Machine tool processing anomaly diagnosis method, system, device and medium based on lof

Abstract

The application discloses a machine tool machining anomaly diagnosis method, system, device and medium based on LOF, and relates to the technical field of machine tool machining monitoring. The method comprises the following steps: collecting position information, program instruction information and vibration signals of each axis of a machine tool; constructing a mapping unit sequence; constructing a density distribution model based on a local outlier factor algorithm; calculating a local outlier factor of a to-be-tested data point; selecting the to-be-tested data point based on the local outlier factor, and obtaining corresponding position information and program instruction information. The application does not need to manually set a fixed threshold value, does not need to rely on fault samples, realizes spatiotemporal fusion of multi-source data and accurate identification of machining anomalies, can accurately locate the machining position and program instruction corresponding to the anomaly, solves the problems that existing diagnosis methods have poor adaptability, are difficult to land, and anomalies are difficult to trace, guarantees precision machining quality, and adapts to machining requirements in the high-end manufacturing field.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing and industrial monitoring technology, specifically relating to a method, system, equipment and medium for diagnosing machine tool processing anomalies based on LOF. Background Technology

[0002] In high-end manufacturing sectors such as aerospace, precision mold making, and automotive parts manufacturing, as product processing precision requirements continue to increase, CNC machine tools, as core processing equipment, directly determine product quality and production efficiency through the stability of their processing. During long-term continuous operation, machine tools are prone to processing abnormalities due to various factors such as mechanical wear, aging electrical components, and fluctuations in cutting parameters. If these abnormalities are not detected and located in a timely manner, it can lead not only to decreased product precision and increased scrap rates, but also, in severe cases, machine tool malfunctions and downtime, causing significant economic losses such as production stoppages and order delays. Therefore, real-time diagnosis and tracing of machine tool processing abnormalities have become a key technical requirement for ensuring precision machining quality.

[0003] Currently, various technical solutions have been developed in the field of machine tool machining anomaly diagnosis, among which the most widely used include traditional threshold monitoring methods and supervised learning-based intelligent diagnostic methods. However, existing technologies still face many insurmountable limitations in practical applications, failing to meet the actual diagnostic needs of high-precision machining. Firstly, traditional threshold monitoring methods lack adaptability. Due to the complexity of CNC machining processes, data varies significantly under different operating conditions. Manually set fixed thresholds often cannot accommodate all machining scenarios. Setting the threshold too high can lead to missed early minor faults, while setting it too low can cause frequent false alarms. Relying on the personal experience of process engineers makes accurate identification of machining anomalies impossible. Secondly, existing diagnostic methods suffer from the problem of separating multi-source data from spatial location. Simply analyzing machine tool acquisition signals without correlating and fusing multi-source information results in operators only receiving vague anomaly signals when the diagnostic system triggers an alarm, unable to determine the specific location and source of the anomaly, hindering accurate source tracing and causing inconvenience for fault diagnosis and process optimization. Meanwhile, supervised learning-based intelligent diagnostic methods require a large number of fault samples for model training, making it difficult for most supervised learning-based diagnostic algorithms to be widely applied in industrial scenarios and failing to solve the cold start problem of diagnostic systems lacking fault samples.

[0004] In current machine tool machining anomaly diagnosis technologies, traditional threshold methods have poor adaptability, supervised learning methods have difficulty in obtaining samples, and all methods suffer from insufficient multi-source data fusion and inability to accurately trace the source of anomalies. It is difficult to achieve accurate identification and spatial positioning of machining anomalies without manual thresholds or fault samples, and thus cannot meet the stringent requirements of high-end manufacturing for machine tool machining quality. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the low diagnostic efficiency of existing machine tool machining anomaly diagnosis methods, thereby providing a machine tool machining anomaly diagnosis method, system, equipment and medium based on LOF.

[0006] A machine tool machining anomaly diagnosis method based on LOF includes the following steps: Collect position information, program instruction information, and vibration signals of each axis of the machine tool; Based on a unified clock, the position information, the program instruction information, and the vibration signal are used to construct a mapping unit sequence; Based on the mapping unit sequence in the normal state, a density distribution model is constructed using the local outlier factor algorithm; The mapping unit sequence of the state to be tested is input into the density distribution model to calculate the local outlier factor of the data point to be tested relative to its neighborhood points. The test data points are selected based on the local outlier factor, and the location information and program instruction information corresponding to the test data points are obtained.

[0007] Furthermore, the density distribution model is a two-level model, and constructing the density distribution model includes the following steps: The historically qualified mapping unit sequence is divided into subsets according to processing stage and spatial region; For each subset, the probability density is estimated using a Gaussian kernel function; Calculate the density mean and standard deviation of each subset, and form a normal density interval based on the density mean and standard deviation; A density distribution model for each subset is constructed based on the probability density and the normal density interval; The density distribution models of each subset are fused using a weighted voting method, with the fusion weight being the product of the sample size ratio and the anomaly detection accuracy.

[0008] Furthermore, the local outlier factor of the data point to be tested relative to its neighborhood points is calculated, including the following steps: Calculate the local reachability density of the data point to be measured and its neighboring points; The local outlier factor of the data point to be tested is calculated based on the local reachability density and expressed as follows: ; in, This represents the data point to be measured. Represents the data points to be measured Neighborhood set Represents the neighborhood points, Represents the local reachability density of a point.

[0009] Further, calculating the local reachability density of the data point to be measured and its neighboring points includes the following steps: Calculate the data points to be measured up to the first... The distance to the nearest neighbor defines the neighborhood range; Calculate the data points to be measured To its neighboring points within its neighborhood range Reachable distance: Based on the data points to be tested To its neighboring points within its neighborhood range Calculate the reachable distance of the data points to be measured. Locally accessible density within a neighborhood.

[0010] Furthermore, based on the data points to be tested The weighted distance to the neighboring points is used to calculate the data points to be tested. To its neighboring points within its neighborhood range The reachable distance, wherein the weighted distance is expressed as: ; in, Indicates the weight of the spatial location dimension. Indicates the weight of the program instruction dimension. Represents the weight of the vibration feature dimension. This represents the Euclidean distance between two points in space. This represents the line number difference value in G code. This represents the difference in vibration characteristics.

[0011] Furthermore, it also includes: Filter neighboring points within a preset time window; Calculate the temporal continuity score based on the selected neighborhood points; Adjust the number of neighboring points until the temporal continuity score meets the preset requirements.

[0012] Furthermore, the mapping unit sequence is a time sequence, and the features of the mapping unit sequence include the three-dimensional coordinates of the spatial position of the axis, the root mean square value of the vibration signal within a preset time window, and the G-code line number.

[0013] A machine tool machining anomaly diagnosis system based on LOF (List of Elements) is used to diagnose machine tool machining anomalies using the aforementioned LOF-based machine tool machining anomaly diagnosis method, comprising: The data acquisition module is used to collect position information, program instruction information, and vibration signals of each axis of the machine tool; The mapping processing module is used to construct a mapping unit sequence based on a unified clock, using the position information, the program instruction information, and the vibration signal. An unsupervised computation engine is used to input the sequence of mapping units of the state to be tested into the density distribution model and calculate the local outlier factor of the data point to be tested relative to its neighborhood points.

[0014] A computer device includes a processor and a memory connected to the processor. The memory stores one or more programs that are executed by the processor to perform the steps in the LOF-based machine tool machining anomaly diagnosis method described above.

[0015] A computer-readable storage medium storing one or more programs, which are executed by a processor to perform the steps in the LOF-based machine tool machining anomaly diagnosis method described above.

[0016] Beneficial Effects: The LOF-based machine tool machining anomaly diagnosis method provided by this invention eliminates the need for manually setting fixed thresholds. It constructs a density distribution model under normal conditions using a local outlier factor algorithm. Relying on the mapping unit sequence of normal conditions, it autonomously learns the characteristic patterns of normal machining, adapting to normal vibration differences under different tools, cutting depths, and machining materials. This effectively avoids missed reports of minor faults and frequent false alarms, reducing reliance on the personal experience of process engineers and achieving accurate identification of machining anomalies, thus meeting the needs of unmanned and intelligent machining. Simultaneously, based on a unified clock, it constructs a mapping unit sequence from the machine tool's axis position information, program instruction information, and vibration signals, achieving the correlation and fusion of multi-source data and spatial location. This allows anomaly diagnosis to be associated with specific machining positions and program instructions, enabling precise source tracing of anomalies. This facilitates operators in quickly locating the workpiece position where the anomaly occurred and the corresponding program step, improving the efficiency of fault diagnosis and process optimization.

[0017] This invention constructs a density distribution model based on the mapping unit sequence of normal states and employs LOF unsupervised learning, effectively solving the problems of high sample acquisition costs and difficulty in implementation of supervised learning methods. As long as normal production data is available, anomaly diagnosis can be carried out, adapting to the actual production needs of high-value precision machining fields. Furthermore, by calculating the local outlier factor of the test data points to select anomalous data and simultaneously acquiring the corresponding location information and program instruction information, it can quickly pinpoint the machining scenario and parameters corresponding to the anomaly. This provides accurate data support for subsequent anomaly cause analysis and process parameter adjustment, further ensuring machine tool machining quality and reducing scrap rates and production losses. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart illustrating the main method steps of the present invention. Detailed Implementation

[0020] To make the above-mentioned objectives, features, and advantages of this application more apparent and understandable, the specific embodiments of this application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of this application. Therefore, this application is not limited to the specific embodiments disclosed below.

[0021] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0022] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "joining," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0023] Example 1: Reference Figure 1 As shown, this embodiment provides a machine tool machining anomaly diagnosis method based on LOF, including the following steps: Step S1: Collect position information, program instruction information, and vibration signals of each axis of the machine tool; Specifically, the location information includes the three-dimensional coordinates of the spatial position along the corresponding axis. The program instruction information includes the current G-code line number of the corresponding axis, and the vibration signal is the vibration signal of the corresponding axis during the machining process collected by an external sensor; the position information, program instruction information and vibration signal are marked based on a unified high-precision clock reference.

[0024] Step S2: Based on a unified clock, construct a mapping unit sequence by combining the position information, the program instruction information, and the vibration signal; In this embodiment, the mapping unit sequence is a time-series sequence. Vibration signals with different sampling frequencies and position coordinates are aligned and interpolated based on a unified clock. The characteristics of the mapping unit sequence include the three-dimensional spatial coordinates of the axis, the root mean square value of the vibration signal within a preset time window, and the G-code line number, represented as: ; in, Represents the corresponding three-dimensional coordinates of the spatial location. This represents the root mean square value of the corresponding vibration signal. Indicates the corresponding G-code line number. This indicates the corresponding time; in some embodiments of this example, the characteristics of the mapping unit sequence also include the peak-to-peak value and kurtosis of the vibration signal. Specifically, the system synchronously collects the position coordinates inside the CNC and the high-frequency vibrations from external sensors. It uses an interpolation algorithm to solve the problem of inconsistent sampling rates between the two. The spatial position three-dimensional coordinates are sampled at 4ms / time, and the vibration signal is sampled at 100μs / time, accurately binding the vibration characteristics to each micrometer-level machining coordinate.

[0025] Step S3: Based on the mapping unit sequence in the normal state, construct a density distribution model using the local outlier factor algorithm; Specifically, the mapping unit sequence of historical qualified processing processes is selected as the training set, the vibration statistical features of each spatial location are extracted, the local reachability density of each data point in the feature space is calculated using the Local Outlier Factor (LOF) algorithm, and a density distribution model of normal processing state is established.

[0026] Step S4: Input the mapping unit sequence of the state to be tested into the density distribution model, and calculate the local outlier factor of the data point to be tested relative to its neighborhood points.

[0027] Step S5: Select the data points to be tested based on the local outlier factor, and obtain the location information and program instruction information corresponding to the data points to be tested.

[0028] In this embodiment, the local outlier factor is used as an anomaly score. Data points are sorted from highest to lowest anomaly score. Regions with anomaly scores exceeding a preset dynamic boundary are highlighted in the 3D spatial model, and their corresponding physical coordinates and G-code line numbers are output. This allows operators to quickly locate the abnormal position in machine tool processing and the G-code executed during the anomaly. For each data point, the system automatically calculates its anomaly score. A higher score indicates a sparser and more outlier processing state at that point. The final result is directly displayed on the 3D model as a heatmap.

[0029] In some implementations of this embodiment, the test data points are sorted from high to low according to the anomaly score to generate an anomaly list, and the anomaly areas with the highest deviation are displayed first.

[0030] This embodiment uses local outlier factors as anomaly scores. In the modeling of the density distribution model, there is no need to pre-acquire or label non-compliant (NG) sample data. The identification of unknown anomalies can be achieved simply by learning the population distribution characteristics of compliant samples.

[0031] As a further improvement to this embodiment, the density distribution model is a two-level model, and constructing the density distribution model includes the following steps: Step S3.1: Divide the historically qualified mapping unit sequence into subsets according to processing stage and spatial region; Step S3.2: Estimate the probability density for each subset using a Gaussian kernel function; In this embodiment, the Gaussian kernel function is represented as: ; Wherein, σ is determined through cross-validation; Step S3.3: Calculate the density mean μ and standard deviation σ of each subset, form a normal density interval based on the density mean μ and the standard deviation σ, represented as [μ-1.5σ, μ+1.5σ], and update the normal density interval in real time through a sliding window; Step S3.4: Construct a density distribution model for each subset based on the probability density and the normal density interval; Step S3.5: The density distribution models of each subset are fused using a weighted voting method, with the fusion weight being the product of the sample size ratio and the anomaly identification accuracy.

[0032] By designing a two-level model, through sample stratification, kernel density initialization, dynamic boundary calibration, and multi-subset fusion, a density distribution model adapted to different processing scenarios is constructed, which supports incremental updates and reduces model maintenance costs.

[0033] Specifically, in step S4, the local outlier factor of the data point to be tested relative to its neighborhood points is calculated, including the following steps: Step S4.1: Calculate the data points to be measured. To the The distance to the nearest neighbor defines the neighborhood range; Step S4.2: Calculate the data points to be measured To its neighboring points within its neighborhood range Reachable distance: ; in, This indicates the reachable distance from the data point to be tested to the neighboring points. Represents the K-distance between the neighboring points. This represents the weighted distance from the data point to be tested to the neighboring points. By calculating the reachable distance, fluctuations in local density can be smoothed, ensuring the stability of the algorithm. Step S4.3: Based on the data points to be tested To its neighboring points within its neighborhood range Calculate the reachable distance of the data points to be measured. In the neighborhood Locally accessible density within: ; in, This indicates that the data point to be tested is in the neighborhood. The local reachability density within the neighborhood is calculated similarly for the neighboring points in the neighborhood. Locally achievable density within .

[0034] The local outlier factor of the data point to be tested is calculated based on the local reachability density and expressed as follows: ; in, This represents the data point to be measured. Indicates the data points to be measured of Neighborhood set Represents the neighborhood points, Represents the local reachability density of a point.

[0035] In this embodiment, if the corresponding : Description of the data points to be tested The density is similar to its neighbors, falling within the normal range. In some implementations of this embodiment, the determination is based on the corresponding... .

[0036] If corresponding (e.g., > 2.0): This indicates the data points to be tested. The density of the data points to be measured is much smaller than the density of their neighbors. It is an isolated anomaly. In some implementations of this embodiment, the determination is based on the corresponding... .

[0037] In this embodiment, based on the data points to be measured The weighted distance to the neighboring points is used to calculate the data points to be tested. To its neighboring points within its neighborhood range The reachable distance, the weighted distance in step S4.2, is expressed as: ; in, Indicates the weight of the spatial location dimension. Indicates the weight of the program instruction dimension. Represents the weight of the vibration feature dimension. This represents the Euclidean distance between two points in space. This represents the differential quantization value for G code line numbers; 0 is used for the same instruction type, and 1 is used for different types. This represents the difference in vibration characteristics.

[0038] Specifically, the weights of the spatial location dimension are represented as follows: ; in, Let ∑I be the mutual information value between coordinate deviation and vibration anomaly, and let ∑I be the sum of the mutual information values ​​of all dimensions. Program instruction dimension weight After setting the basic weights based on the G-code type, dynamic adjustments are made based on the anomaly occurrence rate; vibration characteristic dimension weights. Determined by the variance proportion under normal processing conditions, and satisfying the following conditions: .

[0039] Through the above design optimizations, a multi-dimensional feature weight dynamic allocation mechanism is adopted to address the multi-dimensional and temporal characteristics of the mapping unit sequence. The correlation between each dimension and processing quality is calculated based on mutual information entropy to determine the weight. At the same time, a temporal constraint neighborhood selection strategy is introduced to ensure that neighborhood points come from the same processing segment, thereby improving the continuity and accuracy of anomaly identification.

[0040] As a further improvement to this embodiment, the method for selecting the neighborhood range k includes: Filter neighboring points within a preset time window: Set the preset time. This corresponds to 4 times the 4ms sampling period, and for the current unit Ui, it only applies in the temporal neighborhood. Filter neighboring points within the range; Calculate the temporal continuity score based on the selected neighborhood points: =Number of consecutive neighboring points / k; Adjust the number of neighboring points until the temporal continuity score meets the preset requirements.

[0041] like The value of k is gradually increased from 15 to 30, with each increment being 2, until the time continuity requirement is met. ).

[0042] Example 2: This embodiment provides a machine tool machining anomaly diagnosis system based on LOF (Logo Flow), used to diagnose machine tool machining anomalies using the LOF-based machine tool machining anomaly diagnosis method described in Embodiment 1, including: The data acquisition module is used to synchronously acquire machine tool servo data and analog data from external sensors; in this embodiment, it is used to acquire position information, program instruction information, and vibration signals of each axis of the machine tool. Specifically, the data acquisition module reads the real-time coordinates of the servo axes from the CNC system via a 1073 acquisition card. The spindle load and current program line number are set, with the sampling frequency set to 250Hz, corresponding to a 4ms cycle. A single-axis or three-axis IEPE accelerometer is mounted on the spindle housing, and the cutting vibration signal is acquired at a sampling rate of 10kHz through a high-speed analog acquisition card.

[0043] The mapping processing module is used to perform spatiotemporal alignment and generate vibration feature data with spatial labels; in this embodiment, it is used to construct a mapping unit sequence based on a unified clock, using the position information, the program instruction information, and the vibration signal. An unsupervised computing engine with a built-in improved LOF algorithm model includes a feature weight allocation unit, a temporal neighborhood filtering unit, and a two-level density modeling unit, used to calculate local outlier factors. In this embodiment, it is used to input the mapping unit sequence of the state to be tested into the density distribution model to calculate the local outlier factor of the data point to be tested relative to its neighboring points. In some embodiments of this example, a human-computer interaction terminal is also included for displaying an abnormal rating sorting list.

[0044] This embodiment also provides a computer device, which includes a processor and a memory. The memory is connected to the processor and stores one or more programs. The one or more programs are executed by the processor to implement the steps in the LOF-based machine tool machining anomaly diagnosis method described above.

[0045] This embodiment also provides a computer-readable storage medium storing one or more programs that are executed by a processor to implement the steps in the LOF-based machine tool machining anomaly diagnosis method described above.

[0046] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0047] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A machine tool machining anomaly diagnosis method based on LOF, characterized in that, Includes the following steps: The system collects position information, program instruction information, and vibration signals from each axis of the machine tool. Based on a unified clock, it constructs a mapping unit sequence using the position information, program instruction information, and vibration signals. Based on the mapping unit sequence in the normal state, it constructs a density distribution model using a local outlier factor algorithm. It inputs the mapping unit sequence in the test state into the density distribution model to calculate the local outlier factor of the test data point relative to its neighboring points. Based on the local outlier factor, it selects the test data point and obtains the corresponding position information and program instruction information. The density distribution model is a two-level model. Constructing the density distribution model includes the following steps: dividing the historically qualified mapping unit sequence into subsets according to processing stage and spatial region; estimating the probability density of each subset using a Gaussian kernel function; calculating the density mean and standard deviation of each subset, and forming a normal density interval based on the density mean and standard deviation; constructing the density distribution model of each subset based on the probability density and the normal density interval; and fusing the density distribution models of each subset using a weighted voting method, with the fusion weight being the product of the sample size proportion and the anomaly identification accuracy.

2. The machine tool machining anomaly diagnosis method based on LOF according to claim 1, characterized in that, Calculating the local outlier factor of a test data point relative to its neighborhood points includes the following steps: calculating the local reachability density of the test data point and its neighborhood points; and calculating the local outlier factor of the test data point based on the local reachability density, expressed as: ； in, This represents the data point to be measured. Represents the data points to be measured Neighborhood set Represents the neighborhood points, Represents the local reachability density of a point.

3. The machine tool machining anomaly diagnosis method based on LOF according to claim 2, characterized in that, Calculate the data point to be tested and its neighboring points. The locally reachable density includes the following steps: calculating the distance from the measured data point to the first... The distance to the nearest neighbor is used to define the neighborhood range; the distance to the data point to be measured is calculated. To its neighboring points within its neighborhood range Reachable distance: based on the data points to be measured To its neighboring points within its neighborhood range Calculate the reachable distance of the data points to be measured. Locally accessible density within a neighborhood.

4. The machine tool machining anomaly diagnosis method based on LOF according to claim 3, characterized in that, Based on the data points to be measured To the neighboring point The weighted distance is used to calculate the data points to be measured. To its neighboring points within its neighborhood range The reachable distance, wherein the weighted distance is expressed as: ； in, Indicates the weight of the spatial location dimension. Indicates the weight of the program instruction dimension. Represents the weight of the vibration feature dimension. This represents the Euclidean distance between two points in space. This represents the line number difference value in G code. This represents the difference in vibration characteristics.

5. The machine tool machining anomaly diagnosis method based on LOF according to claim 1, characterized in that, Also includes: Filter neighboring points within a preset time window; Calculate the temporal continuity score based on the selected neighborhood points; Adjust the number of neighboring points until the temporal continuity score meets the preset requirements.

6. The machine tool machining anomaly diagnosis method based on LOF according to claim 1, characterized in that, The mapping unit sequence is a time sequence, and the features of the mapping unit sequence include the three-dimensional coordinates of the spatial position of the axis, the root mean square value of the vibration signal within a preset time window, and the G-code line number.

7. A machine tool machining anomaly diagnosis system based on LOF, used to diagnose machine tool machining anomalies using the LOF-based machine tool machining anomaly diagnosis method according to any one of claims 1-6, characterized in that, include: The data acquisition module is used to collect position information, program instruction information, and vibration signals of each axis of the machine tool; The mapping processing module is used to construct a mapping unit sequence based on a unified clock, using the position information, the program instruction information, and the vibration signal. An unsupervised computation engine is used to input the sequence of mapping units of the state to be tested into the density distribution model and calculate the local outlier factor of the data point to be tested relative to its neighborhood points.

8. A computer device, characterized in that, The computer device includes a processor and a memory connected to the processor. The memory stores one or more programs, which are executed by the processor to implement the steps in the LOF-based machine tool machining anomaly diagnosis method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, which are executed by a processor to implement the steps in the LOF-based machine tool machining anomaly diagnosis method as described in any one of claims 1-6.

Images