An automatic monitoring system based on carbon fiber multi-axial warp knitting machine

Abstract

The present invention relates to the field of textile technology, and in particular to an automatic monitoring system based on a carbon fiber multi-axial warp knitting machine, comprising: a data acquisition unit; a monitoring and analysis unit for determining a warp knitting state according to an oil coefficient and a vibration amplitude, and responding to the warp knitting state to determine a monitoring method; an optimization analysis unit for determining an optimization method according to a pulse coefficient; an image selection unit for determining a differential selection method according to an analysis condition, wherein the differential selection method is to determine a first selection method according to an influence state of a special point or to determine a second selection method according to a similarity coefficient of a radiation area; an instability analysis unit for determining an instability coefficient determination method according to a sub-region difference coefficient in a differential image; the present invention can improve the accuracy of monitoring the weaving process of a warp knitting machine.

Description

Technical Field

[0001] The invention relates to the technical field of textiles, and in particular to an automatic monitoring system based on a carbon fiber multi-axial warp knitting machine. Background Art

[0002] The inherent breakage characteristics of carbon fiber during the manufacturing process pose serious challenges to the production of carbon fiber fabrics. In addition, the quality of the fabric is affected by a variety of complex quality parameters, making it particularly difficult to accurately monitor and warn of abnormal conditions, posing a serious threat to production quality. Therefore, how to improve the accuracy of weaving process monitoring and timely detect and warn of abnormal fluctuations is a technical problem that needs to be urgently solved by technical personnel in this field.

[0003] Chinese patent publication number CN117721587A discloses a remote monitoring yarn breakage automatic stop system for warp knitting machines, including: using artificial intelligence detection technology based on machine vision, obtaining yarn images taken with a high-resolution camera and extracting features to detect yarn breakage in the warp knitting machine, and if yarn breakage occurs, the machine needs to be stopped in time. It can be seen that the above technical solution has the following problems: the monitoring method is single, and the monitoring method is not adaptively adjusted according to the actual operating status of the warp knitting machine, resulting in poor monitoring effect. Summary of the invention

[0004] To this end, the present invention provides an automated monitoring system based on a carbon fiber multi-axial warp knitting machine, so as to overcome the problem in the prior art that the monitoring mode is not adaptively adjusted according to the actual operating state of the warp knitting machine, resulting in poor monitoring effect.

[0005] To achieve the above object, the present invention provides an automatic monitoring system based on a carbon fiber multi-axial warp knitting machine, comprising:

[0006] A data collection unit, used to collect monitoring data;

[0007] A monitoring and analysis unit connected to the data acquisition unit, for determining a warp knitting state according to an oil coefficient and a vibration amplitude, and responding to the warp knitting state to determine a monitoring mode, wherein the monitoring mode is to determine an optimization mode according to a pulse coefficient, or to determine whether a target monitoring period is in an abnormal state according to an instability coefficient of a monitoring image;

[0008] An optimization analysis unit connected to the monitoring analysis unit is used to determine an optimization method according to the pulse coefficient, wherein the optimization method is to determine whether the target monitoring period is in an abnormal state according to the impact threshold, or to determine whether the target monitoring period is in an abnormal state according to the coupling ratio displacement coefficient and the tension law coefficient;

[0009] An image selection unit, which is connected to the monitoring and analysis unit, and is used to determine a differential selection method according to the analysis conditions, wherein the differential selection method is to determine a first selection method according to the influence state of a special point or to determine a second selection method according to a similarity coefficient of a radiation area;

[0010] An instability analysis unit is respectively connected to the monitoring and analysis unit and the image selection unit, and is used to determine an instability coefficient determination method according to a sub-region difference coefficient of a differential image. The instability coefficient determination method is to determine the instability coefficient according to a domain variation reference value and a sub-influence threshold or to determine the instability coefficient according to a sub-influence threshold.

[0011] Furthermore, the monitoring and analysis unit determines the warp knitting state according to the oil coefficient and the vibration amplitude, and the warp knitting state includes:

[0012] A first warp knitting state in which the oil coefficient is less than a preset oil coefficient and the vibration amplitude is less than a preset vibration amplitude;

[0013] The second warp knitting state in which the oil coefficient is greater than or equal to the preset oil coefficient or the vibration amplitude is greater than or equal to the preset vibration amplitude.

[0014] Further, the monitoring and analysis unit responds to the warp knitting state to determine the monitoring mode;

[0015] The monitoring and analyzing unit responds to the first warp knitting state and determines to execute a monitoring mode that determines an optimization mode according to a pulse coefficient;

[0016] The monitoring and analyzing unit responds to the second warp knitting state and determines whether a monitoring mode of determining whether a target monitoring period is in an abnormal state according to an instability coefficient of a monitoring image is executed.

[0017] Furthermore, the optimization analysis unit determines the optimization method according to the pulse coefficient, wherein:

[0018] If the pulse coefficient is greater than or equal to the preset pulse coefficient, the optimization method is to determine whether the target monitoring period is in an abnormal state according to the impact threshold;

[0019] If the pulse coefficient is less than the preset pulse coefficient, the optimization method is to determine whether the target monitoring period is in an abnormal state based on the coupling ratio displacement coefficient and the tension law coefficient.

[0020] Further, the image selection unit responds to the analysis conditions to determine the differential selection method;

[0021] The analysis condition of the image selection unit response is that the special point ratio is greater than or equal to the preset special point ratio or the special point distribution coefficient is greater than or equal to the preset special point distribution coefficient, and the difference selection method is determined to determine the first selection method according to the influence state of the special point;

[0022] The analysis condition responded by the image selection unit is that the proportion of special points is less than the preset special point proportion and the special point distribution coefficient is less than the preset special point distribution coefficient, and the difference selection method is determined to determine the second selection method according to the similarity coefficient of the radiation area.

[0023] Further, the image selection unit responds to the influence state of the special point to determine the first selection mode;

[0024] The image selection unit responds to the first impact state and determines to execute a first selection method of correlation selection based on the interval reference value and the data correlation coefficient;

[0025] The image selection unit responds to the second impact state and determines to execute a first selection method of selecting a differential point according to a feature threshold;

[0026] The influence states of the special points include: a first influence state in which the interval coefficient is less than a preset interval coefficient, and a second influence state in which the interval coefficient is greater than or equal to the preset interval coefficient.

[0027] Furthermore, the image selection unit determines a second selection method according to a similarity coefficient of the radiation area;

[0028] If the similarity coefficient is greater than or equal to the preset similarity coefficient, the second selection method is to determine the interval of the differential points in the influence area according to the characteristic coefficient of the influence area;

[0029] If the similarity coefficient is less than the preset similarity coefficient, the second selection method is to determine the number of differential points according to the regional evaluation coefficient;

[0030] The difference point interval is negatively correlated with the influence area characteristic coefficient, and the difference point quantity is positively correlated with the area evaluation coefficient.

[0031] Furthermore, the instability analysis unit responds to the setting conditions to determine the region segmentation method of the characteristic region, wherein:

[0032] The setting condition of the instability analysis unit response is that the connected domain distribution coefficient is greater than or equal to the preset connected domain distribution coefficient and the connected similarity threshold is greater than or equal to the preset connected similarity threshold, and the determination area segmentation method is uniform division according to the connected threshold;

[0033] The setting condition of the instability analysis unit response is that the connected domain distribution coefficient is less than the preset connected domain distribution coefficient or the connected similarity threshold is less than the preset connected similarity threshold, and the determination area segmentation method is to perform associated division according to the connected domain difference and the distance reference value;

[0034] The feature region is the smallest rectangle in the difference image that can contain each connected domain.

[0035] Further, the instability analysis unit determines the instability coefficient determination method according to the sub-region difference coefficient of the differential image;

[0036] If the sub-region difference coefficient is greater than or equal to the preset sub-region difference coefficient, the instability coefficient is determined according to the domain variation reference value and the sub-influence threshold value;

[0037] If the sub-region difference coefficient is less than the preset sub-region difference coefficient, the instability coefficient is determined according to the sub-influence threshold.

[0038] Furthermore, the instability analysis unit determines a sub-influence threshold according to the influence coefficient, wherein the influence coefficient is determined according to a reference value of the influence sub-region distribution.

[0039] If the affected sub-region distribution reference value is greater than or equal to the preset affected sub-region distribution reference value, the influence coefficient is determined according to the connectivity reference value;

[0040] If the affected sub-region distribution reference value is less than the preset affected sub-region distribution reference value, the influence coefficient is determined according to the distance influence value.

[0041] Compared with the prior art, the beneficial effect of the present invention lies in that, in the technical scheme of the present invention, the warp knitting state is determined according to the oil coefficient and the vibration amplitude, and the wear degree of the warp knitting machine is effectively reflected by the oil coefficient and the vibration amplitude, and then different monitoring methods are adaptively selected according to the warp knitting state, so that the selection of monitoring method is more in line with the actual application scenario, avoiding the problem of poor monitoring effect caused by large fluctuations in monitoring data when the degree of wear is large, thereby improving the reliability of the monitoring results.

[0042] Furthermore, in the present invention, the pulse coefficient is used to effectively reflect the weaving speed of the warp knitting machine, and then the optimization method is determined according to the pulse coefficient, so that the selection of the optimization method is more in line with the actual monitoring needs, avoiding the problem of poor monitoring effect caused by unstable tension when the weaving speed is fast, and being able to discover potential problems in time, thereby avoiding the increase in maintenance costs caused by the deterioration of potential problems in the operation of the warp knitting machine.

[0043] Furthermore, the present invention effectively reflects the distribution of special points in the feature data through analysis conditions, and then adaptively selects different differential selection methods according to the analysis conditions, so that the selection of the differential selection method can improve the reliability of monitoring, avoid the problem of poor accuracy of the instability coefficient due to the poor representativeness of the differential image corresponding to the differential point, and thus improve the accuracy of monitoring.

[0044] Furthermore, in the present invention, a first selection method is determined according to the influence status of special points, and the influence status of special points is used to effectively reflect the degree of correlation of special points. Different first selection methods are adaptively selected according to the influence status of special points, so that the selection of the first selection method can select differential points according to the actual state of the target monitoring period, and the second selection method is determined according to the similarity coefficient of the radiation area, and the similarity coefficient of the radiation area is used to effectively reflect the similarity of the radiation area. Different second selection methods are adaptively selected according to the similarity coefficient, so that the selection of the second selection method is more consistent with the actual monitoring scene, so that the selected differential points can more accurately reflect the actual situation of the data, thereby reducing errors caused by improper selection.

[0045] Furthermore, in the present invention, the regional segmentation method of the feature area is determined according to the setting conditions, and the similarity of the connected domains is effectively reflected by the setting conditions, and then different regional segmentation methods are adaptively selected according to the setting conditions, so that the selection of the regional segmentation method is more in line with the actual application scenario, avoiding the problem of poor accuracy of instability coefficient determination due to unreasonable sub-region division, thereby improving the accuracy of monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a unit connection diagram of the automatic monitoring system based on the carbon fiber multi-axial warp knitting machine of the present invention;

[0047] Figure 2 This is a flow chart of the present invention for determining a monitoring method according to a warp knitting state;

[0048] Figure 3 This is a flow chart of determining the differential selection method according to the analysis conditions of the present invention;

[0049] Figure 4 The present invention is a flow chart of a method for determining a region segmentation method of a feature region according to setting conditions. DETAILED DESCRIPTION

[0050] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0051] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0052] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0053] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0054] See also Figures 1 to 4 As shown, the present invention provides an automatic monitoring system based on a carbon fiber multi-axial warp knitting machine, comprising:

[0055] A data collection unit, used to collect monitoring data;

[0056] A monitoring and analysis unit connected to the data acquisition unit, for determining a warp knitting state according to an oil coefficient and a vibration amplitude, and responding to the warp knitting state to determine a monitoring mode, wherein the monitoring mode is to determine an optimization mode according to a pulse coefficient, or to determine whether a target monitoring period is in an abnormal state according to an instability coefficient of a monitoring image;

[0057] An optimization analysis unit connected to the monitoring analysis unit is used to determine an optimization method according to the pulse coefficient, wherein the optimization method is to determine whether the target monitoring period is in an abnormal state according to the impact threshold, or to determine whether the target monitoring period is in an abnormal state according to the coupling ratio displacement coefficient and the tension law coefficient;

[0058] An image selection unit, which is connected to the monitoring and analysis unit, and is used to determine a differential selection method according to the analysis conditions, wherein the differential selection method is to determine a first selection method according to the influence state of a special point or to determine a second selection method according to a similarity coefficient of a radiation area;

[0059] An instability analysis unit is respectively connected to the monitoring and analysis unit and the image selection unit, and is used to determine an instability coefficient determination method according to a sub-region difference coefficient of a differential image. The instability coefficient determination method is to determine the instability coefficient according to a domain variation reference value and a sub-influence threshold or to determine the instability coefficient according to a sub-influence threshold.

[0060] The application scenario of the present invention is abnormality monitoring of warp knitting machines. The monitoring data include but are not limited to yarn tension, vibration frequency of bearings, temperature of motors and lateral displacement. The lateral displacement is the displacement of the lateral offset distance of the wire ring and the comb bar in the lateral direction. The yarn tension, vibration frequency of bearings, temperature of motors and lateral displacement are measured by tension sensors, temperature sensors, vibration sensors and displacement sensors respectively. This is easy for technicians in this field to understand and will not be described in detail. A continuous cyclic monitoring cycle is provided in the present invention. The data status is determined once at the end of each monitoring cycle. The duration of the monitoring cycle can be set according to user needs. The greater the user's demand for monitoring accuracy, the shorter the duration of the monitoring cycle. A value of a monitoring cycle is provided. The monitoring cycle is 15 minutes. The target monitoring cycle is the monitoring cycle currently being monitored.

[0061] In the present invention, several historical records are correspondingly provided, and any historical record records at least one iterative fluctuation coefficient, instability coefficient, oil coefficient, vibration amplitude, influence threshold, coupling ratio dislocation coefficient, tension law coefficient, special point distribution coefficient and fluctuation threshold in the historical process of monitoring the target warp knitting machine, and each historical record corresponds to a qualified mark, which records whether the monitoring process of the warp knitting machine meets the user's needs. The qualified mark can be recorded manually. It can be understood that the user can determine whether the monitoring process meets the needs according to the self-set indicators. The self-set indicators can be but not limited to the false alarm rate, which will not be elaborated here. Among them, the false alarm rate is the number of times the abnormal state of the target monitoring cycle is wrongly determined;

[0062] The monitoring points in the present invention are set by the user, and a monitoring point setting method is provided, in which the starting time of a single monitoring cycle is recorded as a monitoring point, and every 1 second is recorded as a monitoring point, that is, monitoring data is recorded once every 1 second.

[0063] Specifically, the monitoring and analysis unit determines the warp knitting state according to the oil coefficient and the vibration amplitude, and the warp knitting state includes:

[0064] A first warp knitting state in which the oil coefficient is less than a preset oil coefficient and the vibration amplitude is less than a preset vibration amplitude;

[0065] The second warp knitting state in which the oil coefficient is greater than or equal to the preset oil coefficient or the vibration amplitude is greater than or equal to the preset vibration amplitude.

[0066] Among them, oil coefficient = fluctuation threshold + oil extreme value, the fluctuation threshold is the standard deviation of the oil reference value corresponding to each monitoring point in the target monitoring period, the oil extreme value is the maximum value of the oil reference value corresponding to each monitoring point in the target monitoring period; the oil reference value is the average value of the concentration values ​​of each metal element in the lubricating oil of the warp knitting machine monitored at a single monitoring point, the elements include but are not limited to iron, copper, chromium, nickel and titanium, the lubricating oil is the oil at the bearing of the warp knitting machine, the concentration value is detected by a method of measuring the oil obtained after sampling by a sampler after ultra-high voltage excitation between the disk and rod electrodes by an oil analysis spectrometer, and the vibration amplitude is the maximum value of the amplitude corresponding to each monitoring point in the target monitoring period; the amplitude is measured by a vibration sensor installed at the bearing of the warp knitting machine, which is easy to understand for those skilled in the art and will not be described in detail;

[0067] The values ​​of the preset oil coefficient and the preset vibration amplitude can be determined by the user according to the actual application scenario. The larger the values ​​of the preset oil coefficient and the preset vibration amplitude are, the greater the user's demand for determining the optimization method based on the pulse coefficient. A value of the preset oil coefficient and the preset vibration amplitude is provided, and the historical records of determining the optimization method based on the pulse coefficient are detected. The average value of the oil coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset oil coefficient, and the average value of the vibration amplitude corresponding to the historical records that can meet the user's needs is recorded as the preset vibration amplitude.

[0068] Specifically, the monitoring and analysis unit responds to the warp knitting state to determine the monitoring mode;

[0069] The monitoring and analyzing unit responds to the first warp knitting state and determines to execute a monitoring mode that determines an optimization mode according to a pulse coefficient;

[0070] The monitoring and analyzing unit responds to the second warp knitting state and determines whether a monitoring mode of determining whether a target monitoring period is in an abnormal state according to an instability coefficient of a monitoring image is executed.

[0071] Among them, in the present invention, the monitoring image is actually collected by an industrial camera installed above the spinning section. The monitoring image is a state image of the yarn, which is used to monitor abnormal conditions such as breakage and defects of the yarn during the production process;

[0072] If the instability coefficient is greater than or equal to the preset instability coefficient, the target monitoring period is determined to be in an abnormal state, and an abnormal reminder is sent to the user; if the instability coefficient is less than the preset instability coefficient, the target monitoring period is determined to be in a normal state;

[0073] The value of the preset instability coefficient can be determined by the user according to the actual application scenario. The greater the user's demand for reducing maintenance costs, the smaller the value of the preset instability coefficient is. A value of the preset instability coefficient is provided, and a historical record of determining whether a target monitoring period is in an abnormal state based on the instability coefficient of the monitoring image is detected. The minimum value of the instability coefficient corresponding to the historical record in which the target monitoring period is in an abnormal state and can meet the user's needs is recorded as the preset instability coefficient.

[0074] Specifically, the optimization analysis unit determines the optimization method according to the pulse coefficient, wherein:

[0075] If the pulse coefficient is greater than or equal to the preset pulse coefficient, the optimization method is to determine whether the target monitoring period is in an abnormal state according to the impact threshold;

[0076] If the pulse coefficient is less than the preset pulse coefficient, the optimization method is to determine whether the target monitoring period is in an abnormal state based on the coupling ratio displacement coefficient and the tension law coefficient.

[0077] Among them, when determining whether the target monitoring period is in an abnormal state according to the impact threshold, if the impact threshold is greater than or equal to the preset impact threshold, the target monitoring period is determined to be in an abnormal state, and an abnormal reminder is sent to the user; if the impact threshold is less than the preset impact threshold, the target monitoring period is determined to be in a normal state;

[0078] When determining whether the target monitoring cycle is in an abnormal state according to the coupling ratio misalignment coefficient and the tension law coefficient, if the coupling ratio misalignment coefficient is greater than or equal to the preset coupling ratio misalignment coefficient or the tension law coefficient is less than the preset tension law coefficient, the target monitoring cycle is determined to be in an abnormal state, and an abnormal reminder is sent to the user; if the coupling ratio misalignment coefficient is less than the preset coupling ratio misalignment coefficient and the tension law coefficient is greater than or equal to the preset tension law coefficient, the target monitoring cycle is determined to be in a normal state;

[0079] Pulse coefficient = 60M / PT, M is the number of photoelectric sensor pulses detected in the target monitoring period, P is the number of pulses generated by one rotation of the wire ring, T is the time length of the target monitoring period, M and P are both monitored by photoelectric sensors installed near the wire ring, which is easy for technicians in this field to understand and will not be described in detail; Impact threshold = lateral displacement reference value + warp yarn speed fluctuation coefficient, lateral displacement reference value is the maximum value of the lateral displacement corresponding to each monitoring point in the target monitoring period, warp yarn speed fluctuation coefficient is the standard deviation of the warp yarn speed corresponding to each monitoring point in the target monitoring period, warp yarn speed is the linear speed of the warp yarn in the warp feeding mechanism, measured by a laser velocimeter, and the warp yarn is the carbon fiber arranged along the length direction of the fabric during the weaving process;

[0080] The coupling ratio offset coefficient is the standard deviation of the coupling ratio corresponding to each monitoring point in the target monitoring period, coupling ratio = factor threshold / yarn tension, factor threshold = angle of encircling angle + angle of yarn guide angle - angle of balloon vertex angle, encircling angle is the angle formed by the yarn on the yarn guide hook, yarn guide angle is the angle between the yarn pulled to the yarn guide hook and the horizontal line, balloon vertex angle is the angle between the balloon yarn and the axis of the tube yarn, the angle of encircling angle, the angle of yarn guide angle and the angle of balloon vertex angle are all measured by the inclination sensor, and the yarn tension is measured by the tension sensor, which is easy to understand for those skilled in the art and will not be described in detail;

[0081] Tension regularity coefficient = length difference coefficient + fluctuation difference coefficient. The yarn tension corresponding to each monitoring point in the target monitoring period is recorded as the tension value to be analyzed. The earliest tension value to be analyzed is recorded as the target data. The tension values ​​to be analyzed outside the target data are recorded as reference data. The reference data and the target data that are the same as the target data are recorded as the same data. The tension values ​​to be analyzed between two adjacent identical data and two adjacent identical data are recorded as a paragraph to be analyzed. Length difference coefficient = the maximum value of the influence quantity corresponding to each paragraph to be analyzed - the minimum value of the influence quantity corresponding to each paragraph to be analyzed. The influence quantity corresponding to a single paragraph to be analyzed is the total amount of tension values ​​to be analyzed contained in the paragraph to be analyzed. Fluctuation difference coefficient = the maximum value of the fluctuation coefficient corresponding to each paragraph to be analyzed - the minimum value of the fluctuation coefficient corresponding to each paragraph to be analyzed. The fluctuation coefficient corresponding to a single paragraph to be analyzed is the standard deviation of each tension value to be analyzed in the paragraph to be analyzed.

[0082] The values ​​of the preset pulse coefficient, the preset impact threshold, the preset coupling ratio offset coefficient and the preset tension law coefficient can be determined by the user according to the actual application scenario. The smaller the value of the preset pulse coefficient is, the greater the user's need to determine whether the target monitoring period is in an abnormal state according to the impact threshold is. A value of the preset pulse coefficient is provided, and the historical record that can meet the user's needs and determine whether the target monitoring period is in an abnormal state according to the impact threshold is recorded as the first reference record, and the average value of the pulse coefficient corresponding to the first reference record is recorded as the preset pulse coefficient. The greater the user's need to reduce maintenance costs, the smaller the values ​​of the preset impact threshold and the preset coupling ratio offset coefficient are, and the preset tension law coefficient is. The larger the value of the number, a preset influence threshold, a preset coupling ratio displacement coefficient and a preset tension law coefficient are provided, the minimum value of the influence threshold corresponding to the first reference record in which the target monitoring period is in an abnormal state is recorded as the preset influence threshold, the historical record that can meet the user's needs and determine whether the target monitoring period is in an abnormal state according to the coupling ratio displacement coefficient and the tension law coefficient is recorded as the second reference record, the minimum value of the coupling ratio displacement coefficient corresponding to the second reference record in which the target monitoring period is in an abnormal state is recorded as the preset coupling ratio displacement coefficient, and the maximum value of the tension law coefficient corresponding to the second reference record in which the target monitoring period is in an abnormal state is recorded as the preset tension law coefficient.

[0083] Specifically, the image selection unit responds to the analysis conditions to determine the differential selection method;

[0084] The analysis condition of the image selection unit response is that the special point ratio is greater than or equal to the preset special point ratio or the special point distribution coefficient is greater than or equal to the preset special point distribution coefficient, and the difference selection method is determined to determine the first selection method according to the influence state of the special point;

[0085] The analysis condition responded by the image selection unit is that the proportion of special points is less than the preset special point proportion and the special point distribution coefficient is less than the preset special point distribution coefficient, and the difference selection method is determined to determine the second selection method according to the similarity coefficient of the radiation area.

[0086] The confirmation method of the special point is to perform point analysis on each feature data, and when performing point analysis on a single feature data, record the feature data as the target feature data, and record the monitoring point whose fluctuation threshold is greater than or equal to the preset fluctuation threshold or whose slope difference coefficient is greater than or equal to the preset slope difference coefficient in the target monitoring period of the target feature data as the analysis point, and continue to perform point analysis on the feature data that has not been analyzed, until the point analysis of each feature data is completed, and record the monitoring point recorded as the analysis point as a special point. It can be understood that the number of analysis points corresponding to a single special point is greater than or equal to 1;

[0087] The characteristic data is the monitoring data whose iteration fluctuation coefficient is greater than the preset iteration fluctuation coefficient. For a single monitoring data, the monitoring data is recorded as the target monitoring data. The iteration fluctuation coefficient corresponding to the target monitoring data is the standard deviation of the value of the target monitoring data corresponding to each monitoring point within the target monitoring period. The value of the preset iteration fluctuation coefficient can be determined by the user according to the actual application scenario. The monitoring data is detected as the historical record of the characteristic data, and the average value of the iteration fluctuation coefficient corresponding to the historical record that can meet the user's needs is recorded as the preset iteration fluctuation coefficient.

[0088] The analysis conditions include a first analysis condition and a second analysis condition, the first analysis condition being that the special point proportion is greater than or equal to a preset special point proportion or the special point distribution coefficient is greater than or equal to a preset special point distribution coefficient, and the second analysis condition being that the special point proportion is less than a preset special point proportion and the special point distribution coefficient is less than a preset special point distribution coefficient;

[0089] Special point proportion = number of special points in the target monitoring cycle / number of monitoring points in the target monitoring cycle. The special point distribution coefficient is the average value of the reference distances corresponding to each special point. For a single special point, the special point is recorded as the target special point, and other special points other than the target special point are recorded as reference special points. The reference distance corresponding to the target special point is the average value of the time intervals from the target special point to each reference special point.

[0090] The fluctuation threshold and the slope difference coefficient are confirmed in the following manner: for a single monitoring point of a target monitoring period of a single characteristic data, the characteristic data is recorded as the target characteristic data, the monitoring point is recorded as the target monitoring point, and the monitoring point adjacent to the target monitoring point is recorded as the adjacent monitoring point. The fluctuation threshold = the maximum value of the difference coefficients corresponding to the adjacent monitoring points / the value of the target characteristic data corresponding to the target monitoring point. The difference coefficient corresponding to a single adjacent monitoring point is the absolute value of the difference between the value of the target characteristic data corresponding to the adjacent monitoring point and the value of the target characteristic data corresponding to the target monitoring point. The slope difference coefficient = the difference coefficient corresponding to the adjacent monitoring point / the time interval from the target monitoring point to the single adjacent monitoring point.

[0091] The values ​​of the preset special point proportion, the preset special point distribution coefficient, the preset fluctuation threshold and the preset slope difference coefficient can be determined by the user according to the actual application scenario. The smaller the values ​​of the preset special point proportion and the preset special point distribution coefficient are, the greater the user's demand for determining the first selection method according to the influence status of the special point. The preset special point proportion is 60%. The historical records of determining the first selection method according to the influence status of the special point are detected, and the average value of the special point distribution coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset special point distribution coefficient. The greater the user's demand for improving the monitoring accuracy, the smaller the values ​​of the preset fluctuation threshold and the preset slope difference coefficient are. A value of a preset fluctuation threshold and a preset slope difference coefficient is provided, and the historical records of setting the monitoring point as a special point are monitored. The average value of the fluctuation threshold corresponding to the historical records that can meet the user's needs is recorded as the preset fluctuation threshold, and the average value of the slope difference coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset slope difference coefficient.

[0092] Specifically, the image selection unit responds to the influence state of the special point to determine the first selection mode;

[0093] The image selection unit responds to the first impact state and determines to execute a first selection method of correlation selection based on the interval reference value and the data correlation coefficient;

[0094] The image selection unit responds to the second impact state and determines to execute a first selection method of selecting a differential point according to a feature threshold;

[0095] The influence states of the special points include: a first influence state in which the interval coefficient is less than a preset interval coefficient, and a second influence state in which the interval coefficient is greater than or equal to the preset interval coefficient.

[0096] Among them, the monitoring image corresponding to each differential point is subjected to differential processing and binarization processing by using a symmetric differential method to obtain a differential image containing only two pixel values ​​(0 and 1). This is easy to understand for those skilled in the art and will not be described in detail.

[0097] When performing association selection according to the interval reference value and the data association coefficient, selection and analysis are performed for each special point. When performing selection and analysis for a single special point, the special point is recorded as a target special point, and special points other than the target special point that are not recorded in the association set are recorded as reference special points. The set of reference special points and target special points whose interval reference value with the target special point is less than the preset interval reference value and whose data association coefficient is greater than the preset data association coefficient is recorded as an association set, and selection and analysis are continued for special points that are not recorded in the association set until all special points are recorded in the association set, then the selection and analysis is stopped; special point selection is performed for each association set. When performing special point selection for a single association set, any special point in the association set is randomly selected as a differential point;

[0098] When selecting differential points according to the feature threshold, special points whose feature threshold is greater than the preset feature threshold are selected as differential points;

[0099] The interval reference value is the time interval between two special points. The method for confirming the data correlation coefficient is that, for two special points, the data correlation coefficient = 1 / the absolute value of the difference between the characteristic thresholds corresponding to the two special points. The characteristic threshold corresponding to a single special point = the number of analysis points corresponding to the special point + the maximum value of the analysis point radiation coefficients of the analysis points corresponding to the special point. The analysis point radiation coefficient corresponding to a single analysis point = the fluctuation threshold + the slope difference coefficient. The interval coefficient is the average value of the interval means corresponding to the special points. For a single special point, the special point is recorded as the target special point, the special points adjacent to the target special point are recorded as the reference special point, and the average value of the time intervals from the reference special points to the target special point is recorded as the interval mean corresponding to the target special point.

[0100] The values ​​of the preset interval reference value, the preset data association coefficient, the preset feature threshold and the preset interval coefficient can be determined by the user according to the actual application scenario. The greater the user's demand for improving the monitoring accuracy, the smaller the value of the preset interval reference value, and the larger the values ​​of the preset data association coefficient and the preset feature threshold. The historical records selected for association according to the interval reference value and the data association coefficient are recorded as reference records, the average value of the interval reference value corresponding to the reference records that can meet the user's needs is recorded as the preset interval reference value, and the average value of the data association coefficient corresponding to the reference records that can meet the user's needs is recorded as the preset data association coefficient. The historical records of selecting differential points according to the feature threshold are detected, and the average value of the feature threshold corresponding to the historical records that can meet the user's needs is recorded as the preset feature threshold. The larger the value of the preset interval coefficient, the greater the user's demand for association selection according to the interval reference value and the data association coefficient. A value of the preset interval coefficient is provided, and the average value of the interval coefficient corresponding to the reference records that can meet the user's needs is recorded as the preset interval coefficient.

[0101] Specifically, the image selection unit determines the second selection method according to the similarity coefficient of the radiation area;

[0102] If the similarity coefficient is greater than or equal to the preset similarity coefficient, the second selection method is to determine the interval of the differential points in the influence area according to the characteristic coefficient of the influence area;

[0103] If the similarity coefficient is less than the preset similarity coefficient, the second selection method is to determine the number of differential points according to the regional evaluation coefficient;

[0104] The difference point interval is negatively correlated with the influence area characteristic coefficient, and the difference point quantity is positively correlated with the area evaluation coefficient.

[0105] Among them, the impact area is to record the earliest special point and the latest special point in the target monitoring cycle as the first special point and the second special point respectively, and record the time period between the time point corresponding to the first special point and the time point corresponding to the second special point as the impact area, and the radiation area includes the first radiation area and the second radiation area. The first radiation area is the time period in the target monitoring cycle where the monitoring time is earlier than the time point corresponding to the first special point, and the second radiation area is the time period in the target monitoring cycle where the monitoring time is later than the time point corresponding to the second special point;

[0106] The differential point interval is the time length between two adjacent differential points in the impact area; the similarity coefficient = 1 / |first radiation coefficient-second radiation coefficient|, the first radiation coefficient is the maximum value of the first characteristic coefficients corresponding to each characteristic data, the first characteristic coefficient corresponding to a single characteristic data is the standard deviation of the value of the characteristic data corresponding to each monitoring point of the characteristic data in the first radiation area, the second radiation coefficient is the maximum value of the second characteristic coefficients corresponding to each characteristic data, the second characteristic coefficient corresponding to a single characteristic data is the standard deviation of the value of the characteristic data corresponding to each monitoring point of the characteristic data in the second radiation area; it should be noted that if the first special point or the second special point is located at the starting monitoring point or the end monitoring point of the target monitoring period, the similarity coefficient is 0;

[0107] The value of the preset similarity coefficient can be determined by the user according to the actual application scenario. The larger the value of the preset similarity coefficient, the greater the user's need to determine the number of differential points according to the regional evaluation coefficient. A value of the preset similarity coefficient is provided, and the historical records of determining the number of differential points according to the regional evaluation coefficient are detected, and the average value of the regional evaluation coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset regional evaluation coefficient;

[0108] The second selection method is to determine the number of differential points according to the regional evaluation coefficient. If the number of differential points is B, the target monitoring period is divided into B equal parts, and each equal point is used as a differential point;

[0109] Regional evaluation coefficient = impact area characteristic coefficient - radiation area similarity coefficient, impact area characteristic coefficient = number of special points in the impact area + average value of characteristic thresholds corresponding to each special point in the impact area.

[0110] Specifically, the instability analysis unit responds to the setting conditions to determine the region segmentation method of the feature region, wherein:

[0111] The setting condition of the instability analysis unit response is that the connected domain distribution coefficient is greater than or equal to the preset connected domain distribution coefficient and the connected similarity threshold is greater than or equal to the preset connected similarity threshold, and the determination area segmentation method is uniform division according to the connected threshold;

[0112] The setting condition of the instability analysis unit response is that the connected domain distribution coefficient is less than the preset connected domain distribution coefficient or the connected similarity threshold is less than the preset connected similarity threshold, and the determination area segmentation method is to perform associated division according to the connected domain difference and the distance reference value;

[0113] The feature region is the smallest rectangle in the difference image that can contain each connected domain.

[0114] The setting condition includes a first setting condition and a second setting condition, the first setting condition is that the connected domain distribution coefficient is greater than or equal to the preset connected domain distribution coefficient and the similarity coefficient is greater than or equal to the preset similarity coefficient, and the second setting condition is that the connected domain distribution coefficient is less than the preset connected domain distribution coefficient or the similarity coefficient is less than the preset similarity coefficient;

[0115] When the region segmentation method is uniform division according to the connectivity threshold, the feature region is divided into a preset number of rectangular regions with equal areas and shapes. The preset number is positively correlated with the connectivity threshold, and a single rectangular region is a sub-region;

[0116] When performing association division according to the difference degree of connected domains and the distance reference value, division detection is performed for each connected domain. When performing division detection for a single connected domain, the connected domain is recorded as a target connected domain, and the connected domains other than the target connected domain that are not recorded in the associated combination are recorded as reference connected domains. The set of the reference connected domains and the target connected domains whose difference degree of connected domains with the target connected domain is less than the preset difference degree of connected domains and whose distance reference value is less than the preset distance reference value is recorded as an associated combination, and the minimum rectangle that can contain each connected domain in the associated combination is recorded as a sub-region, and the division detection is continued for the connected domains that are not recorded in the sub-region until all connected domains are recorded in the sub-region, then the division detection is stopped;

[0117] The connected domain is the clustered area of ​​characteristic pixels in the differential image. The pixels in a single connected domain are all characteristic pixels, and the characteristic pixels are pixels with a pixel value of 1.

[0118] The connected domain distribution coefficient is the average value of the connected distances corresponding to each connected domain. For a single connected domain, the connected domain is recorded as the target connected domain, and other connected domains outside the target connected domain are recorded as reference connected domains. The minimum value of the shortest distances from the center position of the target connected domain to the center position of each reference connected domain is recorded as the connected distance. The center position of a single connected domain is the center of the circumscribed circle of the connection.

[0119] Connectivity similarity threshold = 1 / (area difference + perimeter difference), area difference = maximum value of the area corresponding to each connected domain - minimum value of the area corresponding to each connected domain, perimeter difference = maximum value of the perimeter corresponding to each connected domain - minimum value of the perimeter corresponding to each connected domain. The area and perimeter of the connected domain are calculated by OpenCV, which is easy to understand for those skilled in the art and will not be described in detail; connectivity threshold = connectivity domain distribution coefficient - connectivity similarity threshold;

[0120] The values ​​of the preset connected domain distribution coefficient and the preset connected similarity threshold can be determined by the user according to the actual application scenario. The smaller the values ​​of the preset connected domain distribution coefficient and the preset connected similarity threshold, the greater the user's demand for uniform division according to the connected threshold. A value of the preset connected domain distribution coefficient and the preset connected similarity threshold is provided, and the historical records of uniform division according to the connected threshold are detected, and the average value of the connected domain distribution coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset connected domain distribution coefficient, and the average value of the connected similarity threshold corresponding to the historical records that can meet the user's needs is recorded as the preset connected similarity threshold;

[0121] The method for confirming the difference of connected domains is that, for any two connected domains, the difference of connected domains = the absolute value of the difference in the areas corresponding to the two connected domains + the absolute value of the difference in the perimeters corresponding to the two connected domains; the distance reference value is the shortest distance from the center position corresponding to one connected domain to the center position corresponding to another connected domain;

[0122] The values ​​of the preset connected domain difference and the preset distance reference value can be determined by the user according to the actual application scenario. The higher the user's demand for the degree of association between the connected domains in a single sub-area, the smaller the values ​​of the preset connected domain difference and the preset distance reference value are. A value of the preset connected domain difference and the preset distance reference value is provided, and the historical records that are associated and divided according to the connected domain difference and the distance reference value are detected, and the average value of the connected domain difference corresponding to the historical records that can meet the user's needs is recorded as the preset connected domain difference, and the average value of the distance reference values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset distance reference value.

[0123] Specifically, the instability analysis unit determines the instability coefficient determination method according to the sub-region difference coefficient of the differential image;

[0124] If the sub-region difference coefficient is greater than or equal to the preset sub-region difference coefficient, the instability coefficient is determined according to the domain variation reference value and the sub-influence threshold value;

[0125] If the sub-region difference coefficient is less than the preset sub-region difference coefficient, the instability coefficient is determined according to the sub-influence threshold.

[0126] It can be understood that each difference point corresponds to a difference image, and the sub-region difference coefficient = (the maximum value of the difference coefficients corresponding to each difference image - the minimum value of the difference coefficients corresponding to each difference image) / the maximum value of the difference coefficients corresponding to each difference image;

[0127] For a single differential image, the differential image is recorded as a target differential image, the differential coefficient corresponding to the target differential image = the number of influencing sub-regions in the target differential image + the reference value of the influencing sub-region distribution in the target differential image, the reference value of the influencing sub-region distribution is the average value of the influencing distances corresponding to each influencing sub-region, for a single influencing sub-region, the influencing sub-region is recorded as a target influencing sub-region, the influencing sub-region outside the target influencing sub-region is recorded as a reference influencing sub-region, and the minimum value of the shortest distances from the center position of the target influencing sub-region to the center position of each reference influencing sub-region is recorded as the influencing distance corresponding to the target influencing sub-region;

[0128] The affected sub-region is a sub-region whose connected area ratio is greater than the preset connected area ratio. For a single sub-region, the connected area ratio = the sum of the areas of the connected domains in the sub-region / the area of ​​the sub-region;

[0129] The values ​​of the preset sub-region difference coefficient and the preset connected area ratio can be determined by the user according to the actual application scenario. The larger the value of the preset sub-region difference coefficient, the greater the user's demand for determining the instability coefficient according to the sub-influence threshold. A value of the preset sub-region difference coefficient is provided, and the historical records of determining the instability coefficient according to the sub-influence threshold are detected. The average value of the sub-region difference coefficient corresponding to the historical records that can meet the user's needs is recorded as the preset sub-region difference coefficient. The greater the user's demand for improving the monitoring accuracy, the smaller the value of the preset connected area ratio is. A value of the preset connected area ratio is provided, and the preset connected area ratio is 70%;

[0130] When the sub-region difference coefficient is greater than or equal to the preset sub-region difference coefficient, the instability coefficient = domain change reference value + sub-influence threshold;

[0131] When the sub-region difference coefficient is less than the preset sub-region difference coefficient, the instability coefficient is positively correlated with the sub-influence threshold;

[0132] The domain-varying reference value is the standard deviation of the differential coefficients corresponding to each differential image.

[0133] Specifically, the instability analysis unit determines the sub-influence threshold according to the influence coefficient, wherein the influence coefficient is determined according to the influence sub-region distribution reference value,

[0134] If the affected sub-region distribution reference value is greater than or equal to the preset affected sub-region distribution reference value, the influence coefficient is determined according to the connectivity reference value;

[0135] If the affected sub-region distribution reference value is less than the preset affected sub-region distribution reference value, the influence coefficient is determined according to the distance influence value.

[0136] Among them, the sub-influence threshold is the average value of the maximum influence coefficients corresponding to each differential image, the maximum influence coefficient is the maximum value of the influence coefficients corresponding to each influencing sub-region in a single differential image, and the method for confirming the influence coefficient is as follows:

[0137] If the influence sub-region distribution reference value is greater than or equal to the preset influence sub-region distribution reference value, the influence coefficient is positively correlated with the connectivity reference value;

[0138] If the impact sub-region distribution reference value is less than the preset impact sub-region distribution reference value, the impact coefficient is negatively correlated with the distance impact value.

[0139] The connectivity reference value is the maximum value of the sub-connectivity coefficients corresponding to each connected domain in a single sub-influence area. The method for confirming the sub-connectivity coefficient corresponding to a single connected domain is to record the minimum rectangle that can contain the connected domain as the reference rectangle, and the sub-connectivity coefficient = the length of the reference rectangle + the width of the reference rectangle. The distance influence value is the average value of the reference distance coefficients corresponding to each connected domain in a single sub-influence area. For a single connected domain in a single sub-influence area, the sub-influence area is recorded as the target sub-influence area, the connected domain is recorded as the first target connected domain, and other connected domains in the target sub-influence area except the first target connected domain are recorded as the first reference connected domain. The minimum value of the shortest distance from the center position of the first target connected domain to the center position of each reference connected domain is recorded as the reference distance coefficient.

[0140] The value of the preset influence sub-region distribution reference value can be determined by the user according to the actual application scenario. The larger the value of the preset influence sub-region distribution reference value is, the greater the user's need to determine the influence coefficient based on the distance influence value is. A value of the preset influence sub-region distribution reference value is provided, and the historical records of determining the influence coefficient based on the distance influence value are detected. The average value of the influence sub-region distribution reference value corresponding to the historical records that can meet the user's needs is recorded as the preset influence sub-region distribution reference value.

[0141] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0142] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An automatic monitoring system based on a carbon fiber multi-axial warp knitting machine, characterized in that: include: A data collection unit, used to collect monitoring data; A monitoring and analysis unit connected to the data acquisition unit, for determining a warp knitting state according to an oil coefficient and a vibration amplitude, and responding to the warp knitting state to determine a monitoring mode, wherein the monitoring mode is to determine an optimization mode according to a pulse coefficient, or to determine whether a target monitoring period is in an abnormal state according to an instability coefficient of a monitoring image; An optimization analysis unit connected to the monitoring analysis unit is used to determine an optimization method according to the pulse coefficient, wherein the optimization method is to determine whether the target monitoring period is in an abnormal state according to the impact threshold, or to determine whether the target monitoring period is in an abnormal state according to the coupling ratio displacement coefficient and the tension law coefficient; An image selection unit, which is connected to the monitoring and analysis unit, and is used to determine a differential selection method according to the analysis conditions, wherein the differential selection method is to determine a first selection method according to the influence state of a special point or to determine a second selection method according to a similarity coefficient of a radiation area; an instability analysis unit, which is connected to the monitoring and analysis unit and the image selection unit respectively, and is used to determine an instability coefficient determination method according to the sub-region difference coefficient of the differential image, wherein the instability coefficient determination method is to determine the instability coefficient according to the domain variation reference value and the sub-influence threshold value or to determine the instability coefficient according to the sub-influence threshold value; The monitoring and analysis unit determines the warp knitting state according to the oil coefficient and the vibration amplitude, and the warp knitting state includes: A first warp knitting state in which the oil coefficient is less than a preset oil coefficient and the vibration amplitude is less than a preset vibration amplitude; A second warp knitting state in which the oil coefficient is greater than or equal to a preset oil coefficient or the vibration amplitude is greater than or equal to a preset vibration amplitude; The monitoring and analysis unit responds to the warp knitting state to determine the monitoring mode; The monitoring and analyzing unit responds to the first warp knitting state and determines to execute a monitoring mode that determines an optimization mode according to a pulse coefficient; The monitoring and analyzing unit responds to the second warp knitting state and determines whether a monitoring mode of determining whether a target monitoring period is in an abnormal state according to an instability coefficient of a monitoring image is executed.

2. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 1 is characterized in that: The optimization analysis unit determines the optimization method according to the pulse coefficient, wherein: If the pulse coefficient is greater than or equal to the preset pulse coefficient, the optimization method is to determine whether the target monitoring period is in an abnormal state according to the impact threshold; If the pulse coefficient is less than the preset pulse coefficient, the optimization method is to determine whether the target monitoring period is in an abnormal state based on the coupling ratio displacement coefficient and the tension law coefficient.

3. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 2 is characterized in that: The image selection unit responds to the analysis conditions to determine the differential selection method; The analysis condition of the image selection unit response is that the special point ratio is greater than or equal to the preset special point ratio or the special point distribution coefficient is greater than or equal to the preset special point distribution coefficient, and the difference selection method is determined to determine the first selection method according to the influence state of the special point; The analysis condition responded by the image selection unit is that the proportion of special points is less than the preset special point proportion and the special point distribution coefficient is less than the preset special point distribution coefficient, and the difference selection method is determined to determine the second selection method according to the similarity coefficient of the radiation area.

4. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 3 is characterized in that: The image selection unit responds to the influence state of the special point to determine the first selection mode; The image selection unit responds to the first impact state and determines to execute a first selection method of correlation selection based on the interval reference value and the data correlation coefficient; The image selection unit responds to the second impact state and determines to execute a first selection method of selecting a differential point according to a feature threshold; The influence states of the special points include: a first influence state in which the interval coefficient is less than a preset interval coefficient, and a second influence state in which the interval coefficient is greater than or equal to the preset interval coefficient.

5. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 4 is characterized in that: The image selection unit determines a second selection method according to a similarity coefficient of the radiation area; If the similarity coefficient is greater than or equal to the preset similarity coefficient, the second selection method is to determine the interval of the differential points in the influence area according to the characteristic coefficient of the influence area; If the similarity coefficient is less than the preset similarity coefficient, the second selection method is to determine the number of differential points according to the regional evaluation coefficient; The difference point interval is negatively correlated with the influence area characteristic coefficient, and the difference point quantity is positively correlated with the area evaluation coefficient.

6. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 5, characterized in that: The instability analysis unit responds to the setting conditions to determine the region segmentation method of the characteristic region, wherein: The setting condition of the instability analysis unit response is that the connected domain distribution coefficient is greater than or equal to the preset connected domain distribution coefficient and the connected similarity threshold is greater than or equal to the preset connected similarity threshold, and the determination area segmentation method is uniform division according to the connected threshold; The setting condition of the instability analysis unit response is that the connected domain distribution coefficient is less than the preset connected domain distribution coefficient or the connected similarity threshold is less than the preset connected similarity threshold, and the determination area segmentation method is to perform associated division according to the connected domain difference and the distance reference value; The feature region is the smallest rectangle in the difference image that can contain each connected domain.

7. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 6, characterized in that: The instability analysis unit determines the instability coefficient determination method according to the sub-region difference coefficient of the differential image; If the sub-region difference coefficient is greater than or equal to the preset sub-region difference coefficient, the instability coefficient is determined according to the domain variation reference value and the sub-influence threshold value; If the sub-region difference coefficient is less than the preset sub-region difference coefficient, the instability coefficient is determined according to the sub-influence threshold.

8. The automatic monitoring system based on carbon fiber multi-axial warp knitting machine according to claim 7, characterized in that: The instability analysis unit determines a sub-influence threshold according to an influence coefficient, wherein the influence coefficient is determined according to an influence sub-region distribution reference value, If the affected sub-region distribution reference value is greater than or equal to the preset affected sub-region distribution reference value, the influence coefficient is determined according to the connectivity reference value; If the affected sub-region distribution reference value is less than the preset affected sub-region distribution reference value, the influence coefficient is determined according to the distance influence value.

Images