Resistance trimming machine equipment cluster running state management method based on cloud platform

Abstract

The invention belongs to the technical field of equipment cluster operation state management, and particularly discloses a resistance trimming machine equipment cluster operation state management method based on a cloud platform, and the method comprises the steps: screening qualified equipment through self-inspection parameters, and carrying out the initial task distribution; performing dynamic parameter adjustment on a qualified resistance trimming machine cluster according to a fusion analysis result of the ablation coordinates, the ablation line width and the ablation depth of the ablation points and the corresponding standard ablation parameters, judging the equipment state based on the path overlap ratio, the line width and the depth threshold, and finally realizing redistribution of abnormal equipment tasks through load rate sorting. According to the invention, through a dynamic screening mechanism, the self-checking parameters of the trimmer devices are compared with the preset self-checking standard parameters, the reliability and stability of the devices are analyzed and calculated in combination with the dynamic weight, and the qualified trimmer cluster is screened for initial task allocation, so that the problem of resistance value out-of-tolerance is avoided; and meanwhile, quantitative analysis of coordinate offset, line width deviation and depth error is carried out.

Description

Technical Field

[0001] The present invention belongs to the technical field of the operation status management of equipment clusters, and relates to a method for managing the operation status of a laser trimming machine equipment cluster based on a cloud platform. Background Technique

[0002] A laser trimming machine is a core device used to adjust the resistance value of a resistor. It realizes resistance value matching by physically or chemically changing the resistor body structure. Among them, a laser trimming machine uses a focused laser beam on the resistor body (such as thick film and thin film materials) to adjust the resistance value by changing the resistance geometry. And the resistance value change is measured in real time during the process, and the cutting is automatically stopped after reaching the target value to ensure accuracy.

[0003] There are still the following deficiencies in the operation status management of the laser trimming machine equipment cluster: 1. The current task allocation of the laser trimming machine cluster mostly depends on the fixed parameters of the equipment, thus ignoring the dynamic performance changes of the equipment during the self-check process. At the same time, during the task execution process of the laser trimming machine equipment, the operation parameters of the equipment cannot be detected in real time for operation regulation, resulting in the lag of operation optimization.

[0004] 2. Currently, in the operation qualification determination of the laser trimming machine equipment cluster, the final resistance value of the resistor is generally used as the core evaluation index, and the attention to the operation path of the laser trimming machine equipment during the operation process is insufficient, ignoring the influence of the dynamic operation path accuracy of the laser trimming machine equipment on the product quality, and thus unable to guarantee the resistance value uniformity and long-term stability of the resistor. Summary of the Invention

[0005] In view of this, in order to solve the problems raised in the above background technique, a method for managing the operation status of a laser trimming machine equipment cluster based on a cloud platform is proposed.

[0006] The object of the present invention can be achieved by the following technical solutions: The present invention provides a method for managing the operation status of a laser trimming machine equipment cluster based on a cloud platform, including: S1. Initial task allocation: comparing the self-check parameters of each laser trimming machine equipment with the preset self-check standard parameters, screening out qualified laser trimming machine clusters and allocating initial tasks.

[0007] S2. Task operation collection: collecting ablation images and ablation depths in the time series during the execution of the initial task, and calculating the ablation coordinates and ablation line widths of the ablation points through edge detection and contour extraction.

[0008] S3. Task operation optimization: dynamically adjusting the parameters of the qualified laser trimming machine cluster according to the fusion analysis results of the ablation coordinates, ablation line widths and ablation depths of the ablation points and their corresponding standard ablation parameters.

[0009] S4. Qualification determination of operation: After adjustment, an ablation path including ablation depth and ablation line width is generated in real time by aligning timestamps. Meanwhile, it is determined whether the equipment is operating abnormally by combining the resistance values of the resistors corresponding to each trimming machine device.

[0010] S5. Secondary task allocation: According to the equipment operation parameters of the trimming machine devices, the trimming machine devices with abnormal operation are shut down. Meanwhile, for the number of unfinished task points of the trimming machine devices with abnormal operation, secondary task allocation is carried out.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) Through the dynamic screening mechanism, the present invention compares the self-check parameters of each trimming machine device with the preset self-check standard parameters, combines the dynamic weight analysis to calculate the reliability and stability of the equipment, and screens out a qualified trimming machine cluster for initial task allocation, reducing the problem of out-of-tolerance resistance values caused by the dynamic performance fluctuations of the equipment.

[0012] (2) By collecting ablation images and ablation depths in a time series, the present invention uses an edge detection algorithm to extract the ablation point contours, and combines with the standard ablation path to perform quantitative analysis of coordinate offset, line width deviation and depth error, generating dynamic parameter adjustments including trajectory correction parameters, power compensation values and speed compensation values, improving the response speed of operation optimization.

[0013] (3) By performing multi-dimensional fusion determination on the ablation path including ablation depth and ablation line width and the resistance value, the present invention breaks through the limitations of traditional single resistance value determination, improves the attention to the accuracy of the operation path of the trimming machine, and further improves the resistance value uniformity and long-term stability of the resistor.

[0014] (4) By real-time collecting the number of unfinished task points of the trimming machine with abnormal operation, and combining with the operation parameters of the trimming machine with qualified operation to calculate the load rate, and sorting the qualified equipment in ascending order of the load rate, preferentially allocating tasks to the equipment with high remaining processing capacity, the present invention realizes the dynamic load balancing of resources within the cluster, avoiding overload or idleness of some equipment. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings based on these drawings without creative efforts.

[0016] Figure 1 It is a schematic diagram of the connection of each module of the system of the present invention.

[0017] Figure 2 It is a schematic diagram of the connection of the screening steps of the qualified trimming machine cluster of the present invention.

[0018] Figure 3 This is a connection schematic diagram of the dynamic parameter adjustment analysis steps of the present invention. Detailed implementation manners

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Please refer to Figure 1 As shown, the present invention provides a method for managing the operating status of a resistor trimming machine equipment cluster based on a cloud platform. The method includes: S1. Initial task allocation: Compare the self-check parameters of each resistor trimming machine equipment with the preset self-check standard parameters, screen the qualified resistor trimming machine cluster, and allocate the initial tasks.

[0021] It should be added that the self-check parameters include: resistance value and self-check path. Among them, the resistance value is obtained through the built-in resistance measurement module installed on the resistor trimming head of the resistor trimming machine equipment, and the self-check path is detected by the laser displacement sensor installed beside the resistor trimming head. The resistance value reflects the result accuracy of the equipment, but cannot reveal the hidden defects during the processing (such as local stress concentration caused by path deviation), while the self-check path analysis focuses on the process reliability and can detect potential risks in advance (such as path deviation may lead to future resistance value out-of-tolerance).

[0022] It should be added that the self-check standard parameters include: standard resistance value and standard self-check path. Among them, the standard resistance value is set by collecting the historical self-check resistance values of the resistor trimming machine equipment in the recent period (such as setting 3 days), and statistically calculating the average value of the historical self-check resistance values as the standard resistance value. The standard self-check path is generated by collecting the path data of multiple self-checks in the fault-free operating state of the resistor trimming machine equipment in the recent period (synchronized with the resistance value statistical period), aligning the multiple paths, and finding out the common trajectory features, such as the position of key points, the curvature of the path, etc.

[0023] It should be added that the allocation steps for initial task allocation include: dividing the resistor trimming machine equipment into a first-level resistor trimming machine cluster, a second-level resistor trimming machine cluster, and a third-level resistor trimming machine cluster based on the self-check qualification degree of the resistor trimming machine equipment in the qualified resistor trimming machine cluster.

[0024] Extract the task types of each task from the cloud platform. The task types include: high-precision tasks, high-speed tasks, and ordinary tasks.

[0025] It should be added that high-precision tasks refer to those with extremely high requirements for processing precision and stability, requiring the equipment to have sub-micron-level control capabilities and tolerate a low error tolerance. For example, the adjustment error of resistance value ≤ ±0.1%, and the repeat positioning accuracy ≤ 1μm. High-speed tasks refer to those that emphasize processing speed and throughput, allowing a moderate sacrifice of precision, and requiring the equipment to have high-frequency response capabilities. For example, the single-task processing time ≤ 50ms, the task throughput per unit time ≥ 200 pieces / minute, and the dynamic response delay < 10ms. Ordinary tasks refer to those with no special requirements for precision and speed, focusing on cost-effectiveness and resource utilization, and allowing a certain degree of performance fluctuation. For example, the adjustment error of resistance value ≤ ±1%, the single-task processing time ≤ 200ms, and the equipment load balance degree ≥ 80%.

[0026] All high-precision tasks are evenly distributed to each trimming machine device in the first-level trimming machine cluster, then all high-speed tasks are distributed to each trimming machine device in the second-level trimming machine cluster, and finally all ordinary tasks are distributed to each trimming machine device in the third-level trimming machine cluster.

[0027] Please refer to Figure 2 As shown, exemplarily, the screened qualified trimming machine cluster includes: Q1. Calculate the deviation between the resistance value of each trimming machine device during each self-check in the self-check parameters and the standard resistance value to obtain the reliability qualification degree of each trimming machine device.

[0028] Furthermore, the analysis of the reliability qualification degree of each trimming machine device includes: Q1-1. Take the difference between the resistance value and the standard resistance value as the resistance deviation of each trimming machine device during each self-check.

[0029] Q1-2. Count the number of self-checks with a resistance deviation less than the preset resistance deviation and the total number of self-checks, and take the ratio of the two as the reliability qualification degree of each trimming machine device.

[0030] Q2. Compare and analyze the self-check path of each trimming machine device during each self-check in the self-check parameters with its standard self-check path to obtain the stability qualification degree of each trimming machine device.

[0031] Furthermore, the analysis of the stability qualification degree of each trimming machine device includes: Q2-1. Extract the self-check coordinate sequence and the standard coordinate sequence from the self-check path of each trimming machine device during each self-check.

[0032] Q2-2. Calculate the minimum cumulative distance of the sequence through the dynamic time warping algorithm.

[0033] Q2-3. If the minimum cumulative distance is less than the reference standard distance, then determine that the stability qualification degree of this self-check is 1, otherwise, determine that the stability qualification degree of this self-check is 0.

[0034] It should be added that the method for obtaining the reference standard distance is as follows: collect the self-check data of multiple stable-running and performance-qualified resistor trimming machines, and calculate the cumulative distance between the self-check coordinate sequence and the standard coordinate sequence. Through statistical methods (such as calculating the mean + 2 times the standard deviation), determine a distance value that can cover the data of the vast majority of resistor trimming machines, and set it as the reference standard distance.

[0035] Q2-4. Statistically calculate the mean value of the stability qualification degree of each self-check as the stability qualification degree of each resistor trimming machine device.

[0036] In a specific embodiment, assume that the coordinates collected during the self-check of a certain resistor trimming machine are: P1 = [(0, 0), (1, 0.5), (2, 1), (3, 1.5)], and the preset standard path is: P2 = [(0, 0), (1, 1), (2, 2), (3, 3)]. The analysis steps for its stability qualification degree are as follows: (1) Calculate the DTW minimum cumulative distance: Align the time stamps of the self-check path and the standard path and calculate the Euclidean distance D(P1, P2) between each pair of points of the self-check path and the standard path: D(1, 1) = 0, D(1, 2) = 1.41, D(2, 1) = 1.12, D(2, 2) = 0.5,....

[0037] (2) Calculate the cumulative distance matrix through the recurrence formula C(i, j) = D(i, j) + min{C(i - 1, j), C(i, j - 1), C(i - 1, j - 1)}, where C(i, j) represents the minimum cumulative distance when aligning the first i points of the self-check path with the first j points of the standard path. i represents the i-th point in the actual path (P1) (for example, the point (0, 0) in P1 is the first point, (1, 0.5) is the second point, and so on). j represents the j-th point in the standard path (P2) (for example, the point (0, 0) in P2 is the first point, (1, 1) is the second point, and so on).

[0038] Initialize C(1, 1) = 0, and gradually fill the matrix. For example: C(2, 2) = 0.5 + min{1.12, 1.41, 0} = 0.5, and then obtain the final cumulative distance C(4, 4) = 2.0.

[0039] (3) Determine the stability qualification degree: Assume that the reference standard distance is d1 = 3.0, and the final cumulative distance d2 = 2.0 < 3.0, then it is determined that the stability qualification degree of this self-check is 1.

[0040] (4) Statistically calculate the mean value of the stability qualification degree: Assume that the device has completed 5 self-checks, and the stability qualification degrees are [1, 1, 0, 1, 1] respectively. Then calculate the mean value of the stability qualification degrees to obtain the stability qualification degree of 0.8.

[0041] Q3. Calculate the weighted average of the reliability qualification degree and the stability qualification degree through the preset dynamic weights to obtain the self-inspection qualification degree of each trimming machine device.

[0042] Q4. Generate a set of candidate trimming machine devices for the trimming machine devices with a self-inspection qualification degree greater than or equal to the preset self-inspection qualification degree.

[0043] Q5. Extract the required number of trimming machine devices from the cloud platform, and select a qualified device cluster from the candidate trimming machines according to the required number of trimming machine devices.

[0044] It should be added that the step of selecting a qualified device cluster from the candidate trimming machines according to the required number of trimming machine devices includes: if the required number of trimming machines is less than or equal to the number of the candidate trimming machine device group, then sort the trimming machine devices in the candidate trimming machine device group in descending order according to the self-inspection qualification degree, and select the trimming machine devices with the number of the required trimming machines before sorting to generate a qualified device cluster.

[0045] If the required number of trimming machines is greater than the number of the candidate trimming machine device group, then use the difference between the required number of trimming machines and the number of the candidate trimming machine device group as the required trimming machine difference.

[0046] Sort the trimming machine devices with a self-inspection qualification degree less than the preset self-inspection qualification degree in descending order according to the self-inspection qualification degree, and then select the trimming machine devices corresponding to the required trimming machine difference, and form a qualified device cluster with the candidate trimming machine device group.

[0047] In the embodiment of the present invention, through the dynamic screening mechanism, the self-inspection parameters of each trimming machine device are compared with the preset self-inspection standard parameters, and the reliability and stability of the device are analyzed and calculated by combining the dynamic weights, and a qualified trimming machine cluster is selected for the initial task allocation, reducing the problem of resistance out-of-tolerance caused by the dynamic performance fluctuation of the device.

[0048] S2. Task operation acquisition: Collect the ablation image and ablation depth in the time series during the execution of the initial task, and calculate the ablation coordinates and ablation line width of the ablation point through edge detection and contour extraction.

[0049] It should be added that the ablation image is collected by a camera installed beside the laser head or a lens on the top of the device aiming at the ablation area of the resistor chip, and the ablation depth is detected by a laser displacement sensor installed on the laser head assembly or independently installed on the robotic arm perpendicular to the laser ablation direction.

[0050] Exemplarily, calculating the ablation coordinates and ablation line width of the ablation point through edge detection and contour extraction includes: obtaining the edge detection image of the ablation point from the ablation image through the edge detection algorithm, and then performing contour extraction on the edge detection image of the ablation point to obtain the contour of the ablation point and each pixel point within the contour.

[0051] Perform weighted average calculation on each pixel point within the contour to obtain the ablation coordinates of the ablation point.

[0052] In a specific embodiment, the process of obtaining the ablation coordinates by performing weighted average calculation on each pixel point within the contour is as follows: (1) If the gray matrix of the ablation image is I(x, y), with a size of W×H and a pixel value range of [0, 255], extract all pixel sets P = {(x1, y1), (x2, y2),..., (xN, yN)} within the ablation point contour.

[0053] (2) Eliminate the illumination difference through the gray normalization formula where , and are respectively the maximum gray value, the minimum gray value of the ablation image, and the standard maximum value of the ablation image.

[0054] (3) Calculate the pixel value of the pixel point through the Gaussian filtering formula where, and are respectively the standard deviation of the Gaussian kernel and the kernel radius, is the offset of the pixel point in the horizontal direction (x-axis), is the offset of the pixel point in the vertical direction (y-axis), is the natural constant.

[0055] (4) Calculate the weight of each pixel point through the formula where is the pixel point number, , is the number of pixel points, has the same meaning as the kernel radius in the Gaussian filtering formula and is used to calculate the pixel point weight.

[0056] (5) Through the formulas and , obtain the ablation coordinates of the ablation point.

[0057] Extract the pixel spacing between the left and right edges of the ablation point contour along the normal direction of the ablation point as the ablation line width of the ablation point.

[0058] S3, Task Running Optimization: Dynamically adjust the parameters of the qualified trimming machine cluster according to the fusion analysis results of the ablation coordinates, ablation line width, and ablation depth of the ablation point and their corresponding standard ablation parameters.

[0059] Please refer to Figure 3As shown, exemplarily, the dynamic parameter adjustment analysis of the qualified resistor trimming machine cluster includes: W1. Extract the standard ablation path of each qualified resistor trimming machine device from the cloud platform, obtain the standard coordinates matching the ablation coordinates in the standard path through the nearest neighbor algorithm, perform the difference operation between the ablation coordinates and the standard coordinates, and generate the real-time path offset.

[0060] It should be added that the nearest neighbor algorithm is a simple and intuitive nearest neighbor algorithm. It will find the coordinate with the closest distance in the standard path for each ablation coordinate as the matching point. Its working principle is that for each point on the ablation path, calculate its distance from all points on the standard path (usually using the Euclidean distance), and then select the point with the smallest distance as the standard coordinate corresponding to the ablation point. The nearest neighbor algorithm directly finds the nearest standard coordinate for the ablation coordinates without complex calculations, improving the response timeliness of trajectory correction. At the same time, the standard ablation path may contain non-linear, multi-twisted or high-density coordinate points, and the nearest neighbor algorithm does not make any prior assumptions about the data distribution or path form, and flexibly adapts to complex path forms. Therefore, the nearest neighbor algorithm is selected to obtain the standard path.

[0061] W2. Analyze the proportional relationship between the real-time path offset and the set reference deviation threshold, and obtain the trajectory correction parameters of each qualified resistor trimming machine device through the adjustment value formula. The adjustment value formula is , where is the trajectory correction parameter, and are the deviation threshold of the real-time path offset and the set reference respectively, is the proportionality coefficient.

[0062] W3. Extract the standard ablation depth of each qualified resistor trimming machine device from the cloud platform.

[0063] W4. Compare and analyze the ablation depth of each qualified resistor trimming machine device with the standard ablation depth to obtain the power compensation value of each qualified resistor trimming machine device.

[0064] Furthermore, comparing and analyzing the ablation depth of each qualified resistor trimming machine device with the standard ablation depth includes: W4-1. Subtract the ablation depth from the standard deviation ablation depth to obtain the ablation depth difference of each qualified resistor trimming machine device;

[0065] W4-2. Import the ablation depth difference into the power compensation formula to obtain the power compensation value of the qualified resistor trimming machine device, and then obtain the power compensation value of each qualified resistor trimming machine device. The power compensation formula is , where is the power compensation value, is the ablation depth difference, is the power compensation proportionality coefficient.

[0066] It should be added that It is obtained through experiments. The power adjustment amount required under different depth differences is tested, and then obtained through data fitting the value. For example, in the case of a known depth difference, the power is adjusted until the standard ablation depth is reached, and the corresponding power change is recorded, so as to calculate the average value as the power compensation ratio coefficient.

[0067] W5. The ablation line width and the standard ablation line width are imported into the line width deviation formula to obtain the speed compensation values of each qualified trimming machine device. The line width deviation formula is , where is the speed compensation value,[[]]END]] and are the ablation line width and the standard line width respectively,[[]]END]] is the initial speed,[[]]END]] is the line width deviation ratio coefficient. Among them, the initial speed is set based on the initial allocation task of the trimming machine device, and the line width deviation ratio coefficient is obtained through gradient descent optimization experimental verification of the historical data set of the operation status of the device cluster.

[0068] When it is a positive deviation, that is, the line width is too wide, and the speed can be increased to reduce the ablation time. When it is a negative deviation, that is, the line width is too narrow, and the speed can be reduced to increase the ablation time. Proportional adjustment mechanism: The adjustment amplitude of the speed is proportional to the relative value of the line width deviation ( ). For example: the standard line width is 50μm, and the actual line width is 55μm (+10% deviation), then the speed is increased by 10%. The adjusted speed is based on the initial speed to avoid vibration or overshoot of the trimming machine device caused by sudden speed changes.

[0069] W6. Based on the trajectory correction parameters, power compensation values and speed compensation values of each qualified trimming machine device, dynamic parameter adjustment is performed.

[0070] In the embodiment of the present invention, by collecting ablation images and ablation depths in time series, using edge detection algorithms to extract the ablation point contours, and combining with the standard ablation path for quantitative analysis of coordinate offsets, line width deviations and depth errors, dynamic parameter adjustments including trajectory correction parameters, power compensation values and speed compensation values are generated, improving the response speed of operation optimization.

[0071] S4. Operation qualification determination: After adjustment, an ablation path including ablation depth and ablation line width is generated in real time through timestamp alignment, and at the same time, combined with the resistance values of the corresponding resistors of each trimming machine device, it is determined whether the device is operating abnormally.

[0072] It should be added that the resistance values of the resistors are collected through resistance sensors.

[0073] Exemplarily, the generation of the ablation path including the ablation depth and ablation line width includes: constructing the actual ablation path of each trimming machine device based on the ablation coordinates of each trimming machine device in a time series.

[0074] Perform timestamp alignment on the ablation coordinates, ablation depth, and ablation line width of each ablation point in the actual ablation path.

[0075] Mark the ablation depth and ablation line width in the actual ablation path of each trimming machine device in chronological order, so as to obtain the ablation path including the ablation depth and ablation line width.

[0076] Exemplarily, the determination of whether the device is operating abnormally includes: using the starting points of the actual ablation path and the standard ablation path as the coincidence points for coincidence comparison, and then extracting the coincidence path length and the standard ablation path length therefrom, and taking the ratio of the two as the path coincidence ratio.

[0077] Extract the maximum ablation line width and the minimum ablation line width from the ablation line widths of each ablation point, and take the difference between the two to obtain the ablation line width difference.

[0078] Analyze and obtain the ablation depth difference in the same way as the analysis method of the ablation line width difference.

[0079] Extract the required resistance value of the resistor from the cloud platform, and calculate the absolute difference between it and the resistance value to obtain the resistance difference of the resistor.

[0080] Respectively compare the path coincidence ratio, ablation line width difference, ablation depth difference, and the resistance difference of the corresponding resistor with their set reference thresholds. Determine the qualified trimming machine devices with parameters greater than their reference thresholds as operating abnormally, and vice versa, determine them as operating normally.

[0081] The embodiment of the present invention performs multi-dimensional fusion determination through the ablation path including the ablation depth and ablation line width and the resistance value, breaks through the limitations of traditional single-resistance-value determination, improves the attention to the accuracy of the operation path of the trimming machine, and further improves the resistance value uniformity and long-term stability of the resistor.

[0082] S5. Task secondary allocation: Perform a shutdown operation on the abnormally operating trimming machine device according to the device operation parameters of the trimming machine device, and at the same time, perform task secondary allocation on the number of uncompleted task points of the abnormally operating trimming machine device.

[0083] It should be added that the equipment operation parameters of the resistor trimming machine include: the initial allocated task points, the single-task processing time, the total available time, and the uncompleted task points. The initial allocated task points are obtained based on the initial task allocation result. The single-task processing time is recorded by a timer equipped in the resistor trimming machine itself. The timer is used to record the time required to complete a single task, and the average value is taken as the single-task processing time after multiple recordings.

[0084] The total available time is obtained by analyzing the operation log recorded by the resistor trimming machine, obtaining information such as the startup time, shutdown time, and maintenance time of the resistor trimming machine, and then calculating the total available time. For example, if the resistor trimming machine is powered on for 8 hours every day and 1 hour is used for equipment maintenance, the total available time is 7 hours.

[0085] The uncompleted task points are obtained by calculating the difference between the initial allocated task points and the completed task points.

[0086] Exemplarily, the secondary task allocation includes: using the uncompleted task points of each resistor trimming machine with abnormal operation as the task points to be reallocated.

[0087] It should be added that the task points to be reallocated = the initial task points of the resistor trimming machine with abnormal operation - the processed task points of the resistor trimming machine with abnormal operation.

[0088] Take the product calculation result of the initial allocated task points and the single-task processing time in the equipment operation parameters of the qualified resistor trimming machine as the initial task processing time, and take the ratio of it to the total available time in the equipment operation parameters as the load rate of the qualified resistor trimming machine.

[0089] Sort the qualified resistor trimming machines in ascending order of the load rate to generate a sequence of qualified resistor trimming machines.

[0090] Take the calculation result of the difference between the total available time and the initial task processing time as the current available time, and compare and analyze the current available time with the single-task processing time to obtain the remaining processing capacity of each qualified resistor trimming machine.

[0091] Based on the sequence of qualified resistor trimming machines, allocate the task points to be reallocated in turn, and the task points allocated to the qualified resistor trimming machine do not exceed its remaining processing capacity until all the task points to be reallocated are completely allocated.

[0092] In a specific embodiment, the equipment operation parameters of the resistor trimming machine are shown in Table 1:

[0093] Table 1: Schematic table of equipment operation parameters of the resistor trimming machine

[0094]

[0095] The task secondary allocation steps of the resistor trimming machine equipment include:

[0096] (1) Calculate the number of task points to be reallocated: The number of task points to be reallocated = the initial task points of the resistor trimming machine equipment with abnormal operation - the processed task points of the resistor trimming machine equipment with abnormal operation. Then, the number of uncompleted task points of equipment A is: 50 - 20 = 30, and the number of uncompleted task points of equipment D is: 60 - 35 = 25. Furthermore, the number of task points to be reallocated is 30 + 25 = 55.

[0097] (2) Calculate the load rates of the resistor trimming machine equipment with qualified operation: The load rate of equipment B is 60 / 240 = 0.25, the load rate of equipment C is 120 / 300 = 0.40, and the load rate of equipment E is 20 / 180 = 0.11.

[0098] (3) Sort the resistor trimming machine equipment with qualified operation in ascending order of load rate: The sorted sequence of resistor trimming machine equipment with qualified operation: Equipment E (0.11) → Equipment B (0.25) → Equipment C (0.40).

[0099] (4) Calculate the remaining processing capabilities of each resistor trimming machine equipment with qualified operation: The remaining processing capability of equipment B is 180 / 2 = 90, the load rate of equipment C is 180 / 3 = 60, and the load rate of equipment E is 160 / 1 = 160.

[0100] (5) The remaining processing capability of equipment E before allocation is 160 points, which is greater than the number of task points to be reallocated, 55. Then, allocate the number of task points to be reallocated to equipment E.

[0101] In the embodiment of the present invention, by collecting in real time the number of uncompleted task points of the resistor trimming machine with abnormal operation, calculating the load rate in combination with the operation parameters of the resistor trimming machine with qualified operation, sorting the qualified equipment in ascending order of load rate, and preferentially allocating tasks to the equipment with high remaining processing capabilities, the dynamic load balancing of resources within the cluster is achieved, and the overload or idleness of some equipment is avoided.

[0102] The above content is only an example and illustration of the concept of the present invention. Those skilled in the art of this technology can make various modifications or supplements to the described specific embodiments or use similar methods to replace them. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A method for managing the operating status of a cluster of trimming machines based on a cloud platform, characterized in that: The method includes: S1. Initial task allocation: Compare the self-check parameters of each trimming machine device with the preset self-check standard parameters, screen the qualified trimming machine cluster and allocate the initial task; S2. Task running and data collection: During the execution of the initial task, collect ablation images and ablation depths in the time series, and calculate the ablation coordinates and ablation line widths of the ablation points through edge detection and contour extraction; S3. Task running optimization: Dynamically adjust the parameters of the qualified trimming machine cluster according to the fusion analysis results of the ablation coordinates, ablation line widths and ablation depths of the ablation points and their corresponding standard ablation parameters; S4. Judgment of running qualification: After adjustment, generate an ablation path including ablation depth and ablation line width in real time through timestamp alignment, and at the same time, combine the resistance values of the corresponding resistors of each trimming machine device to determine whether the device runs abnormally; S5. Secondary task allocation: Perform a shutdown operation on the trimming machine device with abnormal operation according to the device operation parameters of the trimming machine device, and at the same time, perform secondary task allocation on the number of unfinished task points of the trimming machine device with abnormal operation.

2. The method for managing the operating status of a trimming machine device cluster based on a cloud platform according to claim 1, wherein: The screening of the qualified trimming machine cluster includes: Q1. Calculate the deviation between the resistance value of each trimming machine device at each self-check in the self-check parameters and the standard resistance value to obtain the reliability qualification degree of each trimming machine device; Q2. Compare and analyze the self-check path of each trimming machine device at each self-check in the self-check parameters with its standard self-check path to obtain the stability qualification degree of each trimming machine device; Q3. Perform weighted average calculation on the reliability qualification degree and the stability qualification degree through the preset dynamic weight to obtain the self-check qualification degree of each trimming machine device; Q4. Generate a group of candidate trimming machine devices for the trimming machine devices with self-check qualification degree greater than or equal to the preset self-check qualification degree; Q5. Extract the required number of trimming machine devices from the cloud platform, and select a qualified device cluster from the candidate trimming machines according to the required number of trimming machine devices.

3. The method for managing the operation status of a trimming machine equipment cluster based on a cloud platform according to claim 2, wherein: The analysis of the reliability qualification degree of each trimming machine device includes: Take the difference between the resistance value and the standard resistance value as the resistance deviation of each trimming machine device at each self-check; Count the number of self-checks with resistance deviation less than the preset resistance deviation and the total number of self-checks, and take the ratio of the two as the reliability qualification degree of each trimming machine device.

4. The method for managing the operation status of a trimming machine device cluster based on a cloud platform according to claim 1, characterized in that: The analysis of the stability qualification degree of each trimming machine device includes: Extract the self-check coordinate sequence and the standard coordinate sequence from the self-check path of each trimming machine device at each self-check; Calculate the minimum cumulative distance of the sequence through the dynamic time warping algorithm; If the minimum cumulative distance is less than the reference standard distance, it is determined that the stability qualification degree of this self-check is 1, otherwise, it is determined that the stability qualification degree of this self-check is 0; Count the average value of the stability qualification degrees of each self-check as the stability qualification degree of each trimming machine device.

5. The method for managing the operating status of a trimming machine equipment cluster based on a cloud platform according to claim 1, characterized in that: Calculating the ablation coordinates and ablation line widths of the ablation points through edge detection and contour extraction includes: Obtain the edge detection image of the ablation point from the ablation image through the edge detection algorithm, and then perform contour extraction on the edge detection image of the ablation point to obtain the contour of the ablation point and each pixel point within the contour; Perform weighted average calculation on each pixel point within the contour to obtain the ablation coordinates of the ablation point; Extract the pixel spacing between the left and right edges of the ablation point profile along the normal direction of the ablation point as the ablation line width of the ablation point.

6. The method for managing the operation status of a trimming machine device cluster based on a cloud platform according to claim 5, characterized in that: The dynamic parameter adjustment analysis of the qualified trimming machine cluster includes: W1. Extract the standard ablation path of each qualified trimming machine device from the cloud platform, obtain the standard coordinates matching the ablation coordinates in the standard path through the nearest neighbor algorithm, and perform the difference operation between the ablation coordinates and the standard coordinates to generate the real-time path offset. W2. Analyze the proportional relationship between the real-time path offset and the set reference deviation threshold, and obtain the trajectory correction parameters of each qualified trimming machine device through the adjustment value formula. W3. Extract the standard ablation depth of each qualified trimming machine device from the cloud platform. W4. Compare and analyze the ablation depth of each qualified trimming machine device with the standard ablation depth to obtain the power compensation value of each qualified trimming machine device. W5. Import the ablation line width and the standard ablation line width into the line width deviation formula to obtain the speed compensation value of each qualified trimming machine device. W6. Perform dynamic parameter adjustment based on the trajectory correction parameters, power compensation values, and speed compensation values of each qualified trimming machine device.

7. The method for managing the operation status of a trimming machine device cluster based on a cloud platform according to claim 6, wherein: The comparison and analysis of the ablation depth of each qualified trimming machine device with the standard ablation depth includes: Subtract the ablation depth from the standard deviation ablation depth to obtain the ablation depth difference of each qualified trimming machine device. Import the ablation depth difference into the power compensation formula to obtain the power compensation value of the qualified trimming machine device, and then obtain the power compensation values of each qualified trimming machine device.

8. The method for managing the operating state of a trimming machine device cluster based on a cloud platform according to claim 1, characterized in that: The generation of the ablation path including the ablation depth and the ablation line width includes: Based on the ablation coordinates of each trimming machine device in the time series, construct the actual ablation path of each trimming machine device. Align the time stamps of the ablation coordinates, ablation depth, and ablation line width of each ablation point in the actual ablation path. Mark the ablation depth and the ablation line width in chronological order in the actual ablation path of each trimming machine device, so as to obtain the ablation path including the ablation depth and the ablation line width.

9. The method for managing the operation status of a trimming machine device cluster based on a cloud platform according to claim 7, wherein: The determination of whether the device is operating abnormally includes: Use the starting points of the actual ablation path and the standard ablation path as the coincidence points for coincidence comparison, and then extract the coincidence path length and the standard ablation path length from them, and use the ratio of the two as the path coincidence ratio. Extract the maximum ablation line width and the minimum ablation line width from the ablation line widths of each ablation point, subtract the two to obtain the ablation line width difference. Analyze the ablation depth difference in the same way as the analysis method of the ablation line width difference. Extract the required resistance value of the resistor from the cloud platform, and calculate the absolute difference between it and the resistance value to obtain the resistance difference of the resistor. Compare the path coincidence ratio, ablation line width difference, ablation depth difference, and the resistance difference of the corresponding resistor with their set reference thresholds respectively, and determine the qualified trimming machine devices with parameters greater than their reference thresholds as operating abnormally, otherwise, determine them as operating normally.

10. The method for managing the operation status of a trimming machine device cluster based on a cloud platform according to claim 1, wherein: The performance of the secondary task allocation includes: Take the number of uncompleted task points of each abnormally operating trimming machine device as the number of task points to be reallocated. Calculate the product of the initial allocated task points and the single-task processing time in the equipment operation parameters of the qualified running resistor trimming machine as the initial task processing time, and take the ratio of it to the total available time in the equipment operation parameters as the load rate of the qualified running resistor trimming machine equipment; Sort the qualified running resistor trimming machine equipment in ascending order of the load rate to generate a sequence of qualified running resistor trimming machine equipment; Calculate the difference between the total available time and the initial task processing time as the current available time, and compare and analyze the current available time with the single-task processing time to obtain the remaining processing capacity of each qualified running resistor trimming machine equipment; Based on the sequence of qualified running resistor trimming machine equipment, allocate the task points to be reallocated in turn, and the task points allocated to the qualified running resistor trimming machine equipment do not exceed its remaining processing capacity until all the task points to be reallocated are completely allocated.

Images