Solder skipping detection system and method

Abstract

The invention discloses a missing welding detection system and method. The missing welding detection system comprises a data acquisition assembly, an anomaly detection assembly, a time sequence analysis assembly, a welding quality detection assembly and a welding control assembly. The data acquisition assembly acquires parameters in the welding process and transmits data to the anomaly detection assembly. The anomaly detection assembly performs preliminary analysis on the received data, identifies potential solder skips and sends the information to the time sequence analysis assembly. And after the time sequence analysis assembly receives the potential solder skips, abnormal trend analysis is carried out in combination with historical welding data. And the welding quality detection assembly further evaluates the quality of the welding points, if it is confirmed that welding skips or quality problems exist, an abnormal signal is sent to the welding control assembly, and the welding control assembly adjusts welding parameters or actions according to the abnormal signal so as to achieve welding correction. According to the invention, the structure is simple, solder skips can be rapidly identified, complex solder skips which are influenced by various factors can be identified through trend judgment, and the detection reliability is obviously improved.

Description

Technical Field

[0001] The invention relates to the field of welding, and in particular to a welding leak detection system and method. Background Art

[0002] Welding leak detection technology is widely used in the field of mechanical manufacturing. Welding leak detection is a detection method used to determine whether there are problems such as missing welding points and insufficient welding quality during welding. The purpose is to ensure the structural strength of welded parts and avoid welding defects that cause product failures during use. In the prior art, welding leak detection methods mainly include image processing, manual recognition and simple electrical detection methods. Image processing methods usually use cameras to collect images of welding points and detect welding defects through image algorithms. However, due to the complex welding environment, such as changes in lighting conditions and smoke interference, the accuracy of image processing is easily affected, resulting in missed detection or false detection. The manual recognition method relies on the experience of the operator to manually judge the welding quality, but this method is labor-intensive and subject to subjective influences, and cannot achieve standardized detection. Simple electrical detection methods identify welding problems by monitoring the basic changes in current, but their sensitivity is low and it is difficult to detect smaller defects. Therefore, in order to solve the above problems, a welding leak detection system and method are needed. Summary of the invention

[0003] In view of the shortcomings of the prior art, the present invention proposes a welding leak detection system and method, wherein a specific technical solution of the welding leak detection system is as follows:

[0004] A welding leak detection system, characterized in that:

[0005] It includes data acquisition component, anomaly detection component, timing analysis component, welding quality detection component and welding control component;

[0006] The data acquisition component collects parameters in the welding process through a sensor network, the abnormality detection component is connected to the data acquisition component, and the data acquisition component transmits the collected data to the abnormality detection component;

[0007] The abnormality detection component is used for preliminary data analysis to identify potential solder leaks. The timing analysis component is connected to the abnormality detection component, and the abnormality detection component sends the identified potential solder leaks to the timing analysis component.

[0008] The timing analysis component detects abnormal trends in combination with welding history data after receiving potential leaky solder joints;

[0009] The welding quality detection component is connected to the timing analysis component, and the welding quality detection component is used to evaluate the quality of the welding point;

[0010] The welding control component receives the abnormal signal of the welding quality detection component, and the welding control component adjusts the welding action according to the abnormal signal to correct the welding.

[0011] In order to better implement the present invention, it can be further provided that: the welding quality detection component is connected to the timing analysis component, and the welding quality detection component is used to perform quality assessment on welding points.

[0012] Furthermore: the welding parameters collected by the sensor network include welding current, welding temperature, welding speed, and welding point coordinates.

[0013] Further: the anomaly detection component uses an isolation forest model to perform preliminary analysis on the data.

[0014] Further: after receiving the potential solder leak point, the timing analysis component calculates the mean value and standard deviation of the data point through a sliding window method to detect abnormal trends.

[0015] A working method of a leak welding detection system, the specific technical scheme is as follows:

[0016] A working method of a welding leak detection system, characterized in that:

[0017] The steps include:

[0018] S1: The data acquisition component monitors various parameters of the welding process in real time through a sensor network, and the data acquisition component transmits the collected parameter data to the abnormality detection component;

[0019] S2: the anomaly detection component performs a preliminary analysis on the received data, identifies potential solder leaks through a preset isolation forest model, and then sends the detected potential solder leak data to the timing analysis component;

[0020] S3: After receiving the potential solder leak point, the timing analysis component performs context analysis on the potential solder leak point by using a sliding window method to further confirm whether it is an abnormal point;

[0021] S4: After the timing analysis component confirms the abnormality, the welding quality detection component performs quality assessment on the abnormal welding point;

[0022] S5: The welding quality detection component evaluates the quality of the welding point by comparing the abnormal point with the standard welding parameters;

[0023] S6: When the welding quality detection component confirms that there is a leak or the welding quality is insufficient, an alarm signal is sent to the alarm component and fed back to the welding control component at the same time;

[0024] S7: The welding control component automatically adjusts the welding parameters or controls the working state of the welding equipment according to the feedback information, and corrects the welding action in time.

[0025] To better implement the present invention, it can be further provided that: the preset isolation forest model includes an isolation forest model of t isolated trees constructed using the data set X;

[0026] The S2 comprises the following steps:

[0027] S2-1: the anomaly detection component receives the parameters collected by the data collection component, and the parameters constitute the data point x;

[0028] S2-2: The anomaly detection component obtains the data point x in the isolated tree T i The path length h(x,T i );

[0029] S2-3: The anomaly detection component detects the path length h(x,T i ) to get the average path length h(x) of data point x:

[0030]

[0031] S2-4: The anomaly detection component calculates the normalization factor c(n) of the path length through the concordance number H(n-1):

[0032]

[0033] S2-5: The anomaly detection component calculates the anomaly score of the data point x by the average path length h(x) and the normalization factor c(n):

[0034]

[0035] S2-6: The anomaly detection component compares the anomaly score s(x,n) with a preset threshold p. If s(x,n)>p, the data point x is judged to be a potential solder leak point, and the potential solder leak point is sent to the timing analysis component.

[0036] Further: S3 comprises the following steps:

[0037] S3-1: The timing analysis component locates potential leaky solder joints x in the time series data according to the detection requirements. i , extract a fixed length N of continuous historical data to form a sliding window:

[0038]

[0039] Among them, x ij For data point x iThe j-th eigenvalue of ;

[0040] S3-2: The time series analysis component performs a calculation on the data point x in the sliding window N. i Calculate the mean μ for each feature j j :

[0041]

[0042] Where N is the number of data points in the window;

[0043] S3-3: The time series analysis component calculates the data point x in the sliding window. i The standard deviation σ j :

[0044]

[0045] S3-4: Every time a new data point x is received n When the timing analysis component calculates the average value μ based on the sliding window j and standard deviation σ j To set range boundaries:

[0046]

[0047] Where k is the coefficient of the adjustment range width;

[0048] S3-5: For each new data point x n , for each eigenvalue x nj Test each one to determine whether it is within the normal range;

[0049] S3-6: If x nj ∈[μ j -k·σ j ,μ+k·σ j ], then the data point x n Mark normal values ​​if Then the data point x n Mark as abnormal;

[0050] S3-7: If data point x is found n abnormal, then judge that the data point x n For a potential solder leak, the timing analysis component sends the abnormal data point x n To the welding quality detection component.

[0051] The beneficial effects of the present invention are as follows: the present invention has a simple structure and can quickly identify leaky welds. By setting a data acquisition component, an abnormality detection component and a timing analysis component, the data acquisition component can realize real-time monitoring of welding process parameters. The abnormality detection component performs preliminary analysis after data acquisition and can quickly locate possible leaky welds. After receiving potential leaky welds, the timing analysis component further combines historical welding data to analyze the context trend of potential leaky welds, thereby enhancing the accuracy of leaky weld judgment. Not only can isolated anomalies be identified, but also complex leaky welds affected by multiple factors can be identified through trend judgment, significantly improving the reliability of detection. The welding control component can adjust the welding action in real time according to the abnormal signal provided by the timing analysis component, and promptly correct defects in the welding process. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is the overall framework diagram of the present invention;

[0053] Figure 2 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0054] 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 only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] A leak detection system such as Figure 1 As shown:

[0056] A leak welding detection system includes a data acquisition component, an abnormality detection component, a timing analysis component, a welding quality detection component and a welding control component. The data acquisition component collects parameters in the welding process through a sensor network. Specifically, the welding parameters collected by the sensor network include welding current, welding temperature, welding speed, and welding point coordinates.

[0057] The anomaly detection component is connected to the data collection component, and the anomaly detection component uses the isolation forest model to perform preliminary analysis on the data. The data collection component transmits the collected data to the anomaly detection component.

[0058] The anomaly detection component is used for preliminary data analysis to identify potential leaky solder joints. The timing analysis component is connected to the anomaly detection component. The anomaly detection component sends the identified potential leaky solder joints to the timing analysis component. After receiving the potential leaky solder joints, the timing analysis component combines the welding history data to detect abnormal trends. Specifically, after receiving the potential leaky solder joints, the timing analysis component calculates the mean and standard deviation of the data points through a sliding window method to detect abnormal trends.

[0059] The welding quality detection component is connected to the timing analysis component, and the welding quality detection component is used to evaluate the quality of the welding point to confirm the welding quality. The welding control component receives the abnormal signal of the welding quality detection component, and the welding control component adjusts the welding action according to the abnormal signal to correct the welding.

[0060] A working method of a leak welding detection system Figure 2 As shown:

[0061] A working method of a leak welding detection system comprises the following steps:

[0062] S1: The data acquisition component monitors various parameters of the welding process in real time through a sensor network, and the data acquisition component transmits the collected parameter data to the abnormality detection component;

[0063] S2: the anomaly detection component performs a preliminary analysis on the received data, identifies potential solder leaks through a preset isolation forest model, and then sends the detected potential solder leak data to the timing analysis component;

[0064] Specifically, the preset isolation forest model includes an isolation forest model of t isolated trees constructed using the data set X;

[0065] The S2 comprises the following steps:

[0066] S2-1: the anomaly detection component receives the parameters collected by the data collection component, and the parameters constitute the data point x;

[0067] S2-2: The anomaly detection component obtains the data point x in the isolated tree T i The path length h(x,T i );

[0068] S2-3: The anomaly detection component detects the path length h(x,T i ) to get the average path length h(x) of data point x:

[0069]

[0070] S2-4: The anomaly detection component calculates the normalization factor c(n) of the path length through the concordance number H(n-1):

[0071]

[0072] S2-5: The anomaly detection component calculates the anomaly score of the data point x by the average path length h(x) and the normalization factor c(n):

[0073]

[0074] S2-6: The anomaly detection component compares the anomaly score s(x,n) with a preset threshold p. If s(x,n)>p, the data point x is judged to be a potential solder leak point, and the potential solder leak point is sent to the timing analysis component.

[0075] S3: After receiving the potential solder leak point, the timing analysis component performs context analysis on the potential solder leak point by using a sliding window method to further confirm whether it is an abnormal point;

[0076] Specifically, S3 includes the following steps:

[0077] S3-1: The timing analysis component locates potential leaky solder joints x in the time series data according to the detection requirements. i , extract a fixed length N of continuous historical data to form a sliding window:

[0078]

[0079] Among them, x ij For data point x i The j-th eigenvalue of ;

[0080] S3-2: The time series analysis component performs a calculation on the data point x in the sliding window N. i Calculate the mean μ for each feature j j :

[0081]

[0082] Where N is the number of data points in the window;

[0083] S3-3: The time series analysis component calculates the data point x in the sliding window. i The standard deviation σ j :

[0084]

[0085] S3-4: Every time a new data point x is received n When the timing analysis component calculates the average value μ based on the sliding window j and standard deviation σ j To set range boundaries:

[0086] [μ j -k·σ j ,μ+k·σ j ];

[0087] Where k is the coefficient of the adjustment range width;

[0088] S3-5: For each new data point x n , for each eigenvalue x nj Test each one to determine whether it is within the normal range;

[0089] S3-6: If x nj ∈[μ j -k·σ j ,μ+k·σ j ], then the data point x n Mark normal values ​​if Then the data point x n Mark as abnormal;

[0090] S3-7: If data point x is found n abnormal, then judge that the data point x n For a potential solder leak, the timing analysis component sends the abnormal data point x n To the welding quality detection component.

[0091] S4: After the timing analysis component confirms the abnormality, the welding quality detection component performs quality assessment on the abnormal welding point;

[0092] S5: The welding quality detection component evaluates the quality of the welding point by comparing the abnormal point with the standard welding parameters;

[0093] S6: When the welding quality detection component confirms that there is a leak or the welding quality is insufficient, an alarm signal is sent to the alarm component and fed back to the welding control component at the same time;

[0094] S7: The welding control component automatically adjusts the welding parameters or controls the working state of the welding equipment according to the feedback information, and corrects the welding action in time.

[0095] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0096] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A welding leak detection system, characterized in that: It includes data acquisition component, anomaly detection component, timing analysis component, welding quality detection component and welding control component; The data acquisition component collects parameters in the welding process through a sensor network, the abnormality detection component is connected to the data acquisition component, and the data acquisition component transmits the collected data to the abnormality detection component; The abnormality detection component is used for preliminary data analysis to identify potential solder leaks. The timing analysis component is connected to the abnormality detection component, and the abnormality detection component sends the identified potential solder leaks to the timing analysis component. The timing analysis component detects abnormal trends in combination with welding history data after receiving potential leaky solder joints; The welding quality detection component is connected to the timing analysis component, and the welding quality detection component is used to evaluate the quality of the welding point; The welding control component receives the abnormal signal of the welding quality detection component, and the welding control component adjusts the welding action according to the abnormal signal to correct the welding.

2. A welding leak detection system according to claim 1, characterized in that: The welding parameters collected by the sensor network include welding current, welding temperature, welding speed, and welding point coordinates.

3. A welding leak detection system according to claim 2, characterized in that: The anomaly detection component uses the isolation forest model to perform preliminary analysis of the data.

4. A welding leak detection system according to claim 3, characterized in that: After receiving potential solder leaks, the timing analysis component calculates the mean and standard deviation of the data points through a sliding window method to detect abnormal trends.

5. A working method of a welding leak detection system according to claim 4, characterized in that: The steps include: S1: The data acquisition component monitors various parameters of the welding process in real time through a sensor network, and the data acquisition component transmits the collected parameter data to the abnormality detection component; S2: the anomaly detection component performs a preliminary analysis on the received data, identifies potential solder leaks through a preset isolation forest model, and then sends the detected potential solder leak data to the timing analysis component; S3: After receiving the potential solder leak point, the timing analysis component performs context analysis on the potential solder leak point by using a sliding window method to further confirm whether it is an abnormal point; S4: After the timing analysis component confirms the abnormality, the welding quality detection component performs quality assessment on the abnormal welding point; S5: The welding quality detection component evaluates the quality of the welding point by comparing the abnormal point with the standard welding parameters; S6: When the welding quality detection component confirms that there is a leak or the welding quality is insufficient, an alarm signal is sent to the alarm component and fed back to the welding control component at the same time; S7: The welding control component automatically adjusts the welding parameters or controls the working state of the welding equipment according to the feedback information, and corrects the welding action in time.

6. The working method of the welding leak detection system according to claim 5, characterized in that: The preset isolation forest model includes an isolation forest model of t isolated trees constructed using the data set X; The S2 comprises the following steps: S2-1: the anomaly detection component receives the parameters collected by the data collection component, and the parameters constitute the data point x; S2-2: The anomaly detection component obtains the data point x in the isolated tree T i The path length h(x,T i ); S2-3: The anomaly detection component detects the path length h(x,T i ) to get the average path length h(x) of data point x: S2-4: The anomaly detection component calculates the normalization factor c(n) of the path length through the concordance number H(n-1): S2-5: The anomaly detection component calculates the anomaly score of the data point x by the average path length h(x) and the normalization factor c(n): S2-6: The anomaly detection component compares the anomaly score s(x,n) with a preset threshold p. If s(x,n)>p, the data point x is judged to be a potential solder leak point, and the potential solder leak point is sent to the timing analysis component.

7. The working method of the welding leak detection system according to claim 6, characterized in that: The S3 comprises the following steps: S3-1: The timing analysis component locates potential leaky solder joints x in the time series data according to the detection requirements. i , extract a fixed length N of continuous historical data to form a sliding window: x i =[x i1 ,x i2 ,…,x id ]; Among them, x id For data point x i The d-th eigenvalue of ; S3-2: The time series analysis component performs a calculation on the data point x in the sliding window N. i Calculate the mean μ for each feature j j : Where N is the number of data points in the window; S3-3: The time series analysis component calculates the data point x in the sliding window. i The standard deviation σ j : S3-4: Every time a new data point x is received n When the timing analysis component calculates the average value μ based on the sliding window j and standard deviation σ j To set range boundaries: [m j -k·s j ,μ+k·s j ]; Where k is the coefficient of the adjustment range width; S3-5: For each new data point x n , for each eigenvalue x nj Test each one to determine whether it is within the normal range; S3-6: If x nj ∈[μ j -k·σ j ,μ+k·σ j ], then the data point x n Mark normal value if Then the data point x n Mark as abnormal; S3-7: If data point x is found n abnormal, then judge that the data point x n For a potential solder leak, the timing analysis component sends the abnormal data point x n To the welding quality detection component.

Images