Real-time data monitoring method for laser cutting machine

Abstract

The invention relates to the field of data processing, in particular to a data real-time monitoring method of a laser cutting machine. The method comprises the steps that data points in the operation process of the laser cutting machine are collected, the data points comprise laser power at different moments, and several neighbor data points are selected from historical data points; calculating the local fluctuation degree of the neighbor data points; then calculating the abnormal degree of the current data point; and finally, the abnormal score of the current data point is calculated, and when the abnormal score is larger than a threshold value, the laser cutting machine breaks down. The problem that the abnormal state of the laser cutting machine cannot be timely and accurately monitored is effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and more specifically, to a method for real-time data monitoring of a laser cutting machine. Background Art

[0002] In the manufacturing fields such as mechanical processing, automobile industry and household appliances, metal cutting process is a key step to realize the manufacturing of complex shapes and structural parts, and has extremely strict requirements on the precision and quality of cutting. Laser cutting machines are widely used in these fields due to their advantages such as high precision, high efficiency and low pollution. The laser cutting machine irradiates the material with a laser beam at an appropriate power density, so that the material absorbs a large amount of energy in a very short time and heats up rapidly, reaching the melting point or even the boiling point, thereby achieving precise cutting. In this process, laser power is a key parameter affecting the cutting effect. If the laser power is too small, the material may not reach a sufficient temperature, resulting in incomplete cutting or even damage to the cutting tool; if the laser power is too large, the material will be over-melted, generating too much slag and heat, resulting in rough cutting surface, material deformation or thermal cracking and other problems. Therefore, the present invention monitors the key parameters in the working process of the laser cutting machine in real time, finds abnormal data in time and notifies relevant personnel to make adjustments, which is very important for ensuring cutting quality and production efficiency.

[0003] At present, the patent application document with publication number CN118710254A and name “Monitoring and maintenance method and system for steel plate laser cutting equipment” discloses a method for establishing a task set for steel plate laser cutting equipment, configuring a basic data set to obtain operating parameters and response parameters, and then performing response parameter supervision to generate a first abnormal value, followed by equipment feature adaptation to generate a second abnormal value, and simultaneously inputting a real-time decision network to generate a monitoring and maintenance decision result, and then performing additional data collection, and finally completing abnormal warning management.

[0004] The above method uses a real-time decision network that requires a large amount of data for training and is computationally complex. It also does not directly monitor laser power data but collects other indirect data. Therefore, it is unable to monitor abnormal conditions of the laser cutting machine in a timely and accurate manner. Summary of the invention

[0005] In order to solve the above-mentioned problem that the abnormal state of the laser cutting machine cannot be monitored timely and accurately, the present invention provides solutions in the following aspects.

[0006] The present invention provides a method for real-time monitoring of data of a laser cutting machine, comprising: collecting data points during the operation of the laser cutting machine, wherein the data points include laser power at different times, calculating the absolute difference between the laser power of historical data points and current data points, and sorting the historical data points in order of absolute difference from small to large, wherein the first K data points are used as neighboring data points; calculating the degree of local fluctuation , ,in, Represents the average time interval between the current data point and all neighboring data points, Represents the neighboring data point The laser power at each data point, Represents the neighboring data point The laser power of each data point; calculate the abnormality of the current data point , ,in, () indicates normalization, represents the approximate normal value of the laser power in the historical data points, represents the total number of neighboring data points, Represents the neighboring data point The laser power at each data point, Indicates the degree of local fluctuation; calculates the abnormal score of the current data point. When the abnormal score is greater than the threshold, the laser cutting machine has a fault. The abnormal score is positively correlated with the abnormal degree, the Euclidean distance between the current data point and each neighboring data point.

[0007] By calculating the absolute difference between the current data point and the historical data point and selecting the most relevant neighboring data point, the historical data similar to the current working condition can be effectively identified. By calculating the local fluctuation degree, the change of laser power in a short period of time can be sensitively captured, and the potential abnormality of the laser cutting machine can be identified in time, ensuring the real-time and accuracy of abnormality monitoring. The degree of abnormality is compared with the mode of laser power in the historical data and the mean of the neighboring data points, which further improves the accuracy of abnormality monitoring and can accurately detect small abnormal fluctuations. The Euclidean distance is used to calculate the abnormality score, so that the relationship between the current data point and its neighboring data points can be accurately quantified, effectively improving the robustness and accuracy of abnormality monitoring.

[0008] Preferably, the abnormal score of the current data point includes: , where A represents the anomaly score of the current data point, represents normalization, S represents the abnormality of the current data point, and K represents the number of neighboring data points. Indicates the collection time of the current data point, represents the collection time of the nth data point of the neighboring data point, Indicates the laser power at the current data point, Represents the laser power of the nth data point in the neighborhood.

[0009] By introducing the calculation method of anomaly score and combining Euclidean distance and local fluctuation degree, the accuracy of anomaly monitoring of laser cutting machines can be effectively improved.

[0010] Preferably, the abnormal score of the current data point includes: , where A represents the anomaly score of the current data point, S represents the degree of anomaly of the current data point, and K represents the number of neighboring data points. Indicates the collection time of the current data point, represents the collection time of the nth data point of the neighboring data point, Indicates the laser power at the current data point, Represents the laser power of the nth data point in the neighborhood.

[0011] Generally speaking, the smaller the time interval between the neighboring data point and the current data point, the greater the impact on the anomaly score of the current data point. Therefore, a larger weight can be given to the corresponding Euclidean distance to more accurately calculate the anomaly score of the current data point.

[0012] Preferably, the approximate normal value of the laser power in the historical data points includes: taking the mode of the laser power in the historical data points as the approximate normal value.

[0013] Preferably, the approximate normal value of the laser power in the historical data point includes: dividing the laser power range of the historical data point into multiple equidistant intervals, selecting an interval containing the largest number of laser powers, and taking the median of the interval or the actual laser power closest to the median as the approximate normal value of the laser power in the historical data point.

[0014] By dividing the laser power of historical data points into intervals and selecting the median of the highest frequency interval or the laser power closest to the median as the approximate normal value, a more robust, precise and efficient data analysis method is provided, which enhances the accuracy, stability and adaptability of monitoring.

[0015] Preferably, when the abnormality score is greater than a threshold value, the laser cutting machine has a fault, including: prompting the laser cutting machine to have a fault by using the color of a fault warning light and / or the sound of a fault warning bell.

[0016] By combining the color change of the fault indicator light and the ring tone prompt when the abnormality score exceeds the threshold, the laser cutting machine can be inspected and repaired in time to avoid more serious problems.

[0017] Preferably, the historical data points include: performing missing completion processing on the historical data points.

[0018] By performing missing completion processing on historical data points, missing values ​​or incomplete data can be effectively filled, ensuring the continuity and integrity of data analysis and avoiding erroneous monitoring results caused by missing data.

[0019] Preferably, when the abnormality score is greater than a threshold value, the laser cutting machine has a fault, and the method includes: sending fault information to a remote terminal through a wireless communication module.

[0020] Sending fault information to a remote terminal through a wireless communication module can improve the fault response speed of the laser cutting machine and enhance remote monitoring and diagnostic capabilities.

[0021] The beneficial effects of the present invention are as follows: by calculating the absolute difference between the current data point and the historical data point and selecting the most relevant neighboring data point, the historical data similar to the current working condition can be effectively identified; by calculating the local fluctuation degree, the change of laser power in a short period of time can be sensitively captured, and the potential abnormality of the laser cutting machine can be timely identified, thereby ensuring the real-time and accuracy of abnormality monitoring; the degree of abnormality is compared with the approximate normal value of the laser power in the historical data and the mean of the neighboring data points, thereby further improving the accuracy of abnormality monitoring, being able to accurately detect tiny abnormal fluctuations; and using the Euclidean distance for calculating the abnormality score, so that the relationship between the current data point and its neighboring data points can be accurately quantified, thereby effectively improving the robustness and accuracy of abnormality monitoring of the laser cutting machine. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] 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 It is a flow chart of a method for real-time data monitoring of a laser cutting machine provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] 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.

[0024] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] Figure 1 The present invention provides a method for real-time monitoring of data of a laser cutting machine, which includes the following steps: S101, collecting data points during the operation of the laser cutting machine, wherein the data points include the laser power at different times, calculating the absolute difference in laser power between the historical data points and the current data points, and sorting the historical data points in order of the absolute difference from small to large, wherein the first K data points are used as neighboring data points.

[0026] For example, a laser power monitor can be installed at the cutting head of a laser cutting machine to collect laser power in real time. The current data point collected is , t is the acquisition time corresponding to the current data point, x is the laser power of the current data point, and the laser powers of the historical data points in the past five moments are , , , and The absolute differences are , , , and ,and , if K=3 is preset, obviously, , and The absolute difference in laser power from the current data point is small, and accordingly, , and The corresponding three data points are the neighboring data points of the current data point.

[0027] In the actual data collection process, due to sensor failure, signal interference or data transmission problems, the collected data may be missing. Therefore, in some embodiments, the historical data points include: performing missing completion processing on the historical data points. By completing the missing data points, the integrity of the data set can be improved and the accuracy of subsequent analysis can be ensured. Generally, the missing values ​​can be inferred based on the existing data points through linear interpolation, spline interpolation and other methods.

[0028] S102. Calculate the local fluctuation degree.

[0029] The local fluctuation degree The specific calculation formula is: ,in, Represents the average time interval between the current data point and all neighboring data points, Represents the neighboring data point The laser power at each data point, Represents the neighboring data point Laser power at each data point in, It represents the sum of the absolute differences between the corresponding laser powers of any two neighboring data points of the current data point. The larger the value, the greater the change between the corresponding laser powers of all neighboring data points. It represents the average value of the difference between the collection time of the current data point and the corresponding collection time of all neighboring data points. The smaller the value is, the more the neighboring data points of the current data point appear in a short period of time, which means that these neighboring data points are closer to the current data point in time. In summary, if the sum of the absolute differences between the corresponding laser powers of any two neighboring data points of the current data point is larger and the neighboring data points are closer to the current data point in time, it means that the local fluctuation degree of the neighboring data points of the current data point is greater, which means that the working state of the laser cutting machine has experienced large fluctuations at this moment, and the possibility that it is in an abnormal state is greater. As described in the example in step S101, , and The corresponding three data points are the neighboring data points of the current data point, and the corresponding collection times are , and , the current time is t, the laser power at the current time is x, then , The calculation result is .

[0030] S103: Calculate the abnormality level of the current data point.

[0031] The abnormality level of the current data point , ,in, () indicates normalization, represents the approximate normal value of the laser power in the historical data points, represents the total number of neighboring data points, Represents the neighboring data point The laser power at each data point, Indicates the degree of local fluctuation. During the cutting process of the laser cutting machine, the laser power will normally stabilize at a value or a range. Therefore, it is necessary to use the approximate normal value of the laser power in the historical data points to reflect the laser power under normal conditions. It indicates the absolute difference between the approximate normal value of the laser power in the historical data point and the average value of the laser power of all neighboring data points of the current data point. The larger the value, the greater the difference between the distribution of laser power of all neighboring data points of the current data point and the laser power under normal conditions, indicating that these neighboring data points are in an abnormal state. In addition, the degree of local fluctuation On the other hand, it also reflects the abnormality of the current data point. and local fluctuations It can more comprehensively and accurately reflect the abnormality of the current data point. Through normalization, the calculation of the abnormality is not affected by the scale, which improves the uniformity and comparability of the calculation. No matter how large the data range is, the consistency of the abnormality can be ensured, and a unified scale is provided for the subsequent calculation of abnormality scores.

[0032] Under some normal circumstances, the laser power in the historical data points will stabilize at a value. At this time, the mode of the laser power in the historical data points can be used as an approximate normal value. Under other normal circumstances, the laser power in the historical data points stabilizes at a range rather than a value. At this time, the mode of the laser power in the historical data points directly calculated may not reflect the laser power under normal circumstances. Therefore, the present invention also provides a method for calculating the mode of the laser power in the historical data points, which specifically includes: dividing a plurality of equidistant intervals based on the laser power range of the historical data points, selecting an interval containing the largest number of laser powers, and using the median of the interval or the actual laser power closest to the median as the approximate normal value of the laser power in the historical data points. For example, the laser power range of the historical data points is (a, f], which is divided into a plurality of equidistant intervals such as the intervals (a, b], (c, d], (d, e] and (e, f], wherein the interval (c, d] contains the largest number of laser powers and the median of the interval is , then you can or closest The actual laser power is taken as the approximate normal value of the laser power in the historical data point.

[0033] S104, calculating the abnormality score of the current data point. When the abnormality score is greater than a threshold, the laser cutting machine has a fault. The abnormality score is positively correlated with the abnormality degree, the Euclidean distance between the current data point and each neighboring data point.

[0034] Specifically, the calculation formula for the abnormal score of the current data point is: , where A represents the anomaly score of the current data point, represents normalization, S represents the abnormality of the current data point, and K represents the number of neighboring data points. Indicates the collection time of the current data point, represents the collection time of the nth data point of the neighboring data point, Indicates the laser power at the current data point, represents the laser power of the nth data point of the neighboring data point. In this formula, It represents the mean Euclidean distance between the current data point and all neighboring data points. The larger the value, the more marginal the current data point is, and the higher the anomaly score is. When it is multiplied by the degree of anomaly S, it takes into account that the metal dust, slag and environmental pollutants generated in the actual cutting process are easy to contaminate or damage the optical components, causing continuous abnormal data in the laser power. There are obvious local fluctuations between these continuous abnormal data points. In this case, the mean Euclidean distance between the current data point and all neighboring data points is small, and the degree of local fluctuation is large, so the anomaly score should be higher. Therefore, The abnormal score of the current data point can be more comprehensively and accurately described. When the abnormal score is greater than the threshold, it means that the laser cutting machine has a fault.

[0035] In some embodiments, since the neighboring data points with a smaller time interval with the current data point have a greater impact on the abnormal score of the current data point, a larger weight can be given to the corresponding Euclidean distance, so as to more accurately calculate the abnormal score of the current data point. Therefore, the present invention also provides another method for calculating the abnormal score of the current data point, which is specifically: , where A represents the anomaly score of the current data point, S represents the degree of anomaly of the current data point, and K represents the number of neighboring data points. Indicates the collection time of the current data point, represents the collection time of the nth data point of the neighboring data point, Indicates the laser power at the current data point, represents the laser power of the nth data point in the neighborhood, is the Euclidean distance between the current data point and the nth data point in the neighboring data points, through to adjust its weight.

[0036] When it is determined that the laser cutting machine has a fault, in order to quickly convey the equipment status so that the operator can repair it as soon as possible. Therefore, in some embodiments, it is also necessary to indicate that the laser cutting machine has a fault through the color of the fault prompt light and / or the sound of the fault prompt bell. In other embodiments, the fault information can also be sent to a remote terminal through a wireless communication module. Sending the fault information to a remote terminal through a wireless communication module can improve the fault response speed of the laser cutting machine and enhance the remote monitoring and diagnosis capabilities.

[0037] The present invention can effectively identify historical data similar to the current working condition by calculating the absolute difference between the current data point and the historical data point and selecting the most relevant neighboring data point. By calculating the local fluctuation degree, the present invention can sensitively capture the change of laser power in a short period of time and timely identify the potential abnormality of the laser cutting machine, thereby ensuring the real-time and accuracy of abnormality monitoring. The abnormality degree is compared with the laser power mode in the historical data and the mean of the neighboring data points, thereby further improving the accuracy of abnormality monitoring and being able to accurately detect small abnormal fluctuations. The Euclidean distance is used for the calculation of the abnormality score, so that the relationship between the current data point and its neighboring data points is accurately quantified, thereby effectively improving the robustness and accuracy of abnormality monitoring of the laser cutting machine.

[0038] 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.

[0039] 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 real-time monitoring of data of a laser cutting machine, characterized in that: include: Collect data points during the operation of the laser cutting machine, wherein the data points include laser power at different times, calculate the absolute difference between the laser power of the historical data points and the current data points, and sort the historical data points in ascending order of the absolute difference, wherein the first K data points are used as neighboring data points; Calculate the local fluctuation degree , ,in, Represents the average time interval between the current data point and all neighboring data points. Represents the neighboring data point The laser power at each data point, Represents the neighboring data point Laser power at each data point; Calculate the abnormality of the current data point , ,in, () indicates normalization, represents the approximate normal value of the laser power in the historical data points, represents the total number of neighboring data points, Represents the neighboring data point The laser power at each data point, Indicates the degree of local fluctuation; The abnormality score of the current data point is calculated. When the abnormality score is greater than a threshold, the laser cutting machine has a fault. The abnormality score is positively correlated with the abnormality degree, the Euclidean distance between the current data point and each neighboring data point.

2. The method for real-time data monitoring of a laser cutting machine according to claim 1, characterized in that: The abnormal score of the current data point includes: , where A represents the anomaly score of the current data point, represents normalization, S represents the abnormality of the current data point, and K represents the number of neighboring data points. Indicates the collection time of the current data point, represents the collection time of the nth data point of the neighboring data point, Indicates the laser power at the current data point, Represents the laser power of the nth data point in the neighborhood.

3. The real-time data monitoring method for a laser cutting machine according to claim 1 is characterized in that: The abnormal score of the current data point includes: , where A represents the anomaly score of the current data point, S represents the degree of anomaly of the current data point, and K represents the number of neighboring data points. Indicates the collection time of the current data point, represents the collection time of the nth data point of the neighboring data point, Indicates the laser power at the current data point, Represents the laser power of the nth data point in the neighborhood.

4. The method for real-time data monitoring of a laser cutting machine according to claim 1, characterized in that: The approximate normal value of the laser power in the historical data points includes: taking the mode of the laser power in the historical data points as the approximate normal value.

5. The method for real-time data monitoring of a laser cutting machine according to claim 1, characterized in that: The approximate normal value of the laser power in the historical data point includes: dividing a plurality of equidistant intervals based on the laser power range of the historical data point, selecting an interval containing the largest number of laser powers, and taking the median of the interval or the actual laser power closest to the median as the approximate normal value of the laser power in the historical data point.

6. The method for real-time data monitoring of a laser cutting machine according to claim 1, characterized in that: When the abnormality score is greater than a threshold, the laser cutting machine has a fault, including: prompting the laser cutting machine to have a fault through the color of a fault warning light and / or the sound of a fault warning bell.

7. The method for real-time monitoring of data of a laser cutting machine according to claim 1, characterized in that: The historical data points include: performing missing completion processing on the historical data points.

8. The method for real-time monitoring of data of a laser cutting machine according to claim 1, characterized in that: When the abnormality score is greater than a threshold value, the laser cutting machine has a fault, including: sending fault information to a remote terminal through a wireless communication module.

Images