A production line inspection system, method, electronic device, and storage medium

Abstract

This application discloses a production line inspection system, method, electronic device, and storage medium. The system includes an inspection control platform and at least one inspection robot. The inspection control platform performs the following processing: determining the target inspection scenario and target inspection robot of the production line; determining multiple inspection locations and action parameters; generating inspection tasks; and sending the inspection tasks to the target inspection robot. The target inspection robot performs the following processing: receiving the inspection task and inspecting the target inspection scenario indicated by the inspection task; obtaining the inspection content; calling the target inspection model and inputting the inspection content into the target inspection model to obtain the inspection results for the target inspection locations; and outputting the inspection results based on the inspection results. By adopting the above-mentioned production line inspection system, method, electronic device, and storage medium, the problem of low inspection efficiency and accuracy in the inspection process of cigarette manufacturing production lines is solved.

Description

Technical Field

[0001] This application relates to the field of intelligent inspection technology, and in particular to a production line inspection system, method, electronic device and storage medium. Background Technology

[0002] Currently, the inspection work of cigarette manufacturing production lines in my country is carried out by operators at each process position according to the relevant process quality inspection requirements in the "Cigarette Process Specifications" issued by the State Tobacco Monopoly Administration in 2016. The inspection involves checking and triggering alarms for a large number of detailed and complex tasks, including process parameters, equipment parameters, instrument status, material status, and processing accuracy of various production equipment and processes. Therefore, a large amount of inspection work exists in the actual production process. For example, it involves recording and verifying data in the system with data on the field screens and instruments, inspecting the status of materials on site, identifying debris after the unpacking process, and addressing production problems such as mismatched process and equipment parameter data and instrument parameters exceeding limits.

[0003] In current technology, automated inspection scenarios in tobacco cigarette factories still rely on conventional warehouse inspections and inventory checks. This involves two methods: routine inspections and special inspections. Routine inspections involve the logistics center conducting spot checks on cigarettes and related key locations in the buffer area at the start and end of each day. Special inspections primarily address issues such as cigarette jams, excess or insufficient cigarettes, empty pickup reports, and monthly and annual audit and financial inventory checks.

[0004] However, due to the repetitive nature and high labor intensity of the inspection work on the cigarette manufacturing production line, and the fact that no inspection robots are used to assist operators in process quality inspection in the workshops of cigarette factories in my country's tobacco industry, the current inspection method results in low inspection efficiency and poor inspection accuracy. Summary of the Invention

[0005] In view of this, this application provides a production line inspection system, method, electronic device and storage medium. Its purpose is to configure corresponding inspection robots for the target inspection scenarios on the production line according to their characteristics, and match the corresponding target inspection model for each target inspection location inspection item. This solves the problems of low inspection efficiency and poor inspection accuracy in the inspection process of cigarette manufacturing production line.

[0006] In a first aspect, embodiments of this application provide a production line inspection system including an inspection control platform and at least one inspection robot; wherein, the inspection control platform performs the following processes: determining a target inspection scenario for the production line and a target inspection robot adapted to the target inspection scenario; determining multiple inspection locations in the target inspection scenario and motion parameters at each inspection location; generating an inspection task for the target inspection scenario, the inspection task including multiple inspection locations, motion parameters at each inspection location, and inspection items corresponding to each inspection location; and sending the inspection task to the target inspection robot; wherein, The target inspection robot performs the following processing: After receiving the inspection task, it inspects the target inspection scene indicated by the inspection task; if the target inspection robot reaches the target inspection position in the inspection task, it controls the target inspection robot to perform the corresponding action according to the action parameters at the target inspection position to obtain the inspection content corresponding to the target inspection position; it calls the target inspection model that matches the inspection item corresponding to the target inspection position, inputs the inspection content into the target inspection model to obtain the inspection result for the target inspection position; based on the inspection result, it outputs the inspection result for the target inspection position.

[0007] Optionally, the action parameters may include an action type identifier, an action execution height value, and an action execution angle value. The action type identifier indicates the photo-taking action, and each inspection robot includes a camera. The target inspection robot may also perform the following processing: adjusting the camera's shooting height to the action execution height value, adjusting the camera's shooting angle to the action execution angle value, controlling the camera to perform the photo-taking action indicated by the action type identifier, obtaining an inspection image of the target inspection location, and determining the inspection image as the inspection content corresponding to the target inspection location.

[0008] Optionally, each inspection item may include an identification mark, which is used to indicate the inspection type for the inspection location. Each inspection robot is equipped with multiple types of inspection models. The target inspection robot may also perform the following processing: based on the identification mark of the inspection item corresponding to the inspection location, the inspection model corresponding to the inspection type indicated by the identification mark is determined as the target inspection model.

[0009] Optionally, the inspection results may include the recognition fit degree; wherein, the target inspection robot may also perform the following processing: comparing the recognition fit degree of the target inspection location with a set threshold; if the recognition fit degree is greater than the set threshold, determining that there is an anomaly at the target inspection location; determining the anomaly type of the target inspection location based on the invoked target inspection model; and determining the target inspection location, inspection content, anomaly type, and inspection results as the inspection results for the target inspection location.

[0010] Optionally, the production line inspection system may also include a mobile terminal and a database server; the target inspection robot may also perform the following processing: if an anomaly is determined to exist at the target inspection location, the inspection result is sent to the mobile terminal through the database server.

[0011] Optionally, the database server may perform the following processing: divide the inspection results with the same scene identifier into data related to the same target inspection scene and group them together.

[0012] Secondly, embodiments of this application provide a production line inspection method, including: determining a target inspection scenario for the production line and a target inspection robot adapted to the target inspection scenario; determining multiple inspection locations in the target inspection scenario and motion parameters at each inspection location; generating an inspection task for the target inspection scenario, the inspection task including multiple inspection locations, motion parameters at each inspection location, and inspection items corresponding to each inspection location; and sending the inspection task to the target inspection robot so that the target inspection robot can perform inspection of the target location indicated by the inspection task according to the scenario.

[0013] Optionally, the target inspection robot suitable for the target inspection scenario is determined by the following method: based on the target environment in the target inspection scenario, an inspection robot that supports inspection in the target environment is selected as the target inspection robot. The target environment may include at least one of the following: ground, aerial track and pipeline. The target inspection robot may include at least one of the following: track-suspended intelligent inspection robot, ground mobile intelligent inspection robot and pipeline rolling inspection robot.

[0014] Thirdly, embodiments of this application provide an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the above-described production line inspection method are performed.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when run by a processor, executes the steps of the above-described production line inspection method.

[0016] The embodiments of this application bring the following beneficial effects:

[0017] This application provides a production line inspection system, method, electronic device, and storage medium, including an inspection control platform and at least one inspection robot. The inspection control platform performs the following processes: determining a target inspection scenario for the production line and a target inspection robot adapted to the target inspection scenario; determining multiple inspection locations in the target inspection scenario and motion parameters at each inspection location; generating an inspection task for the target inspection scenario, the inspection task including multiple inspection locations, motion parameters at each inspection location, and inspection items corresponding to each inspection location; and sending the inspection task to the target inspection robot. The process involves a target inspection robot that performs the following steps: upon receiving an inspection task, it inspects the target inspection scenario indicated by the task; if the robot reaches the target inspection location, it executes corresponding actions according to the motion parameters at that location to obtain the inspection content corresponding to that location; it calls the target inspection model matching the inspection items at the target location, inputs the inspection content into the model, and obtains the inspection results for that location; and it outputs the inspection results for that location based on the results. This application addresses the problems of low efficiency and poor accuracy in inspection processes on cigarette manufacturing production lines by configuring appropriate inspection robots for the characteristics of target inspection scenarios on the production line and matching corresponding target inspection models to the inspection items at each target inspection location.

[0018] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the following drawings are some implementation methods of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the production line inspection system provided in an embodiment of this application;

[0021] Figure 2 A flowchart illustrating the processing steps performed in the inspection control platform according to an embodiment of this application;

[0022] Figure 3 A flowchart illustrating the processing steps performed in an inspection robot according to an embodiment of this application;

[0023] Figure 4A flowchart illustrating the production line inspection method provided in this application embodiment;

[0024] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] In current technology, automated inspection scenarios in tobacco cigarette factories still rely on conventional warehouse inspections and inventory checks. This involves two methods: routine inspections and special inspections. Routine inspections involve the logistics center conducting spot checks on cigarettes and related key locations in the buffer area at the start and end of each day. Special inspections primarily address issues such as cigarette pack jams, excess or insufficient cigarettes, empty retrieval reports, and monthly and annual audit and financial inventory checks. However, due to the repetitive and labor-intensive nature of inspection work on cigarette production lines, and the fact that no inspection robots are used to assist operators in process quality inspections within the workshops of tobacco factories in my country, the current inspection methods suffer from low efficiency and poor accuracy.

[0027] Based on this, embodiments of this application provide a production line inspection system, method, electronic device, and storage medium. By configuring corresponding inspection robots for the characteristics of target inspection scenarios on the production line and matching corresponding target inspection models for each target inspection location, the system solves the problems of low inspection efficiency and poor inspection accuracy in the inspection process of cigarette manufacturing production lines.

[0028] To facilitate understanding of this embodiment, a production line inspection method disclosed in this application will first be described in detail. Figure 1 This is a schematic diagram of the production line inspection system provided in the embodiments of this application, as shown below. Figure 1 As shown, the production line inspection system 10 includes an inspection control platform 100 and a target inspection robot 200.

[0029] First, a brief introduction to the names involved in the embodiments of this application will be given.

[0030] In this embodiment, the target inspection robot 200 can be controlled by the inspection control platform 100 to perform inspections on multiple inspection items at multiple inspection locations in the target inspection scenario, wherein:

[0031] Inspection Control Platform 100:

[0032] It refers to a platform that can manage and control the inspection process of multiple target inspection robots. This inspection control platform can integrate and control multiple management and control systems in a unified manner.

[0033] Inspection site:

[0034] This refers to the production workshop of a cigarette factory. The size of the production workshop depends on the actual situation of the cigarette factory. Each production workshop has multiple production lines, and each production line is used to complete the production process that meets the cigarette manufacturing specifications.

[0035] This application provides a production line inspection system that can configure corresponding inspection robots according to the characteristics of the target inspection scenarios on the production line, and match the corresponding target inspection model for each target inspection location. This solves the problems of low inspection efficiency and poor inspection accuracy in the inspection process of cigarette manufacturing production lines.

[0036] The exemplary parts provided in the embodiments of this application will be described below.

[0037] The inspection control platform 100 can perform the following processes:

[0038] Figure 2 This is a flowchart illustrating the processing steps performed in the inspection control platform according to an embodiment of this application. Figure 2 As shown, it includes:

[0039] Step S1001: Determine the target inspection scenario for the production line and the target inspection robot suitable for the target inspection scenario.

[0040] Step S1002: Determine multiple inspection locations in the target inspection scenario and the action parameters at each inspection location;

[0041] Step S1003: Generate an inspection task for the target inspection scenario;

[0042] Step S1004: Send the inspection task to the target inspection robot.

[0043] In step S1001, the target inspection scenario refers to a production line in a cigarette factory. Each production line corresponds to a production process, which, for example, includes but is not limited to: opening the package, re-moistening, and feeding. Each production process includes multiple production devices, which together constitute a production process to achieve the corresponding production function. During the inspection process, it is necessary to check the operation of the production process to ensure that it is in an operating state that meets production requirements. By checking multiple inspection points on the production process, the purpose of monitoring the operating status is achieved. These inspection points are systematically combined to form the target inspection scenario.

[0044] The Target Inspection Robot 200 refers to an intelligent robot capable of autonomous positioning and navigation, using radar for intelligent obstacle avoidance. Equipped with an adjustable camera, its height and angle are adjustable. Based on the inspection path, time, and frequency specified in the inspection task, it takes photos and records videos of the target inspection equipment, achieving fully automated inspection scene monitoring. It also features an alarm function. Target inspection robots come in various types, and the Target Inspection Robot 200 includes, but is not limited to: track-suspended intelligent inspection robots, ground-mobile intelligent inspection robots, and pipeline rolling intelligent inspection robots.

[0045] In an optional embodiment, the target inspection robot 200 adapted to the target inspection scenario is determined by the following method: based on the target environment in which the target inspection scenario is located, an inspection robot that supports inspection in the target environment is selected as the target inspection robot 200. The target environment includes at least one of the following: ground, aerial track and pipeline. The target inspection robot 200 includes at least one of the following: track-suspended inspection robot, ground mobile inspection robot and pipeline rolling inspection robot.

[0046] In practical implementation, for target inspection scenarios where the target environment is on the ground, a ground-mobile intelligent inspection robot can be selected as the target inspection robot 200; for target inspection scenarios where the target environment is on an elevated track, a track-suspended intelligent inspection robot can be selected as the target inspection robot 200; and for target inspection scenarios where the target environment is a pipeline, a pipeline-rolling intelligent inspection robot can be selected as the target inspection robot 200. If the target inspection scenario is in multiple target environments, the target inspection scenario can be inspected by combining multiple types of target inspection robots 200 corresponding to the target environments.

[0047] In step S1002, the inspection location refers to multiple specific locations in the target inspection scenario. These inspection locations are the specific points in the production workshop. Each inspection location has a corresponding inspection number. When the target inspection robot 200 performs the inspection task, it will move to each target inspection location in ascending order of the inspection location number and perform the corresponding inspection actions to obtain the inspection content of the target inspection location.

[0048] Action parameters refer to the parameters required for the target inspection robot 200 to complete the inspection action after reaching the target inspection position. These action parameters include, but are not limited to, action type identifier, action execution height value, and action execution angle value. The action type identifier refers to the identifier of the action type performed by the target inspection robot 200. This action type identifier includes, but is not limited to: taking pictures, taking videos, collecting sound, and collecting amplitude.

[0049] In practice, the inspection control platform 100 determines a set of action parameters corresponding to each inspection location and associates the action parameters of each inspection location with the inspection location number of that inspection location. The inspection task containing the action parameters of all target inspection locations in the target inspection scenario is sent to the target inspection robot 200. The target inspection robot 200 will perform the corresponding action at the target inspection location according to the action parameters in the inspection task.

[0050] In step S1003, the inspection task refers to a task set for the target inspection scenario that can be identified and executed by the target inspection robot 200. The target inspection robot 200 can complete the inspection of multiple inspection locations in the target inspection scenario according to the inspection task. The inspection task specifies the specific inspection behavior of the target inspection robot 200 for each inspection step in the entire inspection process for the target inspection scenario. For example, the inspection task specifies the inspection location number of each inspection location in the target inspection scenario, the specific coordinates of each inspection location, the inspection action required to reach each inspection location, and the inspection items corresponding to that inspection location.

[0051] Here, the inspection task can also be set to the inspection time, inspection frequency and inspection cycle for the target inspection scenario. In this way, intelligent inspection can be achieved, and the inspection time and cycle of the target inspection robot 200 can be precisely controlled. For example, it can be set to inspect the target inspection scenario once per hour at 2 pm every Monday.

[0052] In the specific implementation of step S1004, the inspection control platform 100 can send the inspection task to the target inspection robot 200 in the form of instruction code or instruction file. After receiving the inspection task, the target inspection robot 200 parses and identifies the inspection task in order to complete the inspection action of each target inspection position indicated in the inspection task in sequence.

[0053] The target inspection robot 200 can perform the following processes:

[0054] Figure 3 This is a flowchart illustrating the processing steps performed in an inspection robot, as provided in an embodiment of this application. Figure 3 As shown, the method includes:

[0055] Step S2001: After receiving the inspection task, perform a patrol inspection of the target inspection scenario indicated by the inspection task.

[0056] Step S2002: If the target inspection robot reaches the target inspection position in the inspection task, control the target inspection robot to perform the corresponding action according to the action parameters at the target inspection position, so as to obtain the inspection content corresponding to the target inspection position.

[0057] Step S2003: Call the target inspection model that matches the inspection items corresponding to the target inspection location, and input the inspection content into the target inspection model to obtain the inspection results for the target inspection location.

[0058] Step S2004: Based on the inspection results, output the inspection results for the target inspection location.

[0059] In the specific implementation of step S2001, after receiving the inspection task, the target inspection robot 200 will move to the corresponding inspection position in ascending order of the inspection position number indicated by the inspection task, and then perform the corresponding inspection action.

[0060] In step S2002, the inspection content refers to the pictures, videos or signals captured by the target inspection robot 200. Different inspection content can be obtained by configuring different acquisition devices for the target inspection robot 200.

[0061] In practice, each time the target inspection robot 200 moves to the target inspection location, it performs the corresponding action according to the action parameters of the target inspection location in the inspection task, obtains the inspection content of the target inspection location, and then continues to move to the next target inspection location to obtain the inspection content of the next target inspection location, until the inspection of all target inspection locations in the target inspection scenario is completed.

[0062] In an optional embodiment, the action type identifier indicates the photo-taking action. Each inspection robot includes a camera. The target inspection robot 200 further performs the following processing: adjusting the camera's shooting height to the action execution height value, adjusting the camera's shooting angle to the action execution angle value, controlling the camera to perform the photo-taking action indicated by the action type identifier, obtaining an inspection image for the target inspection location, and determining the inspection image as the inspection content corresponding to the target inspection location.

[0063] Here, the motion execution height value and motion execution angle value are used to adjust the position of the camera so that the camera is aimed at the desired position for taking pictures and to obtain the inspection content of the target inspection position.

[0064] In step S2003, the target inspection model refers to an artificial intelligence model trained using deep learning algorithms. This model includes various types, each capable of analyzing different types of anomalies to obtain the most reasonable analysis results. For example, the types of target inspection models may include, but are not limited to, classification inspection models, target detection inspection models, and image segmentation inspection models. Deployed on the target inspection robot 200, this model can determine whether the target inspection location meets the inspection requirements based on the inspection content. For example, after being established, the target inspection model can inspect the process quality as specified in the "Cigarette Process Specifications" issued by the State Tobacco Monopoly Administration in 2016. The inspection content includes, but is not limited to, the process parameters, equipment parameters, instrument status, material status, and processing accuracy of each production equipment and process. It can also detect and identify abnormal images and provide identification results and classifications.

[0065] Here, the classification inspection model refers to an inspection model capable of identifying the category of a target in an image at a target inspection location. It can be trained on a training dataset using a classification neural network algorithm. Examples of classification neural network algorithms include, but are not limited to, DenseNet and ResNet algorithms. The object detection inspection model refers to an inspection model capable of identifying the objects and their specific locations in an image at a target inspection location. It can be trained on a training dataset using an object detection neural network algorithm. Examples of object detection neural network algorithms include, but are not limited to, FastR-CNN and YOLO algorithms. The image segmentation inspection model refers to an inspection model capable of identifying the edges of each target in an image at a target inspection location to distinguish different individuals. It can be trained on a training dataset using an image segmentation neural network algorithm. Examples of image segmentation neural network algorithms include, but are not limited to, MaskR-CNN.

[0066] It should be noted that the target inspection model needs to be constructed before deploying it on the target inspection robot 200. Here, developers control the target inspection robot 200 to complete set actions based on the image recognition requirements of the target inspection scene. The robot 200 moves to the designated target inspection location according to the pre-set inspection route, inspection time, and inspection frequency, and performs the set actions to take pictures of the target inspection scene. Then, the captured photos are labeled and enhanced to form a training dataset. Different deep learning algorithms are then used to train the training dataset to build different types of target inspection models. Here, the target inspection model will be optimized by adjusting algorithm parameters, including but not limited to: learning frequency, save interval, initial learning rate, number of steps, and number of learning decay rounds. Transfer learning can also be performed by importing pre-trained models to accelerate the convergence speed of the model. Finally, the goodness of fit and generalization of the target inspection model will be evaluated to meet the recognition accuracy requirements. If the generated target inspection model fails to meet the accuracy requirements after model evaluation, the target inspection model will be continuously optimized and trained until it meets the accuracy requirements.

[0067] In one optional embodiment, each inspection item includes an identification mark, which indicates the inspection type for the inspection location. Each inspection robot is equipped with multiple types of inspection models. The target inspection robot 200 further performs the following process: based on the identification mark of the inspection item corresponding to the target inspection location, the inspection model corresponding to the inspection type indicated by the identification mark is determined as the target inspection model.

[0068] Here, the inspection items refer to the specific equipment and inspection types inspected by the target inspection robot 200 after it arrives at the target inspection location. The inspection items include identification tags, which can be used to determine the specific inspection type. For example, the inspection types include, but are not limited to, classification, target detection, and instance segmentation.

[0069] In practical implementation, the inspection task clearly defines the inspection items corresponding to each target inspection location. After the target inspection robot 200 arrives at the target inspection location, it will determine the target inspection model corresponding to the identification mark in the inspection item. For example, if the identification mark indicates that the inspection type is classification, the target inspection robot 200 will select the classification inspection model as the target inspection model from multiple types of inspection models, and input the image corresponding to the target inspection location into the target inspection model to output the inspection result of the target inspection location.

[0070] In step S2004, the inspection result refers to information that indicates an abnormality at the target inspection location.

[0071] In an optional example, the inspection results include the recognition fit degree, and the target inspection robot 200 further performs the following processing: comparing the recognition fit degree of the target inspection location with a set threshold; if the recognition fit degree is greater than the set threshold, determining that there is an anomaly at the target inspection location; determining the anomaly type of the target inspection location based on the invoked target inspection model; and determining the target inspection location, inspection content, anomaly type, and inspection results as the inspection results for the target inspection location.

[0072] Here, the recognition fit is a numerical value used to characterize the similarity between the current image and the target image at the target inspection location, ranging from 0 to 1. The target image refers to the image when an anomaly occurs at the target inspection location. The threshold is a set value that can be determined by those skilled in the art based on actual circumstances; this application does not impose any limitation on it. Anomaly types correspond to recognition identifiers. For example, when the recognition method corresponding to the recognition identifier of the target inspection location is classification, if the recognition fit is greater than the set threshold, it indicates that an anomaly has occurred at the target inspection location, and the anomaly type is classification anomaly. The identifier of the target inspection scene is used to represent which specific inspection scene among multiple target inspection scenes. The identifier of the target inspection scene can be a number, symbol, or text, or a combination of numbers, symbols, or text.

[0073] In another optional example, the target inspection robot 200 may also perform the following processing: comparing the recognition fit of the target inspection location with a set threshold; if the recognition fit is less than or equal to the set threshold, it is determined that there is an anomaly at the target inspection location. In this case, the recognition fit is a value that characterizes the similarity between the current image of the target inspection location and the image when no anomaly occurs in the inspection items at the target inspection location, and the value ranges from 0 to 1.

[0074] It should be noted that whether to determine an anomaly at a target inspection location by identifying a goodness-of-fit greater than a set threshold, or by identifying a goodness-of-fit less than or equal to a set threshold, depends on the difficulty of collecting negative samples in the target inspection scenario. Here, negative samples refer to erroneous samples, i.e., images when the inspection item at the target inspection location is abnormal, and positive samples refer to correct samples, i.e., images when the inspection item at the target inspection location is normal. For example, if collecting negative samples is difficult, positive sample images are used as target images, and an alarm is issued when the similarity between the inspection image and the target image is less than or equal to a set threshold. If collecting negative samples is easy, negative sample images are used as target images, and an alarm is issued when the similarity between the inspection image and the target image is greater than a set threshold.

[0075] In practice, if the recognition fit of the target inspection location is greater than the set threshold, it indicates that the target inspection location is abnormal. At this time, the target inspection robot will determine the inspection result corresponding to the target inspection location to clarify the specific information of the target inspection location where the abnormality occurred.

[0076] In an optional embodiment, the production line inspection system further includes a database server 300 and a mobile terminal 400, wherein the target inspection robot 200 also performs the following processing: if it is determined that there is an abnormality at the target inspection location, the inspection result is sent to the mobile terminal 400 through the database server 300.

[0077] Here, mobile terminal 400 mainly refers to a mobile device used to display inspection results and to operate on the inspection results. Mobile terminal 400 may include any of the following devices: smartphone, tablet computer, or laptop computer. Inspection results refer to information that can describe the inspection status of the target inspection location, including the identification of the target inspection scene, the target inspection location, the inspection content, the anomaly type, and the inspection results.

[0078] In an optional embodiment, the database server 300 performs the following process: dividing the inspection results with the same scene identifier into data related to the same target inspection scene and grouping them together.

[0079] Here, the database server 300 is mainly used to classify and store the inspection results so that staff can query the inspection information. The scene identifier refers to the identifier of the target inspection scene. The identifier can be a number, symbol, or text, or a combination of the above identifiers.

[0080] In practice, the database server 300 will classify and store the received inspection results according to the scene identifier, so that staff can query the inspection results of the target inspection scene corresponding to the scene identifier.

[0081] It should be noted that data analysis software can be deployed on the database server 300 to analyze and summarize the inspection results sent by the target inspection robot 200. During daily inspections, this software can also automatically generate inspection reports and create relevant statistical charts for each inspection area, so that the inspection data analysis can be sent to the administrator. The data analysis software can query historical inspection reports according to a set time period.

[0082] As can be seen, compared with the prior art, this application can achieve the following beneficial effects:

[0083] 1) This system uses a target inspection robot to reach the set target inspection location, performs automatic inspections of relevant inspection scenarios at a specified frequency, and alarms for abnormalities, thereby reducing the number of operators in relevant positions and lowering the company's operating costs.

[0084] 2) The target inspection robot can compare, identify and alarm on inspection scene images based on deep learning models. The standardization is uniform, which reduces the probability of abnormal situations caused by factors such as insufficient personnel status and business capabilities.

[0085] 3) Leveraging the accuracy of the training model, the uniformity of the inspection robot's execution of inspection paths and frequencies helps improve the standardization of the inspection requirements outlined in the State Tobacco Monopoly Administration's "Cigarette Process Specifications." Traditional inspection methods rely primarily on the subjective will of operators to implement relevant regulations, and the discovery of problems during inspections is also related to individual professional skills, often leading to inadequate implementation of the requirements in the "Cigarette Process Specifications." The production line inspection method introduced in this application can intelligently inspect relevant processes according to pre-set rules, and retain and statistically analyze the inspection results. This minimizes the inconsistencies in the implementation of standards in the "Cigarette Process Specifications" due to human factors, thus improving the uniformity of implementation.

[0086] 4) Traditional computer vision recognition algorithms cannot train models to recognize most scene images in the "Cigarette Manufacturing Specifications." However, the deep learning algorithms used in this invention, including classification, semantic segmentation, object detection, and instance segmentation, can model these scene images and train high-precision image recognition models according to the specific needs of the image recognition scene. This enables the recognition and alarm of abnormal scene images, improving the uniformity and accuracy of inspection standards.

[0087] 5) Due to limitations in equipment size, location, and actual space at each cigarette factory, some inspection requirements in the "Cigarette Process Specifications" are difficult for operators to implement in practice. Furthermore, some inspection points cannot be fully implemented due to height, safety, or other reasons. This invention employs a track-suspended intelligent inspection robot, a ground-mobile intelligent inspection robot, and a pipeline-rolling intelligent inspection robot, capable of image acquisition and intelligent inspection in the air, on the ground, and inside pipelines, respectively, thus solving this problem to the greatest extent possible and enabling inspection work in difficult locations.

[0088] 6) Traditional manual inspections rely heavily on operators to manually record inspection data and issue alarms, resulting in low reliability and requiring manual analysis. This application categorizes and statistically analyzes all inspection items and results, storing the results in a relational database. It can also generate Excel spreadsheets based on query requirements and push them to the administrator's mobile app. Furthermore, it can push graphical daily, weekly, and monthly reports to mobile clients on a daily, weekly, and monthly basis.

[0089] Based on the same inventive concept, this application also provides a production line inspection method corresponding to the production line inspection system. Since the principle of the method in this application is similar to that of the production line inspection system described above in this application, the implementation of the method can refer to the implementation of the system, and the repeated parts will not be described again.

[0090] Figure 4 A flowchart illustrating the production line inspection method provided in this application embodiment is shown below. Figure 4 As shown, the method includes:

[0091] Step S501: Determine the target inspection scenario for the production line and the target inspection robot suitable for the target inspection scenario.

[0092] Step S502: Determine multiple inspection locations in the target inspection scenario and the motion parameters at each inspection location;

[0093] Step S503: Generate an inspection task for the target inspection scenario;

[0094] Step S504: Send the inspection task to the target inspection robot so that the target inspection robot can complete the intelligent inspection of the production line according to the inspection task.

[0095] Optionally, the target inspection robot suitable for the target inspection scenario is determined by the following method: based on the target environment in which the target inspection scenario is located, an inspection robot that supports inspection in the target environment is selected as the target inspection robot. The target environment includes at least one of the following: ground, aerial track and pipeline. The target inspection robot includes at least one of the following: track-suspended inspection robot, ground mobile inspection robot and pipeline rolling inspection robot.

[0096] Corresponding to Figure 4 In the production line inspection method, this application also provides a structural schematic diagram of an electronic device 600, as shown in the embodiment. Figure 5 As shown, the electronic device 600 includes a processor 610, a memory 620, and a bus 630. The memory 620 stores machine-readable instructions executable by the processor 610. When the electronic device 600 is running, the processor 610 and the memory 620 communicate via the bus 630. When the machine-readable instructions are executed by the processor 610, the aforementioned production line inspection method can be performed. By configuring corresponding inspection robots for the characteristics of target inspection scenarios on the production line and matching corresponding target inspection models for each target inspection location's inspection items, the problem of low inspection efficiency and poor inspection accuracy in the cigarette manufacturing production line is solved.

[0097] Corresponding to Figure 4 In addition to the production line inspection method described above, this application also provides a computer-readable storage medium storing a computer program, which is executed by a processor to perform the steps of the production line inspection method described above.

[0098] Specifically, the storage medium can be a general-purpose storage medium, such as a portable disk or hard drive. When the computer program on the storage medium is run, it can execute the above-mentioned production line inspection method. By configuring corresponding inspection robots according to the characteristics of the target inspection scenarios on the production line, and matching the corresponding target inspection model for each target inspection location, the problem of low inspection efficiency and poor inspection accuracy in the inspection process of the cigarette manufacturing production line is solved.

[0099] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0100] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0101] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0102] In addition, the functional units in the embodiments provided in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0103] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0104] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0105] Finally, it should be noted that the above-described embodiments are merely specific implementations of this application, used to illustrate the technical solutions of this application, and not to limit them. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the technical scope disclosed in this application. Such modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A production line inspection system, characterized in that, It includes an inspection control platform and at least one inspection robot, with multiple types of inspection models deployed in each inspection robot; The inspection control platform performs the following processing: Determine the target inspection scenarios for the production line and the target inspection robot adapted to the target inspection scenarios; Determine multiple inspection locations in the target inspection scenario and the motion parameters at each inspection location; Generate an inspection task for the target inspection scenario. The inspection task includes multiple inspection locations, action parameters at each inspection location, and inspection items corresponding to each inspection location. Each inspection item includes an identification identifier, which is used to indicate the inspection type for the inspection location. Send the inspection task to the target inspection robot; The target inspection robot performs the following processing: After receiving the inspection task, a patrol inspection is carried out on the target inspection scenario indicated by the inspection task; If the target inspection robot reaches the target inspection position in the inspection task, the target inspection robot is controlled to perform the corresponding action according to the action parameters at the target inspection position, so as to obtain the inspection content corresponding to the target inspection position. Based on the identification identifier of the inspection item corresponding to the target inspection location, a target inspection model that matches the inspection item corresponding to the target inspection location is selected and called. The inspection content is input into the target inspection model to obtain the inspection result for the target inspection location. Based on the inspection results, output the inspection results for the target inspection locations; The target environment includes at least one of the following: ground, aerial track, and pipeline; the target inspection robot includes at least one of the following: track-suspended inspection robot, ground-mobile inspection robot, and pipeline rolling inspection robot; the inspection control platform performs the following processing: Based on the target environment in which the target inspection scenario is located, the inspection robot that supports inspection in the target environment is selected as the target inspection robot.

2. The production line inspection system as described in claim 1, characterized in that, The motion parameters include motion type identifier, motion execution height value, and motion execution angle value. The motion type identifier indicates the photo-taking motion, and each inspection robot includes a camera. The target inspection robot also performs the following processing: Adjust the camera's shooting height to the action execution height value, adjust the camera's shooting angle to the action execution angle value, control the camera to perform the photo-taking action indicated by the action type identifier, obtain the inspection image for the target inspection location, and determine the inspection image as the inspection content corresponding to the target inspection location.

3. The production line inspection system as described in claim 2, characterized in that, The target inspection robot also performs the following processing: The inspection model corresponding to the inspection type indicated by the identification mark is determined as the target inspection model.

4. The production line inspection system as described in claim 1, characterized in that, The inspection results include the degree of fit of the identification; The target inspection robot also performs the following processing: The recognition fit of the target point detection location is compared with a set threshold; If the recognition fit is greater than a set threshold, it is determined that the target point detection position is abnormal; Based on the invoked target inspection model, determine the anomaly type of the target inspection location; The target inspection location, inspection content, anomaly type, and inspection results are determined as the inspection results for the target inspection location.

5. The production line inspection system as described in claim 4, characterized in that, The production line inspection system also includes a mobile terminal and a database server; The target inspection robot also performs the following processing: If an anomaly is determined at the target inspection location, the inspection result is sent to the mobile terminal via the database server.

6. The production line inspection system as described in claim 5, characterized in that, The database server performs the following processing: Inspection results with the same scene identifier are classified as data related to the same target inspection scene and grouped together.

7. A production line inspection method, characterized in that, include: Determine the target inspection scenarios for the production line and the target inspection robot adapted to the target inspection scenarios; Determine multiple inspection locations in the target inspection scenario and the motion parameters at each inspection location; Generate an inspection task for the target inspection scenario. The inspection task includes multiple inspection locations, action parameters at each inspection location, and inspection items corresponding to each inspection location. Each inspection item includes an identification identifier, which is used to indicate the inspection type for the inspection location. The inspection task is sent to the target inspection robot so that the target inspection robot can perform an inspection of the target inspection scenario indicated by the inspection task. The target inspection robot suitable for the target inspection scenario is determined using the following method: Based on the target environment of the target inspection scenario, the inspection robot that supports inspection in the target environment is selected as the target inspection robot. The target environment includes at least one of the following: ground, aerial track and pipeline. The target inspection robot includes at least one of the following: track-suspended inspection robot, ground mobile inspection robot and pipeline rolling inspection robot. The inspection of the target inspection scenario indicated by the inspection task is completed in the following manner: Based on the identification identifier of the inspection item corresponding to the target inspection location, select and call the target inspection model that matches the inspection item corresponding to the target inspection location, and input the inspection content into the target inspection model to obtain the inspection result for the target inspection location; Based on the inspection results, output the inspection results for the target inspection location.

8. An electronic device, characterized in that, It includes a processor, a memory, and a bus. The memory stores machine-readable instructions that the processor can execute. When the electronic device is running, the processor communicates with the memory via the bus, and the machine-readable instructions are executed by the processor to perform the steps of the production line inspection method as described in claim 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, performs the steps of the production line inspection method as described in claim 7.

Images