Abnormal data management system and method for placement machines based on data analysis

Abstract

The invention discloses a chip mounter abnormal data management system and method based on data analysis, and belongs to the field of abnormal monitoring technology. The invention classifies and processes the abnormalities of the chip mounter, calculates the abnormal thresholds and abnormal occurrence points of all abnormal types of the chip mounter; determines the acquisition cycle of the characteristic data of the chip mounter, uses the acquisition cycle to collect the characteristic data of the chip mounter in real time, draws a real-time characteristic data change curve chart, and performs real-time abnormal warning for the chip mounter; calculates the warning response threshold of the chip mounter control personnel; calculates the mutation time of each abnormal warning, judges the mutation time of each warning, and searches for the emergency abnormal warning of the chip mounter; determines the normal point of the emergency abnormal warning; calculates the abnormal threshold advance point of the emergency abnormal warning; calculates the abnormal threshold optimization point of the emergency abnormal warning, and uses the optimization point to replace the abnormal threshold of the emergency abnormal threshold.

Description

Technical Field

[0001] The present invention relates to the technical field of abnormality monitoring, and in particular to a chip placement machine abnormality data management system and method based on data analysis. Background Art

[0002] In the 1980s, surface mount technology (SMT) gradually replaced traditional through-hole technology (THT), and placement machines began to become popular in the electronics manufacturing industry. Early placement machines mainly relied on mechanical control, data management was relatively simple, and exception handling mainly relied on the operator's experience and manual intervention; with the introduction of computer control systems, placement machines began to have basic automation capabilities. Anomaly detection and data management began to rely on simple sensors and logic control systems, but the detection capabilities were limited, and data analysis was mainly performed after the fact. With the development of modern artificial intelligence (AI) and machine learning technologies, the management of abnormal data for placement machines has entered a new stage. By training models, the system is able to identify complex abnormal patterns and even predict potential equipment failures and production problems. Adaptive control systems can dynamically adjust the parameters of placement machines based on real-time data to further improve production stability and efficiency. However, there are still some problems in the use of artificial intelligence to monitor abnormalities in placement machines. The control personnel of the placement machines have become overly dependent on artificial intelligence automation, and some abnormal mutations in the placement machines are extremely fast. After the system detects the risk of abnormalities and issues an early warning, when the control personnel receive the early warning information and respond, the abnormality has already occurred, causing damage to the placement machine. Therefore, under the condition that the control personnel's response time cannot be changed in advance, how to avoid the occurrence of such emergency mutation abnormalities as much as possible and reduce the damage caused by emergency abnormalities to the placement machine is crucial to the abnormal monitoring of the placement machine. Summary of the invention

[0003] The purpose of the present invention is to provide a system and method for managing abnormal data of a placement machine based on data analysis to solve the problems raised in the above-mentioned background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0005] A method for managing abnormal data of a placement machine based on data analysis, the method comprising the following steps:

[0006] S100, collecting all abnormal events that occurred in the placement machine in history, searching for characteristic data reflecting each abnormal event of the placement machine, classifying and processing the abnormalities of the placement machine, and calculating the abnormal thresholds and abnormal occurrence points of all abnormal types of the placement machine using each characteristic data;

[0007] Furthermore, the specific steps for classifying and processing the abnormalities of the placement machine are as follows:

[0008] S101, collect the equipment data of all abnormal events that occurred in the placement machine in history, extract the data value of each equipment data before and after the abnormality occurs, and set the data value of each equipment data before the abnormality occurs in each abnormal event to be , It means extracting the 1st, 2nd, 3rd, ..., nth device data value before the abnormality occurs in each abnormal event, and extracting the data value of each device after the abnormality occurs in each abnormal event. , Indicates extracting the 1st, 2nd, 3rd, ..., nth device data value after the abnormality occurs in each abnormal event, where n is a positive integer;

[0009] S102, using the extracted data values ​​of each device before and after the abnormality occurs in each abnormal event to find the characteristic data of each abnormal event, the formula is: In the formula, Sc represents the data difference between each device data in each abnormal event before and after the abnormality occurs; the data difference obtained by calculating all device data in each abnormal event is , Indicates the difference between the 1st, 2nd, 3rd, ..., nth device data before and after the abnormality occurs in each abnormal event. Compare the differences between the n types of device data and find the data type corresponding to the maximum difference in device data as the characteristic data of the abnormal event; the characteristic data of each abnormal event in the collected history is , represents the characteristic data of the 1st, 2nd, 3rd, ..., mth abnormal events in the collected history, where m is a positive integer; the abnormalities are classified using the characteristic data of each abnormal event, and the abnormal events with the same characteristic data are classified into one category, and finally all the abnormalities that occurred in the history of the placement machine are classified into , Indicates the 1st, 2nd, 3rd, ..., hth abnormality that occurs in the placement machine;

[0010] S103, collect the characteristic data value of each abnormality of the placement machine in the history , It represents the characteristic data of the 1st, 2nd, 3rd, ..., kth occurrence of each anomaly in the collected history, where k is a positive integer. The characteristic data of each anomaly in the collected history when it occurs is used to calculate each anomaly threshold, and the formula is:

[0011] ,

[0012] In the formula, Ina represents the calculated anomaly threshold of each anomaly type, Tz represents the characteristic data collected when each anomaly occurs in the history, k represents the number of characteristic data collected when each anomaly occurs in the history, and st_Tz represents the standard deviation of k characteristic data collected when each anomaly occurs; the calculated anomaly threshold of each anomaly type is , represents the calculated abnormality thresholds of the 1st, 2nd, 3rd, ..., hth abnormalities of the placement machine;

[0013] S104. Calculate the abnormality occurrence point using the characteristic data of each abnormality in the collected history, and the formula is:

[0014] ,

[0015] In the formula, Ha represents each abnormal occurrence point of the calculation, Indicates the minimum value of characteristic data of each anomaly in the collected history, It represents the maximum value of the characteristic data of each anomaly in the collected history; after calculating each anomaly occurrence point using the characteristic data, the occurrence points of all types of anomalies are obtained as

[0016] , Indicates the 1st, 2nd, 3rd, ..., hth abnormality occurrence points of the placement machine.

[0017] After classifying the anomalies using the device data of the abnormalities that occurred in the placement machine in the past, the various anomalies in the placement machine can be handled separately, and when early warning of the anomalies is given, the inaccurate prediction of different anomalies by a single data can be avoided. The characteristic data corresponding to each anomaly can be used to accurately predict each anomaly; and when handling the anomalies, different processing plans can be formulated according to the type of anomaly, which greatly increases the work efficiency of the system.

[0018] S200, collecting historical working data of the placement machine, using the working data to determine the collection period of the feature data of the placement machine, using the collection period to collect the feature data of the placement machine in real time, drawing a real-time feature data change curve, and using the abnormal threshold to issue a real-time abnormal warning to the placement machine;

[0019] Furthermore, the specific steps of using the abnormal threshold to carry out real-time abnormal warning for the placement machine are as follows:

[0020] S201, collect the working data of the placement machine in history, analyze the collected working data of the placement machine, and extract the time difference between each change of the working data of the placement machine , Indicates the time difference of the 1st, 2nd, 3rd, ..., jth change of the working data of the placement machine in the collected history, where j is a positive integer; the average value of the time difference of the jth extracted working data change is calculated, and the average value of the working data change time difference is used as the collection cycle Cy for monitoring each abnormal characteristic data of the placement machine;

[0021] S202, using the calculated collection cycle to collect feature data of each abnormality when the placement machine is working, assuming that the real-time feature data of each abnormality collected is , It represents the 1st, 2nd, 3rd, ..., pth feature data of each abnormality collected in real time, where p is a positive integer; a curve graph is constructed using the collected real-time feature data, the calculated abnormal threshold and the abnormality occurrence are marked in the curve graph, and the constructed curve graph is judged in real time. When the chip mounter is judged to have the corresponding type of abnormal risk and issue an early warning; when When the chip placement machine is working normally, there is no abnormal risk. By collecting the real-time feature data of the chip placement machine, it is determined that the abnormal type that issues the warning is , It indicates the 1st, 2nd, 3rd, ..., qth abnormal warning after judgment, where q is a positive integer.

[0022] The abnormal threshold is calculated using the characteristic data in the history. The abnormal threshold can be used to judge the state of the placement machine when it is working in real time, and each abnormality can be predicted in advance to avoid the occurrence of equipment abnormalities and reduce the probability of equipment damage.

[0023] S300, collecting the response time of the control personnel receiving the warning information to perform maintenance on the placement machine after the placement machine performs abnormal warning in the history, and calculating the warning response threshold of the placement machine control personnel by using the response time of the control personnel in the history;

[0024] Furthermore, the specific steps of calculating the warning response threshold of the control personnel of the placement machine by using the response time of the control personnel in history are as follows:

[0025] S301, after the SMT machine issues an abnormal warning in the acquisition history, the control personnel receives the warning information and performs maintenance on the SMT machine. , It indicates the 1st, 2nd, 3rd, ..., dth response time of the control personnel to perform maintenance on the SMT machine after the SMT machine issues an abnormal warning in the collected history, where d is a positive integer;

[0026] S302, using the collected history of abnormal warnings of the placement machine, the control personnel's response time to receive the warning information and perform maintenance on the placement machine is calculated, and the formula is:

[0027] ,

[0028] In the formula, Itr represents the warning response threshold of the controller. Indicates the minimum response time of the control personnel after receiving the warning information in the collected history. Indicates the maximum value of the response time of the control personnel after receiving the warning information in the collected history; It indicates the standard deviation of the response time of the control personnel after receiving the warning information in the collected history.

[0029] S400, analyzing the characteristic data change curve of the real-time placement machine, calculating the mutation time of each abnormal warning according to the abnormal occurrence point, judging the mutation time of each warning by using the warning response threshold interval, and finding the emergency abnormal warning of the placement machine;

[0030] Furthermore, the specific steps for finding the emergency abnormal warning of the placement machine are:

[0031] S401, extracting the characteristic data curve graph constructed by collecting the real-time characteristic data of each abnormality of the placement machine, and calculating the slope of the curve at each abnormality warning, the formula is:

[0032] ,

[0033] In the formula, sd represents the slope of the curve at each abnormal warning. Indicates the characteristic data value of each abnormal warning in the characteristic data curve graph, Indicates the characteristic data value collected before each abnormal warning in the characteristic data curve diagram; Cy indicates the collection cycle when monitoring the characteristic data;

[0034] S402, calculating the mutation time of each abnormal warning according to the calculated slope of the characteristic data curve of each abnormal warning, the formula is:

[0035] ,

[0036] In the formula, tm represents the mutation time of each abnormal warning, Ha represents the abnormal occurrence point of each abnormal warning, Ina represents the mutation time of each abnormal warning, and sd represents the slope of the curve at each abnormal warning;

[0037] S403: Use the warning response threshold to judge the mutation time of each abnormal warning calculated. When abnormal warning is judged, early prevention can be carried out. When it is judged that the control personnel cannot take timely precautions after the abnormal warning, the abnormal warning is an emergency abnormal warning; all abnormal warnings are judged and all emergency abnormal warnings are found.

[0038] By calculating the response time of the control personnel, the warning response threshold is obtained, and the mutation time of each abnormal warning is judged by the warning response threshold. In this way, the types of emergency abnormalities that the control personnel cannot respond to due to the accelerated abnormal changes even if the system issues an abnormal warning can be found; the found emergency abnormalities can be processed to further reduce the losses caused by the placement machine abnormalities.

[0039] S500, after finding the emergency abnormality warning of the placement machine, determine the type of emergency abnormality warning in the placement machine, collect the characteristic data of the emergency abnormality warning type in the history when it is working normally, and determine the normal point of the emergency abnormality warning; use the real-time characteristic data change curve and the warning response threshold to calculate the abnormal threshold advance point of the emergency abnormality warning;

[0040] Furthermore, the specific steps of calculating the abnormal threshold advance point of the emergency abnormal warning using the real-time characteristic data change curve and the warning response threshold are as follows:

[0041] S501, determine the abnormal type of the emergency abnormal warning as an emergency abnormality according to the searched emergency abnormality warning, and collect the characteristic data of the emergency abnormality in the normal operation of the placement machine in the history. , It represents the 1st, 2nd, 3rd, ..., fth characteristic data of the emergency abnormality in the normal operation of the placement machine in the history, where f is a positive integer; the average value of the f characteristic data of the emergency abnormality collected when the placement machine is working normally is calculated as the normal point nor of the emergency abnormality warning, and the abnormal threshold advance point of the emergency abnormality warning is calculated according to the mutation time of the emergency abnormality warning and the warning response threshold, and the formula is:

[0042] ,

[0043] In the formula, Tt represents the abnormal threshold advance point of the emergency abnormal warning, Ina represents each abnormal threshold calculated, Itr represents the warning response threshold, tm represents the mutation time of each abnormal warning, and Cy represents the slope of the curve at each abnormal warning;

[0044] S502: For all the emergency abnormal warnings found, according to step S501, a normal point and an abnormal threshold advance point corresponding to each emergency abnormal warning are calculated.

[0045] S600: Comprehensively analyze the normal point and the advance point of the abnormal threshold of the emergency abnormal warning, calculate the abnormal threshold optimization point of the emergency abnormal warning, and use the optimization point to replace the abnormal threshold of the emergency abnormal threshold.

[0046] Furthermore, the specific steps of replacing the abnormal threshold of the emergency abnormal threshold with the optimization point are as follows:

[0047] S601, calculate the abnormal threshold optimization point of the emergency abnormal warning according to the calculated normal point and abnormal threshold advance point of each emergency abnormal warning, the formula is:

[0048] ,

[0049] In the formula, Ty represents the abnormal threshold optimization point of the emergency abnormality warning, and the abnormal threshold of the emergency abnormality is optimized and replaced by the calculated abnormal threshold optimization point.

[0050] After comprehensively analyzing and calculating the abnormal threshold optimization point of the emergency abnormal warning and the abnormal threshold advance point, the abnormal threshold is replaced with the optimized point. This not only takes into account the problem that the control personnel cannot respond to the warning in time, but also takes into account the problem that the system may increase the probability of misjudging emergency abnormalities due to the simple use of the advance point. Through comprehensive analysis of the two aspects, the abnormal threshold of the emergency abnormality is optimized, and the time from the abnormal warning to the abnormal occurrence is increased, so that the control personnel can respond in time, further reducing the occurrence of abnormalities in the placement machine.

[0051] The abnormal data management system of the placement machine based on data analysis includes data collection module, abnormal classification module, abnormal warning module, abnormal warning analysis module and optimization module;

[0052] The data collection module is used to collect the equipment data and working data of the placement machine when an abnormality occurs in the history;

[0053] The abnormality classification module is used to analyze the device data collected in the history of the placement machine when the abnormality occurs, find the characteristic data of each abnormal event, and classify the placement machine abnormality;

[0054] The abnormal warning module is used to calculate the abnormal threshold value by using the characteristic data of the placement machine when the abnormality occurs in the history, collect the real-time characteristic data of the placement machine when it is working, and judge whether there is an abnormal risk of the placement machine in real-time operation;

[0055] The abnormal warning analysis module is used to calculate the warning response threshold by using the historical control personnel's response time to the abnormal warning, analyze all real-time abnormal thresholds, calculate the mutation time of each abnormal warning, judge the mutation time, and find emergency abnormal warnings;

[0056] The optimization module is used to calculate the normal point and the advance point of the abnormal warning for the emergency abnormal warning, determine the abnormal threshold optimization point of the emergency abnormal warning according to the normal point and the advance point of the abnormal warning, and use the abnormal threshold optimization point to optimize and replace the abnormal threshold of the emergency abnormality.

[0057] The abnormal warning module includes a threshold calculation unit, a real-time acquisition unit and a warning judgment unit;

[0058] The threshold value calculation unit is used to calculate the abnormality threshold value of each abnormality by using the characteristic data collected in the history when the placement machine has an abnormality;

[0059] The real-time acquisition unit is used to analyze the working data of the placement machine in history, calculate the acquisition cycle, use the acquisition cycle to collect the real-time characteristic data of the placement machine when it is working, and draw a characteristic data curve graph;

[0060] The early warning judgment unit is used to judge the characteristic data collected in real time by using the abnormality threshold of each abnormality, determine whether there is an abnormality risk in the placement machine and issue an early warning.

[0061] The abnormal warning analysis module includes a warning response threshold calculation unit, a mutation time calculation unit and an emergency abnormal warning search unit;

[0062] The warning response threshold value calculation unit is used to calculate the response time of the control personnel in the history when handling the abnormality of the placement machine after receiving the warning information, so as to obtain the warning response threshold value;

[0063] The mutation time calculation unit is used to analyze the curve graph constructed by the real-time collected feature data, determine the slope of the curve during abnormal warning, and calculate the mutation time of each abnormal warning;

[0064] The emergency abnormality warning search unit is used to judge the mutation time by using the warning corresponding threshold value and search for the emergency abnormality warning.

[0065] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0066] The present invention utilizes the device data of abnormalities that occurred in the placement machine in the history to classify the abnormalities, which can avoid the inaccurate prediction of different abnormalities by single data. Each abnormality can be accurately predicted by utilizing the characteristic data corresponding to each abnormality; and when processing the abnormalities, different processing plans can be formulated according to the type of abnormality, which greatly increases the working efficiency of the system.

[0067] The present invention uses the early warning response threshold to judge the mutation time of each abnormal early warning, and can find out the types of emergency abnormalities to which the control personnel cannot respond due to the accelerated abnormal changes even if the system issues an abnormal early warning; the found emergency abnormalities can be processed to further reduce the losses caused by the abnormalities of the placement machine.

[0068] The present invention performs a comprehensive analysis on the normal point and the advance point of the abnormal threshold of the emergency abnormal warning to obtain the abnormal threshold optimization point, and then uses the optimized point to replace the abnormal threshold, thereby optimizing the abnormal threshold of the emergency abnormality, increasing the time from the abnormal warning to the abnormal occurrence, so that the control personnel can respond in time, and further reducing the occurrence of abnormalities in the placement machine. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0070] Figure 1 It is a module distribution diagram of the abnormal data management system of the chip placement machine based on data analysis of the present invention;

[0071] Figure 2 It is a schematic diagram of the steps of the abnormal data management method of the chip placement machine based on data analysis of the present invention;

[0072] Figure 3 It is a real-time characteristic data curve schematic diagram of the abnormal data management method of the placement machine based on data analysis of the present invention. DETAILED DESCRIPTION

[0073] 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.

[0074] See also Figure 1-Figure 3 , the present invention provides a technical solution:

[0075] A method for managing abnormal data of a placement machine based on data analysis, the method comprising the following steps:

[0076] S100, collecting all abnormal events that occurred in the placement machine in history, searching for characteristic data reflecting each abnormal event of the placement machine, classifying and processing the abnormalities of the placement machine, and calculating the abnormal thresholds and abnormal occurrence points of all abnormal types of the placement machine using each characteristic data;

[0077] The specific steps for classifying and processing the abnormalities of the placement machine are as follows:

[0078] S101, collect the equipment data of all abnormal events that occurred in the placement machine in history, extract the data value of each equipment data before and after the abnormality occurs, and set the data value of each equipment data before the abnormality occurs in each abnormal event to be , It means extracting the 1st, 2nd, 3rd, ..., nth device data value before the abnormality occurs in each abnormal event, and extracting the data value of each device after the abnormality occurs in each abnormal event. , Indicates extracting the 1st, 2nd, 3rd, ..., nth device data value after the abnormality occurs in each abnormal event, where n is a positive integer;

[0079] S102, using the extracted data values ​​of each device before and after the abnormality occurs in each abnormal event to find the characteristic data of each abnormal event, the formula is: In the formula, Sc represents the data difference between each device data in each abnormal event before and after the abnormality occurs; the data difference obtained by calculating all device data in each abnormal event is , Indicates the difference between the 1st, 2nd, 3rd, ..., nth device data before and after the abnormality occurs in each abnormal event. Compare the differences between the n types of device data and find the data type corresponding to the maximum difference in device data as the characteristic data of the abnormal event; the characteristic data of each abnormal event in the collected history is , represents the characteristic data of the 1st, 2nd, 3rd, ..., mth abnormal events in the collected history, where m is a positive integer; the abnormalities are classified using the characteristic data of each abnormal event, and the abnormal events with the same characteristic data are classified into one category, and finally all the abnormalities that occurred in the history of the placement machine are classified into , Indicates the 1st, 2nd, 3rd, ..., hth abnormality that occurs in the placement machine;

[0080] S103, collect the characteristic data value of each abnormality of the placement machine in the history , It represents the characteristic data of the 1st, 2nd, 3rd, ..., kth occurrence of each anomaly in the collected history, where k is a positive integer. The characteristic data of each anomaly in the collected history when it occurs is used to calculate each anomaly threshold, and the formula is:

[0081] ,

[0082] In the formula, Ina represents the calculated anomaly threshold of each anomaly type, Tz represents the characteristic data collected when each anomaly occurs in the history, k represents the number of characteristic data collected when each anomaly occurs in the history, and st_Tz represents the standard deviation of k characteristic data collected when each anomaly occurs; the calculated anomaly threshold of each anomaly type is , represents the calculated abnormality thresholds of the 1st, 2nd, 3rd, ..., hth abnormalities of the placement machine;

[0083] S104. Calculate the abnormality occurrence point using the characteristic data of each abnormality in the collected history, and the formula is:

[0084] ,

[0085] In the formula, Ha represents each abnormal occurrence point of the calculation, Indicates the minimum value of characteristic data of each anomaly in the collected history, It represents the maximum value of the characteristic data of each anomaly in the collected history; after calculating each anomaly occurrence point using the characteristic data, the occurrence points of all types of anomalies are obtained as , Indicates the 1st, 2nd, 3rd, ..., hth abnormality occurrence points of the placement machine.

[0086] After classifying the anomalies using the device data of the abnormalities that occurred in the placement machine in the past, the various anomalies in the placement machine can be handled separately, and when early warning of the anomalies is given, the inaccurate prediction of different anomalies by a single data can be avoided. The characteristic data corresponding to each anomaly can be used to accurately predict each anomaly; and when handling the anomalies, different processing plans can be formulated according to the type of anomaly, which greatly increases the work efficiency of the system.

[0087] S200, collecting historical working data of the placement machine, using the working data to determine the collection period of the feature data of the placement machine, using the collection period to collect the feature data of the placement machine in real time, drawing a real-time feature data change curve, and using the abnormal threshold to issue a real-time abnormal warning to the placement machine;

[0088] The specific steps of using abnormal thresholds to provide real-time abnormal warning for placement machines are as follows:

[0089] S201, collect the working data of the placement machine in history, analyze the collected working data of the placement machine, and extract the time difference between each change of the working data of the placement machine , Indicates the time difference of the 1st, 2nd, 3rd, ..., jth change of the working data of the placement machine in the collected history, where j is a positive integer; the average value of the time difference of the jth extracted working data change is calculated, and the average value of the working data change time difference is used as the collection cycle Cy for monitoring each abnormal characteristic data of the placement machine;

[0090] S202, using the calculated collection cycle to collect feature data of each abnormality when the placement machine is working, assuming that the real-time feature data of each abnormality collected is , It represents the 1st, 2nd, 3rd, ..., pth feature data of each abnormality collected in real time, where p is a positive integer; a curve graph is constructed using the collected real-time feature data, the calculated abnormal threshold and the abnormality occurrence are marked in the curve graph, and the constructed curve graph is judged in real time. When the chip mounter is judged to have the corresponding type of abnormal risk and issue an early warning; when When the chip placement machine is working normally, there is no abnormal risk. By collecting the real-time feature data of the chip placement machine, it is determined that the abnormal type that issues the warning is , It indicates the 1st, 2nd, 3rd, ..., qth abnormal warning after judgment, where q is a positive integer.

[0091] The abnormal threshold is calculated using the characteristic data in the history. The abnormal threshold can be used to judge the state of the placement machine when it is working in real time, and each abnormality can be predicted in advance to avoid the occurrence of equipment abnormalities and reduce the probability of equipment damage.

[0092] S300, collecting the response time of the control personnel receiving the warning information to perform maintenance on the placement machine after the placement machine performs abnormal warning in the history, and calculating the warning response threshold of the placement machine control personnel by using the response time of the control personnel in the history;

[0093] The specific steps for calculating the warning response threshold of the control personnel of the placement machine by using the response time of the control personnel in history are as follows:

[0094] S301, after the SMT machine issues an abnormal warning in the acquisition history, the control personnel receives the warning information and performs maintenance on the SMT machine. , It indicates the 1st, 2nd, 3rd, ..., dth response time of the control personnel to perform maintenance on the SMT machine after the SMT machine issues an abnormal warning in the collected history, where d is a positive integer;

[0095] S302, using the collected history of abnormal warnings of the placement machine, the control personnel's response time to receive the warning information and perform maintenance on the placement machine is calculated, and the formula is:

[0096] ,

[0097] In the formula, Itr represents the warning response threshold of the controller. Indicates the minimum response time of the control personnel after receiving the warning information in the collected history. Indicates the maximum value of the response time of the control personnel after receiving the warning information in the collected history; It indicates the standard deviation of the response time of the control personnel after receiving the warning information in the collected history.

[0098] S400, analyzing the characteristic data change curve of the real-time placement machine, calculating the mutation time of each abnormal warning according to the abnormal occurrence point, judging the mutation time of each warning by using the warning response threshold interval, and finding the emergency abnormal warning of the placement machine;

[0099] The specific steps to find the emergency abnormal warning of the placement machine are:

[0100] S401, extracting the characteristic data curve graph constructed by collecting the real-time characteristic data of each abnormality of the placement machine, and calculating the slope of the curve at each abnormality warning, the formula is:

[0101] ,

[0102] In the formula, sd represents the slope of the curve at each abnormal warning. Indicates the characteristic data value of each abnormal warning in the characteristic data curve graph, Indicates the characteristic data value collected before each abnormal warning in the characteristic data curve diagram; Cy indicates the collection cycle when monitoring the characteristic data;

[0103] S402, calculating the mutation time of each abnormal warning according to the calculated slope of the characteristic data curve of each abnormal warning, the formula is:

[0104] ,

[0105] In the formula, tm represents the mutation time of each abnormal warning, Ha represents the abnormal occurrence point of each abnormal warning, Ina represents the mutation time of each abnormal warning, and sd represents the slope of the curve at each abnormal warning;

[0106] S403: Use the warning response threshold to judge the mutation time of each abnormal warning calculated. When abnormal warning is judged, early prevention can be carried out. When it is judged that the control personnel cannot take timely precautions after the abnormal warning, the abnormal warning is an emergency abnormal warning; all abnormal warnings are judged and all emergency abnormal warnings are found.

[0107] By calculating the response time of the control personnel, the warning response threshold is obtained, and the mutation time of each abnormal warning is judged by the warning response threshold. In this way, the types of emergency abnormalities that the control personnel cannot respond to due to the accelerated abnormal changes even if the system issues an abnormal warning can be found; the found emergency abnormalities can be processed to further reduce the losses caused by the placement machine abnormalities.

[0108] S500, after finding the emergency abnormality warning of the placement machine, determine the type of emergency abnormality warning in the placement machine, collect the characteristic data of the emergency abnormality warning type in the history when it is working normally, and determine the normal point of the emergency abnormality warning; use the real-time characteristic data change curve and the warning response threshold to calculate the abnormal threshold advance point of the emergency abnormality warning;

[0109] The specific steps of calculating the abnormal threshold advance point of the emergency abnormal warning using the real-time characteristic data change curve and the warning response threshold are as follows:

[0110] S501, determine the abnormal type of the emergency abnormal warning as an emergency abnormality according to the searched emergency abnormality warning, and collect the characteristic data of the emergency abnormality in the normal operation of the placement machine in the history. , It represents the 1st, 2nd, 3rd, ..., fth characteristic data of the emergency abnormality in the normal operation of the placement machine in the history, where f is a positive integer; the average value of the f characteristic data of the emergency abnormality collected when the placement machine is working normally is calculated as the normal point nor of the emergency abnormality warning, and the abnormal threshold advance point of the emergency abnormality warning is calculated according to the mutation time of the emergency abnormality warning and the warning response threshold, and the formula is:

[0111] ,

[0112] In the formula, Tt represents the abnormal threshold advance point of the emergency abnormal warning, Ina represents each abnormal threshold calculated, Itr represents the warning response threshold, tm represents the mutation time of each abnormal warning, and Cy represents the slope of the curve at each abnormal warning;

[0113] S502: For all the emergency abnormal warnings found, according to step S501, a normal point and an abnormal threshold advance point corresponding to each emergency abnormal warning are calculated.

[0114] S600: Comprehensively analyze the normal point and the advance point of the abnormal threshold of the emergency abnormal warning, calculate the abnormal threshold optimization point of the emergency abnormal warning, and use the optimization point to replace the abnormal threshold of the emergency abnormal threshold.

[0115] The specific steps of replacing the abnormal threshold of the emergency abnormal threshold with the optimization point are as follows:

[0116] S601, calculate the abnormal threshold optimization point of the emergency abnormal warning according to the calculated normal point and abnormal threshold advance point of each emergency abnormal warning, the formula is:

[0117] ,

[0118] In the formula, Ty represents the abnormal threshold optimization point of the emergency abnormality warning, and the abnormal threshold of the emergency abnormality is optimized and replaced by the calculated abnormal threshold optimization point.

[0119] After comprehensively analyzing and calculating the abnormal threshold optimization point of the emergency abnormal warning and the abnormal threshold advance point, the abnormal threshold is replaced with the optimized point. This not only takes into account the problem that the control personnel cannot respond to the warning in time, but also takes into account the problem that the system may increase the probability of misjudging emergency abnormalities due to the simple use of the advance point. Through comprehensive analysis of the two aspects, the abnormal threshold of the emergency abnormality is optimized, and the time from the abnormal warning to the abnormal occurrence is increased, so that the control personnel can respond in time, further reducing the occurrence of abnormalities in the placement machine.

[0120] The abnormal data management system of the placement machine based on data analysis includes data collection module, abnormal classification module, abnormal warning module, abnormal warning analysis module and optimization module;

[0121] The data collection module is used to collect the equipment data and working data of the placement machine when an abnormality occurs in the history;

[0122] The abnormality classification module is used to analyze the device data collected in the history of the placement machine when the abnormality occurs, find the characteristic data of each abnormal event, and classify the placement machine abnormality;

[0123] The abnormal warning module is used to calculate the abnormal threshold value by using the characteristic data of the placement machine when the abnormality occurs in the history, collect the real-time characteristic data of the placement machine when it is working, and judge whether there is an abnormal risk of the placement machine in real-time operation;

[0124] The abnormal warning analysis module is used to calculate the warning response threshold by using the historical control personnel's response time to the abnormal warning, analyze all real-time abnormal thresholds, calculate the mutation time of each abnormal warning, judge the mutation time, and find emergency abnormal warnings;

[0125] The optimization module is used to calculate the normal point and the advance point of the abnormal warning for the emergency abnormal warning, determine the abnormal threshold optimization point of the emergency abnormal warning according to the normal point and the advance point of the abnormal warning, and use the abnormal threshold optimization point to optimize and replace the abnormal threshold of the emergency abnormality.

[0126] The abnormal warning module includes a threshold calculation unit, a real-time acquisition unit and a warning judgment unit;

[0127] The threshold value calculation unit is used to calculate the abnormality threshold value of each abnormality by using the characteristic data collected in the history when the placement machine has an abnormality;

[0128] The real-time acquisition unit is used to analyze the working data of the placement machine in history, calculate the acquisition cycle, use the acquisition cycle to collect the real-time characteristic data of the placement machine when it is working, and draw a characteristic data curve graph;

[0129] The early warning judgment unit is used to judge the characteristic data collected in real time by using the abnormality threshold of each abnormality, determine whether there is an abnormality risk in the placement machine and issue an early warning.

[0130] The abnormal warning analysis module includes a warning response threshold calculation unit, a mutation time calculation unit and an emergency abnormal warning search unit;

[0131] The warning response threshold value calculation unit is used to calculate the response time of the control personnel in the history when handling the abnormality of the placement machine after receiving the warning information, so as to obtain the warning response threshold value;

[0132] The mutation time calculation unit is used to analyze the curve graph constructed by the real-time collected feature data, determine the slope of the curve during abnormal warning, and calculate the mutation time of each abnormal warning;

[0133] The emergency abnormality warning search unit is used to judge the mutation time by using the warning corresponding threshold value and search for the emergency abnormality warning.

[0134] Embodiment 1: Analyze the device data when abnormalities occur in the history of the placement machine, and classify the abnormalities of the placement machine into mechanical failure and software failure; the characteristic data of the game are amplitude and device response time;

[0135] The abnormal thresholds of the two abnormalities calculated through historical data are 70μm and 1s respectively; the occurrence point is 100μm and 3s; the collection cycle of the design calculation is 1s, and the characteristic data of the placement machine when working is collected in real time and a curve chart is constructed as shown in the figure. Figure 3 As shown, through analyzing the curve chart, the placement machine has two types of failure risks: mechanical and software;

[0136] The warning response threshold is 3s based on the historical response time of the control personnel. The mutation time of the two abnormal warnings is calculated using the formula:

[0137] ,

[0138] The mutation time of the two abnormal warnings is calculated to be 5 and 2.8; the emergency abnormal warning is judged to be a software failure; the normal point of the software failure is determined to be 0.2s through historical data, and the advance point is calculated using the formula:

[0139] ,

[0140] The calculated advance point of software failure is 0.86, the calculated abnormality threshold optimization point of emergency abnormality warning is 0.53, and the new emergency abnormality threshold after replacement is 0.53.

[0141] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0142] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for managing abnormal data of a placement machine based on data analysis, characterized in that: The method comprises the following steps: S100, collecting all abnormal events that occurred in the placement machine in history, searching for characteristic data reflecting each abnormal event of the placement machine, classifying the abnormalities of the placement machine, and calculating the abnormal thresholds and abnormal occurrence points of all abnormal types of the placement machine using each characteristic data; S200, collecting historical working data of the placement machine, using the working data to determine the collection period of the feature data of the placement machine, using the collection period to collect the feature data of the placement machine in real time, drawing a real-time feature data change curve, and using the abnormal threshold to issue a real-time abnormal warning to the placement machine; S300, collecting the response time of the control personnel receiving the warning information to perform maintenance on the placement machine after the placement machine performs abnormal warning in the history, and calculating the warning response threshold of the placement machine control personnel by using the response time of the control personnel in the history; S400, analyzing the characteristic data change curve of the real-time placement machine, calculating the mutation time of each abnormal warning according to the abnormal occurrence point, judging the mutation time of each warning by using the warning response threshold interval, and finding the emergency abnormal warning of the placement machine; S500, after finding the emergency abnormality warning of the placement machine, determine the type of emergency abnormality warning in the placement machine, collect the characteristic data of the emergency abnormality warning type in the history when it is working normally, and determine the normal point of the emergency abnormality warning; use the real-time characteristic data change curve and the warning response threshold to calculate the abnormal threshold advance point of the emergency abnormality warning; S600: Comprehensively analyze the normal point and the advance point of the abnormal threshold of the emergency abnormal warning, calculate the abnormal threshold optimization point of the emergency abnormal warning, and use the optimization point to replace the abnormal threshold of the emergency abnormal threshold.

2. The method for managing abnormal data of a placement machine based on data analysis according to claim 1, characterized in that: The specific steps of classifying and processing the abnormalities of the placement machine in S100 are: S101, collect the equipment data of all abnormal events that occurred in the placement machine in history, extract the data value of each equipment data before and after the abnormality occurs, and set the data value of each equipment data before the abnormality occurs in each abnormal event to be , It means extracting the 1st, 2nd, 3rd, ..., nth device data value before the abnormality occurs in each abnormal event, and extracting the data value of each device after the abnormality occurs in each abnormal event. , Indicates extracting the 1st, 2nd, 3rd, ..., nth device data value after the abnormality occurs in each abnormal event, where n is a positive integer; S102, using the extracted data values ​​of each device before and after the abnormality occurs in each abnormal event to find the characteristic data of each abnormal event, the formula is: In the formula, Sc represents the data difference between each device data in each abnormal event before and after the abnormality occurs; the data difference obtained by calculating all device data in each abnormal event is , Indicates the difference between the 1st, 2nd, 3rd, ..., nth device data before and after the abnormality occurs in each abnormal event. Compare the differences between the n types of device data and find the data type corresponding to the maximum difference in device data as the characteristic data of the abnormal event; the characteristic data of each abnormal event in the collected history is , represents the characteristic data of the 1st, 2nd, 3rd, ..., mth abnormal events in the collected history, where m is a positive integer; the characteristic data of each abnormal event is used to classify the abnormalities, and the abnormal events with the same characteristic data are classified into one category, and finally all the abnormalities that occurred in the placement machine history are classified into , Indicates the 1st, 2nd, 3rd, ..., hth abnormality that occurs in the placement machine; S103, collect the characteristic data value of each abnormality of the placement machine in the history , It represents the characteristic data of the 1st, 2nd, 3rd, ..., kth occurrence of each anomaly in the collected history, where k is a positive integer. The characteristic data of each anomaly in the collected history when it occurs is used to calculate each anomaly threshold, and the formula is: , In the formula, Ina represents the calculated anomaly threshold of each anomaly type, Tz represents the characteristic data collected when each anomaly occurs in the history, k represents the number of characteristic data collected when each anomaly occurs in the history, and st_Tz represents the standard deviation of k characteristic data collected when each anomaly occurs; the calculated anomaly threshold of each anomaly type is , represents the calculated abnormality thresholds of the 1st, 2nd, 3rd, ..., hth abnormalities of the placement machine; S104. Calculate the abnormality occurrence point using the characteristic data of each abnormality in the collected history, and the formula is: , In the formula, Ha represents each abnormal occurrence point of the calculation. Indicates the minimum value of characteristic data of each anomaly in the collected history, It represents the maximum value of the characteristic data of each anomaly in the collected history; after calculating each anomaly occurrence point using the characteristic data, the occurrence points of all types of anomalies are obtained as , Indicates the 1st, 2nd, 3rd, ..., hth abnormality occurrence points of the placement machine.

3. The method for managing abnormal data of a placement machine based on data analysis according to claim 2, characterized in that: The specific steps of using the abnormal threshold to provide real-time abnormal warning to the placement machine in S200 are: S201, collect the working data of the placement machine in history, analyze the collected working data of the placement machine, and extract the time difference between each change of the working data of the placement machine , Indicates the time difference of the 1st, 2nd, 3rd, ..., jth change of the working data of the placement machine in the collected history, where j is a positive integer; the average value of the time difference of the jth extracted working data change is calculated, and the average value of the working data change time difference is used as the collection cycle Cy for monitoring each abnormal characteristic data of the placement machine; S202, using the calculated collection cycle to collect feature data of each abnormality when the placement machine is working, assuming that the real-time feature data of each abnormality collected is , It represents the 1st, 2nd, 3rd, ..., pth feature data of each abnormality collected in real time, where p is a positive integer; a curve graph is constructed using the collected real-time feature data, the calculated abnormal threshold and the abnormality occurrence are marked in the curve graph, and the constructed curve graph is judged in real time. When the chip mounter is judged to have the corresponding type of abnormal risk and issue an early warning; when When the chip placement machine is working normally, there is no abnormal risk. By collecting the real-time feature data of the chip placement machine, it is determined that the abnormal type for which the warning is issued is , It indicates the 1st, 2nd, 3rd, ..., qth abnormal warning after judgment, where q is a positive integer.

4. The method for managing abnormal data of a placement machine based on data analysis according to claim 3 is characterized in that: The specific steps of calculating the warning response threshold of the control personnel of the placement machine by using the historical response time of the control personnel in S300 are: S301, after the SMT machine issues an abnormal warning in the acquisition history, the control personnel receives the warning information and performs maintenance on the SMT machine. , It indicates the 1st, 2nd, 3rd, ..., dth response time of the control personnel to perform maintenance on the SMT machine after the SMT machine issues an abnormal warning in the collected history, where d is a positive integer; S302, using the collected history of abnormal warnings of the placement machine, the control personnel's response time to receive the warning information and perform maintenance on the placement machine is calculated, and the formula is: , In the formula, Itr represents the warning response threshold of the controller. Indicates the minimum response time of the control personnel after receiving the warning information in the collected history. Indicates the maximum value of the response time of the control personnel after receiving the warning information in the collected history; It indicates the standard deviation of the response time of the control personnel after receiving the warning information in the collected history.

5. The method for managing abnormal data of a placement machine based on data analysis according to claim 4 is characterized in that: The specific steps of searching for the emergency abnormality warning of the placement machine in S400 are: S401, extracting the characteristic data curve graph constructed by collecting the real-time characteristic data of each abnormality of the placement machine, and calculating the slope of the curve at each abnormality warning, the formula is: , In the formula, sd represents the slope of the curve at each abnormal warning. Indicates the characteristic data value of each abnormal warning in the characteristic data curve graph, Indicates the characteristic data value collected before each abnormal warning in the characteristic data curve diagram; Cy indicates the collection cycle when monitoring the characteristic data; S402, calculating the mutation time of each abnormal warning according to the calculated slope of the characteristic data curve of each abnormal warning, the formula is: , In the formula, tm represents the mutation time of each abnormal warning, Ha represents the abnormal occurrence point of each abnormal warning, Ina represents the mutation time of each abnormal warning, and sd represents the slope of the curve at each abnormal warning; S403: Use the warning response threshold to judge the mutation time of each abnormal warning calculated. When abnormal warning is judged, early prevention can be carried out. When it is judged that the control personnel cannot take timely precautions after the abnormal warning, the abnormal warning is an emergency abnormal warning; all abnormal warnings are judged and all emergency abnormal warnings are found.

6. The method for managing abnormal data of a placement machine based on data analysis according to claim 5, characterized in that: The specific steps of calculating the abnormal threshold advance point of the emergency abnormal warning by using the real-time characteristic data change curve and the warning response threshold in S500 are: S501, determine the abnormal type of the emergency abnormal warning as an emergency abnormality according to the searched emergency abnormality warning, and collect the characteristic data of the emergency abnormality in the normal operation of the placement machine in the history. , It represents the 1st, 2nd, 3rd, ..., fth characteristic data of the emergency abnormality in the normal operation of the placement machine in the history, where f is a positive integer; the average value of the f characteristic data of the emergency abnormality collected when the placement machine is working normally is calculated as the normal point nor of the emergency abnormality warning, and the abnormal threshold advance point of the emergency abnormality warning is calculated according to the mutation time of the emergency abnormality warning and the warning response threshold, and the formula is: , In the formula, Tt represents the abnormal threshold advance point of the emergency abnormal warning, Ina represents each abnormal threshold calculated, Itr represents the warning response threshold, tm represents the mutation time of each abnormal warning, and Cy represents the slope of the curve at each abnormal warning; S502: For all the emergency abnormal warnings found, according to step S501, a normal point and an abnormal threshold advance point corresponding to each emergency abnormal warning are calculated.

7. The method for managing abnormal data of a chip mounter based on data analysis according to claim 6, characterized in that: The specific steps of replacing the abnormality threshold of the emergency abnormality threshold with the optimization point in S600 are: S601, calculate the abnormal threshold optimization point of the emergency abnormal warning according to the calculated normal point and abnormal threshold advance point of each emergency abnormal warning, the formula is: , In the formula, Ty represents the abnormal threshold optimization point of the emergency abnormality warning, and the abnormal threshold of the emergency abnormality is optimized and replaced by the calculated abnormal threshold optimization point.

8. The abnormal data management system of the chip placement machine based on data analysis is characterized by: The abnormal data management system of the placement machine includes data collection module, abnormal classification module, abnormal warning module, abnormal warning analysis module and optimization module; The data collection module is used to collect the equipment data and working data of the placement machine when an abnormality occurs in the history; The abnormality classification module is used to analyze the device data collected in the history of the placement machine when the abnormality occurs, find the characteristic data of each abnormal event, and classify the placement machine abnormality; The abnormal warning module is used to calculate the abnormal threshold value by using the characteristic data of the placement machine when the abnormality occurs in the history, collect the real-time characteristic data of the placement machine when it is working, and judge whether there is an abnormal risk of the placement machine in real-time operation; The abnormal warning analysis module is used to calculate the warning response threshold by using the historical control personnel's response time to the abnormal warning, analyze all real-time abnormal thresholds, calculate the mutation time of each abnormal warning, judge the mutation time, and find emergency abnormal warnings; The optimization module is used to calculate the normal point and the advance point of the abnormal warning for the emergency abnormal warning, determine the abnormal threshold optimization point of the emergency abnormal warning according to the normal point and the advance point of the abnormal warning, and use the abnormal threshold optimization point to optimize and replace the abnormal threshold of the emergency abnormality.

9. The abnormal data management system for chip placement machines based on data analysis according to claim 8, characterized in that: The abnormal warning module includes a threshold calculation unit, a real-time acquisition unit and a warning judgment unit; The threshold value calculation unit is used to calculate the abnormality threshold value of each abnormality by using the characteristic data collected in the history when the placement machine has an abnormality; The real-time acquisition unit is used to analyze the working data of the placement machine in history, calculate the acquisition cycle, use the acquisition cycle to collect the real-time characteristic data of the placement machine when it is working, and draw a characteristic data curve graph; The early warning judgment unit is used to judge the characteristic data collected in real time by using the abnormality threshold of each abnormality, determine whether there is an abnormality risk in the placement machine and issue an early warning.

10. The abnormal data management system for chip placement machines based on data analysis according to claim 8, characterized in that: The abnormal warning analysis module includes a warning response threshold calculation unit, a mutation time calculation unit and an emergency abnormal warning search unit; The warning response threshold value calculation unit is used to calculate the response time of the control personnel in the history when handling the abnormality of the placement machine after receiving the warning information, so as to obtain the warning response threshold value; The mutation time calculation unit is used to analyze the curve graph constructed by the real-time collected feature data, determine the slope of the curve during abnormal warning, and calculate the mutation time of each abnormal warning; The emergency abnormality warning search unit is used to judge the mutation time by using the warning corresponding threshold value and search for the emergency abnormality warning.

Images