State recognition method and device of suction equipment, suction equipment and chip mounter

Abstract

The invention relates to a state recognition method and device of suction equipment, the suction equipment and a chip mounter, and relates to the technical field of industrial digital intelligence. The data set comprises material throwing data of the suction equipment, the number of used times of the suction equipment and a detection result of a product in process, and the detection result of the product in process indicates whether the material sucked by the suction equipment is correctly mounted on the product in process or not; and taking the data set as an input parameter of a first model, and predicting a first equipment state of the suction equipment. By using the state recognition method of the suction equipment, the equipment state of whether the suction equipment is abnormal or not can be obtained under the condition that monitoring equipment is not added.

Description

Technical Field

[0001] The present disclosure relates to the field of industrial digital intelligence technology, and in particular to a state recognition method and device for a suction device, a suction device, and a placement machine. Background Art

[0002] With the deep integration of information technology and manufacturing, the development of intelligent manufacturing has achieved results. In some manufacturing fields, it is gradually moving towards the direction of digitalization and generating new models of intelligent manufacturing. In the field of intelligent manufacturing, there are suction devices, which are used to suck materials and mount them on the designated position of the work-in-progress to complete the product production. In the production process of the work-in-progress, the suction equipment is prone to malfunctions due to contamination, blockage, wear and tear after extensive use, etc., resulting in the inability of the suction equipment to normally suck and mount materials, causing the material to be thrown away.

[0003] In the related art, additional monitoring equipment is installed on the production line to monitor the equipment status of the suction equipment, but the additional configuration of the monitoring equipment will bring equipment costs. Summary of the invention

[0004] In order to overcome the problems existing in the related art, the present disclosure provides a state recognition method and device of a suction device, a suction device and a placement machine.

[0005] According to a first aspect of an embodiment of the present disclosure, a method for identifying a state of a suction device is provided, comprising:

[0006] Acquire a data set; the data set includes the material throwing data of the suction device, the number of times the suction device has been used, and the work-in-progress detection result, the work-in-progress detection result indicating whether the material sucked by the suction device is correctly mounted on the work-in-progress;

[0007] The data set is used as an input parameter of a first model to predict a first device state of the suction device.

[0008] Optionally, the data set further includes at least one of the following:

[0009] The number of times the suction device is to be used, the material number of the material, and the device image of the suction device.

[0010] Optionally, a first weight corresponding to the discarded material data in the data set and a second weight corresponding to the work-in-progress inspection result are both greater than weights corresponding to other data in the data set.

[0011] Optionally, the data set includes material throwing data, the number of times used and work-in-progress detection results at multiple time points; using the data set as an input parameter of the first model to predict the first device state of the suction device includes:

[0012] The data set is used as an input parameter of the first model to predict a first device state of the suction device at the multiple time points.

[0013] Optionally, the method further comprises:

[0014] The data set and the first device state of the suction device at the multiple time points are used as input parameters of the second model to predict the future second device state of the suction device.

[0015] Optionally, the method further comprises:

[0016] Obtaining an abnormal value of the suction device according to a first device state of the suction device at the plurality of time points, a second device state of the suction device in the future, and the material throwing data of the suction device;

[0017] According to multiple abnormal values ​​of multiple suction devices, early warning maintenance information of different priorities is output to the multiple suction devices; wherein, the higher the abnormal value of the suction device, the higher the priority of the early warning maintenance information.

[0018] Optionally, the early warning maintenance information includes a first device state of the suction device, a second device state of the suction device, and material throwing data of the suction device.

[0019] Optionally, using the data set and the first device state of the suction device at the multiple time points as input parameters of the second model to predict the future second device state of the suction device includes:

[0020] For any type of suction device among multiple types, the data set related to the suction device and the first device state of the suction device are used as input parameters of the second model to predict the future second device state of the suction device; wherein the data sets related to different types of suction devices are different.

[0021] Optionally, acquiring the data set includes:

[0022] Acquire identification information corresponding to the suction action of the suction device;

[0023] A data set corresponding to the identification information is determined.

[0024] According to a second aspect of an embodiment of the present disclosure, a state recognition device for a suction device is provided, comprising:

[0025] An acquisition module is configured to acquire a data set; the data set includes the material throwing data of the suction device, the number of times the suction device has been used, and the work-in-progress detection result, and the work-in-progress detection result indicates whether the material sucked by the suction device is correctly mounted on the work-in-progress;

[0026] The first prediction module is configured to use the data set as an input parameter of a first model to predict a first device state of the suction device.

[0027] According to a third aspect of an embodiment of the present disclosure, there is provided a suction device, comprising:

[0028] processor;

[0029] a memory for storing processor-executable instructions;

[0030] Wherein, the processor is configured to:

[0031] Execute the steps of the state recognition method of the suction device provided in the first aspect of the embodiment of the present disclosure.

[0032] According to a fourth aspect of an embodiment of the present disclosure, a placement machine is provided, on which the suction device provided by the third aspect of an embodiment of the present disclosure is provided.

[0033] According to a fifth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of the state identification method of the suction device provided in the first aspect of the present disclosure are implemented.

[0034] According to a sixth aspect of an embodiment of the present disclosure, a computer program product is provided, including a computer program, which, when executed by a processor, implements the steps of the state recognition method of the suction device provided in the first aspect of the present disclosure.

[0035] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0036] Since the discarded material data, number of uses and work-in-process inspection results in the data set are data that can be obtained without the need for additional monitoring equipment, there is no need to configure additional monitoring equipment for monitoring in the process of obtaining the first device status of the suction device based on this part of the data set, which can save equipment costs.

[0037] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0039] Figure 1 The figure is a flow chart of a method for identifying a state of a suction device according to an exemplary embodiment.

[0040] Figure 2 The figure is a flow chart of a method for identifying a state of a suction device according to an exemplary embodiment.

[0041] Figure 3 The figure is a flow chart of a method for identifying a state of a suction device according to an exemplary embodiment.

[0042] Figure 4 The diagram is a schematic diagram showing a prediction of an abnormality of a suction device and an actual abnormality of the suction device according to an exemplary embodiment.

[0043] Figure 5 The figure is a flow chart of a method for identifying a state of a suction device according to an exemplary embodiment.

[0044] Figure 6 The figure is a block diagram of a state recognition device for a suction device according to an exemplary embodiment.

[0045] Figure 7 The figure is a block diagram of a state recognition device for a suction device according to an exemplary embodiment.

[0046] Figure 8 is a block diagram of a chip system according to an exemplary embodiment. DETAILED DESCRIPTION

[0047] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0048] It should be noted that all actions of acquiring signals, information or data in the present disclosure are carried out in compliance with the relevant data protection laws and policies of the country where the device is located and with the authorization given by the owner of the corresponding device.

[0049] Figure 1 is a flow chart of a method for identifying a state of a suction device according to an exemplary embodiment. Figure 1 As shown, the following steps are included.

[0050] In step S10, a data set is obtained.

[0051] The data set includes the material throwing data of the suction equipment, the number of times the suction equipment has been used, and the inspection results of the work-in-progress products.

[0052] For the discarded material data of the suction device, the discarded material data includes at least one of the discarded material quantity and the discarded material rate. The discarded material quantity refers to the quantity of materials that are discarded due to failure to mount the materials correctly during the process of the suction device sucking and mounting the materials on the work-in-progress; the discarded material rate refers to the ratio of the quantity of discarded materials to the total quantity of materials input. The discarded material rate includes the upper limit value and lower limit value of the historical discarded material rate of the suction device and the discarded material rate within the preset time before the current time point. Taking the preset time of 10 minutes as an example, it can be the discarded material rate within 10 minutes before the current time point.

[0053] As for the number of times the suction device has been used, the number of times used refers to the number of suctions the suction device has performed during a period from the last maintenance of the suction device to the current time point.

[0054] For the work-in-process inspection result, the work-in-process inspection result indicates whether the material sucked by the suction device is correctly mounted on the work-in-process. The work-in-process inspection result can be indicated by a first inspection result and a second inspection result. The first inspection result is, for example, 1, indicating that the material is correctly mounted on the work-in-process, and the second inspection result is, for example, 0, indicating that the material is not correctly mounted on the work-in-process. Among them, whether the material is correctly mounted on the work-in-process includes: whether the material is mounted on the work-in-process in a specified direction, whether the material is correctly mounted at the specified position of the work-in-process. Whether the material is correctly mounted at the specified position of the work-in-process, for example, is that the material is rotated, offset, or the material is missing compared to the specified position.

[0055] Among them, the work-in-process inspection results may include a first work-in-process inspection result and a second work-in-process inspection result. The first work-in-process inspection result is a preliminary inspection result obtained by an automatic optical inspection device (Automatic Optical Inspection, AOI) to inspect the work-in-process, and the second work-in-process inspection result is a secondary inspection result obtained by manually re-inspecting the work-in-process.

[0056] Optionally, the suction device can be a suction nozzle installed on a placement machine. For example, when the material is a component such as a capacitor or a board, and the work-in-progress is a circuit board, the suction nozzle is used to suck the component and mount the component on the circuit board.

[0057] In step S20, the data set is used as an input parameter of a first model to predict a first device state of the suction device.

[0058] The first model is a model for detecting the current and / or historical first device state of the suction device, and the first model may be an anomaly detection model (Anomaly Transformer). The anomaly detection model may include a supervised model, a semi-supervised learning model, and an unsupervised learning model.

[0059] For supervised models, supervised models refer to models that are trained with labeled data. For example, the equipment status of the suction equipment at a certain point in time is used as label data, and the material throwing data, the number of times used, and the inspection results of the work-in-progress at that point in time are used as training samples to train the anomaly detection model. Label data includes normal label data and abnormal label data. Normal label data can be normal equipment status, and abnormal label data can be abnormal equipment status. By training the anomaly detection model with normal label data and abnormal label data, the anomaly detection model can be able to distinguish between normal label data and abnormal label data.

[0060] For example, when training an outlier detection model, the thrown material data, the number of times used, and the work-in-process inspection results can be used as training samples of the outlier detection model, and the actual device status of the suction device can be used as a label to train the outlier detection model. For example, the training samples are input into the outlier detection model to obtain the predicted device status; then the loss function between the predicted device status and the actual device status is determined, and the network parameters such as the weight and activation function of the outlier detection model are updated using the loss function to obtain the trained outlier detection model. In this way, the trained outlier detection model will learn the potential influence relationship between various factors such as the thrown material data, the number of times used, and the work-in-process inspection results, and the device status of the suction device, and then accurately judge whether the suction device is abnormal based on various factors.

[0061] It can be understood that after the data set is input into the outlier detection model, the outlier detection model predicts the probability of whether the suction device is abnormal. In order to obtain the first device state of whether the suction device is abnormal, the probability output by the outlier detection model can also be compared with the preset probability. If the probability output by the outlier detection model is greater than or equal to the preset probability, the first device state of the suction device is set to the first label, indicating that the suction device is normal. If the probability output by the outlier detection model is less than the preset probability, the first device state of the suction device is set to the second label, indicating that the suction device is abnormal.

[0062] For the semi-supervised learning model, the semi-supervised learning model can be trained using a small number of training samples with labeled data and a large number of training samples without labeled data, such as labeling a small number of training samples in abnormal device states, or labeling training samples in normal device states, etc. If a small number of training samples in abnormal device states are labeled, the labeled training samples can be used to train the classifier first, and then training samples without labeled data with high confidence can be gradually added for iterative training; if training samples in normal device states are labeled, iterative training can be performed based on distance or linear model methods.

[0063] For an unsupervised learning model, an unsupervised learning model refers to a model trained without labeled data. The unsupervised learning model detects anomalies by analyzing the intrinsic characteristics of each training sample.

[0064] The first device status is used to indicate whether the suction device is abnormal, for example, whether the suction device is abnormal at a historical time point and / or a current time point. The first device status includes a first label and a second label, the first label indicates that the suction device is normal, and the second label indicates that the suction device is abnormal.

[0065] For the material throwing data of the suction device in the data set, the reason for using the material throwing data as the input parameter of the first model is that the more the material throwing amount and / or the higher the material throwing rate in the material throwing data of the suction device, it means that the suction device cannot normally absorb the material due to contaminant blockage or wear and tear, resulting in part of the material being discarded. Therefore, the more the material throwing amount and / or the higher the material throwing rate in the material throwing data, the more abnormal the suction device is. Therefore, the material throwing data can be used as the input parameter of the first model, so that the first model can learn the potential relationship between the more the material throwing amount and / or the higher the material throwing rate in the material throwing data, and the higher the degree of abnormality of the suction device, so as to accurately obtain the abnormal situation of the suction device under different material throwing data.

[0066] As for the number of times the suction device in the data set has been used, the reason for using the number of times the suction device has been used as the input parameter of the first model is that the higher the number of times the suction device has been used, the higher the degree of wear of the suction device, which may cause the suction device to be more abnormal. Therefore, the number of times the suction device has been used can be used as the input parameter of the first model, so that the first model can learn the potential relationship between the more times the suction device has been used and the higher the degree of abnormality of the suction device, so as to accurately obtain the abnormal situation of the suction device under different times of use.

[0067] For the WIP inspection results in the data set, the reason for using the WIP inspection results as the input parameters of the first model is that if the material sucked by the suction device is not correctly mounted on the WIP, it means that the suction device cannot correctly mount the material due to wear or other reasons, and it can also indicate that the suction device is abnormal. Therefore, the WIP inspection results can be used as the input parameters of the first model, allowing the first model to learn the potential relationship between the failure of the material to be correctly mounted on the WIP and the abnormality of the suction device, so as to accurately obtain the abnormal situation of the suction device under different WIP inspection results.

[0068] Optionally, identification information corresponding to the suction action of the suction device may be obtained; and a data set corresponding to the identification information may be determined. The identification information may be a code number corresponding to the suction action of the suction device.

[0069] For the suction equipment, each suction action of the suction equipment has unique identification information. By obtaining the identification information, the data set corresponding to the identification information can be obtained. It can also be understood that by obtaining the identification information, the throwing data, the number of times used and the work-in-process inspection results related to the suction equipment corresponding to the identification information can be obtained. These data can be directly obtained from the material management system without the need for additional configuration of monitoring equipment for monitoring.

[0070] Optionally, in addition to using the discarded material data, the number of times used and the work-in-process inspection results as input parameters of the first model, the data set can also use the number of times the suction device is to be used, the material number of the material and the device image of the suction device as input parameters of the first model to predict the first device state of the suction device.

[0071] For the number of times to be used of the suction device, the number of times to be used is the number of times the suction device needs to absorb materials in this production process. For example, if the materials to be absorbed in this production process are 1000, then the number of times to be used is 1000. The reason for taking the number of times to be used of the suction device as the input parameter of the first model is that when the throwing rate of the suction device is the same, the degree of use of the suction device is also different. Although the throwing rate of the suction device absorbing 10,000 times of materials and the throwing rate of the suction device absorbing 1,000 times of materials may be the same, the degree of use of the suction device absorbing 10,000 times of materials is much greater than the degree of use of the suction device absorbing 1,000 times of materials. Therefore, taking the number of times to be used of the suction device as the input parameter of the first model, the first model can learn the influence of the number of times to be used on the device state of the suction device, and learn the potential relationship between the more times to be used, the higher the abnormal degree of the suction device, and the lower the number of times to be used, the lower the abnormal degree of the suction device, so as to accurately obtain the abnormal situation of the suction device under different times to be used, and when it is predicted that there is an abnormality of the suction device under the number of times to be used of a certain production, the suction device is maintained in time.

[0072] For the material number of the material sucked by the suction device, the material number is associated with the material attributes such as the size, type, weight and shape of the material. When the suction device has been used for the same number of times, the degree of wear of the suction device when sucking lighter materials is lighter, and the degree of wear of the suction device when sucking heavier materials is heavier. Therefore, using the material number of the suction device as the input parameter of the first model can allow the model to learn the influence of the material weight corresponding to the material number on the device status of the suction device, and learn the potential relationship between the heavier the material weight, the higher the abnormality of the suction device, and the lighter the material weight, the lower the abnormality of the suction device, so as to accurately obtain the abnormal situation of the suction device when sucking materials of different weights.

[0073] For the device image of the suction device, the device image shows the degree of wear of the suction device, such as the number of millimeters of wear of the suction device. The higher the degree of wear of the suction device, the higher the abnormality of the suction device. Therefore, the device image of the suction device can be used as an input parameter of the first model, so that the model can learn the influence of the degree of wear of the suction device on the device state of the suction device, and learn the potential relationship between the higher the degree of wear of the suction device and the higher the degree of abnormality of the suction device, so as to accurately obtain the abnormal conditions of the suction device under different degrees of wear.

[0074] In the related art, the following solutions can be used to maintain the suction equipment:

[0075] (1) Predictive maintenance: by configuring various monitoring devices to collect performance data of the suction equipment during operation, a deep learning model is used to predict the remaining life of the suction equipment based on the performance data, and the suction equipment is maintained according to the prediction results. However, the predictive maintenance solution requires additional monitoring equipment, which brings additional equipment costs.

[0076] (2) Preventive maintenance: determine the number of times the suction device has been used, and when the number of times it has been used is greater than a set threshold, perform maintenance on the suction device.

[0077] First, a fixed set threshold value will cause waste of the suction equipment. For example, if the set threshold value is 15,000 times, some suction equipment sucks lighter materials. Therefore, when the suction equipment has been used 15,000 times, the suction equipment can still be used normally. At this time, if the suction equipment is replaced for maintenance, it will lead to waste of the suction equipment, and replacing the suction equipment will also bring waste of manpower.

[0078] Secondly, a fixed set threshold will also increase the risk of failure in the placement process. For example, if the set threshold is 15,000 times, some suction devices suck heavier materials, so when the suction device has not been used for 15,000 times, the suction device can no longer work properly. If the suction device is replaced and maintained only after 15,000 times, the risk of failure in the placement process will increase, and the yield of the work-in-progress mounted by the suction device will be low or the material throwing data of the suction device will be high.

[0079] (3) Post-maintenance: after the absorption equipment fails, the absorption equipment will be maintained and replaced.

[0080] Firstly, the production process will be interrupted when the suction equipment is maintained, resulting in part of the production process not being able to proceed smoothly; secondly, frequent failures of the suction equipment will also accelerate the wear of the suction equipment and shorten its service life.

[0081] Through the above technical solution, compared with the predictive maintenance solution, since the discarded material data, the number of times used and the work-in-process inspection results can be directly obtained from the material management system through identification information, without the need for additional monitoring equipment for detection, the equipment cost is saved.

[0082] Compared with the preventive maintenance plan, the first model can predict the first device state of the suction device in real time based on the thrown material data, the number of times it has been used, and the work-in-process inspection results. The predicted first device state will change adaptively according to the weight of the material sucked by the suction device. When the material weight is heavy, the first model predicts that the degree of abnormality of the suction device will be higher, so it will promptly prompt the abnormality of the suction device and maintain the suction device in time, reducing the risk of failure; when the material weight is light, the first model predicts that the degree of abnormality of the suction device will be lower, so the probability of prompting the abnormality of the suction device is lower, and it is not necessary to maintain and replace the suction device with a lower degree of abnormality, reducing the waste of suction equipment and labor.

[0083] Compared with the post-maintenance solution, the first model can predict the first device status of the suction device in real time, so as to perform maintenance and replacement of the suction device in advance, reduce the interruption of the production process, and improve production efficiency.

[0084] The following is an exemplary embodiment involved in the present disclosure, which is used to explain that the first model pays more attention to the material throwing data of the suction device and the work-in-progress detection results, so that the first device state of the suction device finally obtained is strongly correlated with the material throwing data and the work-in-progress detection results.

[0085] In the data set, the first weight corresponding to the discarded material data and the second weight corresponding to the work-in-process inspection result can be configured to be greater than the weights corresponding to the remaining data in the data set. The weights corresponding to the remaining data in the data set include the weight corresponding to the number of times used, the weight corresponding to the number of times to be used, the weight corresponding to the material number, and the weight corresponding to the device image.

[0086] The first model includes an attention mechanism (Anomaly-Attention). Anomaly-Attention is a method that uses a learnable Gaussian kernel function to capture the correlation information between adjacent elements in a sequence, compares it with the information calculated by the standard attention mechanism, and filters out outliers based on the KL divergence value of the two types of information. The attention mechanism has a query vector / input vector (Q), a key vector (K), and a value vector (V). Each data in the above data set represents an input vector, and the weight corresponding to each data represents a key vector, so each input vector has its own attention score. Among them, the attention score of each input vector can be calculated by the following formula:

[0087]

[0088] In formula (1), Attention(Q,K,V) represents the attention score of the input vector; Q represents the input vector, K represents the weight corresponding to the input vector, and V represents the value vector.

[0089] It can be seen from the above formula (1) that each input vector has its own corresponding weight, so the weights of some input vectors can be selectively set higher, so that the attention scores corresponding to these input vectors are also higher, and the first model will pay more attention to the impact of input vectors with higher weights on the first device state of the absorption device.

[0090] Therefore, the weights corresponding to the input vectors corresponding to the two data, the thrown material data and the work-in-process inspection results, can be set higher than the weights corresponding to the rest of the data in the data set. Correspondingly, the attention scores corresponding to the thrown material data and the work-in-process inspection results will also be higher, so that the first model can pay more attention to the impact of the thrown material data and the work-in-process inspection results on the first device state of the suction device. This means that the higher the thrown material rate and / or the number of thrown materials in the thrown material data, and the higher the work-in-process inspection results detect that the material is not correctly mounted on the work-in-process, the probability that the first model outputs the first device state indicating that the suction device is abnormal will be greater.

[0091] When the first model learns the time series characteristics of the data set, an inductive bias that pays more attention to the discarded material data and the work-in-process inspection results can be introduced, so that the first model pays more attention to abnormal work-in-process conditions when determining whether the suction equipment is abnormal.

[0092] Through the above technical scheme, since the material throwing data and the work-in-process inspection results can intuitively reflect whether the suction equipment is abnormal, for example, the material throwing rate and / or the amount of material throwing in the material throwing data are high, indicating that the suction equipment is abnormal, and the work-in-process inspection results indicate that the material is not correctly mounted on the work-in-process, indicating that the suction equipment is abnormal, the weights of the material throwing data and the work-in-process inspection results can be set larger, so that the first model pays more attention to the impact of the material throwing data and the work-in-process inspection results on the suction equipment, thereby predicting a more accurate first equipment status.

[0093] Figure 2 is an exemplary embodiment involved in the above step S20, which is used to explain that the first device state predicted by the first model is a device state under the time series feature, and includes the following steps:

[0094] In step S21, the data set is used as an input parameter of the first model to predict the first device state of the suction device at the multiple time points.

[0095] Among them, the discarded material data, the number of times used and the inspection results of the work-in-progress in the data set are the discarded material data, the number of times used and the inspection results of the work-in-progress at multiple time points, and the data in the data set are time series data.

[0096] The multiple time points include the current time point and multiple consecutive historical time points before the current time point.

[0097] For example, taking the current time point as 15:25 and the multiple consecutive historical time points before the current time point including 15:20, 15:15, 15:10, 15:05, and 15:00 as an example, the material throwing data, the number of times used, and the work-in-process inspection results at the multiple time points of 15:00, 15:05, 15:10, 15:15, 15:20, and 15:25 can be input into the first model, and the first model predicts the first device status of the suction device at the multiple time points of 15:00, 15:05, 15:10, 15:15, 15:20, and 15:25, respectively, and the predicted first device status is also the first device status at multiple time points.

[0098] It can be understood that the first model is an anomaly detection model for detecting time series data. The first model can calculate the degree of correlation between data at multiple different time points through the attention mechanism. Therefore, it can accurately locate and distinguish anomalies from the first device states at multiple time points, that is, identify abnormal first device states from the first device states at multiple time points.

[0099] Through the above technical solution, since the data in the data set are time series data at multiple time points, and the first model is an anomaly detection model, after the first model receives the time series data at multiple time points and predicts the first device states at multiple time points, it can accurately locate the first device state indicating an abnormality from the first device states at multiple time points, and then determine whether the first device state at the current time point indicates an abnormality.

[0100] Figure 3 is an exemplary embodiment of the present disclosure, which is used to explain an exemplary solution of obtaining a second device state in the future of an absorption device based on the first device state at multiple time points by the second model, and includes the following steps:

[0101] In step S30, the data set and the first device state of the suction device at the multiple time points are used as input parameters of the second model to predict the future second device state of the suction device.

[0102] The second model may be a Prentice, Williams, and Peterson Total Time (PWP-TT) model, which can predict data for a period of time in the future based on data at multiple time points.

[0103] For example, the second model can first integrate the discarded material data, number of uses, work-in-process inspection results, number of uses, material numbers and equipment images at multiple time points in the data set into a first curve that changes with time, and then connect the first equipment status at multiple time points into a second curve that changes with time, and then integrate the first curve and the second curve into a target curve; finally, based on the target curve, predict the second equipment status of the suction equipment at a future time point.

[0104] During this process, the second model can analyze recurrence events at multiple time points (for example, abnormal first equipment status, material throwing data, number of uses, work-in-progress inspection results, number of times to be used, material numbers and equipment images, etc. at multiple time points), and then consider the order and hierarchy between multiple recurrence events, paying attention to the total time of each recurrence event, so as to predict the second equipment status of the suction equipment at a future time point.

[0105] For example, taking the current time point as 15:25 and the multiple consecutive historical time points before the current time point including 15:20, 15:15, 15:10, 15:05, and 15:00 as an example, the throwing data, number of uses, work-in-process inspection results, number of uses to be made, material numbers and equipment images at multiple time points of 15:00, 15:05, 15:10, 15:15, 15:20, and 15:25, and the first equipment states at multiple time points of 15:00, 15:05, 15:10, 15:15, 15:20, and 15:25 can be input into the second model, and the second model can predict the second equipment states of the suction equipment at multiple time points of 15:30, 15:35, 15:40, 15:45, 15:50, and 15:55 in the future based on these historical data.

[0106] It can be understood that both the first device status and the second device status indicate whether the suction device is abnormal, but the first device status indicates the device status of the suction device at a historical time point and / or a current time point, while the second device status indicates the device status of the suction device at a future time point.

[0107] Optionally, for any type of suction device among multiple types, a data set related to the suction device and a first device state of the suction device are used as input parameters of the second model to predict the future second device state of the suction device; wherein different types of suction devices have different data sets related.

[0108] Since the relevant throwing data, number of times used, work-in-process inspection results and other data of different suction devices are different, for any suction device, the respective associated data sets and the first device status will be used as input parameters of the second model to predict the future second device status of different suction devices. After multiple predictions by the second model, the second model corresponding to each suction device will be different. In this way, different second models can be used to predict more accurate second device status of different suction devices.

[0109] Through the above technical solution, the second model can predict the state of the second device at multiple time points in the future based on the discarding data, the number of times used and the work-in-process inspection results at multiple time points, as well as the state of the first device at multiple time points. It can predict in advance whether the suction equipment will be abnormal in the future, so as to maintain the suction equipment in advance, and reduce the production interruption and shortened service life caused by the maintenance of the suction equipment after the abnormality in the above-mentioned post-maintenance plan.

[0110] See also Figure 4 As shown, Figure 4 The horizontal axis is the time point, and the vertical axis is the abnormal value of the absorption device. Figure 4 The black dots in the figure are the second model issuing an alarm, warning that an abnormality of the suction device will occur at these future time points. The vertical lines indicate that an abnormality / nozzle error actually occurred in the suction device. Figure 4 It can be seen that before each abnormality actually occurs in the suction equipment, the second model has issued an alarm in advance to warn the suction equipment of an abnormality. This scheme of early warning of suction equipment abnormalities can greatly improve maintenance efficiency and reduce the occurrence of actual abnormalities in the suction equipment. After the abnormalities of the suction equipment are reduced, the generation of abnormal work-in-progress can also be reduced. By issuing an alarm in advance, the first-time yield of work-in-progress can be increased by 0.5%, the material loss can be reduced by 0.2%, and the online abnormality debugging time can be reduced by 13 minutes / shift, thereby improving production efficiency.

[0111] Figure 5 This is an exemplary embodiment of the present disclosure, which is used to explain that after obtaining the first device status and the second device status, a priority early warning scheme is performed on multiple suction devices, including the following steps:

[0112] In step S40, an abnormal value of the suction device is obtained according to the first device state of the suction device at the multiple time points, the second device state of the suction device in the future, and the material throwing data of the suction device.

[0113] The sum of the score value when the first device status indicates an abnormality, the score value when the second device status indicates an abnormality, and the score value when the thrown material data indicates an abnormality can be used as the abnormality value of the suction device.

[0114] The score value increased when the second device status indicates an abnormality is greater than the score value increased when the first device status indicates an abnormality, and the score value when the first device status indicates an abnormality is greater than the score value when the throwing material data is abnormal.

[0115] For example, when the second device status indicates an abnormality, 10 points can be added to the suction device; when the first device status indicates an abnormality, 5 points can be added to the suction device; when the throwing material data is abnormal (for example, the throwing material rate is greater than the preset throwing material rate, and the throwing material quantity is greater than the preset throwing material quantity), 1 point can be added to the suction device. The final abnormal value is the sum of the three, 16 points.

[0116] In step S50, according to the multiple abnormal values ​​of the multiple suction devices, early warning maintenance information of different priorities is output to the multiple suction devices.

[0117] Multiple abnormal values ​​of multiple suction devices can be arranged in order from large to small. The higher the priority of the suction device in front, the more urgent the early warning maintenance information will be output. The higher the abnormal value of the suction device, the higher the risk level indicated by the early warning maintenance information, and the maintenance personnel will be reminded to maintain the suction device earlier.

[0118] The early warning maintenance information includes a first device state of the suction device, a second device state of the suction device, and material throwing data of the suction device.

[0119] For example, the output early warning maintenance information indicates that the first device state of the suction device at the current time point is abnormal, the second device state of the suction device at a future time point is abnormal, and the throwing data of the suction device is abnormal. These three pieces of information are the causes of the abnormality of the suction device, thereby providing maintenance personnel with a reference for maintenance of the suction equipment.

[0120] Through the above technical scheme, multiple suction devices can be prioritized according to the abnormal values ​​of the suction devices, so as to output early warning maintenance information with different priorities. For the suction device with a higher abnormal value, it is considered that the risk value of the suction device is higher, and the suction device will be maintained earlier; for the suction device with a lower abnormal value, it is considered that the risk value of the suction device is lower, and the suction device will be maintained later, thereby generating a maintenance plan in which high-risk suction devices are maintained first and low-risk devices are maintained later.

[0121] Furthermore, the reason for the abnormality of each abnormal suction device is provided, thereby providing a reference for the maintenance personnel to maintain the suction device and reducing the blindness of the maintenance personnel in maintaining the suction device.

[0122] Figure 6 is a block diagram of a state recognition device for a suction device according to an exemplary embodiment. The state recognition device 600 for a suction device is configured to execute the steps of the state recognition method for a suction device proposed in the present disclosure.

[0123] Optionally, refer to Figure 6 The state recognition device 600 of the suction device includes an acquisition module 610 and a first prediction module 620 .

[0124] The acquisition module 610 is configured to acquire a data set; the data set includes the material throwing data of the suction device, the number of times the suction device has been used, and the work-in-progress detection result, and the work-in-progress detection result indicates whether the material sucked by the suction device is correctly mounted on the work-in-progress;

[0125] The first prediction module 620 is configured to use the data set as an input parameter of a first model to predict a first device state of the suction device.

[0126] Optionally, the data set further includes at least one of the following:

[0127] The number of times the suction device is to be used, the material number of the material, and the device image of the suction device.

[0128] Optionally, a first weight corresponding to the discarded material data in the data set and a second weight corresponding to the work-in-progress inspection result are both greater than weights corresponding to other data in the data set.

[0129] Optionally, the data set includes discarded material data, number of uses and work-in-progress inspection results at multiple time points; the first prediction module 620 is also configured to use the data set as an input parameter of the first model to predict the first device state of the suction device at the multiple time points.

[0130] Optionally, the state recognition device 600 of the suction device includes a second prediction module;

[0131] The second prediction module is configured to use the data set and the first device state of the suction device at the multiple time points as input parameters of the second model to predict the second device state of the suction device in the future.

[0132] Optionally, the state recognition device 600 of the suction device includes an abnormal value module and an output module;

[0133] An abnormal value module is configured to obtain an abnormal value of the suction device according to a first device state of the suction device at the plurality of time points, a second device state of the suction device in the future, and the material throwing data of the suction device;

[0134] The output module is configured to output early warning maintenance information of different priorities to the multiple suction devices according to the multiple abnormal values ​​of the multiple suction devices; wherein, the higher the abnormal value of the suction device, the higher the priority of the early warning maintenance information.

[0135] Optionally, the early warning maintenance information includes a first device state of the suction device, a second device state of the suction device, and material throwing data of the suction device.

[0136] Optionally, the second prediction module is also configured to predict the future second device state of the suction device for any type of suction device among multiple types, using the data set related to the suction device and the first device state of the suction device as input parameters of the second model; wherein the data sets related to different types of suction devices are different.

[0137] Optionally, the acquisition module 610 includes:

[0138] An acquisition submodule, configured to acquire identification information corresponding to the suction action of the suction device;

[0139] The mapping submodule is configured to determine a data set corresponding to the identification information.

[0140] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0141] The present disclosure also provides a computer-readable storage medium on which computer program instructions are stored. When the program instructions are executed by a processor, the steps of the state identification method of the suction device provided by the present disclosure are implemented.

[0142] Figure 7 7 is a block diagram of a state recognition device 700 for a suction device according to an exemplary embodiment. For example, the device 700 may be a suction device such as a suction nozzle, and the suction device is configured to be in a placement machine, and the placement machine may be a placement machine of the SMT (Surface Mount Technology) segment, and the present disclosure does not limit this.

[0143] Reference Figure 7, the device 700 may include one or more of the following components: a processing component 702 , a memory 704 , a power component 706 , a multimedia component 708 , an audio component 710 , an input / output interface 712 , a sensor component 714 , and a communication component 716 .

[0144] The overall operation of the processing component 702 control device 700 is usually, such as the operation associated with display, phone call, data communication, camera operation and recording operation. The processing component 702 can include one or more processors 720 to execute instructions, to complete all or part of the steps of the state recognition method of the above-mentioned suction equipment. In addition, the processing component 702 can include one or more modules, which is convenient for the interaction between the processing component 702 and other components. For example, the processing component 702 can include a multimedia module to facilitate the interaction between the multimedia component 708 and the processing component 702.

[0145] The memory 704 is configured to store various types of data to support operations on the device 700. Examples of such data include instructions for any application or method operating on the device 700, contact data, phone book data, messages, pictures, videos, etc. The memory 704 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0146] The power supply component 706 provides power to the various components of the device 700. The power supply component 706 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the device 700.

[0147] The multimedia component 708 includes a screen that provides an output interface between the device 700 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundaries of the touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 708 includes a front camera and / or a rear camera. When the device 700 is in an operating mode, such as a shooting mode or a video mode, the front camera and / or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

[0148] The audio component 710 is configured to output and / or input audio signals. For example, the audio component 710 includes a microphone (MIC), and when the device 700 is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in the memory 704 or sent via the communication component 716. In some embodiments, the audio component 710 also includes a speaker for outputting audio signals.

[0149] The input / output interface 712 provides an interface between the processing component 702 and the peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include but are not limited to: a home button, a volume button, a start button, and a lock button.

[0150] The sensor assembly 714 includes one or more sensors for providing various aspects of status assessment for the device 700. For example, the sensor assembly 714 can detect the open / closed state of the device 700, the relative positioning of components, such as the display and keypad of the device 700, the sensor assembly 714 can also detect the position change of the device 700 or a component of the device 700, the presence or absence of user contact with the device 700, the orientation or acceleration / deceleration of the device 700, and the temperature change of the device 700. The sensor assembly 714 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 714 may also include an optical sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 714 may also include an accelerometer, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

[0151] The communication component 716 is configured to facilitate wired or wireless communication between the device 700 and other devices. The device 700 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 716 receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 716 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

[0152] In an exemplary embodiment, the device 700 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to execute the above-mentioned state recognition method of the suction device.

[0153] In an exemplary embodiment, a non-transitory computer-readable storage medium including instructions is also provided, such as a memory 704 including instructions, and the instructions can be executed by the processor 720 of the apparatus 700 to complete the state identification method of the above-mentioned suction device. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0154] In another exemplary embodiment, a computer program product is also provided. The computer program product includes a computer program executable by a programmable device, and the computer program has a code portion for executing the above-mentioned state recognition method of the suction device when executed by the programmable device.

[0155] Some embodiments of the present disclosure also provide a chip system, such as Figure 8 As shown, the chip system includes at least one processor 801 and at least one interface circuit 802. The processor 801 and the interface circuit 802 can be interconnected through a line. For example, the interface circuit 802 can be used to receive signals from other devices (such as a memory of an electronic device). For another example, the interface circuit 802 can be used to send signals to other devices (such as processor 801). Exemplarily, the interface circuit 802 can read instructions stored in the memory and send the instructions to the processor 801. When the instructions are executed by the processor 801, the state recognition device of the suction device can perform the various steps in the above embodiments. Of course, the chip system can also include other discrete devices, and some embodiments of the present disclosure are not specifically limited to this.

[0156] In some embodiments of the present disclosure, the interface circuit 802 can obtain data, program instructions and / or information from the internal storage area of ​​the chip system; it can also obtain data, program instructions and / or information from outside the chip system.

[0157] Optionally, the chip system also includes a memory 803, and the memory 803 is used to store necessary computer programs and data.

[0158] Those skilled in the art may also understand that the various illustrative logical blocks and steps listed in the embodiments of the present application may be implemented by electronic hardware, computer software, or a combination of the two. Whether such functions are implemented by hardware or software depends on the specific application and the design requirements of the entire system. Those skilled in the art may use various methods to implement the functions described for each specific application, but such implementation should not be understood as exceeding the scope of protection of the embodiments of the present application.

Claims

1. A method for identifying the state of a suction device, characterized in that: include: Get the dataset; The data set includes the material throwing data of the suction device, the number of times the suction device has been used, and the work-in-progress detection result, wherein the work-in-progress detection result indicates whether the material sucked by the suction device is correctly mounted on the work-in-progress; The data set is used as an input parameter of a first model to predict a first device state of the suction device.

2. The method according to claim 1, characterized in that The data set also includes at least one of the following: The number of times the suction device is to be used, the material number of the material, and the device image of the suction device.

3. The method according to claim 1, characterized in that In the data set, the first weight corresponding to the discarded material data and the second weight corresponding to the work-in-progress inspection result are both greater than the weights corresponding to the remaining data in the data set.

4. The method according to claim 1, characterized in that The data set includes the material throwing data, the number of times used and the product inspection results at multiple time points; the data set is used as the input parameter of the first model to predict the first device state of the suction device, including: The data set is used as an input parameter of the first model to predict a first device state of the suction device at the multiple time points.

5. The method according to claim 4, characterized in that The method further comprises: The data set and the first device state of the suction device at the multiple time points are used as input parameters of the second model to predict the future second device state of the suction device.

6. The method according to claim 5, characterized in that The method further comprises: Obtaining an abnormal value of the suction device according to a first device state of the suction device at the plurality of time points, a second device state of the suction device in the future, and the material throwing data of the suction device; According to multiple abnormal values ​​of multiple suction devices, early warning maintenance information of different priorities is output to the multiple suction devices; wherein, the higher the abnormal value of the suction device, the higher the priority of the early warning maintenance information.

7. The method according to claim 6, characterized in that The early warning maintenance information includes a first device state of the suction device, a second device state of the suction device, and material throwing data of the suction device.

8. The method according to claim 5, characterized in that The step of using the data set and the first device state of the suction device at the plurality of time points as input parameters of the second model to predict the future second device state of the suction device includes: For any type of suction device among multiple types, the data set related to the suction device and the first device state of the suction device are used as input parameters of the second model to predict the future second device state of the suction device; wherein the data sets related to different types of suction devices are different.

9. The method according to claim 1, characterized in that: The step of acquiring a data set comprises: Acquire identification information corresponding to the suction action of the suction device; A data set corresponding to the identification information is determined.

10. A state recognition device for a suction device, characterized in that: include: An acquisition module is configured to acquire a data set; the data set includes the material throwing data of the suction device, the number of times the suction device has been used, and the work-in-progress detection result, and the work-in-progress detection result indicates whether the material sucked by the suction device is correctly mounted on the work-in-progress; The first prediction module is configured to use the data set as an input parameter of a first model to predict a first device state of the suction device.

11. A suction device, characterized in that: include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to: Execute the steps of the method according to any one of claims 1 to 9.

12. A chip mounter, characterized in that: The placement machine is provided with the suction device according to claim 11.

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

14. A computer program product, characterized in that The invention comprises a computer program, which implements the steps of the method according to any one of claims 1 to 9 when being executed by a processor.

Images