Fault detection method and device, electronic equipment and storage medium

Abstract

The invention discloses a fault detection method and device, electronic equipment and a storage medium. The method comprises the following steps: determining a plurality of process nodes on a cigarette production line, for each process node, acquiring form data corresponding to process produced cigarettes through a sensor corresponding to the process node, and determining the number of loose-end cigarettes corresponding to the process node according to the form data; under the condition that the number of the multiple loose-end cigarettes meets the loose-end quality inspection condition, loose-end change data between every two process nodes is determined according to the number of the loose-end cigarettes; and the number of the loose-end cigarettes and the at least one loose-end change data are input into a detection model to determine a detection result, and the detection result comprises a first result representing whether a fault node exists in the process nodes or not and / or a second result representing a component adjustment parameter corresponding to the fault node. Based on the technical scheme of the invention, fault detection of the process node can be carried out from two links of the empty head quality inspection condition and the detection model, and the accuracy of fault detection of the process node is improved.

Description

Technical Field

[0001] The present invention relates to the field of computer application technology, and in particular to a fault detection method, device, electronic equipment and storage medium. Background Art

[0002] With the development of automation, in the tobacco industry, the cigarette production line has developed into a highly automated process. The cigarette production line involves multiple process links, each of which is deployed with mechanical parts for production. Mechanical parts need to be inspected and repaired to avoid production accidents caused by abnormal mechanical parts.

[0003] In the prior art, mechanical parts are usually regularly inspected and maintained by maintenance personnel. However, due to the uncertainty of the maintenance personnel's professional capabilities, it is often difficult to ensure the accuracy of fault detection of mechanical parts in each process and each link. Summary of the invention

[0004] The present invention provides a fault detection method, device, electronic equipment and storage medium to solve the problem that it is currently difficult to ensure the accuracy of fault detection of mechanical parts in various processes and links of cigarette production.

[0005] According to one aspect of the present invention, a fault detection method is provided, the method comprising:

[0006] Determine multiple process nodes on a cigarette production line, collect morphological data corresponding to cigarettes produced by each process node through a sensor corresponding to the process node, and determine the number of empty cigarettes corresponding to the process node based on the morphological data;

[0007] In the case where the number of multiple empty cigarettes meets the empty quality inspection condition, determining the empty change data between every two process nodes according to the number of empty cigarettes;

[0008] Input multiple numbers of empty cigarettes and at least one empty cigarette change data into the detection model to determine the detection result; wherein, the detection model is obtained by training the boosting learning model based on the training sample set; the detection result includes a first result and / or a second result corresponding to the first result, the first result characterizes whether there is a fault node in the process node, and the second result characterizes the component adjustment parameter corresponding to the fault node.

[0009] According to another aspect of the present invention, there is provided a fault detection device, the device comprising:

[0010] A multi-node data acquisition module is used to determine multiple process nodes on a cigarette production line, collect morphological data corresponding to the cigarettes produced by each process node through a sensor corresponding to the process node, and determine the number of empty cigarettes corresponding to the process node based on the morphological data;

[0011] A data analysis module, for determining the empty cigarette change data between every two process nodes according to the number of empty cigarettes when the number of empty cigarettes meets the empty cigarette quality inspection condition;

[0012] A fault prediction module is used to input the multiple numbers of empty cigarettes and at least one empty cigarette change data into a detection model to determine a detection result; wherein the detection model is obtained by training a boosting learning model based on a training sample set; the detection result includes a first result and / or a second result corresponding to the first result, the first result characterizing whether there is a fault node in the process node, and the second result characterizing a component adjustment parameter corresponding to the fault node.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the fault detection method described in any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the fault detection method described in any embodiment of the present invention when executed.

[0018] The technical solution of the embodiment of the present invention determines multiple process nodes on the cigarette production line, collects the morphological data corresponding to the cigarettes produced in the process for each process node through the corresponding sensor of the process node, and determines the number of empty cigarettes corresponding to the process node according to the morphological data; when the number of empty cigarettes meets the empty quality inspection conditions, the empty change data between each two process nodes is determined according to the number of empty cigarettes; the number of empty cigarettes and at least one empty change data are input into the detection model to determine the detection result; wherein the detection model is obtained by training the boosting learning model based on the training sample set; the detection result includes the first result and / or the second result corresponding to the first result, the first result characterizes whether there is a fault node in the process node, and the second result characterizes the component adjustment parameter corresponding to the fault node. The present invention achieves the effect of detecting the fault of process components by detecting the number of empty cigarettes. Based on the technical solution of the present invention, the fault detection of process nodes is performed from the two links of empty quality inspection conditions and detection model, which can effectively improve the accuracy of fault detection of process nodes.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a flow chart of a fault detection method provided according to Embodiment 1 of the present invention;

[0022] Figure 2 is a flow chart of a fault detection method provided according to Embodiment 2 of the present invention;

[0023] Figure 3 is an architecture diagram of a fault detection system provided according to an embodiment of the present invention;

[0024] Figure 4 is a structural schematic diagram of a fault detection device provided according to Embodiment 3 of the present invention;

[0025] Figure 5 It is a structural schematic diagram of an electronic device for implementing the fault detection method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0028] Embodiment 1

[0029] Figure 1 A flowchart of a fault detection method is provided for Embodiment 1 of the present invention. This embodiment is applicable to the situation of fault detection of mechanical components corresponding to multiple process links on a cigarette production line. This method can be executed by a fault detection device, which can be implemented in the form of hardware and / or software, and the fault detection device can be configured in a computer. As Figure 1 shown, the method includes:

[0030] S110. Determine multiple process nodes on the cigarette production line. For each of the process nodes, collect the form data of the cigarettes produced in the process through the corresponding sensors of the process nodes, and determine the number of empty-headed cigarettes corresponding to the process nodes according to the form data.

[0031] Among them, the cigarette production line can be understood as the production line of cigarettes. The process node can be understood as the node corresponding to the process link of cigarette production. It should be understood that there are multiple process nodes on the cigarette production line. Optionally, the multiple process nodes include a first node corresponding to a cutting drum, a second node corresponding to a rubbing plate, a third node corresponding to a turning drum, a fourth node corresponding to a conveyor belt, and a fifth node corresponding to a cigarette discharging channel.

[0032] In the embodiments of the present invention, corresponding sensors are installed at each of the process nodes. The types of sensors corresponding to different process nodes may be different or the same. The types of sensors may include image sensors, laser sensors, optoelectronic sensors, etc.

[0033] The cigarette rods produced in the process can be understood as the cigarette rods produced at the current process node. The morphological data can characterize the morphological characteristics of the cigarette rods produced at the current production node. Optionally, the morphological data may be image data, optoelectronic data, laser data, etc. collected by sensors. Generally, the morphologies of the cigarette rods produced at different process nodes may be different.

[0034] The number of hollow cigarette rods can be understood as the number of cigarette rods with the morphological defect of being hollow. In the embodiments of the present invention, the number of hollow cigarette rods is related to the actual detection situation and is not specifically limited herein. Exemplarily, if the number of cigarette rods produced in the target batch is 1000, the number of hollow cigarette rods may be 10, 15, or 50, etc.

[0035] S120. When the numbers of hollow cigarette rods at multiple process nodes meet the hollow quality inspection conditions, determine the hollow change data between every two process nodes according to the numbers of hollow cigarette rods.

[0036] Among them, the hollow quality inspection conditions can be used to judge whether the numbers of hollow cigarette rods corresponding to each process node are qualified. Optionally, the hollow quality inspection conditions may be that the numbers of hollow cigarette rods corresponding to each process node do not exceed the corresponding hollow quantity thresholds. Among them, the hollow quantity thresholds may be thresholds related to the numbers of hollow cigarette rods. In the embodiments of the present invention, the hollow quantity thresholds corresponding to each process node may be the same or different. The specific values of the hollow quantity thresholds corresponding to each process node may be related to the scenario requirements and are not specifically limited herein.

[0037] The hollow change data can characterize the change situation between the numbers of hollow cigarette rods at every two process nodes. Optionally, the hollow change data may be the difference between the numbers of hollow cigarette rods at every two process nodes. Exemplarily, the number of hollow cigarette rods at process node A is 25, and the number of hollow cigarette rods at process node B is 30. The hollow change data between process node A and process node B is 5, which indicates that compared with the previous process node A of process node B, 5 more hollow cigarette rods have been added.

[0038] Optionally, after determining the number of hollow cigarette rods corresponding to the process node according to the morphological data, it further includes:

[0039] When the number of empty cigarettes does not satisfy the empty cigarette quality inspection condition, a fault node is determined, and a fault alarm is issued for the fault node through an alarm device.

[0040] The target fault node can be understood as a process node with a fault. Specifically, the process node where the number of empty cigarettes exceeds the corresponding empty number threshold is determined as the target fault node. In the embodiment of the present invention, there can be one or more target fault nodes.

[0041] The alarm device can be understood as a device with an alarm function. In the embodiment of the present invention, the alarm device can be preset according to the scene requirements, and is not specifically limited here. Exemplarily, the alarm device can be a speaker or an alarm light.

[0042] Based on the above embodiment, when the number of empty cigarettes at any process node exceeds the standard, the process node can be regarded as a fault node, and a targeted fault alarm can be performed on the fault node.

[0043] S130, inputting the plurality of empty cigarette quantities and at least one empty cigarette change data into a detection model to determine a detection result.

[0044] In an embodiment of the present invention, the detection model is obtained by training a boosting learning model based on a training sample set; the detection result includes a first result and / or a second result corresponding to the first result, the first result characterizing whether there is a fault node in the process node, and the second result characterizing the component adjustment parameter corresponding to the fault node.

[0045] Wherein, the detection model can be used to identify a faulty node in a process node. Optionally, the faulty node includes a mechanical component with a fault. The detection model can also be used to determine the component adjustment parameters corresponding to the mechanical component with a fault in the faulty node. As an optional embodiment, specifically, a plurality of the empty cigarette numbers and at least one of the empty change data are input into the detection model, so that the detection model analyzes the change trend of the number of empty cigarettes between a plurality of process nodes based on the plurality of the empty change data to obtain the faulty node; and the detection model analyzes the number of empty cigarettes at the faulty node to obtain the component adjustment parameters corresponding to the faulty node.

[0046] The first result characterizes whether there is a faulty node in the process node. Optionally, the first result may include the presence of a faulty node or the absence of a faulty node. In the case where the first result is that there is no faulty node, the detection result may only include the first result. In the case where the first result is that there is a faulty node, the detection result may include the first result and the second result. The second result may characterize the component adjustment parameters corresponding to the faulty node. The component adjustment parameters may be used to adjust the parameters of the mechanical components in the faulty node. From the perspective of the detection model, it is believed that the production of cigarettes through mechanical components with the component adjustment parameters can reduce the number of cigarettes with empty defects.

[0047] The training sample set can be used to train the boosting learning model to obtain the detection model.

[0048] The boosted learning model can be understood as a strong learner obtained by integrating multiple weak learners.

[0049] Optionally, after inputting the plurality of empty cigarette numbers and at least one empty cigarette change data into the detection model to determine the detection result, the method further includes:

[0050] In the case where the detection result includes the first result and the second result, a trend change graph is generated based on the process node, the number of empty cigarettes and the empty cigarette change data, and the first information corresponding to the fault node and the second information corresponding to the component adjustment parameter are marked on the trend change graph to obtain a fault marking graph;

[0051] The fault marking diagram is displayed by a display device; and a fault warning is issued by a warning device.

[0052] The trend change graph may represent the number of empty cigarettes between multiple process nodes. In the embodiment of the present invention, the format of the trend change graph is not specifically limited and can be set according to the scene requirements. Optionally, the trend change graph may be a line graph, a bar graph, or a scattered single graph.

[0053] The fault marking diagram may be marked with relevant information corresponding to the detection result. The relevant information may include first information and second information.

[0054] The display device can be understood as a device with a data display function. Optionally, the display device can be a display screen of a detection terminal. Exemplarily, the detection terminal can be a computer or a mobile phone.

[0055] The warning device can be understood as a device with a warning function. In the embodiment of the present invention, the warning device can be preset according to the scene requirements, and is not specifically limited here. Exemplarily, the warning device can be a speaker or an alarm light.

[0056] Based on the above-mentioned embodiment scheme, it is possible to display the number of empty cigarettes corresponding to each process node, the empty cigarette change data and the detection results to the customer in a visual manner, so that the customer can clearly understand the empty cigarette situation of each process node, the empty cigarette change situation between each process node and the fault situation of each process node.

[0057] Optionally, after inputting the plurality of empty cigarette numbers and at least one empty cigarette change data into the detection model to determine the detection result, the method further includes:

[0058] For each of the fault nodes, parameters of the cigarette production component corresponding to the fault node are adjusted according to the component adjustment parameters.

[0059] Among them, the cigarette production components can be understood as mechanical components related to cigarette production.

[0060] Based on the above embodiment scheme, after the detection result is determined by the detection model, the effect of automatically adjusting the parameters of the mechanical components of the faulty node based on the detection result can be achieved without manual intervention, which can effectively reduce labor costs.

[0061] Optionally, the training sample set includes a plurality of sample subsets; the boosting learning model includes a plurality of weak learners; and before inputting the plurality of empty cigarette numbers and at least one empty cigarette change data into the detection model, the method further includes:

[0062] For each sample subset, the corresponding weak learner is trained by using the sample subset to obtain the trained weak learner;

[0063] Determine the learner weight corresponding to each of the weak learners that have completed training;

[0064] The detection model is determined according to the multiple weak learners that have been trained and the multiple learner weights.

[0065] The sample subset may be used to train the weak learner, which may be understood as a basic deep learning model.

[0066] The learner weight can be understood as the corresponding weight of the weak learner. Optionally, the learner weight corresponding to each weak learner can be determined based on the corresponding sample subset or based on the training situation of the corresponding weak learner. In an embodiment of the present invention, the learner weights corresponding to different weak learners can be different or the same.

[0067] Based on the above embodiment scheme, by first training a weak learner and then constructing the detection model based on a weighted integration of multiple trained weak learners, the accuracy of the integrated detection model can be made higher.

[0068] The technical solution of the embodiment of the present invention determines multiple process nodes on the cigarette production line, collects the morphological data corresponding to the cigarettes produced in the process for each process node through the corresponding sensor of the process node, and determines the number of empty cigarettes corresponding to the process node according to the morphological data; when the number of empty cigarettes meets the empty quality inspection conditions, the empty change data between each two process nodes is determined according to the number of empty cigarettes; the number of empty cigarettes and at least one empty change data are input into the detection model to determine the detection result; wherein the detection model is obtained by training the boosting learning model based on the training sample set; the detection result includes the first result and / or the second result corresponding to the first result, the first result characterizes whether there is a fault node in the process node, and the second result characterizes the component adjustment parameter corresponding to the fault node. The present invention achieves the effect of detecting the fault of process components by detecting the number of empty cigarettes. Based on the technical solution of the present invention, the fault detection of process nodes is performed from the two links of empty quality inspection conditions and detection model, which can effectively improve the accuracy of fault detection of process nodes.

[0069] Embodiment 2

[0070] Figure 2 This is a flowchart of a fault detection method provided in the second embodiment of the present invention. This embodiment is to collect the morphological data corresponding to the cigarettes produced in the process through the corresponding sensors of each process node in the above embodiment. Figure 2 As shown, the method includes:

[0071] S210. Determine multiple process nodes on a cigarette production line.

[0072] Optionally, the plurality of process nodes include a first node corresponding to the cutting drum, a second node corresponding to the washboard, a third node corresponding to the U-turn drum, a fourth node corresponding to the conveyor belt, and a fifth node corresponding to the smoke lowering channel.

[0073] Among them, the cutting drum, the corrugated plate, the turning drum, the conveyor belt, and the lower cigarette channel can be understood as five key components on the cigarette production line.

[0074] Under normal circumstances, the cutting drum can be used to accurately cut double-length cigarettes into two independent cigarettes. The corrugated plate can be used to complete the assembly and forming of the filter tip and the cigarette. The turning drum can be used to turn the cigarette 180° between the filter tip end and the tobacco end. The conveyor belt can be used to convey the produced cigarettes from the cigarette making machine to the packaging machine at high speed for buffering. The lower cigarette channel can be used to make the cigarettes enter the packaging station in an orderly manner.

[0075] S220. For the first node, collect the first laser data of the cigarettes produced in the process through the first laser sensor, and use the first laser data as the morphological data of the first node.

[0076] Among them, the first laser sensor can be installed at the first node and is used to collect the laser data of the cigarettes produced in the process processed by the cutting drum, that is, the first laser data. In the embodiment of the present invention, the specific form of the laser data is not specifically limited. Optionally, the laser data can be a laser image or point cloud data, etc.

[0077] S230. For the second node, collect the first image data of the cigarettes produced in the process through the first image sensor, and use the first image data as the morphological data of the second node.

[0078] Among them, the first image sensor can be installed at the second node and is used to collect the image data of the cigarettes produced in the process processed by the corrugated plate, that is, the first image data.

[0079] S240. For the third node, collect the second laser data of the cigarettes produced in the process through the second laser sensor; use the second laser data as the morphological data of the third node.

[0080] Among them, the second laser sensor can be installed at the third node and is used to collect the laser data of the cigarettes produced in the process processed by the turning drum, that is, the second laser data.

[0081] S250. For the fourth node, collect the second image data of the cigarettes produced in the process through the second image sensor, and use the second image data as the morphological data of the fourth node.

[0082] Among them, the second image sensor can be installed at the fourth node and is used to collect the image data of the cigarettes produced in the process conveyed by the conveyor belt, that is, the second image data.

[0083] S260. For the fifth node, collect first photoelectric data of the cigarettes produced in the process through a first photoelectric sensor, and use the first photoelectric data as the morphological data of the fifth node.

[0084] Among them, the first photoelectric sensor can be installed at the fifth node, and is used to collect photoelectric data of the cigarettes produced in the process and transmitted through the down-smoke channel, that is, the first photoelectric data.

[0085] Optionally, the plurality of process nodes further include a sixth node corresponding to a rolling component, a seventh node corresponding to a hollow end detection component, and an eighth node corresponding to a packaging machine; for each of the process nodes, collecting the morphological data corresponding to the cigarettes produced in the process through the sensor corresponding to the process node also includes:

[0086] For the sixth node, humidity data of the cigarettes produced in the process is collected by a humidity sensor, and the humidity data is used as the morphological data of the sixth node;

[0087] For the seventh node, collecting second photoelectric data of the cigarettes produced in the process through a second photoelectric sensor, and using the second photoelectric data as the morphological data of the seventh node;

[0088] For the eighth node, third image data of the cigarettes produced in the process are collected by a third image sensor, and the third image data are used as the morphological data of the eighth node.

[0089] Among them, the rolling component, the empty end detection component and the packaging machine can be understood as the other three key components on the cigarette production line. Commonly, the process relationship of the above eight components is from first to last: rolling component, cutting drum, rubbing board, empty end detection component, U-turn drum, conveyor belt, cigarette down channel, and packaging machine.

[0090] Generally, the rolling component can be used to combine the three elements of tobacco, cigarette paper and filter into a complete cigarette. The empty detection component can be used to monitor the empty ends of cigarettes. The empty detection component can be an original matching component installed on the production line. The corresponding sensor of the empty detection component can be related to the matching standard of the component, which is not specifically limited here. The packaging machine can be used for aluminum foil lining packaging, trademark paper wrapping, transparent paper packaging and hard box forming of cigarettes.

[0091] The humidity sensor may be installed at the sixth node to collect humidity data of cigarettes produced by the process of the rolling component. It should be understood that the humidity of the shredded tobacco affects the number of empty cigarettes.

[0092] The second photoelectric sensor may be installed at the seventh node, and is used to collect the photoelectric data of the cigarettes produced by the process detected by the empty detection component, namely, the second photoelectric data.

[0093] The third image sensor may be installed at the eighth node, and is used to collect image data of cigarettes produced through the process of the packaging machine, namely, the third image data.

[0094] Based on the above embodiment, it is possible to obtain

[0095] S270, determining the number of empty cigarettes corresponding to the process node according to the morphological data.

[0096] S280. When the number of multiple empty cigarettes meets the empty quality inspection conditions, determine the empty change data between every two process nodes according to the number of empty cigarettes.

[0097] S290, inputting the plurality of empty cigarette quantities and at least one empty cigarette change data into a detection model to determine a detection result.

[0098] The technical solution of the embodiment of the present invention is that, for the first node, the first laser data of the cigarette produced in the process is collected by the first laser sensor, and the first laser data is used as the morphological data of the first node; for the second node, the first image data of the cigarette produced in the process is collected by the first image sensor, and the first image data is used as the morphological data of the second node; for the third node, the second laser data of the cigarette produced in the process is collected by the second laser sensor, and the second laser data is used as the morphological data of the third node; for the fourth node, the second image data of the cigarette produced in the process is collected by the second image sensor, and the second image data is used as the morphological data of the fourth node; for the fifth node, the first photoelectric data of the cigarette produced in the process is collected by the first photoelectric sensor, and the first photoelectric data is used as the morphological data of the fifth node. The present invention realizes the comprehensiveness and diversity of the acquired morphological data by installing sensors with different functions at multiple process nodes on the cigarette production line, so as to realize the subsequent detection model to perform cigarette empty detection and cigarette component detection according to different forms of data, and enrich the dependent data for the detection model to perform related calculations.

[0099] Figure 3 is an architecture diagram of a fault detection system provided according to an embodiment of the present invention. Figure 3 , the fault detection method is further explained as follows.

[0100] Based on the sensors installed at different positions on the production line, the weight sensors or density sensors originally installed on the cigarette weight and density monitoring components on the production line, the photoelectric sensors or suction resistance sensors originally installed on the short end detection components on the production line, and the image sensors originally installed on the packaging machines on the production line, the changing trend of the number of short end defective cigarettes in each process of the production line is analyzed, the process nodes where the short end defects are aggravated are determined, and the production parameters of the relevant components at the process nodes are adjusted.

[0101] The humidity data of the tobacco at the end of the cigarette is collected by the humidity sensor on the rolling component, and the number of empty cigarettes at the process node is determined based on the collected humidity data. In addition to the humidity sensor, the sensor on the process node can also be installed with a weight sensor and a density sensor, so as to determine the number of empty cigarettes at the process node by combining the data obtained by multiple sensors.

[0102] The laser sensor installed at the cutting drum collects the laser data of the cigarettes produced there, and the cutting quality of the produced cigarettes is analyzed through the collected laser data to determine the number of empty cigarettes at this process node. Among them, the cutting quality is related to parameters such as the bluntness of the cutting drum blade, burrs on the blade edge, or abnormal rotation speed.

[0103] The image sensor installed at the washboard collects the image data of the produced cigarettes, and the degree of concavity of the appearance of the produced cigarettes is analyzed through the collected image data to determine the number of empty cigarettes at the process node. Among them, the degree of concavity is related to parameters such as the rubbing strength of the washboard.

[0104] The photoelectric sensor installed on the original empty detection component collects the photoelectric data of the cigarettes produced at the location, and the number of empty cigarettes at the process node is determined based on the collected photoelectric data. In addition to the photoelectric sensor, the sensor on the process node can also be installed with a draw resistance sensor and an image sensor, etc., to combine the data obtained by multiple sensors to determine the number of empty cigarettes at the process node.

[0105] The laser sensor installed at the U-turn drum collects the laser image of the cigarettes produced there, and the number of empty cigarettes in the cigarette inversion process link is determined by the collected laser image. Specifically, the number of empty cigarettes at the process node can be determined by analyzing the brightness distribution of the laser image.

[0106] The image data of the process node is collected by the image sensor at the cigarette conveyor belt, and the number of empty cigarettes in the cigarette conveying process link is determined by the collected image data. Among them, the conveyor belt resistance in the cigarette conveying process link may cause the cigarette to break or twist, so that the number of empty cigarettes increases.

[0107] The photoelectric data of the process node is collected by the photoelectric sensor installed at the cigarette downflow channel of the packaging machine to determine the number of empty cigarettes in this process link.

[0108] The image sensor originally installed on the packaging machine collects image data of cigarettes produced by the process being processed by the packaging machine, and the collected images are analyzed to determine the number of empty cigarettes in this process link.

[0109] The technical solution of the present invention combines the newly added sensors with the original sensors on the production line to realize the detection of the number of empty cigarettes in each key process link on the production line. It improves the comprehensiveness of the detection of the number of empty cigarettes, which serves as the basis for the subsequent accurate positioning of faults in the process links.

[0110] The technical solution of the present invention is further described as follows.

[0111] Data fusion and correlation analysis. Multidimensional data fusion, specifically, different types of data obtained by various sensors, such as cigarette weight, density, draw resistance, humidity, temperature, and the number of empty ends removed by cigarette making and packaging machines, are fused through data fusion technology to form a total data set, thereby eliminating noise and errors from different data sources and improving the credibility of the data. Position difference analysis, specifically, by comparing sensor data at different locations, the differences between different process links are analyzed.

[0112] Trend analysis and identification of short defects. Trend detection and prediction, specifically, can use time series analysis and sliding window methods to predict the trend of various indicators of cigarettes and monitor the changes of short defects in real time. Through deep learning models, identify the occurrence patterns of short defects in different process links. Intelligent prediction model, specifically, based on machine learning technology, such as regression analysis, decision tree or neural network, establish a prediction model for short defects. This model can predict the development trend of short defects based on the currently collected data and discover potential problems in advance.

[0113] Anomaly detection and cause analysis. Multi-dimensional anomaly detection, specifically, by setting a reasonable tolerance range, detect abnormal data in each process link in real time. Real-time analysis of the time or location of the short defect to determine whether it is caused by parameter fluctuations in a specific link. Multi-factor cause analysis, specifically, combining sensor data with production line equipment status to analyze the specific cause of the short defect. Through methods such as multivariate regression or causal inference analysis, identify whether the key factors that exacerbate the short are related to factors such as equipment failure, process fluctuations, or raw material changes.

[0114] Intelligent early warning and feedback mechanism. Intelligent early warning, specifically, when it detects that the number of short defects increases and the trend is abnormal, it automatically sends out an early warning signal to notify the operator. The early warning information includes the location and time of the defect, and can also provide possible analysis reasons to help the operator take corresponding measures in time. Automatic feedback and adjustment, specifically, automatically adjust relevant production parameters, such as equipment operating speed, heating temperature and pressure, to eliminate or reduce the occurrence of short defects. Intelligent parameter adjustment is realized, reducing manual intervention.

[0115] Closed-loop optimization and intelligent decision support. Dynamic adjustment and optimization. Specifically, through real-time monitoring and intelligent algorithms, the production process is continuously adjusted to optimize parameters. For example, if the defect rate of a certain link is high, the parameter adjustment strategy is intelligently recommended based on historical data to optimize the production process. Decision support system. Specifically, detailed production reports and defect analysis reports are generated to provide decision support for production managers. The report can include the root cause of the defect, risk points in the production process, or future improvement measures.

[0116] Visualization and data display. Real-time data visualization, specifically, presents the sensor data and analysis results of each link through a graphical interface, so that operators and engineers can intuitively understand the quality status of each process link and quickly determine the occurrence and trend of short defects. Defect distribution diagram and trend diagram, specifically, display the distribution of short defects, analyze the defect incidence rate of each production link, and show the changing trend of defects through trend diagrams to help production managers predict future quality fluctuations.

[0117] The present invention combines sensor technology with existing production data, and through multi-dimensional data fusion and trend analysis, monitors the empty head defects in the cigarette production process in real time, accurately determines the cause of the empty head defect, and identifies whether the source is a mechanical failure, process fluctuation or other unknown factors, thereby providing accurate fault diagnosis and early warning. It realizes real-time feedback of the operating status of the production line through intelligent data processing and big data analysis, supports automatic adjustment of production parameters, effectively reduces the occurrence of empty head defects, and improves production efficiency and product quality. According to the real-time data changes in the production process, potential problems can be quickly discovered, and through early warning, defect expansion and waste can be avoided, thereby maximizing production efficiency.

[0118] The boost learning model described in this embodiment is further described as follows. The mathematical model related to the boost learning model and the training process of the model are as follows:

[0119] Model initialization. Given a training dataset {(x1, y1), (x2, y2), …, (x n ,y n )}, where x i is a feature, yi is a label. Among them, in each round of training to improve the learning model, the goal is to minimize the loss function of the model, and the loss function can be as follows:

[0120]

[0121] Among them, represents the model prediction error, including squared error or logarithmic loss; Ω(h t ) represents the regularization term, which is used to control the complexity of the tree; F(x i ) represents the predicted value of the model on the sample x i ; h t represents the t-th tree; T represents the number of trees.

[0122] The training steps of the boosting learning model are to gradually train multiple weak learners and weight-sum their prediction results to obtain the final model. Residual calculation: At the t-th round of training, the residual is the gradient of the objective function. Specifically, train a new weak learner h t (x) and make it fit the residual as well as possible. The implementation formula is as follows:

[0123]

[0124] Among them, represents the objective (residual) of the model at the t-th round, that is, the difference between the true value y i and the predicted value F t-1 (x).

[0125] Update the model. Add the new weak learner to the current model to obtain the updated model F t (x):

[0126] F t (x) = F t-1 (x) + ηh t (x)

[0127] Among them, η represents the learning rate, which is used to control the step size of each round of update and prevent overfitting; F t (x) represents the updated model.

[0128] Weight-sum all the iterative models to obtain the detection model. The implementation formula is as follows:

[0129]

[0130] Among them, F0(x) represents the initialized model; h t (x) represents the decision tree obtained in the t-th round of training; F(x) represents the detection model.

[0131] Based on the technical solution of the present invention, a detection model that is more suitable for fault detection of mechanical components corresponding to multiple process links on a cigarette production line can be trained to improve the accuracy of fault detection.

[0132] Embodiment 3

[0133] Figure 4 This is a schematic diagram of the structure of a fault detection device provided in Embodiment 3 of the present invention. Figure 4 As shown, the device includes: a multi-node data acquisition module 310, a data analysis module 320 and a fault prediction module 330.

[0134] Among them, the multi-node data acquisition module 310 is used to determine multiple process nodes on the cigarette production line, and for each of the process nodes, the morphological data corresponding to the cigarettes produced by the process node are collected through the corresponding sensors of the process node, and the number of empty cigarettes corresponding to the process node is determined according to the morphological data; the data analysis module 320 is used to determine the empty cigarette change data between every two process nodes according to the number of empty cigarettes when the number of empty cigarettes meets the empty quality inspection conditions; the fault prediction module 330 is used to input the number of empty cigarettes and at least one empty cigarette change data into the detection model to determine the detection result; wherein the detection model is obtained by training the boosting learning model based on the training sample set; the detection result includes a first result and / or a second result corresponding to the first result, the first result characterizing whether there is a faulty node in the process node, and the second result characterizing the component adjustment parameter corresponding to the faulty node.

[0135] The technical solution of the embodiment of the present invention determines multiple process nodes on the cigarette production line, collects the morphological data corresponding to the cigarettes produced in the process for each process node through the corresponding sensor of the process node, and determines the number of empty cigarettes corresponding to the process node according to the morphological data; when the number of empty cigarettes meets the empty quality inspection conditions, the empty change data between each two process nodes is determined according to the number of empty cigarettes; the number of empty cigarettes and at least one empty change data are input into the detection model to determine the detection result; wherein the detection model is obtained by training the boosting learning model based on the training sample set; the detection result includes the first result and / or the second result corresponding to the first result, the first result characterizes whether there is a fault node in the process node, and the second result characterizes the component adjustment parameter corresponding to the fault node. The present invention achieves the effect of detecting the fault of process components by detecting the number of empty cigarettes. Based on the technical solution of the present invention, the fault detection of process nodes is performed from the two links of the empty quality inspection conditions and the detection model, which can effectively improve the accuracy of fault detection of process nodes.

[0136] Optionally, the plurality of process nodes include a first node corresponding to the cutting drum, a second node corresponding to the washboard, a third node corresponding to the U-turn drum, a fourth node corresponding to the conveyor belt, and a fifth node corresponding to the smoke lowering channel;

[0137] The multi-node data acquisition module 310 includes: a first data acquisition unit, a second data acquisition unit, a third data acquisition unit, a fourth data acquisition unit, and a fifth data acquisition unit;

[0138] The first data acquisition unit is used to collect first laser data of cigarettes produced in the process through a first laser sensor for the first node, and use the first laser data as the morphological data of the first node;

[0139] The second data acquisition unit is used for acquiring first image data of the cigarette produced in the process through a first image sensor for the second node, and using the first image data as the morphological data of the second node;

[0140] The third data acquisition unit is used to collect second laser data of the cigarettes produced in the process through a second laser sensor for the third node; and use the second laser data as the morphological data of the third node;

[0141] The fourth data collection sheet is used to collect second image data of the cigarettes produced in the process through a second image sensor for the fourth node, and use the second image data as the morphological data of the fourth node;

[0142] The fifth data collection unit is used to collect the first photoelectric data of the cigarettes produced in the process through the first photoelectric sensor for the fifth node, and use the first photoelectric data as the morphological data of the fifth node.

[0143] Optionally, the plurality of process nodes further include a sixth node corresponding to a rolling component, a seventh node corresponding to a blank detection component, and an eighth node corresponding to a packaging machine;

[0144] The multi-node data acquisition module 310 includes: a sixth data acquisition unit, a seventh data acquisition unit and an eighth data acquisition unit;

[0145] The sixth data collection unit is used to collect humidity data of cigarettes produced in the process through a humidity sensor for the sixth node, and use the humidity data as the morphological data of the sixth node;

[0146] The seventh data collection unit is used for collecting second photoelectric data of the cigarettes produced in the process through a second photoelectric sensor for the seventh node, and using the second photoelectric data as the morphological data of the seventh node;

[0147] The eighth data acquisition unit is used to acquire the third image data of the cigarettes produced in the process by using the third image sensor for the eighth node, and use the third image data as the morphological data of the eighth node.

[0148] Optionally, the fault detection device further includes: a trend graph marking module and a display warning module;

[0149] The trend diagram annotation module is used to generate a trend change diagram based on the process node, the number of empty cigarettes and the empty cigarette change data after the multiple empty cigarette numbers and at least one empty cigarette change data are input into the detection model to determine the detection result, and annotate the first information corresponding to the fault node and the second information corresponding to the component adjustment parameter on the trend change diagram to obtain a fault annotation diagram when the detection result includes the first result and the second result;

[0150] The display warning module is used to display the fault marking diagram through a display device; and to issue a fault warning through a warning device.

[0151] Optionally, the fault detection device also includes: a component parameter adjustment module, which is used to adjust the parameters of the cigarette production component corresponding to the fault node according to the component adjustment parameters for each fault node after inputting the multiple numbers of empty cigarettes and at least one empty cigarette change data into the detection model to determine the detection result.

[0152] Optionally, the fault detection device also includes: a fault alarm module, which is used to determine the fault node after determining the number of empty cigarettes corresponding to the process node according to the morphological data, and to issue a fault alarm to the target fault node through an alarm device when the number of empty cigarettes does not meet the empty quality inspection conditions.

[0153] Optionally, the training sample set includes multiple sample subsets; the boost learning model includes multiple weak learners; the fault detection device further includes: a weak learner training module, a weight determination module and a detection model determination module;

[0154] The weak learner training module is used to train the corresponding weak learner for each sample subset by using the sample subset before inputting the plurality of empty cigarette numbers and at least one empty cigarette change data into the detection model to obtain the trained weak learner;

[0155] The weight determination module is used to determine the learner weight corresponding to each weak learner that has completed training;

[0156] The detection model determination module is used to determine the detection model according to the multiple weak learners and the multiple learner weights that have been trained.

[0157] The fault detection device provided in the embodiment of the present invention can execute the fault detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0158] Embodiment 4

[0159] Figure 5 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0160] like Figure 5 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0161] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disc, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0162] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the fault detection method.

[0163] In some embodiments, the fault detection method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the fault detection method described above can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the fault detection method by any other suitable means (e.g., by means of firmware).

[0164] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), application-specific standard products (ASSP), systems-on-chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0165] A computer program for implementing the method of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The computer programs may be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine, or entirely on the remote machine or server.

[0166] In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium may be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0167] In order to provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0168] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), blockchain network, and the Internet.

[0169] A computing system can include a client and a server. The client and the server are generally far from each other and typically interact through a communication network. The client-server relationship is created by computer programs running on respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0170] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is imposed herein.

[0171] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A fault detection method, characterized in that, include: Determine multiple process nodes on a cigarette production line, collect morphological data corresponding to cigarettes produced by each process node through a sensor corresponding to the process node, and determine the number of empty cigarettes corresponding to the process node based on the morphological data; In the case where the number of multiple empty cigarettes meets the empty quality inspection condition, determining the empty change data between every two process nodes according to the number of empty cigarettes; Input multiple numbers of empty cigarettes and at least one empty cigarette change data into the detection model to determine the detection result; wherein, the detection model is obtained by training the boosting learning model based on the training sample set; the detection result includes a first result and / or a second result corresponding to the first result, the first result characterizes whether there is a fault node in the process node, and the second result characterizes the component adjustment parameter corresponding to the fault node.

2. The method according to claim 1, wherein The plurality of process nodes include a first node corresponding to the cutting drum, a second node corresponding to the washboard, a third node corresponding to the U-turn drum, a fourth node corresponding to the conveyor belt, and a fifth node corresponding to the smoke lowering channel; For each of the process nodes, collecting the shape data corresponding to the cigarettes produced in the process through the sensor corresponding to the process node includes: For the first node, first laser data of cigarettes produced in the process is collected by a first laser sensor, and the first laser data is used as the morphological data of the first node; For the second node, first image data of the cigarette produced in the process is collected by a first image sensor, and the first image data is used as the morphological data of the second node; For the third node, collecting second laser data of the cigarettes produced in the process through a second laser sensor; using the second laser data as the morphological data of the third node; For the fourth node, collecting second image data of the cigarettes produced in the process by a second image sensor, and using the second image data as the morphological data of the fourth node; For the fifth node, the first photoelectric data of the cigarettes produced in the process is collected by a first photoelectric sensor, and the first photoelectric data is used as the morphological data of the fifth node.

3. The method according to claim 2, wherein The plurality of process nodes also include a sixth node corresponding to a rolling component, a seventh node corresponding to a hollow head detection component, and an eighth node corresponding to a packaging machine; for each of the process nodes, the morphological data corresponding to the cigarettes produced in the process are collected through the sensors corresponding to the process nodes, and further include: For the sixth node, humidity data of the cigarettes produced in the process is collected by a humidity sensor, and the humidity data is used as the morphological data of the sixth node; For the seventh node, collecting second photoelectric data of the cigarettes produced in the process through a second photoelectric sensor, and using the second photoelectric data as the morphological data of the seventh node; For the eighth node, third image data of the cigarettes produced in the process are collected by a third image sensor, and the third image data are used as the morphological data of the eighth node.

4. The method according to claim 1, wherein After inputting the quantities of multiple empty-tipped cigarettes and at least one empty-tip change data into the detection model to determine a detection result, the following steps are further included: When the detection result includes a first result and a second result, generate a trend change graph based on the process node, the quantity of empty-tipped cigarettes, and the empty-tip change data, and mark the first information corresponding to the fault node and the second information corresponding to the component adjustment parameter on the trend change graph to obtain a fault annotation graph; Display the fault annotation graph through a display device; and perform a fault warning through a warning device.

5. The method according to claim 1, wherein After inputting the quantities of multiple empty-tipped cigarettes and at least one empty-tip change data into the detection model to determine a detection result, the following steps are further included: For each fault node, adjust the parameters of the cigarette production component corresponding to the fault node according to the component adjustment parameter.

6. The method according to claim 1, characterized in that After determining the quantity of empty-tipped cigarettes corresponding to the process node according to the form data, the following steps are further included: When the quantity of empty-tipped cigarettes does not meet the empty-tip quality inspection condition, determine a fault node, and perform a fault alarm on the target fault node through an alarm device.

7. The method according to claim 1, characterized in that The training sample set includes multiple sample subsets; the boosting learning model includes multiple weak learners; Before inputting the quantities of multiple empty-tipped cigarettes and at least one empty-tip change data into the detection model, the following steps are further included: For each sample subset, train the corresponding weak learner with the sample subset to obtain the trained weak learner; Determine the learner weights corresponding to each trained weak learner; Determine the detection model according to the multiple trained weak learners and the multiple learner weights.

8. A fault detection device, characterized in that, Including: A multi-node data acquisition module, configured to determine multiple process nodes on a cigarette production line. For each process node, collect form data of the cigarettes produced in the process through the sensor corresponding to the process node, and determine the quantity of empty-tipped cigarettes corresponding to the process node according to the form data; A data analysis module, configured to determine the empty-tip change data between every two process nodes according to the quantity of empty-tipped cigarettes when the quantities of multiple empty-tipped cigarettes meet the empty-tip quality inspection condition; A fault prediction module, configured to input the quantities of multiple empty-tipped cigarettes and at least one empty-tip change data into a detection model to determine a detection result; wherein, the detection model is obtained by training a boosting learning model based on a training sample set; the detection result includes a first result and / or a second result corresponding to the first result, the first result indicating whether there is a fault node in the process node, and the second result indicating the component adjustment parameter corresponding to the fault node.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and when the computer program is executed by the at least one processor, enables the at least one processor to perform the fault detection method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for implementing the fault detection method according to any one of claims 1-7 when the computer instructions are executed by a processor.

Images