Intelligent manufacturing method and system based on centralized service platform industrial internet of things

Abstract

The application discloses an intelligent manufacturing method based on a centralized service platform industrial internet of things, which comprises the following steps: inputting parameter configuration information by a user platform, decomposing a process into multiple groups of configuration data groups by a service platform, storing and processing received configuration data groups by a management platform, and storing the configuration data groups as reference data groups; an object platform sends its sensing information as a verification data group at a set time interval; the management platform receives and processes the verification data group, and sends the reference data group and the verification data group to the service platform; the service platform receives the reference data group and the verification data group, and compares them; if the comparison exceeds a set threshold range, a stop running instruction is generated and sent to the object platform to control the production line equipment to stop running. The application further discloses a system for implementing the intelligent manufacturing method based on the centralized service platform industrial internet of things. The application can be applied to various intelligent manufacturing production lines, and can quickly control the production line equipment to stop running when an abnormality occurs in production.

Description

TECHNICAL FIELD

[0001] The present specification relates to the technical field of intelligent manufacturing, and particularly relates to an intelligent manufacturing method and system based on a centralized service platform industrial internet of things. BACKGROUND

[0002] In recent years, the industrial internet of things technology combining information technology and operation technology has developed rapidly, and its application in the intelligent manufacturing industry is also highly expected. At present, when the industrial internet of things technology is applied in the intelligent manufacturing industry, different types of data acquisition devices, control systems and the like are generally connected and integrated to increase diversity and scale. When applied, the real-time state information and production data of the field equipment are collected through the data acquisition devices, and the real-time data such as the required equipment and production progress are provided to the user, so as to solve the problems of automatic data collection, processing, statistics, analysis and the like, and further improve the production efficiency, optimize the resource allocation and management efficiency. However, the existing industrial internet of things technology suitable for intelligent manufacturing is generally developed for a single production project, and is compatible with a certain type of control system or a specific control system used by a user, lacks universality, and cannot take timely and effective measures when production abnormalities occur, which seriously affects its popularization and application. For example, there are multiple processes for processing a blank, but the existing industrial internet of things technology for intelligent processing of the blank only controls a certain process, and cannot be applied to other processes.

[0003] Therefore, it is hoped that an intelligent manufacturing method based on a centralized service platform industrial internet of things can be provided to control multiple processes of a blank. This makes it possible to better monitor the progress of the blank, and to take timely measures when process abnormalities occur. SUMMARY

[0004] One or more embodiments of the specification provide an intelligent manufacturing method based on a centralized service platform industrial Internet of Things, comprising a user platform, a service platform, a management platform, a sensing network platform and an object platform that interact in turn; wherein: the user platform is configured as a terminal device that interacts with a user, receives user input information to generate an instruction and sends it to the service platform, and shows the information sent by the service platform to the user; the service platform is configured as a first server that receives the instruction sent by the user platform and sends it to the management platform after processing, and obtains the information required by the user from the management platform and sends it to the user platform; the management platform is configured as a second server that receives the instruction sent by the service platform and controls the operation of the object platform according to the instruction, receives and stores the sensing information sent by the object platform; the sensing network platform is configured as a communication network and a gateway that interact between the management platform and the object platform; the object platform is configured as a production line device that executes manufacturing and a production line sensor that executes data collection, receives the instruction of the management platform to run, and sends sensing information to the management platform through the sensing network platform; the service platform adopts a centralized arrangement, which means that the platform uniformly receives data, uniformly processes data and uniformly sends data; the management platform and the sensing network platform both adopt a post-division platform arrangement, which means that the management platform and the sensing network platform are both provided with a total platform and multiple sub-platforms, control information and object platform parameter configuration information are transmitted from the sub-platforms to the total platform, and sensing information is transmitted from the total platform to the sub-platforms; the method comprises: when configuring the object platform parameters, the user inputs object platform parameter configuration information through the user platform, the service platform receives the object platform parameter configuration information sent by the user platform and disassembles it into multiple groups of configuration data according to the process and sends it to the management platform; the management platform stores and processes the received configuration data groups and sends them to the sensing network platform, and the stored configuration data groups are used as reference data groups; the sensing network platform stores and processes the received configuration data groups and sends them to the object platform, and the object platform completes the object platform parameter configuration according to the object platform parameter configuration information; when the production line device of the object platform runs, the object platform sends its sensing information as a verification data group to the management platform through the sensing network platform at a set time interval; the management platform receives the verification data group sent by the object platform and processes it, and sends the reference data group corresponding to the object platform that sent the verification data group and the verification data group to the service platform;The service platform receives the reference data set and the check data set for comparison. If all data in the check data set is compared with the reference data set within a set threshold range, the data is cleared and not processed subsequently. If there is data in the check data set that is compared with the reference data set beyond the set threshold range, a stop running instruction is generated and sent to the management platform and fed back to the user platform. The management platform sends to the corresponding object platform through the sensor network platform to control the production line equipment to stop running.

[0005] The one or more embodiments of the specification provide a system for intelligent manufacturing method based on centralized service platform industrial Internet of Things, comprising user platform, service platform, management platform, sensing network platform and object platform which interact in turn; wherein: the user platform is configured as a terminal device interacting with users, receives user input information to generate instructions and sends them to the service platform, and shows the information sent by the service platform to the user; the service platform is configured as a first server, receives the instructions sent by the user platform and sends them to the management platform after processing, and obtains the information required by the user from the management platform and sends it to the user platform; the management platform is configured as a second server, receives the instructions sent by the service platform and controls the operation of the object platform according to the instructions, receives and stores the sensing information sent by the object platform; the sensing network platform is configured as a communication network and gateway interacting between the management platform and the object platform; the object platform is configured as production line equipment for execution of manufacturing and production line sensors for data collection, receives the instructions of the management platform and sends the sensing information to the management platform through the sensing network platform; the service platform adopts centralized arrangement, which means that the platform uniformly receives data, uniformly processes data and uniformly sends data; the management platform and the sensing network platform both adopt the after-division platform arrangement, which means that the management platform and the sensing network platform are both provided with a total platform and multiple sub-platforms, and the transmission of control information and object platform parameter configuration information is from the sub-platform to the total platform, and the transmission of sensing information is from the total platform to the sub-platform; when configuring the object platform parameters, the user inputs the object platform parameter configuration information through the user platform, the service platform receives the object platform parameter configuration information sent by the user platform and disassembles it into multiple groups of configuration data according to the process and sends them to the management platform; the management platform stores and processes the received configuration data groups and sends them to the sensing network platform, and the stored configuration data groups are used as reference data groups; the sensing network platform stores and processes the received configuration data groups and sends them to the object platform, and the object platform completes the configuration of object platform parameters according to the object platform parameter configuration information; when the production line equipment of the object platform runs, the object platform sends its sensing information as verification data groups to the management platform through the sensing network platform at a set time interval; the management platform receives the verification data groups sent by the object platform and processes them, and sends the reference data groups corresponding to the object platform sending the verification data groups and the verification data groups to the service platform;The service platform receives reference data set and check data set for comparison, if all data in the check data set is compared with the reference data set within the set threshold range, the data is cleared and not processed subsequently, if there is data in the check data set compared with the reference data set exceeding the set threshold range, a stop running instruction is generated and sent to the management platform and fed back to the user platform, the management platform is sent to the corresponding object platform through the sensing network platform to control the production line equipment to stop running. BRIEF DESCRIPTION OF DRAWINGS

[0006] The present specification will be further illustrated in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not restrictive, in these embodiments, the same numbers represent the same structures, wherein:

[0007] Figure 1 is an exemplary module diagram of the intelligent manufacturing system based on the centralized service platform industrial Internet of Things according to some embodiments of the present specification;

[0008] Figure 2 is a flowchart of configuring the object platform parameters in one specific embodiment of the present application;

[0009] Figure 3 is a flowchart of checking the sensing information in one specific embodiment of the present application;

[0010] Figure 4 is an exemplary flowchart of the intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to some embodiments of the present specification;

[0011] Figure 5 is an exemplary schematic diagram of applying the intelligent manufacturing method based on the centralized service platform industrial Internet of Things to rough casting according to some embodiments of the present specification;

[0012] Figure 6 is an exemplary schematic diagram of applying the intelligent manufacturing method based on the centralized service platform industrial Internet of Things to rough forging according to some embodiments of the present specification;

[0013] Figure 7 is an exemplary schematic diagram of applying the intelligent manufacturing method based on the centralized service platform industrial Internet of Things to rough stamping according to some embodiments of the present specification;

[0014] Figure 8 is an exemplary flowchart of the method of intelligently manufacturing the rough based on the performance of the rough according to some embodiments of the present specification. DETAILED DESCRIPTION

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the drawings needed to be used in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some examples or embodiments of the present specification, and for those skilled in the art, the present specification can also be applied to other similar scenarios without creative labor. Unless it is obvious from the language environment or otherwise stated, the same reference numbers in the drawings represent the same structure or operation.

[0016] Flowcharts are used in the present specification to illustrate the operations performed by the system according to the embodiments of the present specification. It should be understood that the preceding or subsequent operations are not necessarily performed in sequence. On the contrary, each step can be processed in reverse order or simultaneously. Meanwhile, other operations can also be added to these processes, or one or more steps of operation can be removed from these processes.

[0017] Figure 1 is an exemplary module diagram of the intelligent manufacturing system based on the centralized service platform industrial Internet of Things according to some embodiments of the present specification. As shown in Figure 1 , the system 100 includes a user platform, a service platform, a management platform, a sensing network platform and an object platform that interact in sequence, wherein the user platform, the service platform, the management platform, the sensing network platform and the object platform are connected by means of sequential communication to realize interaction.

[0018] The user platform is configured as a terminal device for interacting with the user, receives user input information to generate instructions and sends them to the service platform, and displays the information sent by the service platform to the user.

[0019] The service platform is configured as a first server, receives the instructions sent by the user platform and sends them to the management platform after processing, and obtains the information required by the user from the management platform and sends it to the user platform. The service platform is also used to judge whether the process is abnormal based on the processed verification data set, and the processed verification data set at least includes the standard degree score of the process executed by the production line equipment; if the process is abnormal, a stop instruction is sent to the management platform, and the management platform controls the production line equipment to stop processing the blank.

[0020] The management platform is configured as a second server, receives the instructions sent by the service platform and controls the object platform to run according to the instructions, receives and stores the sensing information sent by the object platform. The management platform is also used to identify the process and obtain the verification data set corresponding to the process. The verification data set is the data obtained for verifying whether the process is executed normally; and the verification data set is processed by an abnormality judgment model to obtain a processed verification data set.

[0021] The sensing network platform is configured as a communication network and gateway for interaction between the management platform and the object platform.

[0022] The object platform is configured to execute the production line equipment of manufacturing and the production line sensor of data collection, receives the instruction operation of the management platform, and sends the sensing information to the management platform through the sensor network platform. The production line equipment can include but is not limited to casting equipment, pouring equipment, pressure equipment, etc. The production line sensor can include but is not limited to temperature sensor, pressure sensor, hardness sensor, etc.

[0023] In the specific implementation of the embodiment, the user platform adopts a smart electronic device such as a desktop computer, a tablet computer, a notebook computer, a mobile phone, etc. for data processing and data communication, which is not limited here. The service platform of the embodiment adopts a centralized arrangement, which means that the platform uniformly receives data, uniformly processes data and uniformly sends data. The management platform and the sensor network platform of the embodiment both adopt a back split platform arrangement, which means that the management platform and the sensor network platform are both provided with a total platform and multiple split platforms, the transmission of control information and object platform parameter configuration information is from the split platform to the total platform, and the transmission of sensing information is from the total platform to the split platform. In the embodiment, the total platform of the management platform is configured as a second master server, and its split platform is configured as a second sub-server. The total platform of the management platform receives and processes data based on the second master server, and the split platform of the management platform receives and processes data based on the second sub-server. The total platform of the sensor network platform is configured as a gateway master server, and its split platform is configured as a gateway sub-server. The total platform of the sensor network platform receives and processes data based on the gateway master server, and the split platform of the sensor network platform receives and processes data based on the gateway sub-server. The data processing process mentioned in the embodiment can be processed by the processor of the terminal device and the server, and the server is equipped with a corresponding database for storing data, which can be stored on the storage device of the server, such as a hard disk or the like. The object platform parameter configuration information includes production line equipment operation data, production line equipment operation data, manufacturing process data, and product data of corresponding blanks, semi-finished products, finished products, etc. at each stage. The sensing information obtained by the object platform includes production line equipment operation data, production line equipment operation data, manufacturing process data and data collected by the sensor.

[0024] In the specific implementation of the embodiment, based on the difference of the production line equipment and the production line sensor applied to different processes, the object platform is divided into multiple, each of which is provided with corresponding production line equipment and production line sensor.

[0025] The intelligent manufacturing method based on the centralized service platform industrial internet of things comprises the steps that the control information and object platform parameter configuration information are transmitted by the user platform, the service platform, the management platform, the sensing network platform and the object platform in sequence, and the sensing information is transmitted by the object platform, the sensing network platform, the management platform, the service platform and the user platform in sequence. When the instruction is transmitted, the receiving and processing of the instruction by each server is specifically the data packet format of the set next level receiving object for facilitating identification.

[0026] It should be understood that Figure 1 The system and its modules shown can be implemented in various ways.

[0027] It should be noted that the above description of the intelligent manufacturing system based on the centralized service platform industrial internet of things and its modules is for convenience of description, and cannot limit the present specification to the scope of the embodiments. It can be understood that, for those skilled in the art, after understanding the principle of the system, the modules can be combined arbitrarily or connected with other modules to form a subsystem without departing from the principle. In some embodiments, Figure 1 The user platform, the service platform, the management platform, the sensing network platform and the object platform disclosed in the present specification can be different modules in a system, or one module can realize the functions of two or more modules. For example, the modules can share one storage module, or each module can have its own storage module. Such variations are within the protection scope of the present specification.

[0028] Figure 2 The flow chart for configuring the object platform parameters in one embodiment of the present application.

[0029] As Figure 2As shown, the flow 200 of the intelligent manufacturing method further comprises: when the object platform parameter is configured, the user inputs object platform parameter configuration information through the user platform, the service platform receives the object platform parameter configuration information sent by the user platform and disassembles the object platform parameter configuration information into multiple groups of configuration data groups according to a process, and sends the configuration data groups to the management platform; the management platform stores and processes the received configuration data groups and sends the configuration data groups to the sensing network platform, and the stored configuration data groups serve as reference data groups; the sensing network platform stores and processes the received configuration data groups and sends the configuration data groups to the object platform, and the object platform completes object platform parameter configuration according to the object platform parameter configuration information. In the embodiment, when the user platform sends parameter configuration information to the object platform, each sub-platform of the management platform stores and processes a group of configuration data groups, the total platform of the management platform stores and processes the configuration data groups processed by all sub-platforms of the management platform after the configuration data groups are aggregated, and sends multiple groups of configuration data groups to the sub-platforms of the sensing network platform one by one; each sub-platform of the sensing network platform stores and processes the received configuration data groups, and the total platform of the sensing network platform stores and processes the configuration data groups processed by all sub-platforms of the sensing network platform after the configuration data groups are aggregated, and sends multiple groups of configuration data groups to the object platform one by one.

[0030] Figure 3 A flowchart for verifying sensing information in an embodiment of the application.

[0031] As shown in the figure, Figure 3 As shown, the flow 300 of the intelligent manufacturing method further comprises: when the production line equipment of the object platform is running, the object platform sends sensing information thereof to the management platform through the sensing network platform as a verification data group at a set time interval; the management platform receives the verification data group sent by the object platform and processes the verification data group, and sends a reference data group corresponding to the object platform sending the verification data group to the service platform; the service platform receives the reference data group and the verification data group for comparison, if all data in the verification data group is within a set threshold range compared with the reference data group, the verification data is cleared and no subsequent processing is performed, if there is data in the verification data group that exceeds the set threshold range compared with the reference data group, a stop running instruction is generated and sent to the management platform and fed back to the user platform, and the management platform sends a stop running instruction to the corresponding object platform through the sensing network platform to control the production line equipment to stop running. The set time interval at which the object platform obtains sensing information and the set threshold range at which parameters are compared are preset according to user requirements.

[0032] In the embodiment, during the running of the production line equipment, the running parameters and the products corresponding to each process are verified in real time, the influence of process deviation on product manufacturing can be quickly avoided, and resources can be saved.

[0033] In some embodiments, when the service platform compares the reference data set with the verification data set and finds that there is data in the verification data set that exceeds the set threshold range compared with the reference data set, the service platform further generates a sensing information acquisition instruction and sends it to the management platform before generating the stop operation instruction. The management platform receives the instruction sent by the service platform and controls the corresponding object platform to send the current sensing information according to the instruction. The object platform feeds back the current sensing information as a verification data set to the management platform through the sensing network platform. The management platform receives the verification data set fed back by the object platform, processes it and then sends it to the service platform. The service platform receives the verification data set and compares it with the reference data set. If all the data in the verification data set is within the set threshold range compared with the reference data set, the data in the verification data set is cleared and no further processing is performed. If there is still data in the verification data set that exceeds the set threshold range compared with the reference data set, the service platform generates the stop operation instruction. The service platform generates the sensing information acquisition instruction immediately when it confirms that there is data in the verification data set that exceeds the set threshold range compared with the reference data set. The time taken for the service platform to send the instruction to the object platform is less than the time taken for the object platform to send the sensing information again after the comparison exceeds the set threshold range.

[0034] In this embodiment, the verification is performed again after the verification, which can avoid the misoperation caused by the data transmission deviation.

[0035] In some embodiments, after the service platform generates the stop running instruction, the service platform also automatically generates a sensing information acquisition instruction of the current process cascade process and sends it to the management platform. The management platform receives the instruction sent by the service platform and controls the cascade process object platform of the current corresponding object platform to send the current sensing information according to the instruction. The cascade process is the previous process or the next process of the current process. The cascade process object platform feeds back the current sensing information to the management platform as a verification data set through the sensing network platform. The management platform receives the verification data set fed back by the cascade process object platform and processes it. Then, the management platform sends the reference data set corresponding to the cascade process object platform that sends the verification data set to the service platform. The service platform receives the verification data set and the reference data set and compares them. If all the data in the verification data set is within the set threshold range compared with the reference data set, the data in the verification data set is cleared and not processed subsequently. If there is data in the verification data set that exceeds the set threshold range compared with the reference data set, a stop running instruction is generated and sent to the management platform and fed back to the user platform. The management platform sends it to the corresponding cascade process object platform through the sensing network platform to control the cascade process production line equipment to stop running. In the intelligent manufacturing process, it may involve blank manufacturing, part manufacturing, whole machine assembly, product testing, etc. In specific processing, it may involve casting, forging, punching, etc. For the blank, it needs to go through a series of consecutive processes to get the finished product. In this embodiment, when a problem is found in a certain process, the cascade process is checked immediately. In this way, the quality of the products manufactured by the cascade process can be ensured.

[0036] In some embodiments, when the cascade procedure is the next procedure of the current procedure, the service platform further generates a cascade procedure perception information acquisition instruction and sends it to the management platform before generating the stop running instruction, the management platform receives the instruction sent by the service platform and controls the corresponding cascade procedure object platform to send the current perception information according to the instruction; the cascade procedure object platform feeds back the current perception information to the management platform as a verification data set through the sensing network platform, the management platform receives the verification data set fed back by the cascade procedure object platform and sends it to the service platform after processing, the service platform receives the verification data set and compares it with the reference data set, if all the data in the verification data set is within the set threshold range compared with the reference data set, the data in the verification data set is cleared and no subsequent processing is performed, if there is still data in the verification data set that exceeds the set threshold range compared with the reference data set, the service platform generates a stop running instruction to control the cascade procedure production line equipment to stop running. For products processed in sequence through procedures, the product processed in the next procedure is further processed on the product manufactured in the previous procedure, when the previous procedure of the current procedure fails, the processed product may also have problems, this embodiment performs preliminary verification when the next cascade procedure has no problem, and then performs further verification, thereby ensuring the quality of the processed product.

[0037] In some embodiments, after the service platform disassembles the parameter configuration information into multiple groups of configuration data sets according to the procedures, the service platform further establishes a product model of the corresponding procedure according to the data in each group of configuration data sets and sends it to the user platform, the user sends a confirmation instruction to the service platform after confirming that the product model is correct through the user platform, and the service platform receives the confirmation instruction and processes it before sending the confirmed configuration data set to the management platform. In this embodiment, the service platform establishes a product model of the corresponding procedure according to the data in each group of configuration data sets, which can be specifically implemented based on existing modeling software such as Solidworks, UG, 3DS Max, etc. In this embodiment, the product model is established for the user to confirm, so that the user can correct the input product parameters in time when there is a deviation, thereby avoiding the influence of incorrect reference data on the correct operation of the system during subsequent verification and validation.

[0038] In some embodiments, the service platform pre-stores product general local design pre-stored information, and before the service platform disassembles the object platform parameter configuration information into multiple groups of configuration data groups according to the process, the service platform further includes screening the general local design information in the object platform parameter configuration information and comparing it with the product general local design pre-stored information. If all comparison results are within the set design deviation threshold range, the disassembly processing is performed. If the comparison result exceeds the set design deviation threshold range, it is sent to the user platform. If the user confirms that the input information is correct through the user platform, a confirmation instruction is sent to the service platform, and the service platform performs disassembly processing again. If the user modifies the input information through the user platform, the service platform compares again until the user platform confirms that the service platform performs disassembly processing again. In this way, when the present embodiment is applied, product general design data comparison can be provided to provide search and inspection functions for the user during early data entry, and the intelligent performance of the system can be improved.

[0039] In some embodiments, when the object platform sends its sensing information as verification data groups to the management platform through the sensing network platform at a set time interval, different object platforms send sensing information in an interleaved manner. In this way, the present embodiment can reduce the work load of the first server, the second server, and the gateway server at the same time, and effectively improve the work efficiency.

[0040] In some embodiments, multiple object platforms corresponding to each process are provided, and when the production line equipment of one object platform stops running, the service platform checks the production condition information of the object platforms of the same process as the object platform, and adjusts the production quantity of the object platforms of the same process as the object platform according to the production demand. In the intelligent manufacturing process, production planning is often made according to the required product quantity of the customer. The present embodiment can ensure that the production quantity meets the expected requirement through coordination processing. When the faulty production line equipment is confirmed to be able to continue to be put into production after debugging, the present embodiment can adjust the processing quantity of the object platforms corresponding to each process again to ensure efficient production.

[0041] Figure 4 is an exemplary flowchart of an intelligent manufacturing method based on a centralized service platform industrial Internet of Things according to some embodiments of the present specification.

[0042] In some embodiments, the intelligent manufacturing method can be a method for processing a blank, wherein the method for processing a blank includes multiple processes, and the processes for processing a blank can at least include casting, forging, and / or stamping, etc.

[0043] In some embodiments, the system 100 performs the process 400 when the object platform's production line equipment runs a process of processing a blank. The production line equipment can include at least a forging equipment, a pressing equipment, and / or a punch press, etc. For more about the object platform and the production line equipment, see Figure 1 and the related descriptions. As shown in Figure 4 the process 400 includes the following steps.

[0044] At step 410, the management platform identifies the process and obtains a verification data set corresponding to the process.

[0045] In some embodiments, the object platform can be disassembled into multiple types according to the process, and the management platform can identify the process according to the type of the object platform. For example, the type of the object platform A is a casting object platform, and the management platform can identify the process obtained from the object platform A as a process of casting a blank, and the perception information received from the object platform A as a verification data set of the process of casting a blank. The casting object platform refers to the object platform that casts a blank. For more about disassembling the object platform, see Figure 2 and the related descriptions. For more about the management platform, see Figure 1 and the related descriptions.

[0046] The verification data set can be various data obtained for verifying whether the process is normally executed. In some embodiments, the verification data set can be perception information. For example, the process can be casting a blank, and the steps of casting a blank can include at least melting a metal. The verification data set can include the operation data and the operation data of the melting equipment, the process data of the metal melting, and the temperature information of the metal obtained by the temperature sensor, etc. For more about the perception information, see Figure 1 and the related descriptions.

[0047] At step 420, the management platform processes the verification data set by an anomaly judgment model to obtain a processed verification data set.

[0048] In some embodiments, the anomaly judgment model is set in the management platform, and the anomaly judgment model can be used to process the verification data set to obtain a processed verification data set. In some embodiments, the anomaly judgment model can be various machine learning models, including but not limited to random forest rules, logistic regression, support vector machines, etc.

[0049] In some embodiments, the anomaly judgment model can be multiple, and different anomaly judgment models can be set in different sub-platforms of different management platforms to process different types of verification data sets.

[0050] In some embodiments, the management platform can input the reference data set and the verification data set into the anomaly judgment model, and the model outputs a standard degree score of the verification data set. The reference data set can be various data representing normal execution of the process. For example, the reference data set can include, but is not limited to, operation data and operation data of the production line equipment running normally, standard process data of the processed blank, data of the production line equipment running normally obtained by the sensor, etc. The types of reference data can include, but are not limited to, numerical range, image of standard process, etc. For more information about the reference data set, see Figure 2 and the related description. For more information about the anomaly judgment model, see Figure 5 , Figure 6 and Figure 7 and the related description.

[0051] Step 430, the service platform determines whether the process is abnormal based on the processed verification data set.

[0052] The processed verification data set at least includes a standard degree score of the production line equipment executing the process. The standard degree score can be used to represent the standard degree of the production line equipment executing the process. For example, the higher the score, the higher the standard degree, and the lower the score, the lower the standard degree. When the score is lower than the set threshold range, it can be determined that the process is abnormal. The set threshold range can be the range of the standard degree score of the production line equipment working normally. In some embodiments, the set threshold range can be set by the user through the user platform and sent to the service platform. In some embodiments, the set threshold range corresponding to different processes can be different, which can be set according to experience.

[0053] In some embodiments, the service platform can compare the obtained standard degree score with the set threshold range. If the standard degree score is not within the set threshold range, it means that the corresponding process is abnormal.

[0054] Step 440, if it is determined that the process is abnormal, the service platform sends a stop instruction to the management platform, and the management platform controls the production line equipment to stop processing the blank.

[0055] In some embodiments, when it is determined that the process is abnormal, the service platform generates an instruction to stop processing the blank to the management platform and feeds back to the user platform. The management platform receives the stop instruction sent by the service platform and controls the corresponding object platform to stop running.

[0056] Figure 5 is an exemplary schematic diagram of applying the intelligent manufacturing method based on the centralized service platform of industrial Internet of Things to blank casting according to some embodiments of the present specification.

[0057] In some embodiments, the abnormality judgment model can be a first model when the production line equipment casts the blank, and the verification data set can include temperature information. As shown in FIG. 6, the schematic flow 500 includes the following steps. Figure 5

[0058] At step 510, the management platform inputs the temperature information into the first model, and the first model outputs a first score.

[0059] The casting step can at least include metal smelting and pouring the smelted metal into a mold to obtain a casting. The mold can refer to a mold used to pour molten metal to form a casting during casting. The temperature information can be temperature-related information generated during the casting process of the blank. In some embodiments, the temperature information can be divided into smelting temperature information and pouring temperature information according to the casting step.

[0060] The smelting temperature information can refer to temperature information during the smelting of the metal. For example, the smelting temperature information can include temperature information of the smelted metal obtained at a first set time interval during the smelting process.

[0061] The first set time interval can refer to a pre-set interval of time for obtaining smelting temperature information. The first set time interval can be set according to experience.

[0062] The pouring temperature information can refer to temperature information during the process of pouring the smelted metal into a mold to obtain a casting. For example, the pouring temperature information includes temperature information of the casting obtained at a second set time interval after pouring.

[0063] The second set time interval can refer to a pre-set interval of time for obtaining pouring temperature information. The second set time interval can be set according to experience.

[0064] Since temperature has an effect on the hardness of the metal, the deformation caused by forging of castings at different temperatures is different. In some embodiments, in order to avoid the casting deforming severely to form waste products during forging, the temperature information of the casting can include the temperature distribution of multiple positions of the casting. For example, the temperature values of the top, middle and bottom positions of the casting. When the temperature values of multiple positions of the casting are all within the suitable temperature for forging, the casting is forged.

[0065] In some embodiments, the temperature information can be obtained in various feasible ways.

[0066] The first score can be used to represent the standard degree of the casting equipment performing the casting process. In some embodiments, the first score can be represented by a numerical value, and the higher the numerical value, the higher the standard degree of the casting process, and vice versa.

[0067] ​In some embodiments, the first score can be determined using a first model. In some embodiments, the first model can include a trained machine learning model, which can include various models and results, such as a deep learning model (DNN), a recurrent neural network (RNN), etc.

[0068] In some embodiments, the input of the first model includes temperature information, and the output is the first score. For example, the input temperature information can be the melting temperature value and the pouring temperature value at each time point; the first model extracts the temperature value of different temperature information at each time point respectively, obtains the change characteristics of the melting temperature value and the change characteristics of the pouring temperature value; and outputs the first score of the casting blank. The input of the first model can also include other temperature information, and the output and determination of the first score can also take other feasible ways.

[0069] In some embodiments, the method for obtaining the first model includes obtaining at least one first training sample and an initial first model, wherein the first training sample includes sample temperature information labeled with a score; and iteratively updating the parameters of the initial first model based on the at least one first training sample to obtain the first model. For example, a plurality of training samples with labels can be input into the initial first model, a loss function can be constructed by the labels and the prediction results of the initial first model, and the parameters of the initial first model can be iteratively updated based on the loss function. When the loss function of the initial first model meets a preset condition, the model training is completed, wherein the preset condition can be that the loss function converges, the number of iterations reaches a threshold, etc. The method for training the initial first model can be various common methods, such as gradient descent method, etc.

[0070] In step 520, if the first score exceeds the first set threshold range, the service platform determines that the casting process is abnormal, and controls the production line equipment to stop casting the blank.

[0071] The first set threshold range can refer to a range of the first score preset in advance. When the first score is within this range, the possibility of abnormal casting process is low, and it can be considered that the casting process is not abnormal.

[0072] In some embodiments, the first set threshold range can be set by experience. For example, experience shows that when the first score is less than 3, the probability of abnormal casting process is close to 100%. Therefore, the first set threshold range can be determined as a range greater than 3.

[0073] In some embodiments of the present specification, by monitoring the temperature of the casting blank throughout the process and using the first model to determine whether the casting process is abnormal, it can be ensured that the progress of the casting blank can be more safely and orderly.

[0074] Figure 6is an exemplary schematic diagram illustrating application of the intelligent manufacturing method based on the centralized service platform industrial internet of things to rough forging according to some embodiments of the present specification.

[0075] In some embodiments, when the production line equipment is forging the rough, the abnormality judgment model can be a second model; the verification data set can include pre-pressing images and post-pressing images of the rough. As shown in Figure 6 The schematic flow 600 includes the following steps.

[0076] At step 610, the management platform inputs the pre-pressing images, the post-pressing images and the first reference data set into the second model, and the second model outputs a second score.

[0077] The pre-pressing images can refer to images before the rough is pressed. For example, images of the casting, images of the rough before each pressing, etc. The pre-pressing images can be used to represent the shape of the rough before being pressed.

[0078] The post-pressing images can refer to images after the rough is pressed. For example, images of the rough after each pressing. The post-pressing images can be used to represent the shape of the rough after being pressed.

[0079] In some embodiments, the pre-pressing images and the post-pressing images of the rough can be obtained by various image acquisition devices.

[0080] The first reference data set can refer to standard data used to judge whether the rough forging is standard. In some embodiments, the first reference data set includes standard pre-pressing images and standard post-pressing images.

[0081] The standard pre-pressing images can be a standard shape of the rough before pressing which is set in advance. For example, the standard pre-pressing images can be images of the product model before forging. In some embodiments, the standard pre-pressing images include a standard shape before the rough is forged. For example, the size of the casting, the flatness of the casting, etc.

[0082] The standard post-pressing images can be a standard shape of the rough after pressing which is set in advance. For example, the standard post-pressing images can be images of the product model after forging. In some embodiments, the standard post-pressing images include a standard shape after the rough is forged. For example, the size of the forging, the flatness of the forging, etc.

[0083] The second score can be used to represent the standard degree of the forging equipment in performing the forging process. The representation of the second score is similar to the first score, and more details about the second score can be found in Figure 5 and the related description.

[0084] In some embodiments, the second score can be determined using a second model. In some embodiments, the second model can include a trained machine learning model, which can include various models and results, such as a deep neural network (DNN), a convolutional neural network (CNN), etc.

[0085] In some embodiments, the input of the second model can be a pair of images, which can include a pre-forging image and a pre-standard pressing image, and a post-forging image and a post-standard pressing image, and the output can be a second score of the forging blank. For example, image a is an image of casting A, and image d is an image of forging A; image a1 is an image of a pre-forging product model (i.e., an image of a standard casting), and image b1 is an image of a post-forging product model (i.e., an image of a standard forging). Wherein forging A can be obtained by forging casting A. Therefore, the input of the second model can be {(a, a1), (d, b1)}; the second model can extract the features of the input images respectively, and then compare the features of a and a1, and the features of d and b1 respectively, and obtain the second score based on the comparison results.

[0086] In some embodiments, the verification data set includes a sequence of images of multiple pressings of the blank. The pre-standard pressing image can be a standard image of the blank before being pressed. The post-standard pressing image can be a standard image of the blank after being pressed for the last time.

[0087] The multiple pressings of the images can come from the forging process of the same blank, and the sequence of the multiple pressings of the images can be obtained by sorting the multiple pressings of the images in the order of forging. For example, casting A is pressed 3 times to obtain forging A, and image a is obtained before casting A is pressed; image b is obtained after casting A is pressed for the first time; and images c and d are obtained after the second and third pressings in the same way as image b is obtained. The sequence of images of multiple pressings of forging A can be obtained: (a, b, c, d).

[0088] Since the intermediate shape of the blank formed during the forging process is not uniform when the blank is forged, the standard image of the blank before being pressed can be used as the pre-standard pressing image, and the standard image of the blank after being pressed for the last time can be used as the post-standard pressing image. For example, the pre-standard pressing image can be a standard image of a casting. The post-standard pressing image can be a standard image of a forging.

[0089] In some embodiments, the second model can be a spatio-temporal network model (CNN+LSTM). In some embodiments, the input of the second model can include the image sequence of multiple pressings, the standard pre-pressing image and the standard post-pressing image, and the output can be the second score of the forging process. For example, the image sequence of multiple pressings of the forging A is (a, b, c, d), the standard pre-pressing image is a1, and the standard post-pressing image is b1; the input of the second model can be {(a, a1), ((a, b, c, d), b1)}; the second model can obtain the feature e of the forging A based on the image sequence of multiple pressings, and then compare the features of a and a1, e and b1, and obtain the second score based on the comparison result.

[0090] In some embodiments of the present specification, by obtaining the feature of the forging based on the image sequence of multiple pressings, the feature of the forging includes the process information of the forging, and the accuracy of the second score is improved.

[0091] In some embodiments, the method of obtaining the second model includes: obtaining at least one second training sample and an initial second model, wherein the second training sample includes a second image sample group labeled with a score, and the second image sample group at least includes a sample pre-pressing image, a sample post-pressing image, a standard pre-pressing image and a standard post-pressing image; and iteratively updating the parameters of the initial second model based on the at least one second training sample to obtain the second model. For example, a plurality of training samples with labels can be input into the initial second model, a loss function can be constructed by the labels and the prediction results of the initial second model, the parameters of the initial second model can be iteratively updated based on the loss function, and the model training is completed when the loss function of the initial second model meets a preset condition, wherein the preset condition can be that the loss function converges, the number of iterations reaches a threshold, etc. The method of training the initial second model can be various common methods, such as gradient descent method, etc.

[0092] In step 620, if the second score exceeds the second set threshold range, the service platform determines that the forging process is abnormal, and controls the production line equipment to stop forging the blank.

[0093] The second set threshold range can refer to a range of the second score preset in advance. When the second score is within the range, the possibility of abnormality of the forging process is low, and it can be considered that the forging process is not abnormal.

[0094] In some embodiments, the way of obtaining the second set threshold range is similar to the way of obtaining the first set threshold. For more information about the second set threshold, see Figure 5 and related descriptions.

[0095] Some embodiments in the specification can timely obtain the situation of the forging blank by monitoring the forging blank and judging whether the forging process is abnormal using the second model, and timely process the abnormal situation. Avoiding greater loss caused by untimely processing.

[0096] Figure 7 is an exemplary schematic diagram of applying the intelligent manufacturing method based on the centralized service platform industrial Internet of Things to blank stamping according to some embodiments of the specification.

[0097] In some embodiments, when the production line equipment stamps the blank, the abnormality judgment model can be a third model; the verification data set can include a pre-stamping image and a post-stamping image of the blank. As shown in Figure 7 The schematic flow 700 includes the following steps.

[0098] Step 710, the management platform inputs the pre-stamping image, the post-stamping image and the second reference data set into the third model, and the third model outputs a third score.

[0099] The pre-stamping image can refer to an image before the blank is stamped. For example, an image of a forging, an image of the blank before each stamping, etc. The pre-stamping image can be used to represent the shape of the blank before being stamped.

[0100] The post-stamping image can refer to an image after the blank is stamped. For example, an image of the blank after each stamping. The post-stamping image can be used to represent the shape of the blank after being stamped.

[0101] In some embodiments, the way of obtaining the pre-stamping image and the post-stamping image is similar to the way of obtaining the pre-pressing image and the post-pressing image. For more information about obtaining the pre-stamping image and the post-stamping image, see Figure 6 and related descriptions.

[0102] The second reference data set can refer to standard data used to judge whether the stamped blank is standard. In some embodiments, the second reference data set includes a standard pre-stamping image and a standard post-stamping image.

[0103] The standard pre-stamping image can be a standard shape of the blank before stamping. For example, the standard pre-stamping image can be an image of a product model before stamping. In some embodiments, the standard pre-stamping image includes a standard shape before stamping the blank. For example, the size of the forging, the thickness of the forging, etc.

[0104] The standard post-stamping image can be a standard shape of the pre-set post-stamping blank. For example, the standard post-stamping image can be an image of a post-stamping product model. In some embodiments, the standard post-stamping image includes a standard shape of the blank after being stamped. For example, the size of the stamped part, the thickness of the stamped part, etc.

[0105] The third score can be used to represent the standard degree of the stamping device performing the stamping process. The third score is represented in a similar manner to the first score, and more about the third score can be found in Figure 5 and the related description.

[0106] In some embodiments, the third score can be determined using a third model. The type of the third model and the manner in which the third model determines the third score are similar to the second model, and more about the third model can be found in Figure 6 and the related description.

[0107] In some embodiments, the verification data set includes a sequence of images of the blank being stamped multiple times. The standard pre-stamping image can be a standard image of the blank before being stamped. The standard post-stamping image can be a standard image of the blank after being stamped the last time. The standard images include a standard shape of the blank.

[0108] The images of the multiple stamping can come from the stamping process of the same blank, and the sequence of the images of the multiple stamping can be obtained by sorting the images of the multiple stamping in the order of the stamping. The stamping process is similar to the forging process, and more about the sequence of the images of the multiple stamping can be found in Figure 6 and the related description.

[0109] Based on the same reason as the forging blank, the standard image of the blank before being stamped can be used as the standard pre-stamping image, and the standard image of the blank after being stamped the last time can be used as the standard post-stamping image. For example, the standard pre-stamping image can be a standard image of the forged part. The standard post-stamping image can be a standard image of the stamped part.

[0110] In some embodiments, the input of the third model can include the sequence of the images of the multiple stamping, the standard pre-stamping image, and the standard post-stamping image, and the output can be the third score of the stamping. The content about the third model obtaining the third score based on the sequence of the images of the multiple stamping, the standard pre-stamping image, and the standard post-stamping image is similar to the content about the second model obtaining the second score based on the sequence of the images of the multiple stamping, the standard pre-stamping image, and the standard post-stamping image, and more about the third model can be found in Figure 6 and the related description.

[0111] In some embodiments of the present specification, the features of the stamped part are obtained based on the sequence of the images of the multiple stamping, so that the features of the stamped part include the process information of the stamping, and the accuracy of the third score is improved.

[0112] In some embodiments, the method of obtaining the third model comprises obtaining at least one third training sample and an initial third model, wherein the third training sample comprises a third image sample set labeled with a score; the third image sample set at least comprises a sample pre-stamping image, a sample post-stamping image, a standard pre-stamping image, and a standard post-stamping image; and parameters of the initial third model are iteratively updated based on the at least one third training sample to obtain the third model. The method of training the third model is similar to the method of training the second model. For more information about the third model, see Figure 6 and the related description.

[0113] At step 720, if the third score exceeds the third set threshold range, the service platform determines that the stamping process is abnormal, and controls the production line equipment to stop stamping the blank.

[0114] The third set threshold range can refer to a pre-set range of the third score. When the third score is within the range, the possibility of abnormality of the stamping process is low, and it can be considered that the stamping process is not abnormal.

[0115] In some embodiments, the way of obtaining the third set threshold range is similar to the way of obtaining the first set threshold. For more information about the third set threshold, see Figure 5 and the related description.

[0116] In some embodiments of the present specification, by monitoring the stamped blank and using the third model to determine whether the stamping process is abnormal, the condition of the stamped blank can be obtained in time, and the abnormal condition can be handled in time. Avoiding loss due to delayed processing.

[0117] Figure 8 is an exemplary flowchart of a method for intelligent manufacturing of a blank based on the performance of the blank according to some embodiments of the present specification. As Figure 8 shown, the flow 800 includes the following steps.

[0118] At step 810, the management platform can predict the performance of the blank after each process based on the verification data set of the plurality of processes for processing the blank.

[0119] The performance of the blank can include, but is not limited to, the hardness, strength, etc. of the blank.

[0120] In some embodiments, the management platform can input the verification data set into a performance prediction model, and the performance prediction model outputs the performance of the blank. For example, the management platform can input the verification data set of the stamped blank into the performance prediction model, and the model outputs the predicted hardness of the blank after stamping.

[0121] In some embodiments, the performance prediction model can be trained by the fourth training sample. The fourth training sample includes a sample check data set. The sample check data set is labeled with the performance of the corresponding blank. The fourth training sample can be obtained by extracting the historical data of the processed blank. The performance prediction model is trained in a similar manner to the first model. For more information about training the performance prediction model, see Figure 4 and the related description.

[0122] In some embodiments, the management platform can predict the performance of the blank after each process by fitting the check data set and the historical statistical data.

[0123] The historical statistical data can be data obtained by statistical processing of the historical data of the processed blank. The historical statistical data can include the check data set of the historical processed blank and the performance of the blank processed by the production line equipment according to the check data set.

[0124] In some embodiments, the management platform can classify the historical statistical data; obtain the check data set and the historical statistical data of the same type as the check data set, fit the relationship between the check data set and the historical statistical data using the least squares method, and then calculate the performance of the blank of the check data set according to the performance of the blank corresponding to the historical statistical data.

[0125] In some embodiments, the management platform can also adjust and modify the predicted performance of the blank by the actual performance of the blank, and save the adjusted and modified data. By continuously recording the actual performance of the blank and the predicted performance of the blank, a better fitting effect can be achieved.

[0126] In some embodiments of the present specification, the performance of the blank is predicted by fitting the check data set and the historical statistical data, so that the predicted performance of the blank is more accurate.

[0127] Step 820, the service platform determines whether the performance of the blank meets the preset condition based on the reference data set.

[0128] In some embodiments, the reference data set includes standard data of the process. For example, the reference data set includes the performance of the blank after each process. For more information about the reference data set, see Figure 4 and the related description.

[0129] The preset condition can refer to the condition that the performance of the processed blank should meet. In some embodiments, the service platform can determine the standard value range of each performance of the blank after the process according to the data of the reference data set. In some embodiments, the service platform can determine whether the performance of the blank after each process meets the standard according to the reference data set. If it does not meet the standard, it means that the processing process is abnormal.

[0130] Step 830, when it is determined that the performance of the blank does not meet the preset condition, stop processing the blank. For more about stopping processing, see Figure 4 and the related description.

[0131] In some embodiments, the service platform can determine whether the process of processing the blank has an abnormal risk based on the scores of the respective processes; if so, perform actual detection on the performance that does not meet the score, and determine whether there is really an abnormal risk; if so, send a stop instruction to the production equipment whose performance does not meet the score.

[0132] The abnormal risk can refer to the risk of abnormality in the process of processing the blank. The abnormal risk can be represented by high and low.

[0133] In some embodiments, the service platform can determine whether the process of processing the blank has an abnormal risk based on the first score, the second score and the third score. For example, the service platform can determine whether the first score, the second score and the third score are all within the corresponding set threshold range; if so, it is determined that the process of processing the blank has no risk; if not, perform actual detection on the performance that does not meet the score.

[0134] The actual detection can be a detection on the running condition of the object platform. For example, a detection device can be arranged on the object platform, which can be used to detect the object platform that has an abnormal risk. When the service platform determines that the process of processing the blank has an abnormal risk, the service platform actively sends an instruction to obtain a detection verification data set to the management platform. The management platform obtains the detection verification data set from the detection device according to the obtaining instruction; the service platform obtains the data from the management platform, and determines whether the object platform that processes the blank really has an abnormal risk. When the actual detection still indicates that the object platform has an abnormal risk, it is determined that the abnormal risk is true, and the management platform controls the object platform to stop processing the blank. The service platform determines whether the detection verification data set is abnormal in a similar way to determining whether the verification data set is abnormal. For more about determining whether the detection verification data set is abnormal, see Figure 4 and the related description.

[0135] In some embodiments, the service platform can fuse the scores of the respective processes, compare the fused score with a threshold, and determine whether there is an abnormal risk based on the comparison result. The fusion includes fusion based on the weight of each score. For example, the service platform can at least use the arithmetic mean method and the weighted average method to fuse the first score, the second score, and the third score to obtain the fused score. In some embodiments, the service platform can determine the credibility of each score and determine the weight of each score in the fusion based on the credibility. In some embodiments, the credibility can be determined according to the accuracy of the abnormality judgment model during training. For example, the weight can be the accuracy of the abnormality judgment model, and the accuracy can be the ratio of the number of correctly predicted training samples to the total number of training samples. For another example, the weight can be the F value of the abnormality judgment model, and the F value is the harmonic value of the precision and the recall. The F value can be obtained by formula (1):

[0136]

[0137] wherein F1 is the F value; the precision is the ratio of the number of correctly predicted positive samples to the total number of predicted positive samples; and the recall is the ratio of the number of correctly predicted positive samples to the total number of actual positive samples.

[0138] In some embodiments of the present specification, by re-detecting the object platform with an abnormal risk, the number of false positives can be reduced, and the reliability of the system can be improved.

[0139] Since the final purpose of the processing blank is that the blank can be used, in some embodiments of the present specification, whether the process is abnormal is determined based on the performance of the blank, which can more directly determine whether the processed blank can be normally used, so that the prediction result is more in line with the actual situation.

[0140] The above detailed description has described the basic concepts, and it is obvious that the above detailed description is only used as an example and does not limit the present specification. Although it is not explicitly stated herein, those skilled in the art can make various modifications, improvements and corrections to the present specification. Such modifications, improvements and corrections are suggested in the present specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of the present specification.

[0141] Meanwhile, specific terms are used in the present specification to describe the embodiments of the present specification. For example, “one embodiment”, “an embodiment”, and / or “some embodiments” means a certain feature, structure, or characteristic related to at least one embodiment of the present specification. Therefore, it should be emphasized and noted that “one embodiment” or “an embodiment” or “one alternative embodiment” mentioned in different places in the present specification does not necessarily refer to the same embodiment. In addition, some features, structures, or characteristics in one or more embodiments of the present specification can be properly combined.

[0142] Similarly, it should be noted that, in order to simplify the presentation of this specification and thus facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0143] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. An intelligent manufacturing method based on a centralized service platform Industrial Internet of Things, comprising a user platform, a service platform, a management platform, a sensor network platform, and an object platform that interact in sequence; characterized in that: in: The user platform is a terminal device configured to interact with the user, receive user input information, generate instructions and send them to the service platform, and display information sent by the service platform to the user; The service platform is configured as a first server, receives instructions sent by the user platform, processes the instructions, and sends the instructions to the management platform, and obtains information required by the user from the management platform and sends the information to the user platform; The management platform is configured as a second server, receives instructions sent by the service platform and controls the operation of the object platform according to the instructions, and receives and stores the perception information sent by the object platform; A sensor network platform configured as a communication network and gateway for interaction between the management platform and the object platform; The object platform is configured as production line equipment that performs manufacturing and production line sensors that perform data collection, receives instructions from the management platform to operate, and sends perception information to the management platform through the sensor network platform; The service platform adopts a centralized layout, which means that the platform uniformly receives, processes and sends data; the management platform and the sensor network platform both adopt a post-sub-platform layout, which means that the management platform and the sensor network platform are both equipped with a main platform and multiple sub-platforms, and control information and object platform parameter configuration information are transmitted from the sub-platforms to the main platform, and perception information is transmitted from the main platform to the sub-platforms; The intelligent manufacturing method is a method for processing a blank, wherein the process of processing the blank includes at least casting, forging and stamping, and the method includes: When configuring the object platform parameters, the user inputs object platform parameter configuration information through the user platform; the service platform receives the object platform parameter configuration information sent by the user platform and decomposes it into multiple configuration data groups according to the process and sends them to the management platform; the management platform stores and processes the received configuration data groups and sends them to the sensor network platform, with the stored configuration data groups serving as reference data groups; the sensor network platform stores and processes the received configuration data groups and sends them to the object platform, and the object platform completes the object platform parameter configuration according to the object platform parameter configuration information; When the production line equipment of the object platform is running, the object platform sends its perception information as a verification data set to the management platform via the sensor network platform at set time intervals; the management platform identifies the process, obtains the verification data set corresponding to the process, and processes the verification data set using the abnormality judgment model; the service platform determines whether there is an abnormality in the process based on the processed verification data set; Wherein, when the production line equipment casts the blank, the abnormality judgment model is a first model; the verification data group includes temperature information; the management platform inputs the temperature information into the first model, and the first model outputs a first score; When the production line equipment is forging a blank, the abnormality judgment model is a second model; the verification data set includes a pre-pressing image and a post-pressing image of the blank during forging; the management platform inputs the pre-pressing image, the post-pressing image, and the first reference data set into the second model, and the second model outputs a second score; When the production line equipment is stamping a blank, the abnormality judgment model is a third model; the verification data group includes a pre-stamping image and a post-stamping image of the blank; the management platform inputs the pre-stamping image, the post-stamping image, and the second reference data group into the third model, and the third model outputs a third score; The service platform performs weighted fusion processing on the first score, the second score, and the third score to obtain a fused score, wherein the weighted weights of the first score, the second score, and the third score are respectively the F-values ​​of the corresponding anomaly judgment model, and the F-value is the harmonic value of the precision and recall of the corresponding anomaly judgment model; Based on the comparison between the fused score and a threshold, it is determined whether there is an abnormal risk; in response to the existence of the abnormal risk, the management platform controls the object platform to stop processing the blank.

2. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 1 is characterized in that: When the user platform sends parameter configuration information to the object platform, each sub-platform of the management platform stores and processes a corresponding set of configuration data groups. The main platform of the management platform aggregates the configuration data groups processed by all sub-platforms of the management platform, stores and processes them, and sends multiple sets of configuration data groups to the sub-platforms of the sensor network platform one by one. Each sub-platform of the sensor network platform stores and processes the received configuration data group, and the main platform of the sensor network platform aggregates, stores and processes the configuration data groups processed by all sub-platforms of the sensor network platform, and sends multiple groups of configuration data groups to the target platform one by one.

3. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 1 is characterized in that: The management platform sends the reference data group and the verification data group corresponding to the target platform that sent the verification data group to the service platform; the service platform receives the reference data group and the verification data group for comparison; when the service platform compares the reference data group with the verification data group, if there is data in the verification data group that exceeds a set threshold range compared with the reference data group, the service platform also actively generates a perception information acquisition instruction before generating a stop operation instruction and sends it to the management platform; the management platform receives the instruction sent by the service platform and controls the corresponding target platform to send current perception information according to the instruction; The object platform feeds back the current perception information as a verification data group to the management platform through the sensor network platform. The management platform receives the verification data group fed back by the object platform, processes it, and sends it to the service platform. The service platform receives the verification data group and compares it with the reference data group. If all data in the verification data group are within the set threshold range when compared with the reference data group, the data in the verification data group is cleared and no subsequent processing is performed. If there is still data in the verification data group that exceeds the set threshold range when compared with the reference data group, the service platform generates a stop operation instruction.

4. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 3 is characterized in that: After the service platform generates a stop operation instruction, the service platform also actively generates a perception information acquisition instruction for the cascade process of the current process and sends it to the management platform. The management platform receives the instruction sent by the service platform and controls the cascade process object platform of the current corresponding object platform to send the current perception information according to the instruction; the cascade process is the previous level process or the next level process of the current process; The cascade process object platform feeds back the current perception information as a verification data group to the management platform through the sensor network platform. After receiving and processing the verification data group fed back by the cascade process object platform, the management platform sends the reference data group and the verification data group corresponding to the cascade process object platform that sent the verification data group to the service platform. The service platform receives the verification data group and the reference data group and compares the two. If all data in the verification data group are within the set threshold range when compared with the reference data group, the data in the verification data group is cleared and no subsequent processing is performed. If there is data in the verification data group that exceeds the set threshold range when compared with the reference data group, a stop operation instruction is generated and sent to the management platform and fed back to the user platform. The management platform sends it to the corresponding cascade process object platform through the sensor network platform to control the cascade process production line equipment to stop operation.

5. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 4 is characterized in that: When the cascade process is the next-level process of the current process, the service platform actively generates a cascade process perception information acquisition instruction before generating a stop operation instruction and sends it to the management platform. The management platform receives the instruction sent by the service platform and controls the corresponding cascade process object platform to send current perception information according to the instruction; The cascade process object platform feeds back the current perception information as a verification data group to the management platform through the sensor network platform. The management platform receives the verification data group fed back by the cascade process object platform, processes it, and sends it to the service platform. The service platform receives the verification data group and compares it with the reference data group. If all data in the verification data group are within the set threshold range when compared with the reference data group, the data in the verification data group is cleared without subsequent processing. If there is still data in the verification data group that exceeds the set threshold range when compared with the reference data group, the service platform generates a stop instruction to control the cascade process production line equipment to stop running.

6. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 1 is characterized in that: The temperature information includes melting temperature information and pouring temperature information; the melting temperature information includes the temperature information of the melted metal obtained at a first set time interval during the melting process; the pouring temperature information includes the temperature information of the casting obtained at a second set time interval after pouring.

7. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 1 is characterized in that: The first reference data set includes a standard pre-pressing image and a standard post-pressing image; the standard pre-pressing image includes a standard shape of the blank before forging; the standard post-pressing image includes a standard shape of the blank after forging.

8. The intelligent manufacturing method based on the centralized service platform industrial Internet of Things according to claim 1 is characterized in that: The second reference data set includes a standard pre-stamping image and a standard post-stamping image; the standard pre-stamping image includes the standard shape of the blank before stamping; the standard post-stamping image includes the standard shape of the blank after stamping.

9. A system for implementing the intelligent manufacturing method of the industrial Internet of Things based on a centralized service platform as described in any one of claims 1 to 8, comprising a user platform, a service platform, a management platform, a sensor network platform, and an object platform that interact in sequence; characterized in that: in: The user platform is a terminal device configured to interact with the user, receive user input information, generate instructions and send them to the service platform, and display information sent by the service platform to the user; The service platform is configured as a first server, receives instructions sent by the user platform, processes the instructions, and sends the instructions to the management platform, and obtains information required by the user from the management platform and sends the information to the user platform; The management platform is configured as a second server, receives instructions sent by the service platform and controls the operation of the object platform according to the instructions, and receives and stores the perception information sent by the object platform; A sensor network platform configured as a communication network and gateway for interaction between the management platform and the object platform; The object platform is configured as production line equipment that performs manufacturing and production line sensors that perform data collection, receives instructions from the management platform to operate, and sends perception information to the management platform through the sensor network platform; The service platform adopts a centralized layout, which means that the platform uniformly receives, processes and sends data; the management platform and the sensor network platform both adopt a post-sub-platform layout, which means that the management platform and the sensor network platform are both equipped with a main platform and multiple sub-platforms, and control information and object platform parameter configuration information are transmitted from the sub-platforms to the main platform, and perception information is transmitted from the main platform to the sub-platforms; The intelligent manufacturing method is a method for processing a blank, wherein the process of processing the blank includes at least casting, forging and stamping. When configuring the object platform parameters, the user inputs the object platform parameter configuration information through the user platform, the service platform receives the object platform parameter configuration information sent by the user platform and decomposes it into multiple configuration data groups according to the process and sends them to the management platform; the management platform stores and processes the received configuration data groups and sends them to the sensor network platform, and the stored configuration data groups serve as reference data groups; the sensor network platform stores and processes the received configuration data groups and sends them to the object platform, and the object platform completes the object platform parameter configuration according to the object platform parameter configuration information; When the production line equipment of the object platform is running, the object platform sends its perception information as a verification data set to the management platform via the sensor network platform at set time intervals; the management platform identifies the process, obtains the verification data set corresponding to the process, and processes the verification data set using the abnormality judgment model; the service platform determines whether there is an abnormality in the process based on the processed verification data set; Wherein, when the production line equipment casts the blank, the abnormality judgment model is a first model; the verification data group includes temperature information; the management platform inputs the temperature information into the first model, and the first model outputs a first score; When the production line equipment is forging a blank, the abnormality judgment model is a second model; the verification data set includes a pre-pressing image and a post-pressing image of the blank during forging; the management platform inputs the pre-pressing image, the post-pressing image, and the first reference data set into the second model, and the second model outputs a second score; When the production line equipment is stamping a blank, the abnormality judgment model is a third model; the verification data group includes a pre-stamping image and a post-stamping image of the blank; the management platform inputs the pre-stamping image, the post-stamping image, and the second reference data group into the third model, and the third model outputs a third score; The service platform performs weighted fusion processing on the first score, the second score, and the third score to obtain a fused score, wherein the weighted weights of the first score, the second score, and the third score are respectively the F-values ​​of the corresponding anomaly judgment model, and the F-value is the harmonic value of the precision and recall of the corresponding anomaly judgment model; Based on the comparison between the fused score and a threshold, it is determined whether there is an abnormal risk; in response to the existence of the abnormal risk, the management platform controls the object platform to stop processing the blank.

Images