A production management method for a fine cutting device

Abstract

This invention discloses a production control method for precision cutting equipment, specifically in the field of production scheduling and control, to address the difficulty in quantifying and assessing the quality impact of continuous switching between products with different contours in existing production management processes. By establishing historical order execution sequences, extracting product contour features to generate contour cluster files, and constructing contour switching inertia files based on historical contour switching processes, the invention maps pending orders to corresponding contour clusters and generates contour cluster switching sequences. Based on the recovery time constant and attenuation amplitude, the invention calculates the cross-sectional quality acceptance status of each order and constructs a cross-sectional quality migration trajectory. Furthermore, it identifies risk segments formed by the continuous accumulation of switching costs, reconstructs the production sequence, generates candidate production plans, and determines the target production plan. This invention achieves quantitative management of the contour switching impact propagation process, improving the rationality of scheduling and quality stability under mixed-line production conditions for precision cutting equipment.

Description

Technical Field

[0001] This invention relates to the field of production scheduling and control technology, and more specifically, to a production control method for precision cutting equipment. Background Technology

[0002] In the actual production management of precision cutting equipment, to meet the production demands of multiple varieties, small batches, and frequent changes in delivery dates, the production line typically needs to continuously switch between different types of products for processing according to the order plan. When product switching with significant contour differences occurs consecutively in the production plan, the surface roughness, burr height, and surface consistency of subsequent products often fluctuate continuously over several batches. This fluctuation does not disappear immediately after the product switching is completed, but gradually weakens and eventually returns to stability as subsequent orders are produced. In actual production, production managers usually schedule production based on order delivery dates, equipment load, and material availability, lacking quantitative assessment methods for the quality impact caused by switching between products with different contours.

[0003] Furthermore, when multiple products with similar quality impact directions are scheduled for consecutive production, the quality impact left over from previous products may continue to accumulate in subsequent orders. This can cause some high-precision products to enter production segments with significant quality fluctuations, resulting in issues such as product rework, repeated inspections, and adjustments to production schedules. Existing production scheduling schemes lack a quantitative description mechanism for the quality recovery process after contour switching, and also lack production control measures based on historical contour switching patterns to assess the quality acceptance status of future orders. This makes it difficult to effectively identify potential quality migration risk segments in the production plan, leading to production plan adjustments often relying on experience-based judgments, making it difficult to balance product quality stability and production organization efficiency.

[0004] Therefore, there is an urgent need for a production control method for precision cutting equipment to analyze the quality recovery patterns formed by different contour switching in historical production processes, and to optimize production plans based on the continuous impact of contour switching on the quality status of subsequent orders. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a production control method for precision cutting equipment to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A production control method for precision cutting equipment includes the following steps: S1. When the precision cutting equipment production line enters the production scheduling update state, retrieve the product inspection records and production sequence records of historical orders, and establish the historical order execution sequence. S2. Extract the contour turning distribution features, continuous shearing distribution features, and closed contour distribution features of each product in the historical order execution sequence, merge historical orders with similar contour features, and generate contour cluster files. S3. Extract contour cluster switching nodes along the historical order execution sequence, associate the recovery curves of cross-sectional quality indicators after each switching node, and use the time constant and attenuation amplitude of the recovery curve as the switching cost to construct a contour switching inertia archive. S4. Map each pending production order to the contour cluster file, expand the pending production order sequence along the current production plan, and generate the corresponding contour cluster switching sequence; S5. Based on the contour cluster switching sequence, query the contour switching inertial archive sequentially to obtain the switching cost of each switching, accumulate and superimpose the switching cost along the sequence, calculate the cross-sectional quality acceptance status of each order node, and generate the cross-sectional quality migration trajectory. S6. Traverse the cross-section quality migration trajectory, mark the order segments where the cross-section quality acceptance state exceeds the allowable boundary due to the continuous accumulation of the same-direction switching cost, and perform insert-type rearrangement of orders within the segment to interrupt the same-direction switching superposition and generate a candidate production plan set. S7. Compare the cross-sectional quality migration trajectories corresponding to each candidate production plan set, and select the scheme with the smallest comprehensive fluctuation evaluation quantity as the target production plan.

[0008] As a further aspect of the present invention, in step S1, establishing the historical order execution sequence specifically includes: Retrieve product specification inspection records from historical orders, extract the number of turning points, the proportion of continuous shearing edge length, and the perimeter of closed contours for each order's product outline, and form an order outline complexity vector. Retrieve product cross-section inspection records from historical orders, extract the average cross-section roughness and average burr height at the completion time of each order, and form an order cross-section quality vector; Retrieve the production sequence records of historical orders, align the order outline complexity vector and the order section quality vector according to the order execution order, and generate the historical order execution sequence.

[0009] As a further aspect of the present invention, in step S2, generating the contour cluster file specifically includes: Traverse the order contour complexity vectors in the historical order execution sequence, calculate the contour difference between any two order contour complexity vectors, and group orders with contour difference less than a preset difference threshold into the same contour cluster. The distribution range of the center of the quality vector of the corresponding section of the order within each contour cluster is statistically analyzed to determine the reference section quality level of the corresponding contour cluster. The center vector of the contour complexity vector of the order within each contour cluster is extracted, and the contour cluster identifier, center vector and reference section quality level are integrated into a contour cluster file.

[0010] As a further aspect of the present invention, in step S3, constructing the contour switching inertial profile specifically includes: Traverse adjacent orders along the historical order execution sequence. When adjacent orders belong to different contour clusters, record the boundary position of the adjacent orders as the contour cluster switching node. Extract the cross-section quality vectors corresponding to consecutive orders after the switching node from the cross-section detection record, retrieve the contour cluster archive, calculate the deviation between each cross-section quality vector and the reference cross-section quality level of the contour cluster after switching, form a cross-section quality deviation sequence according to the order execution order, fit the cross-section quality deviation sequence, and generate the inertial decay curve of the corresponding contour cluster switching node. The number of orders that the cross-sectional mass deviation must cross to recover to the preset stable range is extracted from the inertial decay curve as the recovery time constant, and the difference between the peak value of the inertial decay curve and the center value of the stable range is extracted as the decay amplitude. The pre-switching contour cluster identifier, post-switching contour cluster identifier, recovery time constant, and attenuation amplitude are combined to form a switching cost record, and then summarized to form a contour switching inertia file.

[0011] As a further aspect of the present invention, in step S4, generating the corresponding contour cluster switching sequence specifically includes: Read the product specification data of the orders to be scheduled to determine the outline cluster to which the orders belong. Read the outline cluster identifier of each order to be scheduled to be scheduled in sequence according to the scheduling order specified in the current production plan. When the outline cluster identifiers of adjacent orders to be scheduled to be scheduled are different, record an outline cluster switch. Combine the outline cluster identifiers of all orders to be scheduled to be scheduled with the switch position to form an outline cluster switch sequence.

[0012] As a further aspect of the present invention, in step S5, generating the cross-sectional quality migration trajectory specifically includes: Traverse the contour cluster switching sequence from the first order to be scheduled for production backwards. When traversing to the contour cluster switching position, extract the contour cluster identifier before switching and the contour cluster identifier after switching to form a switching pair. Search the contour switching inertial archive for the switching cost record that matches the switching pair and obtain the corresponding recovery time constant and attenuation amplitude. Starting from the switching position, an attenuation sequence corresponding to the switching cost is constructed based on the recovery time constant and the attenuation amplitude. The attenuation sequence is then mapped to the consecutive orders after the switching. When a new contour cluster switches, the corresponding attenuation sequence is superimposed with the previous attenuation sequence that has not yet finished attenuation to obtain the cumulative acceptance offset corresponding to each order. Based on the reference section quality level and cumulative acceptance offset of the contour cluster to which each order belongs, calculate the section quality acceptance status of each order, and connect the section quality acceptance status of all orders sequentially along the order sequence to be scheduled to generate the section quality migration trajectory.

[0013] As a further aspect of the present invention, in step S6, generating a candidate production plan set specifically includes: Traverse the cross-sectional quality transfer trajectory of each order and mark the order segments where the cross-sectional quality transfer status of adjacent orders continuously decreases and the decrease exceeds the preset threshold as unidirectional degradation segments. Orders within the unidirectional degradation zone are grouped according to their belonging contour clusters. The switching cost records corresponding to each contour cluster switch are extracted, and the intervals where the switching cost continuously accumulates are identified as inertial superposition intervals. For each inertial overlay interval, query the contour switching inertial file, filter the transition contour clusters that simultaneously satisfy the condition that the switching costs corresponding to both the first and last contour clusters of the interval are lower than the original switching costs, and select orders from the pending production orders outside the inertial overlay interval in the current production plan to insert into the inertial overlay interval, generating a reconstructed production schedule sequence. After inserting an order, the cross-sectional quality migration trajectory is recalculated, and the reconstructed production scheduling sequence with inertial superposition interval elimination is used as a candidate solution. All insertable positions are traversed to generate a candidate production plan set.

[0014] As a further aspect of the present invention, in step S7, selecting the scheme with the smallest comprehensive fluctuation evaluation quantity as the target production plan specifically includes: For the cross-sectional quality migration trajectory corresponding to each candidate production plan set, extract the change in cross-sectional quality acceptance status between each pair of adjacent nodes in the trajectory, and accumulate the absolute values ​​of the changes between each pair of adjacent nodes along the sequence to obtain the cumulative value of quality fluctuation of the candidate scheme. Simultaneously, the recovery time constant for each handover is extracted from the contour cluster handover sequence corresponding to each candidate scheme, and the total handover recovery time of the candidate scheme is obtained by accumulating the recovery time constant along the sequence. The weighted sum of the cumulative quality fluctuation value and the total switching recovery time is used as the comprehensive fluctuation evaluation value, and the candidate production plan with the smallest comprehensive fluctuation evaluation value is selected as the target production plan.

[0015] The technical effects and advantages of the production control method for precision cutting equipment of the present invention are as follows: This invention establishes historical order execution sequences, contour cluster archives, and contour switching inertia archives to continuously record and quantify the cross-sectional quality recovery patterns formed after switching between different contour products. This transforms the traditional experience-based production scheduling management process into a data-driven control process based on historical production patterns. During the production planning stage, this invention can identify the switching inertia relationships between different contour clusters, predict the propagation range and attenuation process of contour switching impacts in subsequent orders, and construct cross-sectional quality migration trajectories. It also allows for advance assessment of the quality acceptance status of each order, thereby identifying potential quality risk areas during the scheduling stage. Furthermore, by identifying the inertia superposition intervals formed by the continuous accumulation of switching costs, this invention introduces orders corresponding to transitional contour clusters to reconstruct the production sequence, weakening the superposition of quality impacts caused by continuous contour switching and ensuring a stable change in the cross-sectional quality acceptance status.

[0016] Compared to production scheduling based solely on order delivery dates and equipment load, this invention balances production organization efficiency with product quality stability, improves the rationality of production scheduling under mixed-line production conditions for products with different profiles, reduces quality fluctuations, repeated inspections, and production adjustments caused by profile switching, and enhances the stability and execution quality of precision cutting equipment production plans. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a production control method for precision cutting equipment according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1 Figure 1 The present invention provides a production control method for precision cutting equipment, which includes the following steps: S1. When the precision cutting equipment production line enters the production scheduling update state, retrieve the product inspection records and production sequence records of historical orders, and establish the historical order execution sequence. S2. Extract the contour turning distribution features, continuous shearing distribution features, and closed contour distribution features of each product in the historical order execution sequence, merge historical orders with similar contour features, and generate contour cluster files. S3. Extract contour cluster switching nodes along the historical order execution sequence, associate the recovery curves of cross-sectional quality indicators after each switching node, and use the time constant and attenuation amplitude of the recovery curve as the switching cost to construct a contour switching inertia archive. S4. Map each pending production order to the contour cluster file, expand the pending production order sequence along the current production plan, and generate the corresponding contour cluster switching sequence; S5. Based on the contour cluster switching sequence, query the contour switching inertial archive sequentially to obtain the switching cost of each switching, accumulate and superimpose the switching cost along the sequence, calculate the cross-sectional quality acceptance status of each order node, and generate the cross-sectional quality migration trajectory. S6. Traverse the cross-section quality migration trajectory, mark the order segments where the cross-section quality acceptance state exceeds the allowable boundary due to the continuous accumulation of the same-direction switching cost, and perform insert-type rearrangement of orders within the segment to interrupt the same-direction switching superposition and generate a candidate production plan set. S7. Compare the cross-sectional quality migration trajectories corresponding to each candidate production plan set, and select the scheme with the smallest comprehensive fluctuation evaluation quantity as the target production plan.

[0020] In step S1, a historical order execution sequence is established.

[0021] Retrieve the 2D contour drawings, process unfolding contour data, or final inspection contour data of the corresponding precision cutting equipment products according to the order number of historical orders, and convert them to the same coordinate reference for analysis. For each product contour, continuously traverse all contour segments along the contour boundary. When the included angle of adjacent contour segments changes and meets the set recognition conditions, the corresponding position is identified as a turning point and accumulated. The set recognition conditions are determined based on the enterprise's inspection data. For example, an included angle change of 15 degrees is used as the turning point recognition condition to avoid statistical distortion of turning points due to measurement noise or slight rounded transitions. Then, extract the length of all boundary segments involved in precision cutting in the contour and calculate the proportion of the continuous shearing edge length to the total length of all shearing edges. The continuous shearing edge is defined as a boundary segment without interruption holes, slotted structures, or independent island-like areas interfering with the continuous change of the contour direction. Then, identify the closed contour regions in the product contour, accumulate the boundary length of each closed contour, and obtain the perimeter data of the closed contour. When a product has multiple closed contours, the perimeters of all closed contours are accumulated. After extracting the number of turning points, the proportion of continuous shearing edge lengths, and the perimeter of the closed contour, the three data items are processed to uniform dimensions and written into the same data structure in a fixed order to form an order contour complexity vector.

[0022] The cross-sectional inspection results for each batch are read sequentially according to the order completion time, and a unique association is established between the inspection timestamp and the order number. For each order, cross-sectional roughness data for all valid inspection locations are extracted from the cross-sectional inspection samples. For inspection locations with duplicate measurements, the final confirmed results are retained, and invalid samples generated during the equipment calibration phase are removed. Subsequently, the cross-sectional roughness of all valid inspection locations within the order is statistically calculated to obtain the average cross-sectional roughness for the corresponding order. After roughness processing, burr inspection results are further read, and the burr heights corresponding to multiple inspection locations within the same order are uniformly summarized. When both automatic inspection data and manually verified data exist in the inspection record, the verified result is used as the final data source. For orders with abnormally discrete burr heights, erroneous data are filtered out based on the re-measurement marker in the inspection record before recalculation. After completing the statistical analysis of the average cross-sectional roughness and average burr height, the two indicators are combined according to a fixed field order to form an order cross-sectional quality vector. All historical orders are sorted according to the actual start and completion times in the production sequence record, with the actual production sequence of the orders serving as the sole sorting criterion. When there are instances of order insertion, rework, or batch splitting within the same time period, the order arrangement is reorganized according to the actual execution order of the equipment. Then, the contour complexity vector and cross-sectional quality vector of the corresponding order are read, and the two types of data are matched one by one according to the order number, and timeline alignment is completed according to the production sequence. For production interruptions caused by machine downtime, mold change, or equipment maintenance, the original order of orders is not changed; only the actual execution sequence is retained. After all orders are matched, the contour complexity vector and cross-sectional quality vector corresponding to each order are combined to form an order status record, and these are arranged sequentially according to the production execution order to form a historical order execution sequence.

[0023] In step S2, a contour cluster file is generated.

[0024] The contour complexity vectors for all orders are read sequentially according to the historical order execution sequence. Each contour complexity vector contains three data points: the number of inflection points, the percentage of consecutive sheared edges, and the perimeter of the closed contour. Since these three data points have different dimensions, a unified scaling process is first performed based on the overall statistical range of historical orders. The number of inflection points for all orders is mapped to a unified evaluation interval according to historical minimum and maximum values. The percentage of consecutive sheared edges is converted to the same evaluation interval as a percentage, and the perimeter of the closed contour is normalized according to the distribution range of the perimeter of historical orders. After scaling is unified, the difference values ​​are calculated item by item for the contour complexity vectors of any two orders, obtaining the difference in inflection points, the difference in the percentage of consecutive sheared edges, and the difference in the perimeter of the closed contour. These three differences are then accumulated to obtain the contour difference degree between the corresponding orders. To avoid distortion of difference judgment due to abnormal fluctuations in a single feature, a proportional accumulation method is used during the accumulation process. After calculating the contour difference degree between all orders, a contour proximity table is established in ascending order. The preset difference threshold is obtained based on historical production data. Specifically, it involves extracting order combinations from historical orders that have been confirmed to have the same cross-sectional quality performance, statistically analyzing the corresponding contour difference distribution range, and using the upper limit of the coverage ratio reaching a set standard as the difference threshold. The upper limit of the difference ratio corresponding to a 90% coverage of historically similar order combinations is used as the difference threshold. When the contour difference between two orders is not greater than this difference threshold, the corresponding orders are grouped into the same contour cluster. When an order simultaneously meets the classification conditions with multiple orders, it is classified according to the principle of minimum contour difference. For newly formed contour clusters, the contour difference relationships between its internal orders and unclassified orders are continuously traversed until no new orders meet the classification conditions, thus forming a structurally stable set of contour clusters.

[0025] Read the cross-sectional quality vectors corresponding to all orders within each contour cluster. Each cross-sectional quality vector consists of the average cross-sectional roughness and the average burr height. Based on the contour cluster division results, project all cross-sectional quality vectors belonging to the same contour cluster onto the roughness and burr height dimensions for statistical analysis. Analyze the continuous distribution intervals with the highest concentration of data in both the roughness and burr height dimensions, defining these intervals as the central distribution intervals. Specifically, divide the statistical segments at fixed intervals, count the number of orders within each segment, and select the continuous segment with the largest proportion of orders as the central distribution interval for the corresponding dimension. Then, determine the center values ​​of the roughness and burr height central distribution intervals and combine them to form the reference cross-sectional quality level for the corresponding contour cluster. After completing the cross-sectional quality statistics, perform the same central statistical processing on the contour complexity vectors corresponding to all orders within the contour cluster. Analyze the central distribution intervals for the number of turning points, the proportion of continuous shearing edge lengths, and the perimeter of the closed contour, and extract the center values ​​of each dimension to form the contour cluster central vector. Finally, a unique contour cluster identifier is assigned to each contour cluster, and the contour cluster identifier, center vector, and reference section quality level are associated and stored to form a contour cluster file.

[0026] In step S3, a contour switching inertial profile is constructed.

[0027] Following the historical order execution sequence, all adjacent orders are traversed one by one, and the contour cluster identifier corresponding to each order is read. When the contour cluster identifiers of two adjacent orders are inconsistent, the boundary position between the end position of the previous order and the start position of the next order is determined as the contour cluster switching node, and the contour cluster identifier before and after the switch is recorded. After completing the switching node identification, the cross-sectional quality vector corresponding to the consecutive orders after the switching node is read from the cross-sectional inspection record, and the reference cross-sectional quality level corresponding to the contour cluster after the switch is read from the contour cluster file. Since the cross-sectional quality vector is composed of the mean cross-sectional roughness and the mean burr height, the deviation between the mean cross-sectional roughness of the current order and the center value of the reference roughness, and the deviation between the mean burr height of the current order and the center value of the reference burr height are calculated respectively. The two deviations are converted to a unified evaluation scale and then accumulated to form the comprehensive deviation of the corresponding order. Subsequently, the comprehensive deviations corresponding to the consecutive orders after the switching node are arranged in order according to the actual production sequence of the orders to form the cross-sectional quality deviation sequence. To ensure the fitting results accurately reflect the attenuation process of the contour switching effect, a smoothing process is performed on the cross-sectional quality deviation sequence. Specifically, a fixed-length sliding window is used to calculate the average deviation of adjacent orders. The sliding window length is determined based on the average batch size of historical orders; for example, a window length covering five consecutive orders is used. After smoothing, a sequence relationship is established according to the order number and the corresponding comprehensive deviation, and an inertial attenuation curve is constructed using a monotonically decaying constraint fitting method. During the fitting process, the curve value corresponding to subsequent orders must not be higher than the curve value corresponding to previous orders. When short-term fluctuations occur locally, the average trend of adjacent intervals is used to replace the abnormal fluctuation points, thereby obtaining an inertial attenuation curve that reflects the gradual weakening process of the contour switching effect.

[0028] Based on the obtained inertia decay curve, the stable interval corresponding to the contour cluster after the switch is determined. The stable interval is determined according to the statistical results of the cross-sectional quality deviation of historical orders belonging to the same contour cluster and without the influence of contour switching. Specifically, the comprehensive deviation distribution interval corresponding to the same contour cluster in the continuous stable production stage is extracted, and the continuous interval with the coverage ratio reaching the set standard is counted as the stable interval. For example, the continuous interval covering the deviation corresponding to 95% of the stable production orders is determined as the stable interval. Then, the center position of the upper and lower boundaries of the stable interval is calculated, and this center position is determined as the center value of the stable interval. After the stable interval is determined, the search is performed point by point along the order direction starting from the starting position of the inertia decay curve. When the curve value first enters the stable interval and the curve values ​​of subsequent consecutive orders remain within the stable interval, the corresponding order position is determined as the recovery completion position. The number of orders spanned from the contour cluster switching node to the recovery completion position is defined as the recovery time constant. For example, if the eighth order after the contour switching node first enters the stable interval and subsequent orders remain stable, the recovery time constant is determined to be eight. Subsequently, the maximum curve value in the inertial decay curve is read, and the difference between this maximum curve value and the center value of the stable interval is calculated. This difference is defined as the decay amplitude. The decay amplitude is used to characterize the maximum deviation corresponding to the contour switching behavior when it has an impact, and the recovery time constant is used to characterize the order span of the continuous propagation of this impact. After completing the extraction of the recovery time constant and decay amplitude, the pre-switching contour cluster identifier, post-switching contour cluster identifier, recovery time constant, and decay amplitude are combined to form a switching cost record for the corresponding contour switching relationship. The above process is repeated for all contour cluster switching nodes in the historical order execution sequence, and the nodes are summarized and stored according to the index relationship established by the pre-switching contour cluster identifier and the post-switching contour cluster identifier, ultimately forming a contour switching inertial archive.

[0029] In step S4, a corresponding contour cluster switching sequence is generated.

[0030] Read the product specification data corresponding to the current pending production orders, and obtain the number of turning points, the proportion of continuous shearing edge lengths, and the perimeter of the closed contour for each pending production order according to the same contour feature extraction rules as historical orders, forming the contour complexity vector of the pending production orders. Then, iterate through all contour cluster center vectors in the contour cluster archive, calculate the contour difference between the pending production order contour complexity vector and the center vector of each contour cluster, and determine the contour cluster with the smallest difference as the corresponding pending production order's assigned contour cluster. When multiple contour clusters have the same difference, further compare the maximum deviation of the pending production order contour complexity vector from the corresponding contour cluster center vector in each dimension, and select the contour cluster with the smallest maximum deviation as the final assigned contour cluster. After classifying all pending production orders, arrange the corresponding assigned contour cluster identifiers sequentially according to the order execution order specified in the current production plan, and iterate along the production scheduling direction starting from the first order. When the corresponding contour cluster identifiers of two adjacent pending production orders are the same, it is determined that the current production process remains within the same contour cluster and continuous production continues, without recording a contour cluster switching event. When the corresponding corresponding contour cluster identifiers of two adjacent pending production orders are different, the boundary position between the end position of the previous order and the start position of the next order is recorded as a contour cluster switching position, and the corresponding contour cluster identifier before and after the switch is also recorded. This process continues until all pending production orders are analyzed. Finally, the corresponding contour cluster identifiers and corresponding contour cluster switching positions of each order are saved sequentially according to the order execution order, forming a complete contour cluster switching sequence. The contour cluster switching sequence not only retains the contour category distribution of pending production orders, but also accurately records the switching order and switching position between different contour clusters, providing a unified data input for subsequent contour switching inertial propagation analysis and cross-sectional quality migration trajectory deduction.

[0031] In step S5, the cross-sectional quality migration trajectory is generated.

[0032] The contour cluster switching sequence is traversed according to the actual execution order of the orders to be scheduled. When the traversal reaches the contour cluster switching position, the corresponding pre-switching contour cluster identifier and post-switching contour cluster identifier are read and combined to form a contour cluster switching pair. Then, the contour cluster switching pair is used as the retrieval condition to query the contour switching inertia archive to obtain the recovery time constant and attenuation amplitude of the corresponding switching relationship. The recovery time constant is used to characterize the order span of the continuous propagation of the influence of the contour switching, and the attenuation amplitude is used to characterize the maximum quality deviation formed by the contour switching in the historical production process. After completing the switching cost reading, the length of the attenuation propagation interval is determined according to the recovery time constant. For example, if the recovery time constant corresponding to a certain contour switching relationship is eight, then the corresponding attenuation propagation interval is formed from the first order after the current contour cluster switching position to the end of the eighth order. Then, the standard inertial attenuation curve corresponding to the contour switching relationship is read. The standard inertial attenuation curve is derived from the center curve record of the inertial attenuation curve corresponding to the same contour switching node in historical orders, and its curve length is consistent with the recovery time constant. Then, the standard inertial attenuation curve is discretized according to the order node to obtain the curve value corresponding to each order position within the attenuation propagation interval. Using the curve value at the starting position of the inertial decay curve as the baseline value, the proportional relationship between the curve value corresponding to each subsequent order position and the baseline value is calculated to form a decay ratio sequence. For example, if the curve value corresponding to a certain order position accounts for 80% of the baseline value, the corresponding decay ratio is 0.8. Then, the decay amplitude is distributed node by node using the decay ratio corresponding to each position, that is, the decay amplitude is multiplied by the decay ratio of the corresponding position to obtain the acceptance offset value corresponding to each order node within the propagation interval. In this way, each order node after the contour switching position obtains a clear acceptance offset value, and this acceptance offset value gradually decreases strictly according to the historical inertial decay law.

[0033] As the process continues traversing the contour cluster switching sequence, if a new contour cluster switch occurs subsequently, a new attenuation propagation interval and its corresponding acceptance offset value sequence are generated in the same manner. When a new attenuation propagation interval overlaps with a previous attenuation propagation interval, the propagation process of the previous contour switching effect is not terminated; instead, the previous attenuation sequence is retained and continues to extend. For example, if an order node is simultaneously covered by two attenuation propagation intervals, the acceptance offset values ​​of the two attenuation sequences at the corresponding positions of the order node are read, and accumulation processing is performed according to a unified evaluation scale. If an order node is simultaneously covered by three attenuation propagation intervals, the acceptance offset values ​​at the corresponding positions of the three attenuation sequences are read, and cumulative summation processing is performed. The cumulative acceptance offset for each order node is obtained in the above manner. The cumulative acceptance offset is used to characterize the comprehensive deviation degree of the current order node after being jointly affected by multiple previous contour switching behaviors. To avoid the difference in the attenuation amplitude dimensions corresponding to different contour switching relationships affecting the superposition result, a unified scale conversion is performed based on the historical statistical range in the contour switching inertial archive before performing accumulation, so that all acceptance offset values ​​are mapped to the same evaluation interval before cumulative calculation. After completing the switching and traversal of all contour clusters, obtain the cumulative acceptance offset record corresponding to each order node in the order sequence to be scheduled.

[0034] After obtaining the cumulative acceptance offset, the cross-sectional quality acceptance status corresponding to each order node is further calculated. Specifically, the reference cross-sectional quality level corresponding to the contour cluster to which the current order belongs is read. The reference cross-sectional quality level is determined by the cross-sectional quality center distribution range formed during the historical stable production stage of the corresponding contour cluster, including the center value of cross-sectional roughness and the center value of burr height. First, the center values ​​of cross-sectional roughness and burr height are mapped to a unified evaluation scale according to the historical statistical range, and a reference quality benchmark value is formed by proportional accumulation. Then, the cumulative acceptance offset corresponding to the current order is read, and the cumulative acceptance offset is subtracted from the reference quality benchmark value to obtain the cross-sectional quality acceptance status corresponding to the current order node. The reference cross-sectional quality status refers to the comprehensive evaluation value obtained after the reference cross-sectional quality level corresponding to the contour cluster to which the current order belongs is corrected by the cumulative acceptance offset, and is used to characterize the degree of deviation of the current order's cross-sectional quality from the reference cross-sectional quality level of the contour cluster. When the cumulative acceptance offset value increases, the cross-sectional quality acceptance status decreases accordingly; when the cumulative acceptance offset gradually decreases, the cross-sectional quality acceptance status gradually approaches the historical stable production state of the corresponding contour cluster. After calculating the cross-sectional quality acceptance status of all order nodes in the same way, the order execution order is used as the horizontal arrangement order, and the cross-sectional quality acceptance status of each order node is connected sequentially to form a continuous trajectory. Each trajectory node corresponds to an order to be scheduled for production, and the line connecting each two adjacent nodes represents the change process of the cross-sectional quality acceptance status during production, thus forming a complete cross-sectional quality migration trajectory.

[0035] In step S6, a candidate production plan set is generated.

[0036] The cross-sectional quality migration trajectory is sequentially traversed. Each trajectory node corresponds to a production order to be scheduled, and the cross-sectional quality acceptance status of that order is recorded. During the traversal, starting from the first order node, the difference in cross-sectional quality acceptance status between adjacent order nodes is calculated sequentially. When the cross-sectional quality acceptance status of the subsequent order node is lower than that of the previous order node, the difference is recorded as a decrease; when the cross-sectional quality acceptance status of the subsequent order node is not lower than that of the previous order node, the current downward trend is terminated. Subsequently, consecutive decrease sequences are statistically analyzed, and the cumulative decrease value of the corresponding decrease sequence is calculated. The preset threshold is obtained based on the statistical analysis of order segments with confirmed quality risks in historical production plans. For example, after statistically analyzing the cumulative decrease value distribution of historical risk segments, the lower limit of the cumulative decrease value covering 90% of historical risk segments is taken as the preset threshold. When the cumulative decrease value corresponding to a consecutive decrease sequence reaches the preset threshold, the order segment covered by the consecutive decrease sequence is marked as a unidirectional degradation segment. After identifying the unidirectional degradation segment, the corresponding contour cluster identifiers for each order within the segment are read and categorized according to these identifiers. Then, the switching cost records corresponding to the switching of each contour cluster within the segment are extracted. These switching cost records include recovery time constants and attenuation amplitudes. Therefore, the propagation contribution value generated by the corresponding switching relationship is further calculated based on the recovery time constant and attenuation amplitude, and a switching contribution sequence is established according to the sequence of contour cluster switching. When the propagation contribution value corresponding to a subsequent switching is consistently greater than that corresponding to a previous switching, the corresponding segment is determined to be in a process of cumulative increase in switching cost. The search continues along the switching contribution sequence until a position where the propagation contribution value decreases or remains constant is reached; this continuously increasing interval is then identified as the inertial superposition interval. Unlike traditional methods that only count the impact of a single contour switching, this embodiment identifies the inertial superposition interval by analyzing the continuous cumulative trend of multiple contour switching behaviors within the same order sequence, thereby locating the key propagation link that leads to the continuous decline in the cross-sectional quality bearing capacity.

[0037] For each identified inertial overlay interval, the identifiers of the first and last contour clusters of the interval are read, and all contour switching records related to them are retrieved from the contour switching inertial archive. Then, starting from the first contour cluster, the switching cost corresponding to switching from the first contour cluster to a candidate contour cluster is calculated; then, starting from the candidate contour cluster, the switching cost corresponding to switching from the candidate contour cluster to the last contour cluster is calculated. Only contour clusters whose switching costs in both cases are lower than the switching cost corresponding to directly switching from the original first contour cluster to the last contour cluster are retained as transition contour clusters. After completing the transition contour cluster selection, orders belonging to the transition contour cluster are retrieved from the pending production orders outside the inertial overlay interval in the current production plan, and an insertable order set is formed according to the original execution order of the orders. Then, each insertable order is inserted into a different position within the inertial overlay interval, forming multiple reconstructed production scheduling sequences. For each reconstructed production scheduling sequence, the processes of contour cluster switching sequence generation, switching cost propagation, cumulative offset calculation, and cross-sectional quality migration trajectory construction are re-executed to obtain the corresponding reconstructed cross-sectional quality migration trajectory. Then, the inertial superposition intervals in the reconstructed trajectory are re-identified. When the continuous growth propagation chain corresponding to the original inertial superposition interval is cut off, and there is no new inertial superposition interval covering the original interval range in the reconstructed trajectory, the corresponding reconstructed production scheduling sequence is determined to have completed the elimination of inertial superposition intervals. The elimination of inertial superposition intervals is not simply the elimination of contour cluster switching behavior, but rather the decomposition of the propagation accumulation relationship formed by multiple continuous growth switching costs into multiple independent propagation segments, so that the cross-sectional quality acceptance state no longer continues to decline along the original interval. Finally, all reconstructed production scheduling sequences that meet the conditions for elimination of inertial superposition intervals are retained as candidate schemes, and all insertable positions and orders corresponding to all transition contour clusters are traversed to form a candidate production plan set.

[0038] In step S7, the scheme with the smallest comprehensive fluctuation evaluation quantity is selected as the target production plan.

[0039] The candidate production plans are evaluated one by one, and the cross-sectional quality migration trajectory corresponding to each candidate production plan is read. Each trajectory node in the cross-sectional quality migration trajectory corresponds to a production order to be scheduled, and the cross-sectional quality acceptance status of that order is recorded. Then, all trajectory nodes are traversed sequentially according to the order execution order, and the change in cross-sectional quality acceptance status between adjacent nodes is calculated starting from the first node. The cross-sectional quality acceptance status of the next node is subtracted from the cross-sectional quality acceptance status of the previous node to obtain the status change value between adjacent orders. Since the cross-sectional quality acceptance status can either increase or decrease, the absolute value of the change value is further processed to uniformly represent the magnitude of the change in cross-sectional quality status between adjacent orders, without distinguishing the direction of change. After completing the calculation of the change amount of all adjacent nodes, the absolute values ​​of each change amount are accumulated and summed according to the trajectory order to obtain the cumulative quality fluctuation value of the corresponding candidate production plan. The cumulative quality fluctuation value is used to describe the overall fluctuation degree of the cross-sectional quality acceptance status during the entire production plan execution process.

[0040] Simultaneously, the contour cluster switching sequences corresponding to each candidate production plan are analyzed. All contour cluster switching relationships are extracted sequentially along the switching sequence, and the contour switching inertia archive is retrieved using the pre-switching and post-switching contour cluster identifiers to read the recovery time constant of the corresponding switching relationship. The recovery time constant characterizes the order span experienced by the corresponding contour switching impact from its generation to its decay to a stable range. Therefore, the larger the recovery time constant, the longer the inertial propagation duration of the corresponding contour switching behavior. Subsequently, according to the sequence of contour cluster switching occurrences, all recovery time constants are summed to obtain the total switching recovery time for the corresponding candidate production plan. After calculating the cumulative value of quality fluctuation and the total switching recovery time, the comprehensive fluctuation evaluation quantity is further calculated. To ensure that the two evaluation indicators participate in the calculation under the same evaluation system, the distribution range of the cumulative value of quality fluctuation and the total switching recovery time for all candidate production plans is first statistically analyzed, and the two indicators are uniformly mapped to the same evaluation interval. Then, the corresponding evaluation weights are determined based on the degree of influence of quality fluctuation events and recovery propagation events on the production plan execution results in historical production management records. For example, in historical statistics, quality fluctuations leading to re-inspection events account for 60% of all abnormal events, and recovery propagation leading to production rhythm adjustments accounts for 40%. Therefore, 0.6 and 0.4 are used as corresponding evaluation weights, respectively. Then, the normalized cumulative value of quality fluctuations and the total switching recovery time are weighted and summed according to their corresponding evaluation weights to obtain the comprehensive fluctuation evaluation value. Finally, the comprehensive fluctuation evaluation values ​​corresponding to all candidate production plans are iterated, sorted in ascending order, and the candidate production plan with the smallest comprehensive fluctuation evaluation value is selected as the target production plan. The selected target production plan not only has a more stable cross-sectional quality acceptance state change process but also corresponds to a shorter profile switching inertial propagation duration, thus ensuring that the production plan achieves optimized results in both quality stability and switching recovery efficiency.

[0041] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0042] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0043] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0044] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0045] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0046] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0047] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0049] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A production control method for precision cutting equipment, characterized in that, Includes the following steps: S1. When the precision cutting equipment production line enters the production scheduling update state, retrieve the product inspection records and production sequence records of historical orders, and establish the historical order execution sequence. S2. Extract the contour turning distribution features, continuous shearing distribution features, and closed contour distribution features of each product in the historical order execution sequence, merge historical orders with similar contour features, and generate contour cluster files. S3. Extract contour cluster switching nodes along the historical order execution sequence, associate the recovery curves of cross-sectional quality indicators after each switching node, and use the time constant and attenuation amplitude of the recovery curve as the switching cost to construct a contour switching inertia archive. S4. Map each pending production order to the contour cluster file, expand the pending production order sequence along the current production plan, and generate the corresponding contour cluster switching sequence; S5. Based on the contour cluster switching sequence, query the contour switching inertial archive sequentially to obtain the switching cost of each switching, accumulate and superimpose the switching cost along the sequence, calculate the cross-sectional quality acceptance status of each order node, and generate the cross-sectional quality migration trajectory. S6. Traverse the cross-section quality migration trajectory, mark the order segments where the cross-section quality acceptance state exceeds the allowable boundary due to the continuous accumulation of the same-direction switching cost, and perform insert-type rearrangement of orders within the segment to interrupt the same-direction switching superposition and generate a candidate production plan set. S7. Compare the cross-sectional quality migration trajectories corresponding to each candidate production plan set, and select the scheme with the smallest comprehensive fluctuation evaluation quantity as the target production plan.

2. The production control method for precision cutting equipment according to claim 1, characterized in that, In step S1, establishing the historical order execution sequence specifically includes: Retrieve product specification inspection records from historical orders, extract the number of turning points, the proportion of continuous shearing edge length, and the perimeter of closed contours for each order's product outline, and form an order outline complexity vector. Retrieve product cross-section inspection records from historical orders, extract the average cross-section roughness and average burr height at the completion time of each order, and form an order cross-section quality vector; Retrieve the production sequence records of historical orders, align the order outline complexity vector and the order section quality vector according to the order execution order, and generate the historical order execution sequence.

3. The production control method for precision cutting equipment according to claim 1, characterized in that, In step S2, generating the contour cluster file specifically includes: Traverse the order contour complexity vectors in the historical order execution sequence, calculate the contour difference between any two order contour complexity vectors, and group orders with contour difference less than a preset difference threshold into the same contour cluster. The distribution range of the center of the quality vector of the corresponding section of the order within each contour cluster is statistically analyzed to determine the reference section quality level of the corresponding contour cluster. The center vector of the contour complexity vector of the order within each contour cluster is extracted, and the contour cluster identifier, center vector and reference section quality level are integrated into a contour cluster file.

4. A production control method for precision cutting equipment according to claim 1, characterized in that, In step S3, constructing the contour switching inertial profile specifically includes: Traverse adjacent orders along the historical order execution sequence. When adjacent orders belong to different contour clusters, record the boundary position of the adjacent orders as the contour cluster switching node. Extract the cross-section quality vectors corresponding to consecutive orders after the switching node from the cross-section detection record, retrieve the contour cluster archive, calculate the deviation between each cross-section quality vector and the reference cross-section quality level of the contour cluster after switching, form a cross-section quality deviation sequence according to the order execution order, fit the cross-section quality deviation sequence, and generate the inertial decay curve of the corresponding contour cluster switching node. The number of orders that the cross-sectional mass deviation must cross to recover to the preset stable range is extracted from the inertial decay curve as the recovery time constant, and the difference between the peak value of the inertial decay curve and the center value of the stable range is extracted as the decay amplitude. The pre-switching contour cluster identifier, post-switching contour cluster identifier, recovery time constant, and attenuation amplitude are combined to form a switching cost record, and then summarized to form a contour switching inertia file.

5. A production control method for precision cutting equipment according to claim 1, characterized in that, In step S4, generating the corresponding contour cluster switching sequence specifically includes: Read the product specification data of the orders to be scheduled to determine the outline cluster to which the orders belong. Read the outline cluster identifier of each order to be scheduled to be scheduled in sequence according to the scheduling order specified in the current production plan. When the outline cluster identifiers of adjacent orders to be scheduled to be scheduled are different, record an outline cluster switch. Combine the outline cluster identifiers of all orders to be scheduled to be scheduled with the switch position to form an outline cluster switch sequence.

6. A production control method for precision cutting equipment according to claim 1, characterized in that, In step S5, generating the cross-sectional quality migration trajectory specifically includes: Traverse the contour cluster switching sequence from the first order to be scheduled for production backwards. When traversing to the contour cluster switching position, extract the contour cluster identifier before switching and the contour cluster identifier after switching to form a switching pair. Search the contour switching inertial archive for the switching cost record that matches the switching pair and obtain the corresponding recovery time constant and attenuation amplitude. Starting from the switching position, an attenuation sequence corresponding to the switching cost is constructed based on the recovery time constant and the attenuation amplitude. The attenuation sequence is then mapped to the consecutive orders after the switching. When a new contour cluster switches, the corresponding attenuation sequence is superimposed with the previous attenuation sequence that has not yet finished attenuation to obtain the cumulative acceptance offset corresponding to each order. Based on the reference section quality level and cumulative acceptance offset of the contour cluster to which each order belongs, calculate the section quality acceptance status of each order, and connect the section quality acceptance status of all orders sequentially along the order sequence to be scheduled to generate the section quality migration trajectory.

7. A production control method for precision cutting equipment according to claim 1, characterized in that, In step S6, generating a candidate production plan set specifically includes: Traverse the cross-sectional quality transfer trajectory of each order and mark the order segments where the cross-sectional quality transfer status of adjacent orders continuously decreases and the decrease exceeds the preset threshold as unidirectional degradation segments. Orders within the unidirectional degradation zone are grouped according to their belonging contour clusters. The switching cost records corresponding to each contour cluster switch are extracted, and the intervals where the switching cost continuously accumulates are identified as inertial superposition intervals. For each inertial overlay interval, query the contour switching inertial file, filter the transition contour clusters that simultaneously satisfy the condition that the switching costs corresponding to both the first and last contour clusters of the interval are lower than the original switching costs, and select orders from the pending production orders outside the inertial overlay interval in the current production plan to insert into the inertial overlay interval, generating a reconstructed production schedule sequence. After inserting an order, the cross-sectional quality migration trajectory is recalculated, and the reconstructed production scheduling sequence with inertial superposition interval elimination is used as a candidate solution. All insertable positions are traversed to generate a candidate production plan set.

8. A production control method for precision cutting equipment according to claim 1, characterized in that, In step S7, selecting the scheme with the smallest comprehensive fluctuation evaluation as the target production plan specifically includes: For the cross-sectional quality migration trajectory corresponding to each candidate production plan set, extract the change in cross-sectional quality acceptance status between each pair of adjacent nodes in the trajectory, and accumulate the absolute values ​​of the changes between each pair of adjacent nodes along the sequence to obtain the cumulative value of quality fluctuation of the candidate scheme. Simultaneously, the recovery time constant for each handover is extracted from the contour cluster handover sequence corresponding to each candidate scheme, and the total handover recovery time of the candidate scheme is obtained by accumulating the recovery time constant along the sequence. The weighted sum of the cumulative quality fluctuation value and the total switching recovery time is used as the comprehensive fluctuation evaluation value, and the candidate production plan with the smallest comprehensive fluctuation evaluation value is selected as the target production plan.

Images