timing data of the to-be-tested timing data at the same moment to obtain the tension mean value. The tension mean value at each moment represents the tension characteristics that should be possessed under normal sewing conditions. Then, sort the tension mean values at all moments according to time to obtain the reference timing data of the to-be-tested timing data. At this time, the reference timing data can be used as the tension timing data under normal sewing conditions, and the tension values and change conditions therein can be regarded as the change characteristics and numerical characteristics that the tension data should have during sewing when the tension is set reasonably.

[0081] Then, under the same thread, align the to-be-tested timing data with the reference timing data at the moments (based on the length of the to-be-tested timing data), and in the to-be-tested timing data and the reference timing data, take the absolute value of the difference between the tension values at the same moment as the corresponding wire tension difference at the same moment. The greater the wire tension difference at this time, the more the tension value in the to-be-tested timing data deviates from the normal tension value, and the greater the degree of abnormality.

[0082] Next, under the same thread, compare the tension values at the same moment in the to-be-tested timing data and the reference timing data. If the tension values are the same, mark them to obtain the reference points; and segment the timing based on the moments corresponding to the reference points to obtain all time periods. In each time period, take the normalized value of the sum of the wire tension differences corresponding to all moments as the fluctuation degree value of each time period. The greater the fluctuation degree value, the greater the deviation between the tension value of the to-be-tested timing data and the normal tension value within the time period, which is regarded as a greater fluctuation amplitude. Normalization is a technical means well-known to those skilled in the art. The selection of the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0083] Finally, the time period with a fluctuation degree value greater than the preset fluctuation threshold is used as the fluctuation segment. Based on the foregoing analysis, it can be known that when the wire is too tight, that is, when the tension is too large, the fluctuation amplitude will be greater; on the contrary, when the wire is too loose, that is, when the tension is too small, the fluctuation amplitude will be smaller. Therefore, among all the fluctuation segments, the product of the sum of the fluctuation degree values of all the fluctuation segments and the total number of the fluctuation segments is normalized to obtain the tension fluctuation index of each offline measured time series data. At this time, the tension fluctuation index may be positive or negative. The larger the positive value of the tension fluctuation index, the greater the tension setting; on the contrary, the smaller the negative value of the tension fluctuation index, the smaller the tension setting.

[0084] It should be noted that in other embodiments of the present invention, other function mapping methods can also be used to normalize the product of the sum of the fluctuation degree values of all the fluctuation segments and the total number of the fluctuation segments, so that the final tension fluctuation index meets the foregoing logic; the value range of the fluctuation threshold is set to (0.5, 1). In this embodiment of the present invention, it can be set to 0.65, and the specific value can be adjusted according to the implementation scenario and is not limited here.

[0085] Step S213: Analyze the change fluctuation of the tension value with time in the measured time series data under the same type of wire, and determine the change clutter index of the measured time series data.

[0086] When the wire is too tight or too loose, it will cause the wire to not maintain a uniform tension during the conveying process, resulting in irregular fluctuations of the tension. Therefore, under each type of wire, the change fluctuation of the tension value with time in the measured time series data can be analyzed to determine the change clutter index of the measured time series data.

[0087] Under each type of wire, obtain the value range of the tension value in the measured time series data, and perform dimensionality reduction segmentation on the tension values within the value range based on the piecewise aggregate approximation algorithm to obtain each tension segmentation point.

[0088] Then, in the measured time series data, segment the measured time series data based on the tension values corresponding to the tension segmentation points in time sequence, so as to obtain a plurality of tension data segments. At this time, the tension values in each tension data segment have relatively similar characteristics, which helps to analyze the change of the tension value in the subsequent process.

[0089] The middle moment of each tension data segment is obtained, and the absolute value of the difference between the middle moments of each two adjacent tension data segments is calculated in the time series as the time deviation value. The larger the time deviation value, the larger the interval between the two adjacent tension data segments. When the intervals between all adjacent tension data segments are more inconsistent, it can be regarded as the more irregular the change of the tension value of the time series data to be tested. Therefore, the variance of all the time deviation values ​​corresponding to the time series data to be tested is normalized and used as the change disorder index of each offline time series data to be tested. The larger the change disorder index, the more unreasonable the setting of the tension value of the time series data to be tested is, and the less it meets the tension arrangement requirements of normal sewing. Normalization is a technical means well known to those skilled in the art, and the selection of the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0090] Step S214: Under each offline line, the tension fluctuation index and the variation disorder index corresponding to the time series data to be tested are integrated to determine the tension characteristic value of the time series data to be tested under each offline line.

[0091] Based on the above steps, it can be known that under each line, the larger the change disorder index corresponding to the measured timing data is, the less the setting of the tension value meets the requirements of normal sewing; the larger the tension fluctuation index is, the positive value indicates that the tension is set too large; the smaller the tension fluctuation index is, the negative value indicates that the tension is set smaller.

[0092] Therefore, the tension fluctuation index of the time series data to be tested is multiplied by the change disorder index. If the product is positive and the larger it is, the more likely it is that the tension setting of each offline time series data to be tested is unreasonable, and the possibility of being too large is higher; conversely, if the product is negative and the smaller it is, the more likely it is that the tension setting of each offline time series data to be tested is unreasonable, but the possibility of being too small is higher; the value after normalization of the obtained product is used as the tension characteristic value of each offline time series data to be tested. At this time, the larger the tension characteristic value, the larger the tension setting, and the smaller the tension characteristic value, the smaller the tension setting. Since the product here may be positive or negative, the normalization can use the Sigmoid() function.

[0093] At this point, the tension deviation index of the current monitoring cycle can be obtained to reflect the rationality of the tension setting; at the same time, the tension characteristic value of the tension timing data of each line under each line can be obtained to reflect whether the tension setting is too large or too small. For different sewing machines, different adjusted parameters, years of use, and different degrees of wear will lead to differences in appropriate tension settings, so the normal line tension timing data in the historical sewing process can be used as a reference to judge and adjust the corresponding appropriate tension. Specifically, the tension deviation index of the current monitoring cycle and the deviation between all tension characteristic values ​​under each line can be combined to determine the tension adjustment degree value of the current monitoring cycle for each line.

[0094] Preferably, in one embodiment of the present invention, the method for obtaining the tension adjustment degree value includes:

[0095] Under each line, the average of the tension characteristic values ​​corresponding to all normal line tension time series data is taken as the standard value. The standard value can provide the characteristics that the tension value should have in a normal sewing process as a reference standard.

[0096] Then, the difference between the tension characteristic value corresponding to the tension time series data of the monitoring line and the standard value is calculated. If the difference is positive and the larger it is, the greater the degree to which the tension characteristic value of the tension time series data of the monitoring line is higher than the reference standard; conversely, if the difference is negative and the smaller it is, the greater the degree to which the tension time series data of the monitoring line is lower than the reference standard, the hyperbolic tangent function is used to process the difference to obtain the deviation factor. At this time, the value range of the deviation factor is (-1,1).

[0097] Given that the larger the tension deviation index of the current monitoring cycle is, the more unreasonable the tension setting of the current monitoring cycle is, and the more it needs to be adjusted; therefore, the deviation factor obtained above is multiplied by the tension deviation index of the current monitoring cycle. At this time, the larger the product is, the more unreasonable the tension setting of the current monitoring cycle is, and the greater the degree of deviation is; conversely, the smaller the product is, the more unreasonable the tension setting of the current monitoring cycle is, but the greater the degree of deviation is, the obtained product is normalized to obtain the tension adjustment degree value of the current monitoring cycle of each line. At this time, the larger the tension adjustment degree value is, the more the tension value needs to be reduced; conversely, the smaller the tension adjustment degree value is, the more the tension value needs to be increased. The normalization here can use the Sigmoid() function.

[0098] Step S3: Based on the tension deviation index of the historical monitoring cycle in the current sewing process and the tension adjustment degree value, the tension adjustment degree value of the current monitoring cycle is adjusted to obtain a new tension adjustment degree value under each line; under each line, based on the new tension adjustment degree value, the current monitoring cycle and the preset line tension in the historical sewing process, the adaptive tension value of the next monitoring cycle is determined.

[0099] In the embodiment of the present invention, in order to ensure the quality of sewing, the current sewing process is divided into multiple monitoring cycles, and the tensions of the upper thread and the bobbin thread in each monitoring cycle are adjusted in real time. Therefore, during the current sewing process, based on the tension deviation index and the change of the tension adjustment degree value in the historical monitoring cycle, as a feedback mechanism, the tension adjustment degree value in the current monitoring cycle can be further adjusted to obtain the new tension adjustment degree value of each thread in the current monitoring cycle. By introducing the data of the historical monitoring cycle in the current sewing process, the adjustment can be made more continuous and stable. Finally, for each thread, based on the new tension adjustment degree value, the current monitoring cycle and the preset tension value in the historical process, the adaptive tension value in the next monitoring cycle is determined.

[0100] Preferably, in an embodiment of the present invention, the method for obtaining the new tension adjustment degree value includes:

[0101] For each thread, in every two adjacent historical monitoring cycles in the current sewing process, the difference between the tension deviation index of the previous historical monitoring cycle and the tension deviation index of the subsequent historical monitoring cycle is used as the first change factor of the subsequent historical monitoring cycle. The larger the first change factor, the closer the tension value in the subsequent historical monitoring cycle is to the normal and reasonable tension value, indicating that the adjustment is more appropriate. Therefore, the participation degree of the comparison before and after the tension adjustment as the subsequent tension modification should be smaller.

[0102] The difference between the absolute value of the tension adjustment degree value of the previous historical monitoring cycle and the absolute value of the tension adjustment degree value of the subsequent historical monitoring cycle is used as the second change factor of the subsequent historical monitoring cycle. Similarly, the larger the second change factor, the more appropriate the adjustment. Then, the ratio of the second change factor of the last historical monitoring cycle to the maximum value of the second change factors of all historical detection cycles is used as the weight factor. The larger the weight factor, the more appropriate the adjustment of the last historical monitoring cycle. Therefore, the participation degree of the comparison before and after the tension adjustment as the subsequent tension modification should be smaller.

[0103] Next, multiply the weight factor by the first change factor of the last historical monitoring period. The larger the resulting product, the smaller the participation degree. Therefore, perform a negative correlation mapping process on the adjustment factor to correct the logical relationship and obtain the adjustment factor of the last historical monitoring period. Finally, normalize the product of the adjustment factor and the tension adjustment degree value of the current monitoring period, and use the normalized value as the new tension adjustment degree value for each type of thread. At this time, the new tension adjustment degree value incorporates the data of the historical monitoring periods during the current sewing process. Therefore, adjusting the tension value of the next monitoring period based on the new tension adjustment degree value will also be more stable and accurate. The negative correlation mapping here can use the formula exp(-x), where exp() represents the exponential function with the natural constant e as the base, and x represents the independent variable.

[0104] It should be noted that the calculation methods of parameters such as the tension deviation index and the tension adjustment degree value of the historical monitoring periods during the current sewing process are the same as those of the current monitoring period, and will not be elaborated here.

[0105] Thus, the new tension adjustment degree value corresponding to each type of thread in the current monitoring period can be obtained. Then, by combining the difference between the current monitoring period and the preset thread tension in the historical process, the adaptive adjustment of the tension value of the next monitoring period can be realized, and the adaptive tension value can be obtained.

[0106] Preferably, in an embodiment of the present invention, the method for obtaining the adaptive tension value includes:

[0107] For each type of thread, take the mean value of the preset thread tensions in all historical sewing processes as the reference value, and take the difference between the preset thread tension of the current monitoring period and the corresponding reference value as the deviation degree value. The larger the positive deviation degree value, the higher the preset tension value of the current monitoring period is higher than the reference value, and it needs to be adjusted smaller; on the contrary, the smaller the negative deviation degree value, the lower the preset tension value of the current monitoring period is lower than the reference value, and it needs to be adjusted larger.

[0108] Finally, take the product of the deviation degree value and the new tension adjustment degree value as the adjustment index. If the adjustment index is positive, the tension adjustment direction is to adjust smaller; if the adjustment index is negative, the tension adjustment direction is to adjust larger. The absolute value of the adjustment index reflects the degree of adjustment. Therefore, take the difference between the preset thread tension of the current monitoring period and the adjustment index as the adaptive tension value of each type of thread in the next monitoring period, that is, set the preset thread tension of each type of thread in the next monitoring period to the corresponding adaptive tension value.

[0109] In summary, the embodiments of the present invention analyze the tension fluctuations over a period of time, so that the influence relationship between the threads and the fluctuations of the thread tension can be taken into account. First, the monitored thread tension time-series data of the current monitoring period in the current sewing process and the normal thread tension time-series data in the historical sewing process are obtained, which provide sufficient data support for the precise adjustment of the thread tension. Comparing the difference between the current monitoring period and the preset thread tension in the historical sewing process to determine the tension deviation index of the current monitoring period helps to initially detect abnormal fluctuations in the thread tension. When the tension of the top thread or the bobbin thread is too large or too small, it will cause irregular fluctuations in the thread during the sewing process, and there are changes in different fluctuation amplitudes. Therefore, under each type of thread, the numerical differences between the thread tension time-series data are compared, and combined with the change of the tension value over time in the thread tension time-series data, the tension characteristic value of each thread tension time-series data is determined to reflect the change of the thread tension. Then, by comprehensively considering the tension deviation index of the current monitoring period and the deviation between all tension characteristic values under each type of thread, the tension adjustment degree value of the current monitoring period under each type of thread is obtained, realizing a comprehensive evaluation of the thread tension. To achieve more precise adjustment of the thread tension, the present invention adjusts the tension adjustment degree value of the current monitoring period based on the tension deviation index and the tension adjustment degree value of the historical monitoring period in the current sewing process, and obtains a new tension adjustment degree value under each type of thread. By introducing the historical monitoring period data, the adjustment becomes more continuous and stable. Finally, under each type of thread, based on the new tension adjustment degree value, the preset thread tension in the current monitoring period and the historical process, the adaptive tension value of the next monitoring period is determined, realizing the dynamic and adaptive adjustment of the thread tension, improving the accuracy of the adjustment, and keeping the thread tension in the sewing process within a suitable range at all times.

[0110] It should be noted that the above sequence of the embodiments of the present invention is only for description and does not represent the advantages or disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0111] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. The key point of each embodiment is to illustrate the differences from other embodiments.

Claims

1. A method for adaptively adjusting the thread tension of a flatbed sewing machine, characterized in that, The method includes: Obtaining the line tension time series data, including the monitored line tension time series data of the current monitoring period in the current sewing process and the normal line tension time series data in the historical sewing process; wherein, the line includes the upper thread and the bobbin thread. Comparing the difference between the current monitoring period and the preset line tension in the historical sewing process to determine the tension deviation index of the current monitoring period; for each type of line, comparing the numerical differences between the line tension time series data, and combining the change of the tension value with time in the line tension time series data to determine the tension characteristic value of each line tension time series data; comprehensively considering the tension deviation index of the current monitoring period and the deviation between all tension characteristic values for each type of line, obtaining the tension adjustment degree value of the current monitoring period for each type of line. Adjusting the tension adjustment degree value of the current monitoring period based on the tension deviation index and the tension adjustment degree value of the historical monitoring period in the current sewing process to obtain the new tension adjustment degree value for each type of line; for each type of line, based on the new tension adjustment degree value, the preset line tension in the current monitoring period and the historical sewing process, determining the adaptive tension value of the next monitoring period.

2. The method for adaptively adjusting the thread tension of a flatbed sewing machine according to claim 1, characterized in that, The method for obtaining the tension deviation index includes: For each type of line, comparing the difference between the preset line tension in the current monitoring period and the historical process to determine the first tension deviation factor corresponding to the current monitoring period. In the current monitoring period and each historical sewing process respectively, taking the difference between the preset line tensions of the two lines as the difference factor, taking the absolute value of the difference between the difference factors in the current monitoring period and each historical sewing process as the difference parameter, and taking the sum value of all difference parameters corresponding to the current monitoring period as the second tension deviation factor corresponding to the current monitoring period. Taking the normalized value of the sum of the first tension deviation factor and the second tension deviation factor corresponding to the current monitoring period as the tension deviation index of the current monitoring period.

3. A thread tension self - adaptive adjustment method for a flatbed sewing machine according to claim 2, characterized in that, The method for obtaining the first tension deviation factor includes: For the same type of line, taking the absolute value of the difference between the preset line tension in the current monitoring period and the preset line tension in each historical sewing process as the difference factor, and taking the sum value of all difference factors corresponding to the current monitoring period as the difference index. Taking the sum value of the difference indexes of the current monitoring period for the two types of lines as the first tension deviation factor corresponding to the current monitoring period.

4. A method for adaptively adjusting the thread tension of a flatbed sewing machine according to claim 1, characterized in that, The method for obtaining the tension characteristic value includes: For the same type of line, randomly selecting a line tension time series data as the time series data to be measured. If the time series data to be measured is the monitored line tension time series data, then all normal line tension time series data are used as the comparison time series data. If the time series data to be measured is the normal line tension time series data, then the remaining normal line tension time series data are used as the comparison time series data. For the same type of line, comparing the numerical differences between the time series data to be measured and all corresponding comparison time series data to determine the tension fluctuation index of the time series data to be measured. For the same type of line, analyzing the change of the tension value with time in the time series data to be measured to determine the change disorder index of the time series data to be measured. The value obtained by normalizing the product of the tension fluctuation index and the change disorder index of the time-series data to be measured is used as the tension eigenvalue of each offline time-series data to be measured.

5. A thread tension self - adaptive adjustment method for a flatbed sewing machine according to claim 4, characterized in that, The method for obtaining the tension fluctuation index includes: Under the same line type, calculate the mean value of the tension values of all comparison time-series data of the time-series data to be measured at the same moment to obtain the tension mean value. Sort the tension mean values at all moments according to time to obtain the reference time-series data of the time-series data to be measured. Under the same line type, in the time-series data to be measured and the reference time-series data, take the absolute value of the difference between the tension values at the same moment as the corresponding line tension difference at the same moment. Under the same line type, compare the tension values at the same moment in the time-series data to be measured and the reference time-series data. If the tension values are the same, mark them to obtain reference points. Based on the moments corresponding to the reference points, segment the time series to obtain all time periods. In each time period, take the value obtained by normalizing the sum value of the line tension differences corresponding to all moments as the fluctuation degree value of each time period. Take the time periods with the fluctuation degree value greater than the preset fluctuation threshold as the fluctuation segments. The value obtained by standardizing the product of the sum value of the fluctuation degree values of all fluctuation segments and the total number of fluctuation segments is used as the tension fluctuation index of each offline time-series data to be measured.

6. A thread tension self - adaptive adjustment method for a flatbed sewing machine according to claim 4, characterized in that, The method for obtaining the change disorder index includes: Under each line type, obtain the value range of the tension values in the time-series data to be measured, and perform dimensionality reduction segmentation on the tension values within the value range based on the piecewise aggregate approximation algorithm to obtain each tension segmentation point. In the time-series data to be measured, segment the time-series data to be measured based on the tension segmentation points in time sequence to obtain multiple tension data segments. Obtain the middle moment of each tension data segment. In the time sequence, calculate the absolute value of the difference between the middle moments of every two adjacent tension data segments as the moment deviation value. The value obtained by normalizing the variance of all moment deviation values corresponding to the time-series data to be measured is used as the change disorder index of each offline time-series data to be measured.

7. A thread tension self - adaptive adjustment method for a flatbed sewing machine according to claim 1, characterized in that, The method for obtaining the tension adjustment degree value includes: Under each line type, take the mean value of the tension eigenvalues corresponding to all normal tension time-series data as the standard value. Process the difference between the tension eigenvalue corresponding to the monitored line tension time-series data and the standard value to obtain the deviation factor, and the value range of the deviation factor is (-1, 1). Normalize the product of the deviation factor and the tension deviation index of the current monitoring period to obtain the tension adjustment degree value of the current monitoring period under each line type.

8. A method for adaptively adjusting the thread tension of a flatbed sewing machine according to claim 1, characterized in that The method for obtaining the new tension adjustment degree value includes: Under each line type, in every two adjacent historical monitoring periods during the current sewing process, take the difference between the tension deviation index of the previous historical monitoring period and the tension deviation index of the next historical monitoring period as the first change factor of the next historical monitoring period, and take the difference between the absolute value of the tension adjustment degree value of the previous historical monitoring period and the absolute value of the tension adjustment degree value of the next historical monitoring period as the second change factor of the next historical monitoring period. The ratio of the second change factor in the last historical monitoring period to the maximum value of the second change factors in all historical detection periods is used as the weight factor; The value obtained by performing a negative correlation mapping on the product of the weight factor and the first change factor in the last historical monitoring period is used as the adjustment factor for the last historical monitoring period; The value obtained by normalizing the product of the adjustment factor and the tension adjustment degree value in the current monitoring period is used as the new tension adjustment degree value.

9. The method for adaptively adjusting the thread tension of a flatbed sewing machine according to claim 1, characterized in that, The method for obtaining the adaptive tension value includes: Under each type of thread, the mean value of the preset thread tensions in all historical sewing processes is used as the reference value, and the difference between the preset thread tension in the current monitoring period and the corresponding reference value is used as the deviation degree value; The product of the deviation degree value and the new tension adjustment degree value is used as the adjustment index, and the difference between the preset thread tension in the current monitoring period and the adjustment index is used as the adaptive tension value of each type of thread in the next monitoring period.

10. A thread tension self - adaptive adjustment method for a flatbed sewing machine according to claim 5, characterized in that, The value range of the preset fluctuation threshold is (0.5, 1).

Images