An Intelligent Fault Diagnosis Method for a Metal 3D Printing Device

Abstract

The present invention relates to the technical field of fault diagnosis of metal 3D printing equipment, and particularly relates to an intelligent fault diagnosis method for metal 3D printing equipment. In this method, the quality state value of the molten pool is calculated by real-time monitoring of the molten pool state information to determine whether it is abnormal. If it is abnormal, a fault self-check instruction is triggered. By analyzing the vibration amplitude and positioning error data of the laser scanning galvanometer, its stability evaluation value is calculated to determine whether it is normal. If it is normal, the power distribution and focus drift distance of each height layer of the laser beam are monitored, and the irradiation evaluation value is calculated to determine whether it is normal. If it is normal, the powder spraying amount of the powder nozzle is monitored, and it is compared and analyzed with the reference value to calculate the powder spraying evaluation value to determine whether it is normal. If all the above components are abnormal, the fault classification situation table is matched to determine the fault level, so as to realize intelligent and accurate real-time fault diagnosis and automatic classification, and improve the printing efficiency and finished product rate.

Description

Technical Field

[0001] The present invention relates to the technical field of metal 3D printing equipment fault diagnosis, and specifically to an intelligent fault diagnosis method for metal 3D printing equipment. Background Art

[0002] A metal 3D printing equipment is an advanced manufacturing equipment that manufactures three-dimensional entities through a layer-by-layer construction method based on the principles of layered manufacturing and discrete accumulation. This technology is widely applicable to various metal materials such as aluminum alloy, titanium alloy, and stainless steel, and can meet the requirements for the manufacture of complex structures in multiple industrial fields such as aerospace, automotive manufacturing, and biomedicine;

[0003] Currently, the quality inspection of metal 3D printing usually relies on means such as X-ray CT scanning to identify printing defects. However, most of these methods perform inspections after printing is completed, making it difficult to detect abnormalities generated during the printing process in a timely manner, resulting in a high waste piece rate and increased production costs. Therefore, it is crucial to achieve real-time monitoring and fault diagnosis of the equipment during the printing process;

[0004] During the metal 3D printing process, the molten pool is the core area where the material melts, and its state directly affects the printing quality and the final forming effect. The state of the molten pool is affected by various factors, among which the states of the laser scanning galvanometer, the laser beam, and the powder nozzle are the key factors leading to molten pool abnormalities. If the abnormalities of these components in the metal 3D printing equipment cannot be detected in a timely manner, it is easy to cause the accumulation of printing defects and affect the quality of the finished product;

[0005] To solve the above defects, a technical solution is provided now. Summary of the Invention

[0006] The present invention aims to solve the problem that existing metal 3D printing equipment lacks accurate analysis of the molten pool state during the printing process to timely identify abnormalities and real-time monitor and diagnose the source of faults, and proposes an intelligent fault diagnosis method for metal 3D printing equipment.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] An intelligent fault diagnosis method for metal 3D printing equipment includes the following steps:

[0009] By monitoring and analyzing the molten pool state information of the printing equipment, calculate the quality state value of the molten pool, and accordingly determine whether the molten pool is in an abnormal state. If the molten pool is determined to be in an abnormal state, trigger a fault self-check instruction;

[0010] Based on the triggered fault self-check instruction, obtain the vibration amplitude data and positioning error data of the laser scanning galvanometer, calculate the vibration amplitude change difference and the positioning error fluctuation value, determine the stability evaluation value of the laser scanning galvanometer, and determine whether it is in a normal state;

[0011] If it is in the normal state, obtain the laser power of each height layer of the laser beam and the laser focus drift distance of the laser beam, calculate the power stability value and the focus drift value, determine the irradiation evaluation value of the laser beam, and determine whether it is in the normal state;

[0012] If it is in the normal state, obtain the powder spraying amount of the powder nozzle, compare and analyze it with the reference powder spraying amount, determine the powder spraying evaluation value of the powder nozzle, and determine whether it is in the normal state;

[0013] If the laser scanning galvanometer, the laser beam, and the powder nozzle are all determined to be in the abnormal state, match the corresponding fault classification situation table to determine the fault level.

[0014] Further, the specific process of solving the quality state value of the molten pool is as follows:

[0015] Obtain the thermal imaging image of the molten pool within the current monitoring period;

[0016] Extract the morphological parameters of the molten pool from the thermal imaging image, and then compare and analyze the morphological parameters of the molten pool with the corresponding preset reference intervals respectively. If a certain morphological parameter is within the preset reference interval, assign a value of 0, otherwise assign a value of 1, and accumulate all the assigned values to calculate the morphological value of the molten pool;

[0017] Extract the color temperature of each detection point in the thermal imaging image, match it with the reference color temperature of the corresponding thermal imaging image of the molten pool, determine the color temperature non - matching points, and integrate each adjacent color temperature non - matching point to form each abnormal area in the thermal imaging image;

[0018] Extract the numerical values of the color temperature of each detection point in each abnormal area for average calculation to obtain the average color temperature of each abnormal area in the thermal imaging image, which is used as the temperature value of each abnormal area in the thermal imaging image;

[0019] Compare and analyze the temperature values of each abnormal area in the thermal imaging image with the temperature threshold to determine the overheated area and the over - cooled area;

[0020] Calculate the total areas of the overheated area and the over - cooled area, and calculate their ratios to the area of the thermal imaging image respectively to obtain the overheat ratio and the over - cool ratio. Multiply the overheat ratio and the over - cool ratio by the corresponding weight coefficients respectively and then add them to obtain the temperature distribution value of the molten pool;

[0021] Identify the defects of the molten pool within the current monitoring period, calculate the total areas of various defect areas, and calculate their ratios to the total area of the molten pool to obtain the defect value of the molten pool;

[0022] Determine the quality state value of the molten pool based on the morphological value, the temperature distribution value, and the defect value.

[0023] Further, the specific process of solving the vibration amplitude change difference is as follows:

[0024] The vibration amplitude data of the laser scanning galvanometer during the current monitoring time period is obtained, and the average is calculated to obtain the vibration amplitude average of the laser scanning galvanometer. The vibration amplitude data of the laser scanning galvanometer at each monitoring moment is subtracted from the vibration amplitude average to obtain the vibration amplitude change value, and the maximum vibration amplitude change value and the minimum vibration amplitude change value are extracted therefrom for difference calculation to obtain the vibration amplitude change difference.

[0025] Furthermore, the specific process of solving the positioning error fluctuation value is as follows:

[0026] Obtain the error between the target position and the actual position of the laser scanning galvanometer under the same instruction in the current monitoring time period, and construct a positioning error sequence;

[0027] The maximum positioning error and the minimum positioning error in the positioning error sequence are eliminated, and then the root mean square error of the error sequence after elimination is calculated to obtain the positioning error fluctuation value.

[0028] Furthermore, the specific process of solving the power stability value is as follows:

[0029] Dividing the laser beam into various height layers according to a preset height to obtain various height layers of the laser beam;

[0030] The laser power at each monitoring time in the current monitoring time period corresponding to each height layer of the laser beam is obtained, and the laser power sequence is obtained by arranging the laser power sequence in descending order, and the mode laser power is extracted from the sequence as the laser power at each height layer of the laser beam corresponding to the current monitoring time period, thereby forming a power data set of the laser beam corresponding to the current monitoring time period;

[0031] Thus, the power stability corresponding to the laser beam is analyzed to obtain the power stability value corresponding to the laser beam.

[0032] Furthermore, the specific process of solving the focus drift value is as follows:

[0033] Obtain the laser focus position of the laser beam at each monitoring moment in the current monitoring time period, and calculate the distance difference between it and the corresponding laser focus target position to obtain the laser focus drift distance of the laser beam in the current monitoring time period, and sort them from large to small to obtain a laser focus drift distance sequence;

[0034] The values of the first one-third of the laser focus drift distance in the laser focus drift distance sequence are extracted and averaged to obtain the focus drift value corresponding to the laser beam.

[0035] Furthermore, the specific process of solving the powder spraying evaluation of the powder nozzle is as follows:

[0036] Obtain the powder spraying amount of the powder nozzle at each monitoring moment corresponding to the current monitoring time period, and obtain the powder spraying amount fpz of the powder nozzle at each monitoring moment corresponding to the current monitoring time period t ;

[0037] Extract the standard powder spraying amount of the powder nozzle corresponding to the current monitoring time period as the reference powder spraying amount fpz of the powder nozzle corresponding to the current monitoring time period * ;

[0038] According to the formula: , obtain the powder spraying evaluation value FPZ of the powder nozzle, where t represents the number of each monitoring moment in the current monitoring time period, n represents the total number of the numbers of each monitoring moment in the current monitoring time period, k represents a natural constant greater than zero, and Fpp t represents the powder spraying amount change rate of the powder nozzle corresponding to the current monitoring time period, and Fpp * represents the reference powder spraying amount change rate, and a1, a2, and λ all represent set correction factors

[0039] Furthermore, the specific process of determining the fault level is as follows:

[0040] Compare and match the stability evaluation value of the laser scanning galvanometer, the irradiation evaluation value of the laser beam, and the powder spraying evaluation value of the powder nozzle with the corresponding fault grading table stored in the cloud database respectively, and obtain the corresponding fault level accordingly

[0041] The technical solution provided by the present invention has the following beneficial effects compared with the known prior art:

[0042] 1. In the present invention, through continuous monitoring of the molten pool state and combined with the analysis of key parameters of the laser scanning galvanometer, laser beam, and powder nozzle, compared with the traditional post-detection method relying on X-ray CT scanning, it can accurately detect abnormalities during the printing process, reduce the accumulation of defects, thereby realizing the efficient fault intelligent detection and diagnosis of metal 3D printing equipment, significantly improving the equipment stability and printing quality, effectively reducing the scrap rate, avoiding the discovery of defects only after printing is completed, thereby reducing material waste and production losses, optimizing the production process, and improving the overall production efficiency

[0043] 2. In the present invention, by adopting the multi-parameter fusion analysis technology, integrating various data such as molten pool thermal imaging analysis, vibration detection, laser power evaluation, and powder spraying amount monitoring, calculating core parameters such as the quality state value of the molten pool, the stability evaluation value of the laser scanning galvanometer, the irradiation evaluation value of the laser beam, and the powder spraying evaluation value of the powder nozzle, thereby realizing accurate fault identification, reducing misjudgment and missed judgment, and improving the accuracy of anomaly detection

[0044] 3. In the present invention, by constructing a cloud database to store the fault classification situation table and automatically matching and analyzing based on the monitoring data, intelligent fault classification is achieved, greatly reducing manual intervention, making fault diagnosis more standardized and efficient, and improving the response speed of equipment maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0046] Figure 1 It is a block diagram of the overall process method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0048] As Figure 1 shown, an intelligent fault diagnosis method for a metal 3D printing device includes the following steps:

[0049] A1: Monitor the molten pool state information of the printing device, and thus perform determination and analysis processing on the molten pool state of the printing device. The specific analysis process is as follows:

[0050] Obtain the thermal imaging image of the molten pool during the current monitoring period through an infrared thermal imager;

[0051] Extract the morphological parameters of the molten pool from the thermal imaging image. The morphological parameters include length, width, and depth. Then, compare and analyze the morphological parameters of the molten pool with the corresponding preset reference intervals respectively. If a certain morphological parameter is within the preset reference interval, assign a value of 0; otherwise, assign a value of 1. Accumulate all the assigned values to calculate the morphological value of the molten pool, and mark it as Rxt;

[0052] Uniformly arrange detection points on the thermal imaging image to obtain each detection point in the thermal imaging image, extract the color temperature of each detection point in the thermal imaging image, and match it with the reference color temperature of the preset thermal imaging image corresponding to the molten pool. If the color temperature of a certain detection point matches the reference color temperature of the preset thermal imaging image corresponding to the molten pool successfully, then mark this detection point as a color temperature compliant point; if the color temperature of a certain detection point does not match the reference color temperature of the preset thermal imaging image corresponding to the molten pool successfully, then mark this detection point as a color temperature non-compliant point;

[0053] Integrate each adjacent color temperature non-compliant point in the thermal imaging image to form each abnormal area in the thermal imaging image;

[0054] Extract the numerical values of the color temperatures of each detection point in each abnormal area for average calculation to obtain the average color temperature of each abnormal area, which is used as the temperature value of each abnormal area in the thermal imaging image;

[0055] Compare and analyze the temperature values of each abnormal area in the thermal imaging image with the preset temperature threshold. If the temperature value of an abnormal area in the thermal imaging image is greater than or equal to the preset temperature threshold, then mark this abnormal area in the thermal imaging image as an overheated area; if the temperature value of an abnormal area in the thermal imaging image is less than the preset temperature threshold, then mark this abnormal area in the thermal imaging image as a cold area;

[0056] Calculate the total areas of the overheated area and the cold area, and respectively calculate their ratios to the area of the thermal imaging image to obtain the overheat ratio and the cold ratio. Multiply the overheat ratio and the cold ratio by the corresponding weight coefficients respectively and then add them to get the temperature distribution value of the molten pool, and mark it as Rwd;

[0057] Identify the defects of the molten pool during the current monitoring period through machine vision technology. The defects include pores, cracks and lack of fusion. Calculate the total areas of various defect areas, and calculate their ratios to the total area of the molten pool to obtain the defect value of the molten pool, and mark it as Rqx;

[0058] According to the formula: , obtain the quality state value RCP of the molten pool, where η1, η2 and η3 respectively represent the weight coefficients of the shape value, the temperature distribution value and the defect value;

[0059] Compare and analyze the quality state value of the molten pool with the preset reference comparison interval. When the quality state value of the molten pool is within the preset reference comparison interval, then determine that the molten pool is in an abnormal state; otherwise, determine that the molten pool is in a normal state;

[0060] If the molten pool is determined to be in an abnormal state, then trigger a fault self-check instruction and execute A2.

[0061] In a specific embodiment, the present invention obtains the thermal imaging image of the molten pool by using an infrared thermal imager, analyzes the morphological parameters and color temperature distribution in the image, can detect the overheated or overcooled areas that may exist in the molten pool in real time, thereby evaluating the temperature distribution state of the molten pool. At the same time, machine vision technology is used to identify the defects of the molten pool, which further enhances the accuracy of fault detection. By comprehensively calculating the quality value of the molten pool in this way, it can efficiently determine whether the molten pool is in an abnormal state. If an abnormal state is detected, a fault self-check instruction will be automatically triggered, so as to prevent defects from being discovered only after printing is completed, providing strong support for timely analyzing and diagnosing the fault state of the key components of the equipment in the follow-up.

[0062] A2: Monitor the stable state information of the laser scanning galvanometer, and perform self-check analysis and processing on the stable state of the laser scanning galvanometer. The specific analysis process is as follows:

[0063] Obtain the vibration amplitude data of the laser scanning galvanometer during the current monitoring time period through a vibration sensor, calculate its mean value to obtain the mean vibration amplitude of the laser scanning galvanometer, calculate the difference between the vibration amplitude data of the laser scanning galvanometer at each monitoring moment during the current monitoring time period and the mean vibration amplitude, obtain the vibration amplitude change value of the laser scanning galvanometer, and extract the maximum vibration amplitude change value and the minimum vibration amplitude change value from it for difference calculation to obtain the vibration amplitude change difference Xzd;

[0064] Obtain the error between the target position and the actual position of the laser scanning galvanometer under the same instruction during the current monitoring time period through a high-precision position sensor, and construct a positioning error sequence;

[0065] Eliminate the maximum positioning error and the minimum positioning error in the positioning error sequence to reduce the influence of extreme values on the calculation results, and then calculate the root mean square error of the error sequence after elimination. The calculation formula is: , to obtain the positioning error fluctuation value Xwz j , where N represents the total number of sampling points, i represents the sampling point number, and Epos(i) represents the positioning error of the i-th sampling point;

[0066] According to the formula: , to obtain the stable evaluation value JXP of the laser scanning galvanometer, where η4 and η5 respectively represent the weight coefficients of the vibration amplitude change difference and the positioning error fluctuation value;

[0067] Compare and analyze the stable evaluation value of the laser scanning galvanometer with the preset stable evaluation threshold. When the stable evaluation value of the laser scanning galvanometer is greater than the preset stable evaluation threshold, the laser scanning galvanometer is determined to be in a normal state, and A3 is executed;

[0068] When the stability evaluation value of the laser scanning galvanometer is less than or equal to the preset stability evaluation threshold, the laser scanning galvanometer is determined to be in an abnormal state, and the stability evaluation value of the laser scanning galvanometer is compared and matched with the fault classification table stored in the cloud database to obtain the fault level of the laser scanning galvanometer;

[0069] Among them, each stability evaluation value of the laser scanning galvanometer corresponds to a fault level, and the fault levels include the first-level fault level, the second-level fault level, and the third-level fault level;

[0070] And the fault level of the laser scanning galvanometer is sent to the management terminal, and the fault level of the laser scanning galvanometer is displayed and notified through the display terminal;

[0071] In a specific embodiment, the present invention calculates the difference in vibration amplitude change and the positioning error fluctuation value to obtain the stability evaluation value of the laser scanning galvanometer, and compares it with the preset threshold to determine whether it is normal. If the stability of the laser scanning galvanometer does not meet the standard, it will automatically match the fault level according to the stability evaluation value of the laser scanning galvanometer and the fault classification table, and send the fault level information to the management terminal for display and notification. Through this real-time monitoring, automated fault diagnosis and fault level analysis, the abnormal state of the laser scanning galvanometer can be accurately identified, providing accurate fault information for maintenance personnel, improving the fault diagnosis efficiency, and reducing the equipment downtime.

[0072] A3: Monitor the irradiation state information of the laser beam, and thus perform self-check analysis and processing on the irradiation state of the laser beam. The specific analysis process is as follows:

[0073] The laser beam is divided into each height layer according to the preset height to obtain each height layer of the laser beam;

[0074] The laser power of each height layer of the laser beam at each monitoring moment in the current monitoring time period is obtained through the laser power sensor, and the laser power of each height layer of the laser beam at each monitoring moment in the current monitoring time period is arranged in descending order to obtain the laser power sequence of each height layer of the laser beam in the current monitoring time period, and the mode laser power is extracted from it as the laser power of each height layer of the laser beam in the current monitoring time period. Furthermore, the power data set of the laser beam in the current monitoring time period is composed of the laser power of each height layer of the laser beam in the current monitoring time period;

[0075] Analyze the power stability corresponding to the laser beam to obtain the power stability value corresponding to the laser beam. The specific analysis is as follows:

[0076] Extract the numerical value of the laser power of each height layer of the laser beam in the current monitoring time period from the power data set of the laser beam in the current monitoring time period and denote it as jgz d, where d represents the number of each height layer, and d = 1, 2, 3…g, and g represents the total number of height layer numbers;

[0077] According to the formula: , the power stability value Zgz corresponding to the laser beam is obtained. Among them, e represents the set natural constant, and u d represents the influence factor corresponding to the set d-th height layer, and jgz d * represents the reference laser power corresponding to the set d-th height layer;

[0078] The laser focus position of each monitoring moment in the current monitoring time period corresponding to the laser beam is obtained through a focus measuring instrument, and the distance difference is calculated with the corresponding laser focus target position to obtain the laser focus drift distance of the laser beam in the current monitoring time period;

[0079] All the laser focus drift distance data in the current monitoring time period corresponding to the laser beam are sorted from largest to smallest to obtain the laser focus drift distance sequence. The specific expression is: H1≥H2≥H3≥…≥H M , where H1 is the maximum laser focus drift distance, and H M is the minimum laser focus drift distance, and M represents the total number of laser focus drift distances;

[0080] Extract the values of the first one-third of the laser focus drift distances in the laser focus drift distance sequence to calculate the average value, obtain the focus drift value corresponding to the laser beam, and mark it as Zjd;

[0081] According to the formula: , the irradiation evaluation value ZSP of the laser beam is obtained. Among them, η6 and η7 respectively represent the weight coefficients of the power stability value and the focus drift value;

[0082] Compare and analyze the irradiation evaluation value of the laser beam with the preset irradiation evaluation threshold. When the irradiation evaluation value of the laser beam is greater than the preset irradiation evaluation threshold, the laser beam is determined to be in a normal state, and A4 is executed;

[0083] When the irradiation evaluation value of the laser beam is less than or equal to the preset irradiation evaluation threshold, the laser beam is determined to be in an abnormal state, and the irradiation evaluation value of the laser beam is compared and matched with the fault classification table stored in the cloud database to obtain the fault level of the laser beam accordingly;

[0084] Among them, each irradiation evaluation value of the laser beam has a corresponding fault level, and the fault levels include the first-level fault level, the second-level fault level, and the third-level fault level;

[0085] And send the fault level of the laser beam to the management end, and thus display and notify the fault level of the laser beam through the display terminal;

[0086] In a specific embodiment, the present invention determines the irradiation evaluation value of the laser beam by calculating the power stability value and the focus drift value of the laser beam, compares it with a preset threshold value to determine whether the laser beam is in a normal state. If the irradiation state of the laser beam does not meet the standard, it will match the fault classification situation table based on the irradiation evaluation value, automatically determine the fault level, and send the information to the management terminal for display and notification. Thus, through the above multi-level data analysis and intelligent evaluation, the accuracy of laser beam anomaly detection is significantly improved, manual intervention is reduced, the stability of the printing process is ensured, and the working efficiency and finished product quality of the metal 3D printing equipment are optimized.

[0087] A4: Monitor the powder spraying state information of the powder nozzle, and thus perform self-check analysis and processing on the powder spraying state of the powder nozzle. The specific analysis process is as follows:

[0088] Obtain the powder spraying amount of the powder nozzle at each monitoring moment during the current monitoring period through a laser particle counter, obtain the powder spraying amount of the powder nozzle at each monitoring moment during the current monitoring period, and mark it as fpz t ;

[0089] Extract the standard powder spraying amount of the powder nozzle during the current monitoring period from the cloud database as the reference powder spraying amount of the powder nozzle during the current monitoring period, and calibrate it as fpz * ;

[0090] According to the formula: , obtain the powder spraying evaluation value FPZ of the powder nozzle, where t represents the number of each monitoring moment during the current monitoring period, n represents the total number of the numbers of each monitoring moment during the current monitoring period, k represents a natural constant greater than zero, Fpp t represents the powder spraying amount change rate of the powder nozzle during the current monitoring period, Fpp * represents the reference powder spraying amount change rate, and a1, a2, and λ all represent set correction factors;

[0091] Compare and analyze the powder spraying evaluation value of the powder nozzle with a preset powder spraying evaluation threshold value. When the powder spraying evaluation value of the powder nozzle is less than the preset powder spraying evaluation threshold value, the powder nozzle is determined to be in a normal state;

[0092] When the powder spraying evaluation value of the powder nozzle is greater than or equal to the preset powder spraying evaluation threshold value, the powder nozzle is determined to be in an abnormal state, and the powder spraying evaluation value of the powder nozzle is compared and matched with the fault classification situation table stored in the cloud database to obtain the fault level of the powder nozzle;

[0093] Among them, each powder spraying evaluation value of the powder nozzle corresponds to a fault level, and the fault levels include a first-level fault level, a second-level fault level, and a third-level fault level;

[0094] Send the fault level of the powder nozzle to the management terminal, and display and notify the fault level of the powder nozzle through the display terminal accordingly;

[0095] In a specific embodiment, the present invention intelligently monitors and diagnoses the powder spraying state of the powder nozzle, obtains the powder spraying amount data through a laser particle counter, compares it with the standard powder spraying amount in the cloud database, calculates the powder spraying evaluation value of the powder nozzle, and compares it with a preset threshold to determine whether the powder nozzle is in a normal state. If the powder spraying state of the powder nozzle does not meet the standard, match the fault classification table, determine the fault level, and send the fault information to the management terminal and notify through the display terminal, thereby improving the accuracy of powder spraying abnormality detection, reducing manual intervention, and ensuring the quality and stability of metal 3D printing.

[0096] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps described in any one of the above are implemented;

[0097] A computer-readable storage medium stores a computer program thereon, and when the computer program is executed by a processor, the steps described in any one of the above are implemented.

[0098] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to only the specific embodiments. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the art in the relevant technical field can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An intelligent fault diagnosis method for a metal 3D printing device, characterized in that, Including the following steps: By monitoring and analyzing the molten pool state information of the printing device, calculate the quality value of the molten pool; Among them, the specific process of solving the quality value of the molten pool is as follows: Obtain the thermal imaging image of the molten pool within the current monitoring period; Extract the morphological parameters of the molten pool from the thermal imaging image, and then compare and analyze the morphological parameters of the molten pool with the corresponding preset reference intervals respectively. If a certain morphological parameter is within the preset reference interval, assign a value of 0, otherwise assign a value of 1, and accumulate all the assigned values to calculate the morphological value of the molten pool; Extract the color temperature of each detection point in the thermal imaging image, match it with the reference color temperature of the corresponding thermal imaging image of the molten pool, determine the color temperature non-conforming points, and integrate each adjacent color temperature non-conforming point to form each abnormal area in the thermal imaging image; Extract the numerical values of the color temperature of each detection point in each abnormal area for average calculation to obtain the average color temperature of each abnormal area, which is used as the temperature value of each abnormal area in the thermal imaging image; Compare and analyze the temperature values of each abnormal area in the thermal imaging image with the temperature threshold to determine the overheated area and the overcooled area; Calculate the total areas of the overheated area and the overcooled area, and calculate their ratios with the area of the thermal imaging image respectively to obtain the overheat ratio and the overcool ratio. Multiply the overheat ratio and the overcool ratio by the corresponding weight coefficients respectively and then add them up to obtain the temperature distribution value of the molten pool; Identify the defects of the molten pool within the current monitoring period, calculate the total areas of various defect areas, and calculate their ratios with the total area of the molten pool to obtain the defect value of the molten pool; Determine the quality value of the molten pool based on the morphological value, the temperature distribution value and the defect value; Accordingly, determine whether the molten pool is in an abnormal state. If the molten pool is determined to be in an abnormal state, trigger a fault self-check instruction; Based on the triggered fault self-check instruction, obtain the vibration amplitude data and the positioning error data of the laser scanning galvanometer, calculate the vibration amplitude change difference and the positioning error fluctuation value, determine the stability evaluation value of the laser scanning galvanometer, and determine whether it is in a normal state; If it is in a normal state, obtain the laser power of each height layer of the laser beam and the laser focus drift distance of the laser beam, calculate the power stability value and the focus drift value, determine the irradiation evaluation value of the laser beam, and determine whether it is in a normal state; If it is in a normal state, obtain the powder spraying amount of the powder nozzle, compare it with the reference powder spraying amount for analysis, determine the powder spraying evaluation value of the powder nozzle, and determine whether it is in a normal state; If the laser scanning galvanometer, the laser beam and the powder nozzle are all determined to be in an abnormal state, match the corresponding fault classification table to determine the fault level.

2. The intelligent fault diagnosis method of a metal 3D printing device according to claim 1, characterized in that, The specific process of solving the vibration amplitude change difference is as follows: Obtain the vibration amplitude data of the laser scanning galvanometer within the current monitoring period, and perform average calculation on it to obtain the average vibration amplitude of the laser scanning galvanometer. Calculate the difference between the vibration amplitude data of the laser scanning galvanometer at each monitoring moment and the average vibration amplitude to obtain the vibration amplitude change value, and extract the maximum vibration amplitude change value and the minimum vibration amplitude change value from it for difference calculation to obtain the vibration amplitude change difference.

3. The intelligent fault diagnosis method of a metal 3D printing device according to claim 1, characterized in that The specific process of solving the positioning error fluctuation value is as follows: Obtain the error between the target position and the actual position of the laser scanning galvanometer under the same instruction during the current monitoring time period, and construct a positioning error sequence; Eliminate the maximum positioning error and the minimum positioning error in the positioning error sequence, and then calculate the root mean square error of the error sequence after elimination to obtain the positioning error fluctuation value.

4. The intelligent fault diagnosis method of a metal 3D printing device according to claim 1, wherein, The specific process of solving the power stability value is as follows: Divide the laser beam into each height layer according to a preset height to obtain each height layer of the laser beam; Obtain the laser power of each height layer of the laser beam corresponding to each monitoring moment during the current monitoring time period, and arrange them in descending order to obtain a laser power sequence, and extract the mode laser power from it as the laser power of each height layer of the laser beam corresponding to the current monitoring time period, thereby constituting a power data set of the laser beam corresponding to the current monitoring time period; Thus, analyze the power stability corresponding to the laser beam to obtain the power stability value corresponding to the laser beam.

5. The intelligent fault diagnosis method of a metal 3D printing device according to claim 1, wherein, The specific process of solving the focus drift value is as follows: Obtain the laser focus position of the laser beam corresponding to each monitoring moment during the current monitoring time period, calculate the distance difference between it and the corresponding laser focus target position to obtain the laser focus drift distance of the laser beam during the current monitoring time period, and sort it in descending order to obtain a laser focus drift distance sequence; Extract the values of the first one-third of the laser focus drift distances in the laser focus drift distance sequence and calculate the average value to obtain the focus drift value corresponding to the laser beam.

6. The intelligent fault diagnosis method of a metal 3D printing device according to claim 1, characterized in that, The specific process of solving the powder spraying evaluation value of the powder nozzle is as follows: Obtain the powder spraying amount of the powder nozzle at each monitoring moment corresponding to the current monitoring time period, and obtain the powder spraying amount fpz of the powder nozzle at each monitoring moment corresponding to the current monitoring time period t ; Extract the standard powder spraying amount of the powder nozzle corresponding to the current monitoring time period as the reference powder spraying amount fpz of the powder nozzle corresponding to the current monitoring time period * ; According to the formula: , the powder spraying evaluation value FPZ of the powder nozzle is obtained, where t represents the number of each monitoring moment in the current monitoring time period, n represents the total number of the numbers of each monitoring moment in the current monitoring time period, k represents a natural constant greater than zero, and Fpp t represents the powder spraying rate of change of the powder nozzle corresponding to the current monitoring time period, and Fpp * represents the reference powder spraying rate of change, and a1, a2, and λ all represent the set correction factors.

7. The intelligent fault diagnosis method of a metal 3D printing device according to claim 1, characterized in that The specific process of determining the fault level is as follows: Compare and match the stability evaluation value of the laser scanning galvanometer, the irradiation evaluation value of the laser beam, and the powder spraying evaluation value of the powder nozzle with the corresponding fault classification table stored in the cloud database respectively, and accordingly obtain the corresponding fault level.

8. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 7.

Images