Workpiece machining data processing method and system

Abstract

The present application relates to the technical field of data processing, and more particularly to a workpiece processing data processing method and system. The method comprises the steps of: obtaining data points in a historical running data set of workpiece processing and cluster centers of each cluster; obtaining an abnormality index of the cluster by a ratio of the number of data points in the cluster to the number of data points in the largest cluster, and obtaining a suspicious degree of the cluster by the Euclidean distance between each data point in the cluster and the cluster center; taking a range value of the data point in the cluster corresponding to the collection time as the credibility of the suspicious degree to obtain a feature weight of the cluster; taking a ratio of the local density of the data point to the feature weight of the cluster where the data point is located as the weighted local density of the data point, calculating an outlier factor of the data point, and realizing workpiece processing data processing by comparing the outlier factor with a preset threshold, thereby effectively improving the accuracy of workpiece processing anomaly monitoring.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for processing workpiece machining data. Background Technology

[0002] The workpieces are used in various manufacturing industries such as wind power, chemical industry, and machinery, and their quality plays a crucial role in the performance and service life of related equipment. During the forging process, after the metal material is heated, it is plastically deformed under fixed pressure using a specialized press, producing workpieces with high strength and high toughness. To ensure workpiece quality, it is necessary to monitor the workpiece processing process.

[0003] Most existing technologies analyze processing parameters collected during the production process of workpieces to obtain monitoring results. For example, patent application CN117495211A discloses a method for predicting the quality of machined workpieces in industrial mother machines based on adaptive cycle detection. This method involves obtaining a sample dataset; randomly dividing the sample dataset into a training set, a validation set, and a test set according to a set ratio; constructing a surface roughness prediction model for the machined workpiece; training the surface roughness prediction model using the training set; and using the trained surface roughness prediction model to obtain the predicted value of the surface roughness of the machined workpiece, thereby achieving the prediction of workpiece quality.

[0004] While the aforementioned existing technologies can improve the efficiency of obtaining workpiece quality results to a certain extent, in actual processing, factors such as internal friction of the machine tool and loosening of parts may cause the hydraulic system to generate continuous abnormal processing parameter data. Such abnormal data will affect the accuracy of workpiece quality prediction.

[0005] Therefore, how to accurately monitor data during the workpiece processing is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] To address the technical problem of accurately monitoring data during workpiece processing, this invention provides a workpiece processing data processing method and system.

[0007] In a first aspect, the present invention provides a workpiece machining data processing method, which adopts the following technical solution:

[0008] A method for processing workpiece machining data includes the following steps:

[0009] A historical operation dataset of workpiece processing is obtained, which contains several data points composed of processing parameters. An iterative self-organizing clustering algorithm is used to obtain the cluster centers of each cluster in the historical operation dataset. The anomaly index of a cluster is obtained by subtracting the ratio of the number of data points in a cluster to the number of data points in the largest cluster from 1. The normalized value of the product of the standard deviation of the Euclidean distance between each data point in the cluster and the cluster center and the anomaly index is recorded as the suspicion level of the cluster. The range of the data points at the corresponding acquisition times in the cluster is used as the confidence level of the suspicion level to obtain the feature weight of the cluster. The local outlier factor algorithm is used to determine the local density of each data point. The ratio of the local density of a data point to the feature weight of the cluster in which the data point belongs is recorded as the weighted local density of the data point. The outlier factor of the data point is obtained through the weighted local density of the data point, and the workpiece processing data is processed by comparing the outlier factor with a preset threshold.

[0010] This invention utilizes a local outlier factor algorithm to obtain anomaly monitoring results in workpiece processing. In this process, the invention considers that factors such as internal machine tool friction and loose parts may cause continuous abnormal processing parameter data in the hydraulic system, which the local outlier factor algorithm might mistake for normal data. Therefore, this invention iteratively clusters historical operating data, determining the feature weights by obtaining the time span of data points within each cluster and the degree of suspicion of the cluster. The local density obtained by the local outlier factor algorithm is then corrected using these feature weights, accurately identifying abnormal data in the operating dataset and effectively improving the accuracy of workpiece processing anomaly monitoring.

[0011] According to a workpiece processing data processing method provided by the present invention, the method for obtaining the historical operation dataset of workpiece processing further includes: setting a preset acquisition frequency, taking the value of each processing parameter obtained at each acquisition moment as a data point, and arranging all data points in chronological order after preprocessing to obtain the historical operation dataset.

[0012] This invention can effectively improve the overall quality of workpiece processing data by preprocessing it, thus preparing the data for subsequent processing.

[0013] According to a workpiece processing data processing method provided by the present invention, the step of obtaining the cluster centers of each cluster in the historical running dataset using an iterative self-organizing clustering algorithm includes: obtaining initial clusters in the historical running dataset using an iterative self-organizing clustering algorithm; obtaining the probability that the latest data point belongs to each initial cluster by the degree of deviation between the processing parameters corresponding to the latest data point and the processing parameters corresponding to the cluster centers of the initial clusters; taking the initial cluster corresponding to the maximum probability that the latest data point belongs to each cluster as the cluster in which the data point is located; and determining the cluster centers of each cluster.

[0014] This invention takes into account the discrepancy between normal and abnormal data in workpiece processing data, although their local density may be similar. Therefore, by dividing the dataset into multiple initial clusters, and further dividing the latest data points according to the degree of deviation between the latest data points and the initial clusters, clusters are obtained, and the data characteristics in each cluster are analyzed to obtain the degree of abnormality.

[0015] According to a workpiece machining data processing method provided by the present invention, the probability of the latest data point belonging to each initial cluster satisfies the following relationship:

[0016] ;

[0017] In the formula, This indicates that the i-th latest data point belongs to the i-th latest data point. The probability of an initial cluster. This indicates the number of processing parameters corresponding to the data point. This represents the value of the j-th processing parameter in the i-th latest data point. Indicates the first The value of the j-th processing parameter in the cluster centers of the initial clusters. Indicates the first The standard deviation of the j-th processing parameter of all data points in the initial cluster. This represents an exponential function with base e. Represents the absolute value symbol.

[0018] According to a workpiece processing data processing method provided by the present invention, the step of taking the range of the acquisition time corresponding to the data points in the cluster as the confidence level of the suspicion level in order to obtain the feature weight of the cluster includes: taking the normalized value of the ratio of the suspicion level of the cluster to its confidence level as the feature weight of the cluster.

[0019] This invention takes into account that anomalies generated during actual processing manifest as continuous anomalous data in the dataset. Therefore, the greater the likelihood that continuous anomalous data generated within a shorter time span is generated by the system itself, which manifests as a concentrated occurrence within a short period in clusters. By combining the time span of clusters with the degree of suspicion, the feature weights of clusters can be accurately obtained.

[0020] According to a workpiece processing data processing method provided by the present invention, the step of obtaining the outlier factor of a data point through the weighted local density of the data point includes: taking the ratio of the average weighted local density of the data point and the weighted local density of the data point to the average weighted local density of the data point as the outlier factor of the data point.

[0021] According to a workpiece processing data processing method provided by the present invention, the workpiece processing data processing is realized by comparing the outlier factor with a preset threshold, including: setting a preset threshold; identifying data points with outliers greater than the preset threshold as abnormal data points; and identifying data points with outliers not greater than the preset threshold as normal data points.

[0022] This invention obtains abnormal data during the workpiece processing by calculating outlier factors, which can accurately process processing data and thus accurately obtain monitoring results of data points.

[0023] Secondly, the present invention provides a workpiece machining data processing system, which adopts the following technical solution:

[0024] A workpiece machining data processing system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the aforementioned workpiece machining data processing method is implemented.

[0025] By adopting the above technical solution, a computer program is generated from the above workpiece processing data processing method and stored in a memory so that it can be loaded and executed by a processor. A terminal device can then be made based on the memory and processor for convenient use.

[0026] The present invention has the following technical effects:

[0027] Based on the above technical solution, when processing workpiece processing data, the present invention can obtain the abnormal monitoring results of workpiece processing through the local outlier factor algorithm. In this process, the present invention iteratively clusters historical operating data, determines its feature weight by obtaining the time span of data points in the cluster and the degree of suspicion of the cluster, and corrects the local density obtained by the local outlier factor algorithm through the feature weight. This can accurately identify abnormal data in the operating dataset, reduce the impact of continuous abnormal processing parameter data generated by the hydraulic system on the monitoring results, and thus effectively improve the accuracy of workpiece processing abnormal monitoring. Attached Figure Description

[0028] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts.

[0029] Figure 1 This is a flowchart illustrating a workpiece machining data processing method provided in an embodiment of the present invention. Detailed Implementation

[0030] 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, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] It should be understood that when the terms "first," "second," etc., are used in the claims, specification, and drawings of this invention, they are only used to distinguish different objects and not to describe a specific order. The terms "comprising" and "including" used in the specification and claims of this invention indicate the presence of the described features, integrals, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof.

[0032] It should be noted that these workpieces are used in various manufacturing industries, including wind power, chemical engineering, and machinery, and their quality plays a crucial role in the performance and service life of related equipment. During the forging process, after the metal material is heated, it is plastically deformed under fixed pressure using a specialized press, producing workpieces with high strength and high toughness. To ensure workpiece quality, it is necessary to monitor any anomalies during the workpiece processing.

[0033] However, in actual processing, due to factors such as internal friction of the machine tool and loosening of parts, the hydraulic system may generate continuous abnormal processing parameter data. The local density of such abnormal processing parameter data is close to that of normal data. When processing, the conventional local outlier factor algorithm may treat abnormal data as normal data because the local density of abnormal data points is not significantly different from the local density of other data points in its neighborhood, thus affecting the accuracy of workpiece processing data processing.

[0034] Based on this, this invention discloses a workpiece machining data processing method. By acquiring the anomaly degree of clusters in the dataset and evaluating the feature weights of data points within the clusters based on the time span, the local density of data points is weighted according to these feature weights. This allows for accurate screening of abnormal data and improves the accuracy of workpiece machining data processing. See details below. Figure 1 As shown, Figure 1 This is a flowchart illustrating a workpiece machining data processing method provided in an embodiment of the present invention. The method specifically includes the following steps S1-S5.

[0035] S1: Obtain the historical operation dataset of workpiece processing, and use the iterative self-organizing clustering algorithm to obtain the cluster center of each cluster in the historical operation dataset.

[0036] The processing parameters can include pressure, speed, etc., and can be set according to actual needs.

[0037] For example, in this embodiment of the invention, before obtaining the processing parameters corresponding to each data point in the historical operation dataset of workpiece processing, the method further includes: setting a preset acquisition frequency, taking the value of each processing parameter obtained at each acquisition moment as a data point, and arranging all data points in chronological order after preprocessing to obtain the historical operation dataset.

[0038] The sampling frequency can be preset to once per second, and can be set according to actual needs. This embodiment of the invention does not impose too many restrictions here.

[0039] For example, data preprocessing can include data cleaning, data format conversion, etc., and the specific settings can be configured according to actual needs. This embodiment of the invention does not impose too many restrictions here.

[0040] It should be noted that during the forging process, the machining parameters are generally relatively stable under normal circumstances. However, due to prolonged use of the forging press, problems such as internal friction and loosening of parts may gradually emerge, affecting the normal operation of the hydraulic system and causing continuous anomalies in the monitored machining parameters. These anomalies typically show some deviation from normal conditions, but exhibit similar characteristics in their localized areas.

[0041] Based on this, the embodiments of the present invention use an iterative self-organizing clustering algorithm to iteratively cluster historical running data, obtain different feature patterns in the distribution of historical data, compare the clustering results after each iteration, and analyze the optimal clustering result.

[0042] For example, in an embodiment of the present invention, obtaining the cluster centers of each cluster in a historical running dataset using an iterative self-organizing clustering algorithm includes: obtaining initial clusters in the historical running dataset using an iterative self-organizing clustering algorithm; obtaining the probability that the latest data point belongs to each initial cluster by the degree of deviation between the processing parameters corresponding to the latest data point and the processing parameters corresponding to the cluster centers of the initial clusters; taking the initial cluster corresponding to the maximum probability that the latest data point belongs to each cluster as the cluster in which the data point is located; and determining the cluster centers of each cluster.

[0043] The specific steps for obtaining the initial clusters in the historical running dataset using the iterative self-organizing clustering algorithm can be implemented using existing technologies, and will not be elaborated here in this embodiment of the invention.

[0044] It is understandable that the cluster center of a cluster is also a data point, and therefore it has corresponding values ​​for each processing parameter.

[0045] For example, in an embodiment of the present invention, the probability of the latest data point belonging to each initial cluster is determined, as can be seen from the following relationship:

[0046] ;

[0047] In the formula, This indicates that the i-th latest data point belongs to the i-th latest data point. The probability of an initial cluster. This indicates the number of processing parameters corresponding to the data point. This represents the value of the j-th processing parameter in the i-th latest data point. Indicates the first The value of the j-th processing parameter in the cluster centers of the initial clusters. Indicates the first The standard deviation of the j-th processing parameter of all data points in the initial cluster. This represents an exponential function with base e. Represents the absolute value symbol.

[0048] In the above formula, Indicates the relationship between the i-th latest data point and the i-th latest data point. The cumulative sum of differences in processing parameters among the cluster centers of the initial clusters; the larger this value, the better the difference between the i-th latest data point and the i-th latest data point. The greater the difference between the cluster centers of the initial clusters, the farther the i-th latest data point is from the i-th cluster. The cluster center of the initial cluster, corresponding to the data point belonging to the first cluster. The smaller the probability of the initial cluster, the better; conversely, the smaller the value, the less likely the i-th latest data point is to be clustered into the i-th cluster. The smaller the difference between the cluster centers of the initial clusters, the closer the i-th latest data point is to the i-th cluster. The cluster center of the initial cluster, corresponding to the data point belonging to the first cluster. The greater the likelihood of an initial cluster, the higher the probability of it forming a cluster.

[0049] Based on the above steps, data points with high similarity are divided into the same cluster, and the clusters are updated in real time according to the probability that the latest data point belongs to each initial cluster. This allows for the rapid identification of patterns and structures in the data as the latest data is continuously input, reflecting the latest characteristics of the data. Based on this clustering result, outliers or isolated points that are significantly different from most data points can be identified, and then the following steps can be performed.

[0050] S2: Subtract the ratio of the number of data points in a cluster to the number of data points in the largest cluster from 1 to obtain the anomaly index of the cluster. The normalized value of the product of the standard deviation of the Euclidean distance between each data point in the cluster and the cluster center and the anomaly index is recorded as the degree of suspicion of the cluster.

[0051] After obtaining the clusters based on the above steps, clusters that may contain anomalies can be identified based on the characteristics of the data points within each cluster. During the forging process, to ensure the workpiece meets relevant standards, corresponding processing parameters are set according to the material standards, resulting in a large amount of normal data generated during production, and the processing parameters for this normal data are relatively stable. If a problem occurs inside the equipment affecting the normal operation of the hydraulic system, continuous abnormal data may be generated, and its distribution within the dataset may be unstable.

[0052] Based on this, embodiments of the present invention determine the degree of anomaly of clusters by obtaining the relative size of each cluster and the standard deviation of data in the clusters.

[0053] It is understandable that the size of a cluster is the number of data points contained in the cluster. The more data points a cluster contains, the larger the cluster is, and the largest cluster is the cluster containing the most data points.

[0054] For example, in an embodiment of the present invention, the degree of suspicion of a cluster is determined, specifically by the following relational expression:

[0055] ;

[0056] In the formula, Indicates the first The degree of suspicion of each cluster, Indicates the first The number of data points in each cluster This indicates the number of data points in the largest cluster. Indicates the first The cluster center of the first cluster is related to the first cluster. In the cluster, the th Euclidean distance between data points Indicates the first The cluster center of the first cluster is related to the first cluster. Mean Euclidean distance between all data points in each cluster This represents the normalization function.

[0057] In the above formula, This indicates anomalies in clustering. The smaller the cluster size in the clustering results, the greater the likelihood that the data within that cluster is anomalously discrete. Therefore, the... The smaller the ratio of the number of data points in the i-th cluster to the number of data points in the largest cluster, the better. The fewer data points a cluster has, the greater the anomaly index of that cluster.

[0058] Indicates the first The standard deviation of the Euclidean distance between each data point in a cluster and the cluster center is a key indicator. The larger this value is, the greater the difference in Euclidean distance between the data points in the cluster, and the more spatially dispersed the data points in the cluster are.

[0059] In summary, the fewer the number of data points in a cluster and the more discrete the spatial distribution of the data points in the cluster, the greater the degree of anomaly of the data points in the cluster, and the greater the suspicion level of the corresponding cluster.

[0060] S3: Use the range of the data points in the cluster at the corresponding collection time as the confidence level of the degree of suspicion to obtain the feature weight of the cluster.

[0061] It should be noted that the above steps cluster the data by using the processing parameters of each data point as samples, thereby obtaining the degree of suspicion of the clusters. However, there is a time span between the sample processing parameter data. Anomalies generated during the actual processing are manifested as continuous anomalous data in the dataset. Therefore, the anomalous data generated in a shorter time span is more likely to be generated by the system itself, and in the cluster, it will appear as concentrated in a short period of time.

[0062] Based on this, the embodiments of the present invention can more comprehensively evaluate the characteristics of each cluster by further obtaining the time span between data points in the cluster as the credibility of the cluster's suspicion level.

[0063] It is understandable that the range of data points within a cluster represents the time span of that cluster. The smaller the range, the more concentrated the data points within the cluster appear in a short period of time.

[0064] For example, in an embodiment of the present invention, when determining the feature weights of a cluster, the range of the cluster can be obtained by the difference between the maximum and minimum values ​​of the data collection time in the cluster, and the range can be used as the confidence level of the degree of suspicion; the normalized value of the ratio of the degree of suspicion of the cluster to its confidence level can be used as the feature weight of the cluster.

[0065] If the time span of data points in a cluster is more concentrated and the cluster is more suspicious, it indicates that the cluster has generated a highly suspicious data aggregation in a short period of time. When identifying abnormal data based on the local outlier factor algorithm, the data characteristics of the cluster should be given more attention than those of other clusters, and the corresponding feature weights should be higher.

[0066] S4: Use the local outlier factor algorithm to determine the local density of each data point, and record the ratio of the local density of the data point to the feature weight of the cluster in which the data point is located as the weighted local density of the data point.

[0067] It should be noted that by obtaining the data characteristics and time span in the clusters based on the above steps, the abnormal data set generated by the system anomaly can be accurately determined from each cluster. However, when directly determining the monitoring results of data points by the local density of data points, such abnormal data may be regarded as normal data.

[0068] Based on this, the embodiments of the present invention use the feature weights of each cluster obtained by the above steps to perform weighted correction on the local density of data points in the cluster, so that the final weighted local density takes into account both the local density distribution of data points and the characteristics of the cluster in which the data points are located.

[0069] The specific steps for determining the local density of each data point using the local outlier factor algorithm can be implemented using existing technologies, and will not be elaborated here in this embodiment of the invention.

[0070] For example, in an embodiment of the present invention, the weighted local density of data points is determined, and the specific relationship can be found in the following formula:

[0071] ;

[0072] In the formula, This represents the weighted local density of the i-th data point. This represents the local density of the i-th data point. This represents the feature weight of the cluster to which the i-th data point belongs.

[0073] After performing weighted correction on the local density of each data point based on the above steps, abnormal data can be accurately obtained based on the corrected local density.

[0074] S5: Obtain the outlier factor of the data points by weighted local density, and realize workpiece processing data processing by comparing the outlier factor with the preset threshold.

[0075] It should be noted that by using the weighted local density of each data point after weighted correction, the outlier factor of each data point can be accurately obtained, thereby accurately obtaining the continuous abnormal data and regular abnormal data generated by the hydraulic system in the running dataset.

[0076] For example, in an embodiment of the present invention, obtaining the outlier factor of a data point through the weighted local density of the data point includes: using the ratio of the average weighted local density of the data point and the weighted local density of the data point to the average weighted local density of the data point as the outlier factor of the data point.

[0077] Specifically, when obtaining other data points within the k-neighborhood of a data point, a value of k can be preset to determine the Euclidean distance between the data point and other data points in the running dataset, and the k data points with the smallest Euclidean distance can be taken as other data points within the k-neighborhood of that data point.

[0078] The value of k can be set to 5; the specific value can be set according to actual needs, and this embodiment of the invention does not impose too many restrictions here.

[0079] For example, in an embodiment of the present invention, workpiece processing data processing is achieved by comparing the outlier factor with a preset threshold, including: setting a preset threshold; identifying data points with outliers greater than the preset threshold as abnormal data points; and identifying data points with outliers not greater than the preset threshold as normal data points.

[0080] The preset threshold can be set to 1; the specific setting can be adjusted according to actual needs.

[0081] It is understandable that if the outlier factor of a data point is greater than a preset threshold, it means that the weighted local density of the data point is lower than the average weighted local density of other data points in its k-neighborhood, and the data point is an outlier anomalous data point.

[0082] By processing each data point in the workpiece operation dataset based on the above steps, the abnormal monitoring results during the workpiece processing can be accurately obtained.

[0083] As can be seen, in this embodiment of the invention, when processing workpiece processing data, a historical operation dataset of workpiece processing can be obtained. The historical operation dataset contains several data points composed of processing parameters. The cluster center of each cluster in the historical operation dataset is obtained using an iterative self-organizing clustering algorithm. The anomaly index of the cluster is obtained by subtracting the ratio of the number of data points in a cluster to the number of data points in the largest cluster from 1. The normalized value of the product of the standard deviation of the Euclidean distance between each data point in the cluster and the cluster center and the anomaly index is recorded as the suspicion level of the cluster. The range of the collection time corresponding to the data point in the cluster is used as the confidence level of the suspicion level to obtain the feature weight of the cluster. The local outlier factor algorithm is used to determine the local density of each data point. The ratio of the local density of the data point to the feature weight of the cluster in which the data point is located is recorded as the weighted local density of the data point. The outlier factor of the data point is obtained through the weighted local density of the data point. The workpiece processing data is processed by comparing the outlier factor with a preset threshold.

[0084] Thus, this embodiment of the invention can obtain the abnormal monitoring results of workpiece processing through the local outlier factor algorithm. In this process, this embodiment considers that factors such as internal machine tool friction and loose parts may cause the hydraulic system to generate continuous abnormal processing parameter data. The local outlier factor algorithm might identify such continuous abnormal data as normal data. Based on this, this embodiment of the invention iteratively clusters historical operating data, obtains the time span of data points in the clusters and the degree of suspicion of the clusters to determine their feature weights, and corrects the local density obtained by the local outlier factor algorithm using these feature weights. This accurately identifies abnormal data in the operating dataset, thereby effectively improving the accuracy of workpiece processing abnormal monitoring.

[0085] This invention also discloses a workpiece machining data processing system, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a workpiece machining data processing method provided by this invention.

[0086] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0087] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as resistive random access memory (RRAM), dynamic random access memory (DRAM), static random access memory (SRAM), enhanced dynamic random access memory (EDRAM), high-bandwidth memory (HBM), hybrid memory cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device.

[0088] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

[0089] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for processing workpiece machining data, characterized in that, include: Obtain the historical operation dataset of workpiece processing. The historical operation dataset contains several data points composed of processing parameters. Use the iterative self-organizing clustering algorithm to obtain the cluster center of each cluster in the historical operation dataset. The anomaly index of a cluster is obtained by subtracting the ratio of the number of data points in a cluster to the number of data points in the largest cluster from 1. The normalized value of the product of the standard deviation of the Euclidean distance between each data point in the cluster and the cluster center and the anomaly index is recorded as the degree of suspicion of the cluster. The range of the data points in the cluster at the corresponding collection time is used as the confidence level of the degree of suspicion, and the normalized value of the ratio of the degree of suspicion of the cluster to its confidence level is used as the feature weight of the cluster. The local density of each data point is determined using the local outlier factor algorithm. The ratio of the local density of a data point to the feature weight of the cluster in which the data point belongs is recorded as the weighted local density of the data point. The outlier factor of the data points is obtained by weighting the local density of the data points, and the workpiece processing data is processed by comparing the outlier factor with the preset threshold.

2. The workpiece machining data processing method according to claim 1, characterized in that, The process of obtaining the historical operation dataset of workpiece processing also includes, prior to: The preset acquisition frequency is used to treat the value of each processing parameter obtained at each acquisition moment as a data point. After preprocessing all data points, they are arranged in chronological order to obtain the historical operation dataset.

3. The workpiece machining data processing method according to claim 2, characterized in that, The method of obtaining the cluster centers of each cluster in the historical running dataset using the iterative self-organizing clustering algorithm includes: The initial clusters in the historical dataset are obtained by using an iterative self-organizing clustering algorithm. The probability of the latest data point belonging to each initial cluster is obtained by the deviation between the processing parameters corresponding to the latest data point and the processing parameters corresponding to the cluster centers of the initial clusters. The initial cluster corresponding to the maximum probability of the latest data point belonging to each cluster is taken as the cluster to which the data point belongs, and the cluster centers of each cluster are determined.

4. The workpiece machining data processing method according to claim 3, characterized in that, The probability that the latest data point belongs to each of the initial clusters satisfies the following relationship: ； In the formula, This indicates that the i-th latest data point belongs to the i-th latest data point. The probability of an initial cluster. This indicates the number of processing parameters corresponding to the data point. This represents the value of the j-th processing parameter in the i-th latest data point. Indicates the first The value of the j-th processing parameter in the cluster centers of the initial clusters. Indicates the first The standard deviation of the j-th processing parameter of all data points in the initial cluster. This represents an exponential function with base e. Represents the absolute value symbol.

5. The workpiece machining data processing method according to claim 1, characterized in that, The process of obtaining the outlier factor of a data point through the weighted local density of the data points includes: The outlier factor of a data point is the ratio of the weighted local density mean of the data point and the weighted local density of the data point itself.

6. The workpiece machining data processing method according to claim 1, characterized in that, The process of processing workpiece machining data by comparing the outlier factor with a preset threshold includes: Preset threshold; Data points with outliers greater than a preset threshold are considered outliers; data points with outliers less than or equal to the preset threshold are considered normal data points.

7. A workpiece machining data processing system, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a workpiece machining data processing method according to any one of claims 1-6.

Images