A method and workpiece machining monitoring system for monitoring workpiece machining

Abstract

The present application relates to a kind of for monitoring workpiece processing method and workpiece processing monitoring system, it relates to the technical field of workpiece processing monitoring, the method includes: the data segmentation processing of same tool, and the spatial coincidence degree of the processing area characterized by these segmented data is calculated, and the data of which coincidence degree is higher than threshold is collected as the first set of corresponding physical area, and process within process label is given to the data therein;For the segmented data of different tools, by spatial position matching, the segmented data of processing area spatial position correspondence is identified, and the segmented data of spatial position correspondence and the sequential order of its tool type comply with the preset process relationship is collected as second set, and inter-process process label is given to the data therein;Based on the above two labels, each segmented data is matched with monitoring strategy;When processing similar workpiece, relevant matching monitoring strategy is called to carry out abnormal monitoring and alarm, so as to solve the technical problems that the process correlation of existing monitoring product is ignored.

Description

Technical Field

[0001] This invention relates to the technical field of workpiece processing monitoring, and more particularly to a method and system for monitoring workpiece processing. Background Technology

[0002] In the machining of large workpieces in small batches, the workpieces are bulky, the processes are complex, and the machining cycle is long. Any machining anomalies such as tool breakage or overload can easily lead to significant cost losses. Therefore, process monitoring and anomaly tracing are essential. Currently, the industry generally uses machine tool data collected from a complete, normal machining process as the learning basis for monitoring models. By mining process characteristics in offline data, support is provided for the formulation of online monitoring strategies. This model is also a technical path adapted to small-batch production and achieving differentiated monitoring.

[0003] However, existing monitoring model building technologies based on offline data mostly segment signals based on tool numbers or NC program numbers, extracting features and judging anomalies solely based on the segmented signals, while neglecting the crucial process correlation characteristics during machining. These correlations are precisely the core of ensuring monitoring accuracy. On the one hand, there may be significant hierarchical process differences under the same tool machining trajectory, such as significant differences in cutting parameters and load thresholds between roughing and finishing. The lack of hierarchical correlation identification leads to a single monitoring standard, easily causing false alarms or missed alarms in the monitoring system. On the other hand, multiple tools have temporal linkage relationships in continuous machining operations in the same area, such as drilling followed by tapping and reaming. Existing technologies have not explored these tool-to-tool correlations, causing the monitoring system to be unable to adapt to the requirements of process linkage, further reducing the adaptability of the monitoring strategy. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a method and a workpiece processing monitoring system for monitoring workpiece processing, which solves the technical problems of neglecting the process correlation of monitored products and the lack of intuitive alarm location in the prior art.

[0006] (II) Technical Solution

[0007] To achieve the above objectives, the main technical solutions adopted by the present invention include:

[0008] In a first aspect, embodiments of the present invention provide a method for monitoring workpiece processing, comprising: acquiring machining process data of a single normal workpiece; the machining process data including a turntable angle; dividing the machining process data into multiple processing stages based on the step change of the turntable angle; within each processing stage, segmenting the data of the same tool to obtain multiple segmented data, calculating the spatial overlap of the processing areas represented by these segmented data, and aggregating the segmented data with spatial overlap exceeding a first threshold into a first set corresponding to the physical area, and assigning a process label to each segmented data in the first set; Within each machining stage, segmented data for different tools are identified through spatial location matching. Segmented data with matching spatial locations and tool type order conforming to preset process relationships are grouped into a second set, and each segmented data in the second set is assigned an inter-process label. Based on the intra-process and inter-process labels carried by each segmented data, a corresponding monitoring strategy is matched for each segmented data. When machining similar workpieces, the corresponding segmented data is matched according to real-time data, and the strategy matched for that segmented data is invoked for anomaly monitoring and alarm.

[0009] Optionally, the machining process data also includes tool tip coordinates and tool type; segmentation is performed based on the tool type and changes in tool tip coordinates: if the tool type is a milling tool, segmentation is performed based on the step change in the tool tip Z coordinate; if the tool type is a hole machining tool, segmentation is performed based on the step change in the tool tip X or Y coordinate.

[0010] Optionally, the spatial overlap of the machining areas represented by these segmented data is calculated, including: if the tool type of the segmented data is a milling tool, the spatial overlap is determined by calculating the geometric similarity of the projection of the envelope representing the machining area on the XY plane; if the tool type of the segmented data is a hole machining tool, the spatial overlap is determined by calculating the overlap rate of the machining depth range representing the vertical direction; the vertical direction is the Z-axis direction of the machine tool coordinate system.

[0011] Optionally, spatial location matching includes: comparing the machining positions represented by the segmented data of different tools; wherein, for the segmented data of milling tools, the comparison is to calculate the geometric similarity of their machining area envelopes; for the segmented data of hole machining tools, the comparison is to determine whether their hole position coordinates are the same.

[0012] Optionally, the preset process relationship is any of the following tool types in the order of processing: roughing milling tools followed by finishing milling tools, drills followed by reamers, or drills followed by taps.

[0013] Optionally, based on the process tags within and between processes carried by each segment of data, a corresponding monitoring strategy is matched for each segment of data, including: determining whether the process rules contained in the process tags between processes of the segment of data are unidentified processes; if not, then the monitoring strategy is matched based on the process tags between processes; if so, then the monitoring strategy is matched based on the process tags within the segment of data.

[0014] Optionally, the monitoring strategy may include at least whether to enable tool breakage monitoring and different spindle speed-up strategies.

[0015] Optionally, the process label within a process includes at least a process rule number, a process area number, and a sequence number within the process area, and the sequence number within the process area includes the layer number and the hole number within the process area; the process label between processes includes at least an inter-process rule number, an inter-process area number, and a sequence number within the inter-process area, and the sequence number within the inter-process area includes the layer number and the hole number within the inter-process area.

[0016] In a second aspect, embodiments of the present invention provide a workpiece processing monitoring system, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the program to implement the method for monitoring workpiece processing as described in any one of the first aspects.

[0017] Thirdly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for monitoring workpiece processing as described in any of the first aspects.

[0018] (III) Beneficial Effects

[0019] The beneficial effects of this invention are:

[0020] This application provides a method and a workpiece machining monitoring system for monitoring the machining of small batches of large workpieces. By realizing two-level process association identification within the process (single tool) and between processes (multiple tools), it is more in line with the complex machining process requirements of small batches of large workpieces compared to the limitations of existing machining monitoring products that are only modeled based on tool number or NC program number. It also solves the technical problems of neglecting process association and unintuitive alarm location in existing monitoring products. Attached Figure Description

[0021] Figure 1 A flowchart of a method for monitoring workpiece processing provided in an embodiment of this application is shown;

[0022] Figure 2 A schematic diagram of a workpiece processing monitoring system provided in an embodiment of this application is shown. Detailed Implementation

[0023] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0024] To address the shortcomings of existing monitoring products that segment monitoring signals based on tool numbers or NC program numbers, extracting features and identifying anomalies solely from these segments while neglecting the process relationships between different levels and tools, and lacking intuitive spatial positioning of alarm information, this application provides a method and system for monitoring small-batch large workpiece machining. By implementing two-level process relationship identification—both within the process (single tool) and between processes (multiple tools)—this method better meets the complex machining requirements of small-batch large workpieces, compared to the limitations of existing machining monitoring products that rely solely on tool numbers or NC program numbers for modeling. Furthermore, this invention enables load distribution visualization and alarm spatial positioning, providing data support for monitoring model learning, improving the monitoring accuracy and anomaly tracing efficiency of small-batch large workpiece machining, and overcoming the deficiencies of existing technologies.

[0025] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention can be understood more clearly and thoroughly, and that the scope of the present invention can be fully conveyed to those skilled in the art.

[0026] To facilitate understanding of the embodiments of this application, the terms involved in this application are explained below:

[0027] "Small-batch production": This is a strategy designed to produce a limited number of products. Also known as small-batch manufacturing, this method produces anywhere from a few to tens of thousands of units, depending on the technology used. In contrast, mass production creates hundreds of thousands or even millions of identical products.

[0028] "Large workpieces": There is no national standard definition for large workpieces in the machining industry. In most machining scenarios, workpieces with a single dimension (length / width / height / diameter) reaching the meter level can be classified as large workpieces.

[0029] Please see Figure 1 , Figure 1 A flowchart illustrating a method for monitoring workpiece processing according to an embodiment of this application is shown. It should be understood that this method can be executed by an electronic device, and the specific device of the electronic device can be configured according to actual needs; this application embodiment is not limited thereto. For example, the electronic device can be a computer or a server, etc. Specifically, this method can be applied to scenarios involving the processing of small batches of large workpieces; the method includes:

[0030] Step S110: Obtain machining process data for a single normal workpiece. The machining process data includes the turntable angle.

[0031] Specifically, a data acquisition card can be used to synchronously acquire machining process signals of a normal workpiece from the CNC machine tool and spindle power sensor at a specific sampling rate. These signals may include the current tool number, timestamp, tool tip coordinates, cutting status, rotary table angle, and spindle power. The normal workpiece is one of a small batch of large workpieces.

[0032] It should be understood that the specific sampling rate can be set according to actual needs, and the embodiments of this application are not limited thereto.

[0033] For example, the specific sampling rate could be 100 sps, or it could be 120 sps, etc.

[0034] Subsequently, tool information can be associated with all tool numbers appearing during the machining process. This tool information can include tool type and tool specifications; that is, in addition to the above information, the machining process data also includes tool information. The tool type can include milling cutters, drilling cutters, taps, and reamers; the tool specification can be the maximum cutting radius. Milling cutter is an abbreviation for milling tools.

[0035] Step S120: Based on the change in the turntable angle, the machining process data is divided into multiple machining stages. Each machining stage corresponds to a fixed machining orientation (or "machining station") of the workpiece on the machine tool.

[0036] Specifically, the acquired machining process data can be cleaned and segmented. Specifically, non-machining data points are filtered out based on cutting state signals; invalid data segments during rapid movement (idle travel) are identified and eliminated by calculating the Euclidean distance between adjacent points in time sequence and combining it with a threshold; finally, based on whether the turntable angle undergoes a step change exceeding a preset threshold, the continuous machining data stream is divided into multiple independent data blocks corresponding to different workpiece machining orientations, establishing a stable spatial reference for subsequent process analysis. The specific value of the preset threshold can be set according to actual needs, and this embodiment is not limited to this.

[0037] To facilitate understanding of step S120, a specific embodiment will be described below.

[0038] Optionally, firstly, the machining process data can be filtered based on the cutting state, retaining only the data (or sample points) under the cutting state. Secondly, the Euclidean distance between all adjacent data under the cutting state can be calculated based on the tool tip coordinates, using the following formula:

[0039] ;

[0040] In the formula, d i This represents the Euclidean distance between the data under the i-th cutting state and the data under the (i+1)-th cutting state; , and The tool tip coordinates in the data of the (i+1)th cutting state; , and Let be the tool tip coordinates in the data of the i-th cutting state.

[0041] Thus, the Euclidean distance value sequence can be obtained. Furthermore, a placeholder value M, much larger than the fast-shift threshold T, can be added at the beginning and end of the Euclidean distance value sequence to obtain a new Euclidean distance value sequence. The specific values ​​of the rapid traverse threshold T and the occupancy value M can be set according to actual needs, and the embodiments of this application are not limited thereto. For example, the rapid traverse threshold T is 15mm, the occupancy value D is 999mm, and the rapid traverse threshold T can be determined according to the maximum cutting speed set in the machine tool machining program.

[0042] Then, the distance values ​​between each sample point (data point) and its two sides can be calculated. For example, for the first sample point, the distances to its two sides are respectively... And M; for the second sample point, the distances to its two sides are respectively and Similarly, each sample point can be labeled based on the distances to both sides of each sample point. For example, if the distances to both sides exceed the fast-shift threshold T, it can be labeled as a type a feature point; if only the front side exceeds the fast-shift threshold T, it can be labeled as a type b feature point; if only the back side exceeds the fast-shift threshold T, it can be labeled as a type c feature point; and if neither side exceeds the fast-shift threshold T, it can be labeled as a normal point.

[0043] Subsequently, points in the fast-moving process can be eliminated based on the sample point markings. For example, type a feature points are isolated points in the fast-moving process, and all of them are eliminated. Another example is to find a type b feature point as the starting point, and then search for the nearest type c feature point (i.e., the ending point). All points between these two points constitute a candidate fast-moving interval. Then, check whether the total number of points in this candidate fast-moving interval is less than a preset fast-moving interval sample point count threshold T_num. If it is less than the fast-moving interval sample point count threshold T_num, it is identified as a fast-moving interval, and all sample points within this interval are eliminated. The specific value of the sample point count threshold T_num can be set according to actual needs, and this value is determined by the machine tool signal update rate and the signal acquisition hardware capability; this embodiment is not limited to this. For example, the sample point count threshold T_num can be 3.

[0044] Subsequently, after data cleaning, data segmentation and storage based on turntable angle can be performed. This process sequentially scans the cleaned time-series data, monitoring the turntable angle value in real time. When the angle difference between adjacent data points exceeds a preset threshold (or a change in angle difference is detected), a valid workstation switch is determined, and this moment is marked as a segmentation point. The system uses adjacent segmentation points as boundaries to divide all continuous data between them into independent processing stage data blocks, where the turntable angle of all data remains stable, corresponding to a fixed processing orientation of the workpiece. Each data block is then stored as a local independent file or database record, using its corresponding stable turntable angle value as the core identifier. The core purpose of this step is to establish a strict mapping between data and physical space. By using the turntable angle as a physical benchmark, the continuous time-series flow is divided into multiple modules with unified spatial benchmarks, thus laying the foundation for subsequent process correlation analysis. This ensures that all spatial calculations are performed in the correct workpiece coordinate system and naturally isolates process data from different processing surfaces, ultimately supporting accurate alarm spatial positioning capabilities. The specific value of the preset threshold can be set according to actual needs, and this embodiment is not limited to this.

[0045] Step S130: Within each machining stage, the data of the same tool is segmented to obtain multiple segmented data. The spatial overlap of the machining areas represented by these segmented data is calculated. Segmented data with spatial overlap exceeding a first threshold are grouped into a first set corresponding to the physical regions. A process label is assigned to each segmented data in the first set. The specific value of the first threshold can be set according to actual needs, and this embodiment is not limited to this.

[0046] It should be understood that the specific process of segmenting data for the same tool can be set according to actual needs, and the embodiments of this application are not limited thereto.

[0047] Optionally, when the machining process data includes tool tip coordinates and tool type, the segmentation process is performed based on the tool type and changes in tool tip coordinates: if the tool type is a milling tool, segmentation is performed based on the step change in the tool tip Z coordinate; if the tool type is a hole machining tool, segmentation is performed based on the step change in the tool tip X or Y coordinate.

[0048] For example, if the tool type is milling, the data is layered. Specifically: for time-series sample points, when the z-coordinate changes, a layering operation is triggered. At this point, it is determined whether the previous layer meets two conditions. If either condition is not met, it is not considered a layer and is discarded: The current layer has a minimum number of sample points (sample_num >= layer_num), where sample_num is the number of sample points in the current data segment, and layer_num is the minimum number of sample points per layer threshold. The specific value of layer_num can be set according to actual needs. For example, layer_num can be 5; the current layer has actual movement in the X or Y direction (Xmin != Xmax or Ymin != Ymax). These four values ​​refer to the maximum and minimum values ​​of the X and Y coordinates of the current layer, and != indicates that they are not equal.

[0049] For example, if the tool type is a hole machining tool, the data is processed by hole segmentation. Specifically: for time-series sample points, when the x-coordinate or y-coordinate changes, the hole segmentation operation is triggered. At this time, it is determined whether the previous data segment meets two conditions. If one of them is not met, it is not considered as data for a hole and is discarded: the current number of hole sample points meets the minimum requirement (sample_num >= hole_num), where sample_num is the number of sample points in the current data segment, and hole_num is the minimum number of sample points per hole threshold. The specific value of hole_num can be set according to actual needs, such as hole_num being 5; there is an actual machining depth in the Z direction (Zmin != Zmax), where Zmin and Zmax refer to the minimum and maximum values ​​of the z-coordinate in the current data segment, and != indicates that they are not equal.

[0050] It should also be understood that the specific process of calculating the spatial overlap of the processing areas represented by these segmented data, and of grouping the segmented data with spatial overlap exceeding the first threshold into the first set of corresponding physical areas, and assigning process labels to each segmented data in the first set, can also be set according to actual needs, and the embodiments of this application are not limited thereto.

[0051] Optionally, if the tool type of the segmented data is a milling tool, the spatial overlap is determined by calculating the geometric similarity of the projection of the envelope representing the machining area onto the XY plane.

[0052] Specifically, a cleaned data segment with a stable turntable angle (i.e., the same turntable angle or considered as the same turntable angle) and the same tool number within the current machining stage is loaded. Based on the step change in the Z-value of the tool tip coordinate of the milling tool, this data segment is divided into multiple layers (i.e., segmented data), each layer representing a continuous planar milling operation. For each layer, based on its tool tip coordinate point set, a tool diameter compensation method is used to obtain the XY plane envelope boundary of the machining area of ​​that layer, thus representing it as a two-dimensional projected region. Under a unified coordinate system defined by the same stable turntable angle, the geometric similarity of the envelope regions of two temporally adjacent layers is calculated. Specifically, the intersection-union ratio (IUU) method is used, calculating the ratio of the intersection area to the union area of ​​the two regions as a similarity metric. When this similarity is higher than a preset threshold, it is determined that the two layers represent continuous layered milling of the same physical area on the workpiece. The specific value of the preset threshold can be set according to actual needs, and this embodiment is not limited to this. Subsequently, multiple layers of data that are determined to belong to the same physical region will be aggregated and given a unified process label within the process, which includes the region identifier and layer number; at the same time, the processing range and Z value of each layer of data in the XY direction will be recorded as its attributes.

[0053] In other words, for process data with the same tool number at the same turntable angle, in-process process analysis is performed, and each level of data is tagged with an in-process process label and its characteristic attributes are recorded. The hierarchical data is stored in a hierarchical database, and a list of in-process process analysis results is output.

[0054] For example, the XY coordinate trajectory envelope of layer data is obtained based on the tool diameter compensation method. Specifically, this method involves processing the layer data coordinate point set... Compensate the tool radius to obtain the contour point set of the material removal area. The envelope boundary of the material removal region is determined by the extreme values ​​of the contour points. .in, Represents the set of coordinate points of the k-th layer; This represents the x-coordinate of the i-th original coordinate point in the k-th layer, where i represents the index of the point within that layer. This represents the ordinate of the i-th original coordinate point in the k-th layer; This represents the total number of coordinate points in the k-th layer; The two values ​​in the table are the minimum and maximum values ​​of the abscissa of the contour point of the material removal area of ​​the k-th layer, respectively, so as to determine the envelope boundary of the material removal area of ​​the layer in the X direction, and when the index k becomes k+1, it is represented as the relevant data of the k+1-th layer. The two values ​​in the equation are the minimum and maximum values ​​of the ordinate of the contour point of the material removal area in the k-th layer, respectively, which determine the envelope boundary of the material removal area in the Y direction.

[0055] Subsequently, the similarity of the envelope curves of adjacent time-series layers is determined. Data with a similarity exceeding a threshold are considered data from different layers within the same region. The similarity calculation employs the intersection-union method, which first requires calculating the intersection and union areas of the material removal regions in adjacent layers. Specifically:

[0056] ;

[0057] ;

[0058] In the formula, The expression represents the area of ​​the intersection interval; max indicates taking the maximum value. Represent the area of ​​the union interval; This represents the material removal area of ​​the k-th layer.

[0059] Then, the similarity between the two layers can be obtained, and the formula is as follows:

[0060] ;

[0061] Furthermore, if the IoU is greater than a preset value, the planar milling layering process rule is considered to have matched successfully, and the two layers of data are determined to be different layers of data in the same region. The specific value of this preset value can be set according to actual needs, and this embodiment is not limited to this. For example, the preset value can be 0.8.

[0062] Additionally, it can record the process rule number, process area number, and process area layer number for each layer of data, serving as the process technology label for the layer data. The process rule number is for layer data that has been successfully identified in the same area, and is 'multi-layer milling layer process'; otherwise, it is 'unidentified process rule'.

[0063] For the region number within the process, the region corresponding to the first layer data is 1. If the subsequent layer is determined to be in the same region as the previous layer, it inherits its region number; otherwise, a new region number is generated, which is the current maximum region number + 1.

[0064] For layer numbers within a process area, the layer number within a process area is a local sequence number assigned to each layer of data within the same area according to time sequence. When a new area is identified for the first time, the layer number within the area for its first layer of data is 1; for subsequent layers of data identified as belonging to this area, the layer numbers within the area are incremented sequentially according to time sequence; the layer numbers of each area are counted independently, and the layer numbers between different areas do not interfere with each other.

[0065] In other words, the process label within a process includes the process rule number, the process area number, and the layer number within the process area, thus fully illustrating the labeling of the process analysis corresponding to this level of data. For example, '1-3-2' within a process means that this level of data successfully matches rule number 1 (multi-layer milling process) and is the second layer of the third area.

[0066] Furthermore, it can record the X range, Y range, and Z value of each layer of data as attribute values ​​for the layer data.

[0067] Optionally, if the tool category of the segmented data is a hole-machining tool, the spatial overlap is determined by calculating the overlap rate of its characteristic machining depth range in the vertical direction; the vertical direction is the Z-axis direction of the machine tool coordinate system.

[0068] Specifically, for segmented data (i.e., hole data) generated by the same hole-machining tool at the same stable turntable angle, the machining depth range of its Z-coordinate is first obtained. Then, for two adjacent hole data points in time sequence, the overlap length of their machining depth ranges in the Z-axis direction is calculated and divided by their respective total depth lengths to obtain two overlap rates. If both calculated overlap rates are higher than a preset overlap rate threshold, it is determined that these two hole data points represent continuous machining operations on the same hole position on the workpiece and belong to the same machining area. The system assigns a unified in-process process label to all hole data points determined to belong to the same physical hole position. This label includes an in-process rule number, an in-process area number, and a unique hole sequence number within the area. Simultaneously, the precise spatial location (X value, Y value) and machining depth range (Zmin, Zmax) of each hole data point are recorded as its attribute values, thereby establishing a precise and traceable process knowledge structure for the hole machining process. The specific value of T_hole can be set according to actual needs, and this embodiment is not limited to this. For example, T_hole is 0.8.

[0069] For example, the Z-coordinate range of hole data can be obtained, and the overlap rate of adjacent hole-level data in time series can be determined. This involves obtaining the overlap length in the z-direction and dividing it by the length of each hole to obtain the overlap rate. If both overlap rates are higher than the overlap rate threshold T_hole, the corresponding two hole-level data are data from different holes in the same region. Furthermore, the process rule number, process region number, and hole number within the process region are recorded for each hole-level data as the process process label for the hole-level data. Specifically, for hole data whose process rule number is the aforementioned successfully identified hole in the same region, the process rule number is 'Hole processing sub-hole process'; otherwise, the process rule number is 'Unidentified process rule'.

[0070] For the region number within the process, the region corresponding to the first hole data is 1. If the subsequent layer is determined to be the same region as the previous layer, it inherits its region number; otherwise, a new region number is generated, which is the current maximum region number + 1.

[0071] For well numbers within a process region, the well number within that region is a local sequence number assigned to each well data within that region according to time sequence. When a new region is first identified, the layer number of the first well data in that region is 1; for subsequent well data identified as belonging to that region, the well numbers within the region are incremented sequentially according to time sequence; the well numbers in each region are counted independently, and the well numbers in different regions do not interfere with each other.

[0072] In other words, the process label within the process can include the process rule number, the process area number, and the hole number within the process area, thus fully illustrating the labeling of the process analysis corresponding to this hole-level data.

[0073] Furthermore, the X, Y, and Z values ​​of each aperture-level data point can be recorded as attribute values ​​for the aperture-level data.

[0074] Step S140: In each processing stage, for the segmented data of different tools, the segmented data that match the spatial position of the processing area are identified by spatial position matching, and the segmented data that match the spatial position and whose tool type sequence conforms to the preset process relationship are collected into a second set, and each segmented data in the second set is assigned an inter-process label.

[0075] It should be noted that this application actually allows for the intersection of the first and second sets. A hierarchical or aperture-level data can simultaneously conform to both intra-process and inter-process process rules; however, during the strategy matching phase, inter-process process rules have higher priority.

[0076] It should be understood that the specific process of spatial location matching can be set according to actual needs, and the embodiments of this application are not limited thereto.

[0077] Optionally, spatial location matching includes: comparing the machining locations represented by the segmented data of different tools; wherein, for the segmented data of milling tools, the comparison is to calculate the geometric similarity of their machining area envelopes.

[0078] For example, layer data generated by milling cutters with different tool numbers are loaded at the same stable turntable angle. For any two layer data from different tools, spatial location matching is achieved by calculating the geometric similarity of the projection of their machining area envelopes onto the XY plane. This calculation uses the same method as the in-process analysis (such as the intersection-union method), that is, calculating the ratio of the intersection area to the union area of ​​the two layers of material removal area projections. If the similarity is higher than a preset second threshold, it is determined that the two layers of data are highly overlapping in space (i.e., their spatial locations match), corresponding to the machining of the same physical area on the workpiece. This matching process does not depend on the tool number or timing, but is purely based on the spatial coverage relationship of the machining footprints, thus laying the foundation for identifying related processes such as roughing and finishing of the same area by different milling tools. The specific value of the second threshold can be set according to actual needs, and the embodiments of this application are not limited to this.

[0079] Furthermore, the process rule number, process area number, and process area layer number for each layer of data are recorded. Specifically, during step S140 (inter-process identification), the corresponding process rule number, process area number, and process area layer number for each layer of data need to be recorded. The naming logic for this recording rule is consistent with the processing method for milling cutter data in step S130 (intra-process identification). The core difference is that step S130 analyzes and collects layer signals for a single tool, while step S140 integrates layer data from multiple tools within the same analytical framework. Therefore, the implementation of this step can directly reuse the area and layer number division logic established in step S130.

[0080] It is important to note that in the product design scheme, the intra-process process label and the inter-process process label are two independent and parallel identification systems. The intra-process process label includes the intra-process rule number, the intra-process area number, and the intra-process area sequence number; the inter-process process label includes the inter-process rule number, the inter-process area number, and the inter-process area sequence number. The "area number" and "intra-process sequence number" in these two sets of labels are independent of each other and are not related.

[0081] It should be noted here that the specific definition of the rule number within the process can be as follows: '0' indicates an unidentified process; '1' indicates a 'multi-layer milling process'; '2' indicates a 'multi-hole machining process'.

[0082] The specific definitions of the inter-process rule number can be as follows: '0' indicates an unidentified process; '101' indicates 'planar milling association'; '102' indicates 'drilling and reaming association'; '103' indicates 'drilling and tapping association'.

[0083] This type of rule number mapping is mainly used for marking and identifying internal data streams, and its specific value can be adjusted according to the actual system design.

[0084] In addition, the process labels for inter-process analysis have the advantage of separately marking a set of area numbers - serial numbers within the area:

[0085] Each process label is independently marked with a set of area numbers and serial numbers. This clearly defines and records which levels of signals together satisfy a specific process rule. Without this independent record, the system will be unable to trace and identify which specific layers of data constitute a process relationship such as "planar milling association" or "drilling and reaming association".

[0086] For example, both T1 and T2 are milling cutters. T1 processes two layers in one region, and the process labels for the two layers of T1 are '1-1-1' and '1-1-2'. T2 processes two regions with a total of three layers, and the process labels for the two layers are '1-1-1', '1-1-2', and '0-2-1'. After inter-process analysis, it is found that the processing area of ​​T1 overlaps with the second processing area of ​​T2. Therefore, the inter-process labels for the above five layers are '101-1-1', '101-1-2', '101-1-3', '101-1-4', and '0-2-1'. In the subsequent strategy matching stage, the inter-process labels clearly indicate that the first four layers of data (101-1-1 to 101-1-4) should be grouped together. The top or bottom layer is determined by comparing the Z-coordinates of each layer within that group (if inter-process rules are used for strategy matching, the inter-process area number and sequence number are used; if intra-process rules are used, the intra-process area number and sequence number are used). Simultaneously, this independent labeling system also provides a direct source of information for the front-end interface's process association display or for the monitoring model to obtain data relationship information.

[0087] It should be noted that during step S140, after loading layer (hole) level data with the same turntable angle (including different tool numbers), inter-process identification needs to be performed on the data according to the tool category. Specifically: for milling tool data, planar milling association identification is performed; for all drilling and reaming tool data, drilling & reaming association identification is performed; and for all drilling and tapping tool data, drilling & tapping association identification is performed. The inter-process rule number for successfully matched layer (hole) data is 'planar milling association', 'drilling & reaming association', or 'drilling & tapping association'. If the match fails, it is 'unidentified process'.

[0088] In typical industrial settings, if only a smooth hole is needed, drilling is performed before reaming; if only a threaded hole is needed, drilling is performed before tapping. Since these two methods have different uses, they are typically not used together in the same location. Therefore, this invention does not consider scenarios where drilling, reaming, and tapping are combined.

[0089] Optionally, for segmented data of tools classified as hole machining tools, the comparison is used to determine whether their hole position coordinates are the same.

[0090] For example, the system retrieves the attribute values ​​of all hole-level data, identifying holes with the same X and Y coordinate values ​​as corresponding to the same physical hole location on the workpiece (i.e., spatially consistent), and grouping them into a "hole machining area." For each such area, the system records the process rule number, the area number, and the sequence number within that area for each hole-level data, thus establishing the organizational structure of the hole system data. Based on this, the system performs two key types of association identification: first, "drilling & reaming association identification," used to identify the process chain where a drill bit pre-machines a pilot hole, which is then finished by a reamer; second, "drilling & tapping association identification," used to identify the process pair where a drill bit machines a threaded pilot hole, which is then tapped to form the thread. According to the identification results, the system assigns a corresponding process label to each hole-level data (such as "drilling step in the drilling-reaming process chain," "tapping step in threaded hole machining," etc.), thereby clarifying its specific role and stage in the multi-process hole machining flow.

[0091] In addition, it records the inter-process rule number, inter-process area number, and hole number within the inter-process area for each layer of data. Correspondingly, the generation method of the inter-process label related to the hole is similar to the generation method of the inter-process label related to the layer mentioned above, as detailed in the relevant descriptions above.

[0092] It should also be understood that the pre-defined process relationship is the machining sequence of any of the following tool types: roughing milling tools followed by finishing milling tools, drills followed by reamers, or drills followed by taps.

[0093] For example, the process begins with confirmed overlapping machining data pairs (the first set). The system checks each pair of data one by one, examining the cutting tools used and their sequence, and rigorously verifying them against various preset process relationships (such as roughing milling cutter followed by finishing milling cutter, drill bit followed by reamer, drill bit followed by tap). Only when the tool type of a data pair completely matches one of the relationships and the sequence is consistent (e.g., the record shows that the roughing milling cutter was indeed processed before the finishing milling cutter), the data pair is retained and included in the second set; data pairs with reversed sequences or that do not conform to any relationship are filtered out. Subsequently, the system assigns a clear process label to each data segment in the second set, such as labeling the first roughing milling cut step as "roughing" and the subsequent finishing milling cut step as "finishing," and records the process chain relationship between them. Finally, the output second set not only lists all relevant machining steps but also clearly indicates the role of each step in the process chain and its relationship, transforming the raw data into directly understandable process flow information.

[0094] Step S150: Based on the process tags within and between processes carried by each segment of data, match the corresponding monitoring strategy for each segment of data.

[0095] Specifically, it determines whether the inter-process rules contained in the inter-process labels of the segmented data are unidentified processes; if not, it prioritizes matching the monitoring strategy based on the inter-process labels; if so, it matches the monitoring strategy based on the intra-process labels of the segmented data. The monitoring strategy includes at least whether to enable tool breakage monitoring and different spindle speed-up strategies.

[0096] For example, the system first checks the inter-process rule number in the inter-process label. As long as the value of this field is valid and not equal to a specific identifier representing an unrelated process (such as "unidentified process"), the system will prioritize the inter-process label group as the basis for strategy matching. Only when the inter-process rule number is invalid or is "unidentified process" will the system fall back and use the intra-process label group for matching.

[0097] When using inter-process labels (high-priority paths): The system uses the inter-process rule number as the core index to directly query the preset expert rule table (e.g., Table 1 below) and locate the corresponding process rule row (e.g., "planar milling association", "drilling & reaming association"). Then, the system combines the position information of this layer of data in the inter-process dimension, mainly based on the layer (hole) number within the inter-process area and its Z-axis coordinate sorting within the corresponding inter-process area number, to determine whether it is the "top layer", "bottom layer", or "other layers" in the process chain (for milling), or a specific tool role (e.g., "drilling tools", "reaming tools"). Finally, based on the specific clauses at the corresponding positions in the rule row, the system assigns the final monitoring strategy to this layer of data (e.g., "tool breakage monitoring off, speed increase 2 levels" or "tool breakage monitoring on, speed increase off").

[0098] Table 1

[0099]

[0100] Furthermore, the "Hierarchical Process and Strategy Matching Result List" is the culmination of the entire offline analysis process, and its generation spans and concludes in step S150. The construction of this list begins with process identification in steps S130 and S140: S130, through intra-process identification, divides the data into physical regions based on spatial overlap and assigns "region number" and "intra-region sequence number" labels, establishing a hierarchical framework; S140, through inter-process identification, analyzes the spatial and sequential relationships across tools, injecting "inter-process rule number" labels such as "roughing-finishing association" into the data, enriching the hierarchical process semantics. Finally, in step S150, the system, based on a preset expert rule base, comprehensively judges each segment of data labeled with intra-process and inter-process tags, matching and binding specific monitoring strategies one-to-one (e.g., "tool breakage monitoring on, speed up by 1 level"). The output of step S150 is the structured integration of the hierarchical position, process role, and matched monitoring strategies of all segmented data, thereby generating this complete "Hierarchical Process and Strategy Matching Result List." This list serves directly as the decision-making basis for the online monitoring phase (step S160), achieving a precise closed loop from process understanding to monitoring execution.

[0101] Step S160: When processing similar workpieces, match the corresponding segmented data according to the real-time data, and call the strategy for matching the segmented data to perform abnormal monitoring and alarm.

[0102] Specifically, when machining similar workpieces, the system receives machining data streams in real time and dynamically segments them into logical segments corresponding to offline levels based on predetermined segmentation logic (such as tool, turntable angle, and coordinate changes). Subsequently, the system quickly compares the machining characteristics of the real-time segments (such as spatial location and tool type) with the offline "hierarchical process and strategy matching result list" to accurately match the corresponding offline segment data, thereby determining the specific process stage, physical region, and sequence of the current machining operation. Upon successful matching, the system performs a dual invocation: first, based on the matched region number and process label, it invokes a dedicated monitoring model pre-trained for that specific level (e.g., "Region A - 3rd Layer Finishing"). This model, having incorporated data from other levels within the same region and undergone differentiated weighted learning during training, possesses optimal monitoring sensitivity for this specific machining scenario. The monitoring model employs a hierarchical signal database as its learning samples. During the model learning phase, it inputs data from other levels under the same region number and assigns different learning weights. Secondly, it synchronously applies monitoring strategies associated with the segmented data (such as "breakage monitoring on, speed up by 1 level"), dynamically adjusting the parameter thresholds and behavior patterns of the real-time monitoring system. During monitoring execution, the system interface clearly and in real-time displays the process name, region, and specific layer number of the current process, achieving transparent visualization of the processing. Once the dedicated monitoring model detects a deviation from the normal pattern in the real-time signal, the system immediately triggers an alarm precisely correlated with the current process context, enabling operators to quickly locate the problem. This step ultimately achieves a leap from extensive monitoring of "one policy per cutter" to refined, adaptive monitoring of "one policy per layer, one model per layer," significantly improving the safety, efficiency, and traceability of the processing.

[0103] Therefore, by utilizing the above technical solutions, this application embodiment acquires the machine tool signal and power signal of a complete workpiece and records the corresponding tool information. Data cleaning is completed through cutting state filtering and rapid traverse filtering. Based on expert rules, it identifies the hierarchical process associations of different trajectories of a single tool and the regional machining associations between multiple tools, outputting standardized process labels, hierarchical attribute values, and strategy labels. It also supports trajectory load coloring visualization and alarm spatial positioning. Furthermore, this application can be deployed and run independently, providing core data support for adaptive matching of monitoring strategies for small-batch large workpiece machining. It can intuitively present the load distribution offline and accurately locate alarm areas during online monitoring, significantly improving the intelligent monitoring level and problem tracing efficiency of the machining process.

[0104] It should be understood that the above-described method for monitoring workpiece processing is merely exemplary, and those skilled in the art can make various modifications based on the above system, and such modified solutions also fall within the protection scope of this application.

[0105] like Figure 2As shown in the figure, this application embodiment also provides a workpiece processing monitoring system, which includes a processor and a memory. The memory stores a computer program, and when the processor executes the program, it implements the above-described method for monitoring workpiece processing.

[0106] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for monitoring workpiece processing as described above.

[0107] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0108] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0109] It should be noted that the word "a" or "an" preceding a component does not preclude the existence of multiple such components. This invention can be implemented using hardware comprising several different components and using a suitably programmed computer. Among the listed devices, several of these devices may be embodied by the same hardware. The use of terms such as "first," "second," "third," etc., is merely for convenience and does not indicate any order. These terms can be understood as part of the component names.

[0110] Furthermore, it should be noted that in the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0111] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the technical solution should be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0112] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the invention's technical solutions and their equivalents, then this invention should also include these modifications and variations.

Claims

1. A method for monitoring workpiece machining, characterized in that, include: Obtain machining process data for a single normal workpiece; The machining process data includes the turntable angle; Based on the step change in the turntable angle, the machining process data is divided into multiple machining stages; Within each processing stage, the data of the same tool is segmented to obtain multiple segmented data, and the spatial overlap of the processing area represented by these segmented data is calculated. Segmented data with spatial overlap higher than a first threshold are grouped into a first set corresponding to the physical area, and each segmented data in the first set is assigned a process label within the process. Within each processing stage, the segmented data of different tools are identified by spatial location matching, and the segmented data that match the spatial location of the processing area are collected into a second set. The segmented data that match the spatial location and whose tool type sequence conforms to the preset process relationship are assigned an inter-process label to each segmented data in the second set. Based on the process tags within and between processes carried by each segment of data, a corresponding monitoring strategy is matched for each segment of data. When processing similar workpieces, the system matches the corresponding segmented data based on real-time data and calls the strategy that matches the segmented data to perform anomaly monitoring and alarm.

2. The method according to claim 1, characterized in that, The machining process data also includes tool tip coordinates and tool type; the segmentation process is performed based on the tool type and changes in tool tip coordinates: if the tool type is a milling tool, segmentation is performed based on a step change in the tool tip Z coordinate; if the tool type is a hole machining tool, segmentation is performed based on a step change in the tool tip X or Y coordinate.

3. The method according to claim 2, characterized in that, The calculation of the spatial overlap of the processing areas represented by these segmented data includes: If the tool category of the segmented data is the milling tool, the spatial overlap is determined by calculating the geometric similarity of the projection of the envelope representing the machining area onto the XY plane. If the tool category of the segmented data is the hole machining tool, the spatial overlap is determined by calculating the overlap rate of its characteristic machining depth range in the vertical direction; the vertical direction is the Z-axis direction of the machine tool coordinate system.

4. The method according to claim 2, characterized in that, The spatial location matching includes: comparing the machining positions represented by the segmented data of different tools; wherein, for the segmented data of the tool category being the milling tool, the comparison is to calculate the geometric similarity of its machining area envelope; for the segmented data of the tool category being the hole machining tool, the comparison is to determine whether their hole position coordinates are the same.

5. The method according to claim 1 or 4, characterized in that, The preset process relationship is any of the following tool types in the order of processing: roughing milling tools followed by finishing milling tools, drills followed by reamers, or drills followed by taps.

6. The method according to claim 1, characterized in that, The method of matching a corresponding monitoring strategy for each data segment based on the intra-process process tag and the inter-process process tag carried by each data segment includes: Determine whether the inter-process rules contained in the inter-process labels of the segmented data are unidentified processes; If not, the monitoring strategy will be matched based on the process label between processes. If so, the monitoring strategy is matched based on the process label within the segmented data.

7. The method according to claim 6, characterized in that, The monitoring strategy includes at least whether to enable tool breakage monitoring and different spindle speed-up strategies.

8. The method according to claim 1, characterized in that, The process label within a process includes at least a process rule number, a process area number, and a sequence number within the process area, and the sequence number within the process area includes a layer number and a hole number within the process area; the process label between processes includes at least an inter-process rule number, an inter-process area number, and a sequence number within the inter-process area, and the sequence number within the inter-process area includes a layer number and a hole number within the inter-process area.

9. A workpiece processing monitoring system, characterized in that, It includes a processor and a memory, the memory storing a computer program, and the processor executing the program to implement the method for monitoring workpiece processing as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method for monitoring workpiece processing as described in any one of claims 1 to 8.

Images