Method and system for monitoring abnormal running state of high-precision numerical control machine tool

Abstract

The invention relates to the technical field of data processing, in particular to a high-precision numerical control machine tool operation state abnormity monitoring method and system, and the method comprises the steps: determining a plurality of initial clustering clusters when a pre-obtained target sequence is clustered through an ISODATA algorithm, and carrying out the clustering of a plurality of initial clustering clusters during the subsequent iteration clustering; taking the reliability of each sampling point in each initial clustering cluster as a weight, carrying out weighted summation on the positions of all the sampling points in the corresponding initial clustering cluster to obtain a plurality of weighted cluster centers during current clustering until a clustering result converges to obtain a plurality of clustering clusters, and carrying out clustering on the plurality of clustering clusters; and according to a comparison result of the normalized value of the distance between each sampling point in each clustering cluster and the clustering center of the corresponding clustering cluster and a preset threshold value, monitoring the abnormal operation state of the numerical control machine tool. The accuracy of monitoring the abnormal operation state of the numerical control machine tool can be improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a method and system for monitoring abnormal operation status of a high-precision CNC machine tool. Background Art

[0002] As automated equipment with a program control system, CNC machine tools play an important role in the manufacturing industry and can achieve high-precision and high-efficiency parts processing. When the operating status of a CNC machine tool is abnormal, the related operating data will also be abnormal. Therefore, by monitoring the status of the operating data, the abnormal operating status of the CNC machine tool can be effectively identified, so that potential problems can be discovered in time to ensure the stability of the production process and product quality.

[0003] In the related technology, a method and system for monitoring the machining state of a CNC machine tool based on feedback data is disclosed in a patent application document with publication number CN117131425A. The method includes: obtaining a fault data set based on the basic information of the CNC machine tool, simplifying the features of the fault samples, and classifying them using the simplified features to obtain a low-dimensional fault feature set. Screening multi-dimensional feature parameters under different working conditions to generate fault feedback data containing working condition features. Constructing a machining state monitoring model to identify the current working condition of the CNC machine tool and make a fault decision. Finally, tracing the root cause of the fault based on the fault decision result, and determining the operation and maintenance direction of the CNC machine tool.

[0004] However, when extracting low-dimensional fault feature sets, the above scheme needs to first obtain the fault data set, and then simplify and classify its features. In addition, a large amount of annotation information is required in the classification process, which is time-consuming, lacks real-time performance, and does not take into account the influence of noise data. As a result, it may be impossible to accurately monitor the abnormal state of CNC machine tools to a certain extent. Summary of the invention

[0005] In order to solve the problem that the abnormal state of a CNC machine tool cannot be accurately monitored, the present invention provides a method and system for monitoring the abnormal operation state of a CNC machine tool with high precision.

[0006] According to a first aspect of the present invention, there is provided a method for monitoring abnormal operation status of a high-precision CNC machine tool, comprising: Obtain various operating data of the CNC machine tool at each sampling point in the current time period to obtain the target sequence; The target sequence is iteratively clustered using the ISODATA algorithm. In each clustering process, the reliability of the sampling points in each cluster is used as the weight, and the positions of all sampling points in the corresponding cluster are weighted and summed to obtain multiple weighted cluster centers for the current clustering, until the positions of all clusters converge to obtain multiple target clusters. According to the normalized value of the distance between each sampling point in each target cluster and the weighted cluster center of the corresponding target cluster, and the comparison result of the preset threshold, the abnormal operation state of the CNC machine tool is monitored; Among them, reliability , is the degree of abnormal fluctuation of the target sequence, which reflects the abnormal fluctuation of the operating data in the target sequence; The target sequence The abnormal degree of change of a sampling point reflects the abnormality of the change trend of the operating data of the sampling point.

[0007] The method of clustering the target sequence of the present invention can effectively distinguish the different operating states of the CNC machine tool when it is working, and in the process of clustering, the weight of the sampling points with lower reliability can be reduced, thereby ensuring the accuracy of the clustering center determined in each iterative clustering process, thereby ensuring the accuracy of the obtained clustering results, and then based on the clustering results with higher accuracy, the abnormal operating state of the CNC machine tool can be accurately identified.

[0008] Preferably, the method for obtaining the degree of abnormal fluctuation of the target sequence includes: Acquire multiple operating data of each sampling point of the CNC machine tool in multiple historical time periods to obtain multiple reference sequences, the length of the reference sequence is the same as the length of the target sequence, and the category of the operating data in the reference sequence is the same as the category of the operating data in the target sequence; Calculate the variance of each running data in the target sequence and the difference between it and the variance of the corresponding running data in all reference sequences, and take the mean of the normalized differences as the degree of fluctuation anomaly of the target sequence.

[0009] Preferably, the degree of fluctuation anomaly satisfies the following relationship: ; In the formula, is the degree of abnormal fluctuation of the target sequence; The target sequence The variance of the running data; For the The reference sequence The variance of the running data; is the absolute value function; is the normalization function; is the number of reference sequences; is the number of types of operating data; is the summation symbol.

[0010] The present invention limits the range of the fluctuation abnormality degree of the target sequence to 0-1, which can reduce the difficulty of calculation.

[0011] Preferably, the abnormal degree of change satisfies the following relationship: ; In the formula, The target sequence The degree of abnormality of changes at each sampling point; The target sequence The sampling point is the same as the previous sampling point. The difference of the running data; For the The first The sampling point is the same as the previous sampling point. The difference of the running data; The first Type of operation data, The first The absolute value of the Pearson correlation coefficient between the running data, is the normalization function, is the number of reference sequences, is the number of types of operating data, is the absolute value symbol; is the summation symbol.

[0012] The present invention determines the abnormal degree of changes of each sampling point in the target sequence by integrating various data, and can ensure the accuracy of the determined abnormal degree of changes.

[0013] Preferably, the method further comprises: use Function, which maps the reliability of the sampling points of each cluster in the clustering process to a value whose sum is 1.

[0014] The present invention can reduce the amount of calculation required for determining weighted cluster centers during an iterative clustering process.

[0015] Preferably, monitoring the abnormal operation state of the CNC machine tool according to the normalized value of the distance between each sampling point in each target cluster and the weighted cluster center of the corresponding target cluster and the comparison result of the preset threshold value includes: Get the preset threshold value; When the normalized value of the distance between any sampling point in any cluster and the cluster center of any cluster is greater than a preset threshold, it is determined that the operating data of any sampling point is abnormal.

[0016] The present invention can accurately monitor the abnormal operation state of a numerically controlled machine tool.

[0017] According to the second aspect of the present invention, a high-precision CNC machine tool operating status abnormality monitoring system is provided. The high-precision CNC machine tool operating status abnormality monitoring system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the first aspect of the present invention.

[0018] The present invention has the following effects: When the present invention utilizes the ISODATA algorithm to iteratively cluster the target sequence, the reliability of the sampling points is used as a weight, which can reduce the influence of noise, so that in each iterative clustering, the determined weighted cluster center can be close to the actual cluster center, thereby ensuring the accuracy of the clustering result, and then based on the clustering result with high accuracy, the abnormal sampling points can be identified to monitor the abnormal operation state of the numerical control machine tool; and when determining the reliability of the sampling points, the abnormal situation of the change trend of the operation data of the sampling points and the abnormality of the fluctuation of the operation data in the target sequence are combined, so that the influence of noise can be considered from many aspects, and the accuracy of the determined reliability is guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 The present invention is a flowchart of a method for monitoring abnormal operation status of a high-precision CNC machine tool according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0021] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0022] Reference Figure 1 , a method for monitoring abnormal operation status of a high-precision CNC machine tool, comprising steps S1 to S3, specifically as follows: S1: Obtain various operating data of the CNC machine tool at each sampling point in the current time period to obtain the target sequence.

[0023] In an exemplary embodiment of the present invention, the current time period refers to a time period consisting of the current moment and several previous moments; the operating data refers to data that can affect the performance of the CNC machine tool; and the target sequence refers to a collection of multiple operating data collected at various moments within the current time period.

[0024] In an exemplary embodiment of the present invention, the operation data includes spindle speed data, spindle vibration data, spindle temperature data, feed speed data, feed displacement data, guide rail temperature data, and guide rail vibration data.

[0025] Among them, the feed speed data refers to the moving speed of the workbench or tool in the feed direction; the feed displacement data refers to the moving distance of the workbench or tool in the feed direction. Specifically, an encoder can be installed on the motor shaft or lead screw of the feed system to obtain the feed speed data by measuring the rotation angle of the motor or the rotation angle and pitch of the lead screw; a laser displacement sensor can be installed at a fixed position on the machine tool to obtain the feed displacement data by emitting a laser beam and measuring the intensity of the reflected light.

[0026] A magnetoelectric sensor can be fixedly installed on the spindle box housing near the spindle position to obtain the spindle speed data by measuring the frequency of the induced electromotive force; an acceleration sensor can be arranged above the spindle to obtain the sum of the absolute values ​​of the vibration acceleration of the spindle in three directions as the spindle vibration data; a temperature sensor can be installed in the spindle box to collect the spindle temperature data.

[0027] A temperature sensor may be arranged on the surface of the guide rail to obtain the temperature data of the guide rail; an acceleration sensor may be arranged on the surface of the guide rail to obtain the vibration data of the guide rail.

[0028] Furthermore, the sampling frequency of all sensors can be set to 10 Hz to collect the operation data of the CNC machine tool in real time, and the spindle speed data, spindle vibration data, spindle temperature data, feed speed data, feed displacement data, guide rail temperature data and guide rail vibration data collected at each moment in the current time period are used as the target sequence. This embodiment does not specifically limit the length of the target sequence.

[0029] S2: Use the ISODATA algorithm to iteratively cluster the target sequence. In each clustering process, the reliability of the sampling points in each cluster is used as the weight, and the positions of all sampling points in the corresponding cluster are weighted and summed to obtain multiple weighted cluster centers for the current clustering, until the positions of all clusters converge to obtain multiple target clusters.

[0030] It should be noted that when performing abnormal monitoring on the operating status of CNC machine tools, it is difficult to determine how many different operating states exist in the data due to the changeable states of CNC machine tools during operation. ISODATA (Iterative Self-organizing Data Analysis Techniques Algorithm) can automatically adjust the number of clusters according to the distribution of data without pre-specifying the number of clusters. Therefore, using this algorithm to cluster the target sequence can effectively distinguish the different operating states of CNC machine tools during operation.

[0031] It should be further explained that various operation data in the target sequence may be affected by noise during the collection process, resulting in a certain deviation between the obtained cluster center and the actual cluster center during the iterative clustering of the target sequence using the algorithm, making the clustering result less accurate. Therefore, the present invention improves the way the algorithm determines the cluster center during each iterative clustering to ensure that the obtained weighted cluster center is close to the actual cluster center, thereby improving the accuracy of the clustering result. The present invention only improves the way the algorithm determines the cluster center during the iterative clustering process, and does not improve other processes of the algorithm.

[0032] In an exemplary embodiment of the present invention, reliability refers to data that measures the degree to which the sampling points in each cluster are affected by noise, and is used as a weight to reduce the influence of the sampling points that are more likely to be affected by noise on the clustering results. For example, when any sampling point in the target sequence is more likely to be affected by noise, it means that the reliability of the sampling point is low. When the target sequence is iteratively clustered using the ISODATA algorithm, the weight of the sampling point needs to be reduced to obtain the weighted cluster center of the cluster where the sampling point is located, so as to avoid deviation between the obtained cluster center and the actual cluster center, and ensure the accuracy of the clustering results.

[0033] Specifically, calculating the reliability of any sampling point in the target sequence includes the following steps: Step 1: Determine the degree of abnormal fluctuation of the target sequence; In an exemplary embodiment of the present invention, the determination of the degree of abnormal fluctuation of the target sequence can be achieved by the following steps: (1) Obtaining a variety of operating data of each sampling point of a CNC machine tool in multiple historical time periods to obtain multiple reference sequences, the length of the reference sequence is the same as the length of the target sequence, and the category of the operating data in the reference sequence is the same as the category of the operating data in the target sequence; Optionally, a certain number, such as 100 historical time periods, may be selected, and then the spindle speed data, spindle vibration data, spindle temperature data, feed speed data, feed displacement data, guide rail temperature data, and guide rail vibration data collected at each moment in the selected historical time period may be used as a reference sequence, thereby obtaining multiple reference sequences. This embodiment does not specifically limit the length of the reference sequence, as long as it is the same as the length of the target sequence.

[0034] (2) Calculate the variance of each running data in the target sequence and the difference between it and the variance of the corresponding running data in all reference sequences, and take the mean of the normalized differences as the degree of fluctuation anomaly of the target sequence.

[0035] Specifically, the degree of fluctuation anomaly of the target sequence satisfies the following relationship: ; In the formula, is the degree of abnormal fluctuation of the target sequence; The target sequence The variance of the running data; For the The reference sequence The variance of the running data; is the absolute value function; is the normalization function; is the number of reference sequences; is the number of types of operating data; is the summation symbol.

[0036] in, Reflects the target sequence The variance of the running data is The reference sequence The difference between the variances of the two operating data. The larger the value, the more abnormal the fluctuation of the operating data in the target sequence, and the greater the abnormal degree of fluctuation of the corresponding target sequence.

[0037] In another embodiment, other methods may be used for normalization, such as normalization through a deformed exponential function. This embodiment does not specifically limit the normalization method used.

[0038] Step 2: Determine the degree of abnormality of changes in each sampling point in the target sequence.

[0039] Specifically, the abnormal degree of change of any sampling point in the target sequence satisfies the following relationship: ; In the formula, The target sequence The degree of abnormality of changes at each sampling point; The target sequence The sampling point is the same as the previous sampling point. The difference of the running data; For the The first The sampling point is the same as the previous sampling point. The difference of the running data; The first Type of operation data, The first The absolute value of the Pearson correlation coefficient between the running data, is the normalization function, is the number of reference sequences, is the number of types of operating data, is the absolute value symbol; is the summation symbol.

[0040] in, Reflects the first The operating data is The variation trend of the sampling points is similar to that of The first The operating data is The difference in the change trend of the sampling points. The larger the value, the greater the difference in the change trend of the sampling point. This type of operation data is greatly affected by noise, and when the first The running data is the same as the reference sequence. When the similarity between the two types of running data is large, the first This type of operation data is relatively more affected by noise.

[0041] Step 3: Calculate the reliability of any sampling point in the target sequence based on the abnormal degree of change of the sampling point and the abnormal degree of fluctuation of the target sequence.

[0042] Specifically, the reliability of any sampling point in the target sequence satisfies the following relationship: ; In the formula, The first The reliability of each sampling point; is the degree of abnormal fluctuation of the target sequence; The target sequence The abnormal degree of change of each sampling point.

[0043] The specific process of iterative clustering of the target sequence using the ISODATA algorithm is as follows: first, a plurality of sampling points, such as 5, are randomly selected from the target sequence as initial cluster centers, and then the Manhattan distance between each sampling point in the target sequence and each initial cluster center is determined, and each sampling point is divided around the initial cluster center with the smallest Manhattan distance, so as to obtain a plurality of initial cluster clusters. It should be noted that in order to avoid the distribution of each type of operation data in the target sequence being too sparse in space, this embodiment uses the Manhattan distance when measuring the distance between each sampling point and each initial cluster center. Of course, a suitable measurement method can also be selected according to the specific situation, and this embodiment is not particularly limited here.

[0044] Then, in the second clustering process, using Function, respectively maps the reliability of all sampling points in each initial cluster to a value equal to 1, then uses the reliability of each sampling point in each initial cluster as a weight, performs weighted summation on the positions of all sampling points in the corresponding initial cluster, and obtains multiple weighted cluster centers for the second clustering, and so on, until the positions of all clusters converge to obtain multiple target clusters.

[0045] For example, in any clustering process, if any cluster contains sample point 1 (0, 1, 1) and sample point 2 (2, 0, 1), and the weights of sample point 1 and sample point 2 are 0.2 and 0.8 respectively, then the weighted cluster center of the cluster is: (0×0.2+2×0.8, 1×0.2+0×0.8, 1×0.2+1×0.8), that is, (1.6, 0.2, 1). It should be noted that the process of iteratively clustering data using the ISODATA algorithm to obtain multiple clusters is a prior art, and this implementation will not be described in detail here.

[0046] Optionally, the reliability is used to determine the weighted cluster center in each clustering process, which can reduce the influence of noise data, thereby improving the accuracy of the obtained clustering results, and further accurately identifying the abnormal operating state of the CNC machine tool.

[0047] S3: monitoring the abnormal operation state of the CNC machine tool according to the normalized value of the distance between each sampling point in each target cluster and the weighted cluster center of the corresponding target cluster and the comparison result of the preset threshold.

[0048] In an exemplary embodiment of the present invention, the abnormal operation state of a numerically controlled machine tool can be monitored by the following steps: Obtain a preset threshold; when the normalized value of the distance between any sampling point in any target cluster and the weighted cluster center of any target cluster is greater than the preset threshold, it is determined that the operating data of any sampling point is abnormal.

[0049] Optionally, the distance between any sampling point in any target cluster and the weighted cluster center of the target cluster may be Manhattan distance or Euclidean distance. This embodiment does not specifically limit the distance type adopted.

[0050] Optionally, the threshold can be set to 0.8 based on empirical values. When the normalized value of the distance between any sampling point in any target cluster and the weighted cluster center of the target cluster is greater than 0.8, it is determined that there is an abnormality in the operating data of the sampling point, and the sampling point can be marked as abnormal, thereby completing the abnormal monitoring of the operating status of the CNC machine tool.

[0051] The present invention also provides a high-precision CNC machine tool operating status abnormality monitoring system, the system includes a memory and a processor, and a computer program is stored in the memory, the computer program integrates the function of a high-precision CNC machine tool operating status abnormality monitoring method, when the computer program is executed, the accuracy of abnormal operating status monitoring of the CNC machine tool can be improved through a high-precision CNC machine tool operating status abnormality monitoring method.

[0052] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0053] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for monitoring abnormal operation status of a high-precision CNC machine tool, characterized in that: include: Obtain various operating data of the CNC machine tool at each sampling point in the current time period to obtain the target sequence; The target sequence is iteratively clustered using the ISODATA algorithm. In each clustering process, the reliability of the sampling points in each cluster is used as the weight, and the positions of all sampling points in the corresponding cluster are weighted and summed to obtain multiple weighted cluster centers for the current clustering, until the positions of all clusters converge to obtain multiple target clusters; Monitoring the abnormal operation state of the CNC machine tool according to the normalized value of the distance between each sampling point in each of the target clusters and the weighted cluster center of the corresponding target cluster and the comparison result of the preset threshold value; Among them, the reliability , is the degree of abnormal fluctuation of the target sequence, which reflects the abnormal fluctuation of the operating data in the target sequence; The target sequence The abnormal degree of change of a sampling point reflects the abnormality of the change trend of the operation data of the sampling point.

2. A method for monitoring abnormal operation status of a high-precision CNC machine tool according to claim 1, characterized in that: The method for obtaining the degree of abnormal fluctuation of the target sequence includes: Acquire multiple operating data of each sampling point of the CNC machine tool in multiple historical time periods to obtain multiple reference sequences, wherein the length of the reference sequence is the same as the length of the target sequence, and the category of the operating data in the reference sequence is the same as the category of the operating data in the target sequence; The variance of each running data in the target sequence is calculated, and the difference between the variance of the corresponding running data in all reference sequences is calculated, and the mean of the normalized differences is used as the degree of abnormal fluctuation of the target sequence.

3. A method for monitoring abnormal operation status of a high-precision CNC machine tool according to claim 2, characterized in that: The degree of abnormal fluctuation satisfies the following relationship: ； In the formula, is the degree of abnormal fluctuation of the target sequence; The target sequence The variance of the running data; For the The reference sequence The variance of the running data; is the absolute value function; is the normalization function; is the number of reference sequences; is the number of types of operating data; is the summation symbol.

4. A method for monitoring abnormal operation status of a high-precision CNC machine tool according to claim 2, characterized in that: The abnormal degree of the change satisfies the following relationship: ； In the formula, The target sequence The degree of abnormality of changes at each sampling point; The target sequence The sampling point is the same as the previous sampling point. The difference of the running data; For the The first The sampling point is the same as the previous sampling point. The difference of the running data; The first Type of operation data, The first The absolute value of the Pearson correlation coefficient between the running data, is the normalization function, is the number of reference sequences, is the number of types of operating data, is the absolute value symbol; is the summation symbol.

5. The method for monitoring abnormal operation status of a high-precision CNC machine tool according to claim 1, characterized in that: The method further comprises: use Function, which maps the reliability of the sampling points of each cluster in the clustering process to a value whose sum is 1.

6. A method for monitoring abnormal operation status of a high-precision CNC machine tool according to claim 1, characterized in that: The monitoring of the abnormal operation state of the CNC machine tool according to the comparison result of the normalized value of the distance between each sampling point in each target cluster and the weighted cluster center of the corresponding target cluster and the preset threshold comprises: Get the preset threshold value; When the normalized value of the distance between any sampling point in any target cluster and the weighted cluster center of any target cluster is greater than the preset threshold, it is determined that the operating data of any sampling point is abnormal.

7. The method for monitoring abnormal operation status of a high-precision CNC machine tool according to claim 1, characterized in that: The operation data includes spindle speed data, spindle vibration data, spindle temperature data, feed speed data, feed displacement data, guide rail temperature data and guide rail vibration data.

8. A high-precision CNC machine tool operating status abnormality monitoring system, characterized in that: The high-precision CNC machine tool operating status abnormality monitoring system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the method as described in any one of claims 1-7.

Images