Quality prediction method and device, electronic equipment and storage medium

Abstract

According to the quality prediction method and device, the electronic equipment and the storage medium, the detection data of each process of the target product in the manufacturing process is obtained, the quality of the target product is predicted based on the quality prediction model according to the detection data, quality risk prediction driven by data fusion is achieved, and the quality risk prediction efficiency is improved. Therefore, a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product are obtained. Therefore, global and local quality risk prediction is realized, global quality management of the complex manufacturing process of aerospace equipment can be supported based on the second quality prediction result, and quality management of each process can also be supported based on the first quality prediction result.

Description

Technical Field

[0001] This application relates to the technical field of quality management, and in particular, to a quality prediction method, device, electronic device, and storage medium. Background Art

[0002] The complexity of aerospace equipment products includes multiple product levels and involves the intersection of multiple disciplines such as machinery, electronics, and software. This characteristic makes the manufacturing process of aerospace equipment products extremely complex, consisting of multiple processes with different quality control mechanisms combined in series, parallel, and mixed connections at multiple levels.

[0003] When traditional quality management methods are applied to the highly complex series, parallel, and mixed manufacturing systems of aerospace equipment, the method of completely relying on the quality mechanism model to conduct analysis and then carry out quality control in the manufacturing process of aerospace equipment has problems such as an explosion in the workload of the model, low efficiency due to the complex product composition, and inaccurate identification and prediction of quality risks and inaccurate positioning of quality problems due to some mechanism gray boxes. Summary of the Invention

[0004] In view of this, the purpose of this application is to propose a quality prediction method, device, electronic device, and storage medium to solve or partially solve the above problems.

[0005] Based on the above purpose, in the first aspect of this application, a quality prediction method is provided, including:

[0006] Obtain the detection data of each process in the manufacturing process of the target product;

[0007] Based on the detection data, predict the quality of the target product based on the quality prediction model to obtain a prediction result, where the prediction result includes the first quality prediction result corresponding to each process in the manufacturing process and the second quality prediction result of the target product.

[0008] In the second aspect of this application, a quality prediction device is provided, including:

[0009] An acquisition module configured to obtain the detection data of each process in the manufacturing process of the target product;

[0010] A prediction module configured to predict the quality of the target product based on the detection data and the quality prediction model to obtain a prediction result, where the prediction result includes the first quality prediction result corresponding to each process in the manufacturing process and the second quality prediction result of the target product.

[0011] In a third aspect of the present application, there is provided an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where when the processor executes the program, the method described in the first aspect is implemented.

[0012] In a fourth aspect of the present application, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method described in the first aspect.

[0013] As can be seen from the above, a quality prediction method, apparatus, electronic device, and storage medium provided by the present application obtain the detection data of each process in the manufacturing process of a target product, and based on the detection data, predict the quality of the target product based on a quality prediction model, realizing data fusion-driven quality risk prediction, thereby obtaining a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product. In this way, quality risk prediction at both the global and local levels is achieved. It can not only support the overall quality management of the complex manufacturing process of aerospace equipment based on the second quality prediction result, but also support the quality management of each process based on the first quality prediction result. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the present application or related technologies, the following will briefly introduce the drawings required for use in the embodiments or related technology descriptions. Obviously, the drawings in the following description are only embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0015] Figure 1 A schematic diagram of manufacturing activities of an exemplary manufacturing process according to an embodiment of the present application is shown.

[0016] Figure 2 A schematic diagram of a neural network model corresponding to an exemplary manufacturing process according to an embodiment of the present application is shown.

[0017] Figure 3 A schematic diagram of an exemplary serial process according to an embodiment of the present application is shown.

[0018] Figure 4 A schematic diagram of neural network modeling of an exemplary serial process according to an embodiment of the present application is shown.

[0019] Figure 5 A schematic diagram of an exemplary coupled parallel process according to an embodiment of the present application is shown.

[0020] Figure 6 A schematic diagram of an exemplary stacked parallel process according to an embodiment of the present application is shown.

[0021] Figure 7 The flowchart shows an exemplary quality prediction method according to an embodiment of the present application.

[0022] Figure 8 The schematic diagram shows an exemplary quality prediction device according to an embodiment of the present application.

[0023] Figure 9 The schematic diagram shows an exemplary electronic device according to an embodiment of the present application. Detailed implementation manners

[0024] To make the objectives, technical solutions, and advantages of the present application more clear and understandable, the following further describes the present application in detail with reference to specific embodiments and the accompanying drawings.

[0025] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the general meanings understood by those of ordinary skill in the field to which the present application belongs. The terms "first", "second" and similar words used in the embodiments of the present application do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms "connected" or "coupled" and the like are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms "upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0026] The complexity of aerospace equipment products includes multiple product levels and involves the intersection of multiple disciplines such as machinery, electronics, and software. This characteristic makes the manufacturing process of aerospace equipment products extremely complex, which is composed of multiple processes with different quality control mechanisms in a series, parallel, and mixed combination of multiple levels. Different types of quality influencing factors are introduced in each link of the manufacturing process and are superimposed, transmitted, and coupled along the complex manufacturing process, affecting the final product quality of aerospace equipment with a highly complex non-linear effect.

[0027] The multi-level complex composition of aerospace products and the serial, parallel, and hybrid manufacturing processes result in huge economic losses and even major safety problems that pose a threat to personnel safety due to any negligence in process quality control. Therefore, stricter and more meticulous measures must be taken in quality control. However, when traditional quality management methods are applied to the highly complex serial, parallel, and hybrid manufacturing systems of aerospace equipment, the method of relying entirely on quality mechanism models to conduct analysis and then carry out quality control in the manufacturing process of aerospace equipment has problems such as an explosion in model workload, low efficiency due to the complex product composition, and inaccurate identification and prediction of quality risks and inaccurate positioning of quality problems due to some mechanism black boxes.

[0028] In view of the significant progress made in digital research and development and intelligent manufacturing in the manufacturing process of aerospace equipment in recent years, a large amount of quality data has been accumulated in the manufacturing process. By using a data-driven research method and leveraging the rich information contained in the data, mechanism analysis can be supplemented and improved. However, directly encapsulating the complex manufacturing process of an aerospace equipment into a neural network system and then carrying out quality risk early warning will result in a lack of process in risk control, a relatively coarse control granularity, and the ability to only manage the overall input and output, making it difficult to refine to each link in the manufacturing process and unable to meet the requirements of precise quantification and refinement in the quality management of the development of complex aerospace equipment.

[0029] To at least solve the above problems, the present application provides a quality prediction method, device, electronic device, and storage medium. By obtaining the detection data of each process in the manufacturing process of the target product and based on the detection data, predicting the quality of the target product based on a quality prediction model, data fusion-driven quality risk prediction is achieved, thereby obtaining a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product. In this way, quality risk prediction is realized at both the global and local levels. It can not only support the overall quality management of the complex manufacturing process of aerospace equipment based on the second quality prediction result but also support the quality management of each process based on the first quality prediction result.

[0030] The manufacturing process of aerospace equipment, especially the assembly process of relatively complex products, consists of multiple processes connected in series, parallel, and hybrid. The input of a certain process in the manufacturing process is the output of one or several upstream processes and the process technology elements of this process, and its output is also the input of the downstream process. The quality of a manufacturing process refers to the situation where the quality characteristics of the products output by the manufacturing process meet the requirements. In the complex manufacturing process of aerospace equipment, the quality of the process is comprehensively affected by the process technology elements of this process and the quality characteristics of the output of the upstream process.

[0031] Figure 1 Shows a schematic diagram of the manufacturing activities of an exemplary manufacturing process according to an embodiment of the present application.

[0032] AsFigure 1 As shown, each process in the manufacturing process can be described as a unified manufacturing activity, including the input of the manufacturing process, the control of the manufacturing process, the mechanism of the manufacturing process, and the output of the manufacturing process.

[0033] Taking a mechanical assembly process as an example, the input of the manufacturing process generally includes the geometric dimensions, assembly tolerances, assembly tooling, assembly tools, etc. of the parts to be assembled in this process; the control of the manufacturing process generally refers to the assembly process plan of this assembly process, specifically including the assembly sequence, the operation of the assembly tools, the installation and cooperation methods of the parts and the assembly tooling, etc.; the mechanism of the manufacturing process generally refers to the basic principle of this assembly process, such as the interference fit and clearance fit of holes and shafts, etc.; the output of the manufacturing process refers to a series of characteristics of the product after the assembly activity is completed, such as the formed dimensional values, form and position tolerances, etc.

[0034] Figure 2 The figure shows a schematic diagram of a neural network model corresponding to an exemplary manufacturing process according to an embodiment of the present application.

[0035] As Figure 2 shown, in the neural network model, x1,…,x i ,…,x m represent the nodes of the input layer, and these nodes receive input data; b1 represents the bias term of the input layer, which is a constant term used to adjust the flexibility of the model; the hidden layer 1 contains multiple nodes, labeled Each hidden layer node is connected to each node of the input layer through a weight ; the hidden layer 2 contains multiple nodes, labeled Each hidden layer node is connected to each node of the hidden layer 1 through a weight ; y1, y2,..., y j ,..., y n represent the nodes of the output layer, and these nodes generate the final output of the neural network. The output layer nodes are connected to the nodes of the hidden layer 2 through a weight .

[0036] As Figure 2 shown, in some embodiments, the input of the manufacturing process can be mapped to the input layer of the neural network, the manufacturing process activities, mechanisms, and controls can be mapped to the hidden layer, and the output of the manufacturing process can be mapped to the output layer of the neural network. Among them, X=(x1,..., x i ,..., x m ) are the quality characteristics of the upstream process, the process elements, equipment elements, and related parameters of this process, and these characteristics and elements can all be detected using digital detection methods to obtain their measured values. Y=(y1, y2,..., y j ,..., y n) is the quality characteristic output by this process, and these quality characteristics can be detected using digital detection methods to obtain their measured values.

[0037] Taking the assembly process as an example, X = (x1,..., x i ,..., x m ) are the flatness of the docking surface in two docking cabin sections, the hole diameter, hole cylindricity, hole position degree of the assembly holes, the diameter, cylindricity of the connecting pins, and other dimensional characteristics. Usually, digital detection equipment such as laser trackers and industrial cameras can be used to detect these dimensional characteristics as the input of the assembly; Y = (y1, y2,..., y j ,..., y n ) are the quality characteristic values of the rocket cabin section that completes the assembly activity and judges whether this assembly process is qualified, such as the gap, step difference, and contour width of the skin at the docking place, the number and position degree of the rivets, etc.

[0038] Through the above method, a process in the aerospace equipment manufacturing process can be converted into a neural network model. It should be noted that the neural network model can, according to needs, refer to the simulation effects of different neural networks for different manufacturing principle activities, select different types of hidden layers, so as to establish a neural network model more suitable for the manufacturing principle of the process, such as deep neural networks, convolutional neural networks, etc.

[0039] The transfer relationship of the quality characteristics between processes in the aerospace equipment manufacturing process can be divided into serial processes, coupled parallel processes, and superimposed parallel processes. The neural network modeling of these three situations will be introduced below.

[0040] Figure 3 Shows a schematic diagram of an exemplary serial process according to an embodiment of the present application.

[0041] (1) Neural network modeling of serial processes

[0042] Two processes with a series relationship show a clear sequence in time and an upstream-downstream relationship in space. As Figure 3 shown, the activities of manufacturing process 302 include the input X of the manufacturing process, the control of the manufacturing process, the mechanism of the manufacturing process, and the output Y of the manufacturing process. The activities of manufacturing process 304 include the input b g of the manufacturing process, the control of the manufacturing process, the mechanism of the manufacturing process, and the output Z of the manufacturing process. The quality characteristic (for example, output Y) obtained by the previous process (for example, manufacturing process 302) will be used as a quality characteristic influencing factor (for example, input Y), and the input b q and b p of the subsequent process act on the manufacturing activities of the subsequent process (for example, manufacturing process 304), thereby affecting the quality characteristic of the subsequent process and obtaining the final output Z.

[0043] Quality characteristics are specific indicators that describe the quality status of products in this manufacturing process. For machining or assembly processes, they usually include geometric characteristics, such as the aperture diameter, hole depth, and positional tolerance of a hole, etc.; the influencing factors of quality characteristics are the elements that affect the values of product quality characteristics in this manufacturing process. For example, in a drilling manufacturing process, these influencing factors include the wear degree of the drill bit, the spindle speed of the machine tool, the feed rate, the magnification, etc. b q 、b p This refers to this part of the factors.

[0044] Figure 4 FIG. shows a schematic diagram of neural network modeling of an exemplary serial process according to an embodiment of the present application.

[0045] During the modeling process, as Figure 4 shown, in some embodiments, after mapping two processes to neural network models respectively, the output layer of the neural network model corresponding to the previous process can be connected to the quality characteristic nodes in the input layer of the neural network model corresponding to the next process, so as to form a neural network model of a serial process.

[0046] A quality characteristic node is the existence form of a specific quality characteristic in this model. For example, the hole diameter of a hole is a quality characteristic node in this model.

[0047] It can be understood that Figure 4 in the neural network model shown in the hidden layer 3 contains multiple nodes, labeled as z1,z2,…,z j ,…,z n , the connection manner between the hidden layer 3, the hidden layer 4 and the output layer is the same as that of the neural network model shown in the above Figure 2 and will not be elaborated here.

[0048] Figure 5 FIG. shows a schematic diagram of an exemplary coupled parallel process according to an embodiment of the present application.

[0049] (2) Neural network modeling of coupled parallel processes

[0050] As Figure 5As shown, there are two processes with a coupled parallel structure that are concurrent in time and often independent in space. The quality characteristics obtained from the two processes will affect the quality characteristics of subsequent processes in the form of a multiplicative effect. For two processes in a coupled parallel relationship (for example, manufacturing process 502 and manufacturing process 504), two fully connected network units are still established. After aligning the dimensions of the outputs of the two neural networks (for example, output Y and output K), a Hadamard product is performed to couple the two outputs, thereby forming a tail with the Hadamard product as the quality influencing factor (for example, input H), and the input b q of b p acting on the neural network tuple of the subsequent process to obtain the final output Z.

[0051] It should be noted that in addition to the Hadamard product, other methods can also be used to perform the product, such as the Kronecker product, dot product, outer product, etc. The embodiments of the present application do not limit this.

[0052] Figure 6 Shows a schematic diagram of an exemplary superimposed parallel process according to an embodiment of the present application.

[0053] (3) Neural network modeling of superimposed parallel processes

[0054] As Figure 6 shown, there are two parts with a superimposed parallel structure that are concurrent in time and in a parallel relationship in space. The quality characteristics obtained from the two processes will affect the quality characteristics of subsequent links in the form of an additive effect. For two processes in a superimposed parallel relationship (for example, manufacturing process 602 and manufacturing process 604), after aligning the dimensions of the outputs of the corresponding fully connected network units of the two processes (for example, output Y and output K), a linear addition is performed to superimpose the two outputs, thereby forming a tail with the sum as the quality influencing factor (for example, input H), and the input b q of b p acting on the neural network tuple of the subsequent process to obtain the final output Z.

[0055] After constructing the neural network, the training and verification of the network need to be carried out. The training set and test set of the network are historical data of the manufacturing process. According to the momentum-stochastic gradient descent method, a backpropagation algorithm is designed to solve each fully connected neural network unit, realizing the forward propagation and backward propagation of each unit, thereby realizing the algorithm design of the total neural network with series-parallel hybrid connection in the whole chain. In some embodiments, the designed neural network is randomly assigned a training set and a test set according to a ratio of 7:3 of all data samples, and the convergence training of the model loss function and the iterative test of the model accuracy are carried out.

[0056] In order to make the series-parallel hybrid neural network have predictive value, it is necessary to train it to achieve the optimal combination of network parameters. Since this prediction problem is a regression problem, in some embodiments, the mean squared error loss function can be selected as the loss function of the overall neural network; the series-parallel hybrid neural network can continuously iterate and optimize the network parameters through the Stochastic Gradient Descent with Momentum Backpropagation algorithm (SGDM). After multiple forward deductions and backpropagation calculations, the optimal neural network parameter combination can be obtained, so as to achieve the best regression effect of the hybrid neural network prediction model. The regression prediction model obtained by training the series-parallel hybrid neural network can predict the abnormal combination problems of different quality characteristic influencing factors, and obtain the prediction results of the conformity of the product quality characteristics under the corresponding feed quality source characteristics and / or the abnormal patterns of the quality control points of the manufacturing system.

[0057] This application proposes a quality prediction method. Specifically, it is a quality risk prediction method for the manufacturing process of aerospace complex equipment based on a hybrid neural network model. This method maps the processes in the aerospace equipment manufacturing process into a neural network, and connects the input and output of the neural network corresponding to a single process according to the upstream and downstream relationships of the processes, forming a hybrid neural network that is one-to-one mapped with the manufacturing process. On this basis, based on the historical quality characteristic data collected during the aerospace equipment manufacturing process, this neural network is trained. After achieving the training effect, the process quality status of the aerospace equipment manufacturing process is obtained in real time based on the online detection system and input into this hybrid neural network to predict the quality of the equipment manufacturing process.

[0058] The method provided in this application is an integrated modeling method that combines aerospace equipment manufacturing knowledge and manufacturing data. Based on this method, the quality risk prediction problem is converted into a neural network regression problem. In addition, thanks to the mapping relationship between the constructed neural network and the manufacturing process, it is possible to predict the quality risks of the final state of aerospace equipment throughout the manufacturing process, and it is also possible to predict the quality risks of key processes during the manufacturing process.

[0059] Figure 7 A schematic flowchart of an exemplary quality prediction method 700 according to an embodiment of this application is shown. As Figure 7 shown, the method 700 may include the following steps.

[0060] In step 702, the detection data of each process in the manufacturing process of the target product is obtained.

[0061] In step 704, based on the detection data, the quality of the target product is predicted based on the quality prediction model to obtain a prediction result, where the prediction result includes a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product.

[0062] In some embodiments, the quality prediction model is constructed by the following method: mapping each process in the manufacturing process to a neural network model; connecting the transfer relationships between multiple neural network models corresponding to each process in the manufacturing process to obtain the quality prediction model.

[0063] In some embodiments, the transfer relationship includes serial processes. The processes in the manufacturing process include a first process and a second process. The multiple neural network models include a first neural network model corresponding to the first process and a second neural network model corresponding to the second process. The connecting the transfer relationships between multiple neural network models corresponding to each process in the manufacturing process to obtain the quality prediction model further includes: in response to the transfer relationship between the first process and the second process being the serial process, connecting the output layer of the first neural network model to the quality characteristic nodes in the input layer of the second neural network model to obtain the quality prediction model.

[0064] In some embodiments, the transfer relationship includes coupled parallel processes. The processes in the manufacturing process include a third process and a fourth process. The multiple neural network models include a third neural network model corresponding to the third process and a fourth neural network model corresponding to the fourth process. The connecting the transfer relationships between multiple neural network models corresponding to each process in the manufacturing process to obtain the quality prediction model further includes: in response to the transfer relationship between the third process and the fourth process being the coupled parallel process, multiplying the output layers of the third neural network model and the fourth neural network model after dimension alignment to obtain a first target neural network model; obtaining the quality prediction model based on the first target neural network model.

[0065] In some embodiments, the transfer relationship includes superimposed parallel processes. The processes in the manufacturing process include a fifth process and a sixth process. The multiple neural network models include a fifth neural network model corresponding to the fifth process and a sixth neural network model corresponding to the sixth process. The connecting the transfer relationships between multiple neural network models corresponding to each process in the manufacturing process to obtain the quality prediction model further includes: in response to the transfer relationship between the fifth process and the sixth process being the superimposed parallel process, adding the output layers of the fifth neural network model and the sixth neural network model after dimension alignment to obtain a second target neural network model; obtaining the quality prediction model based on the second target neural network model.

[0066] In some embodiments, the manufacturing process further includes a seventh process, and the plurality of neural network models further include a seventh neural network model corresponding to the seventh process. Connecting the transfer relationships between the plurality of neural network models corresponding to each process in the manufacturing process to obtain the quality prediction model further includes: in response to the transfer relationship between the third process and the fourth process in parallel and the seventh process being a serial process, connecting the output layer of the first target neural network model to the quality characteristic nodes in the input layer of the seventh neural network model to obtain the quality prediction model; in response to the transfer relationship between the fifth process and the sixth process in parallel and the seventh process being a serial process, connecting the output layer of the second target neural network model to the quality characteristic nodes in the input layer of the seventh neural network model to obtain the quality prediction model.

[0067] In some embodiments, each process in the manufacturing process includes the input of the manufacturing process, the mechanism of the manufacturing process, the control of the manufacturing process, and the output of the manufacturing process. The neural network model includes an input layer, a hidden layer, and an output layer. Mapping each process in the manufacturing process to a neural network model further includes: mapping the input of the manufacturing process to the input layer of the neural network model; and / or mapping the mechanism and control of the manufacturing process to the hidden layer of the neural network model; and / or mapping the output of the manufacturing process to the output layer of the neural network model.

[0068] A quality prediction method, device, electronic device, and storage medium provided by the present application obtain the detection data of each process in the manufacturing process of the target product, and based on the detection data, predict the quality of the target product based on the quality prediction model, realizing the quality risk prediction driven by data fusion, thereby obtaining the first quality prediction result corresponding to each process in the manufacturing process and the second quality prediction result of the target product. In this way, the quality risk prediction at both the global and local levels is realized. It can not only support the overall quality management of the complex manufacturing process of aerospace equipment based on the second quality prediction result, but also support the quality management of each process based on the first quality prediction result.

[0069] It should be noted that the method of the embodiment of the present application can be executed by a single device, such as a computer or a server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In this case of a distributed scenario, one of the multiple devices can only execute one or more steps of the method of the embodiment of the present application, and these multiple devices will interact with each other to complete the described method.

[0070] It should be noted that some embodiments of the present application have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the above embodiments and still achieve the desired results. Additionally, the processes depicted in the figures do not necessarily require the particular order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0071] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present application further provides a quality prediction device.

[0072] Referring to Figure 8 , the quality prediction device includes:

[0073] An acquisition module 801, configured to acquire the detection data of each process in the manufacturing process of the target product.

[0074] A prediction module 802, configured to predict the quality of the target product based on the detection data and the quality prediction model to obtain a prediction result, where the prediction result includes a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product.

[0075] The device further includes a construction module, configured to map each process in the manufacturing process to a neural network model; connect the transfer relationships between the multiple neural network models corresponding to each process in the manufacturing process to obtain the quality prediction model.

[0076] The construction module is further configured to, in response to the transfer relationship between the first process and the second process being a serial process, connect the output layer of the first neural network model to the quality characteristic nodes in the input layer of the second neural network model to obtain the quality prediction model.

[0077] The construction module is further configured to, in response to the transfer relationship between the third process and the fourth process being a coupled parallel process, align the dimensions of the output layers of the third neural network model and the fourth neural network model and then multiply them to obtain a first target neural network model; obtain the quality prediction model based on the first target neural network model.

[0078] The construction module is further configured to, in response to the transfer relationship between the fifth process and the sixth process being a superposition and parallel process, add the output layers of the fifth neural network model and the sixth neural network model after aligning their dimensions to obtain a second target neural network model; and obtain the quality prediction model based on the second target neural network model.

[0079] The construction module is further configured to, in response to the transfer relationship between the third process and the fourth process in parallel and the seventh process being a serial process, connect the output layer of the first target neural network model to the quality characteristic nodes in the input layer of the seventh neural network model to obtain the quality prediction model; and in response to the transfer relationship between the fifth process and the sixth process in parallel and the seventh process being a serial process, connect the output layer of the second target neural network model to the quality characteristic nodes in the input layer of the seventh neural network model to obtain the quality prediction model.

[0080] The construction module is further configured to map the input of the manufacturing process to the input layer of the neural network model; and / or map the mechanism of the manufacturing process and the control of the manufacturing process to the hidden layer of the neural network model; and / or map the output of the manufacturing process to the output layer of the neural network model.

[0081] For convenience of description, when describing the above device, various modules are described separately according to their functions. Of course, when implementing this application, the functions of each module can be implemented in one or more software and / or hardware.

[0082] The device in the above embodiment is used to implement the corresponding method 700 in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated here.

[0083] Based on the same inventive concept, corresponding to the method in any of the above embodiments, the present application further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, it implements the method 700 described in any of the above embodiments.

[0084] Figure 9 FIG. shows a schematic diagram of an exemplary electronic device according to an embodiment of the present application. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. Among them, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are communicatively connected to each other inside the device through the bus 1050.

[0085] The processor 1010 can be implemented in the form of a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0086] The memory 1020 can be implemented in the form of a ROM (Read Only Memory), a RAM (Random Access Memory), a static storage device, a dynamic storage device, etc. The memory 1020 can store an operating system and other application programs. When implementing the technical solutions provided in the embodiments of this specification through software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0087] The input / output interface 1030 is used to connect to the input / output module to achieve information input and output. The input / output module can be configured as a component in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Among them, the input device can include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device can include a display, a speaker, a vibrator, an indicator light, etc.

[0088] The communication interface 1040 is used to connect to a communication module (not shown in the figure) to achieve communication and interaction between this device and other devices. Among them, the communication module can achieve communication through a wired method (such as USB, network cable, etc.) or through a wireless method (such as a mobile network, WIFI, Bluetooth, etc.).

[0089] The bus 1050 includes a path for transmitting information between various components of the device (such as the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0090] It should be noted that although only the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040, and the bus 1050 are shown in the above device, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, those skilled in the art can understand that the above device may also only include the components necessary to implement the solutions of the embodiments of this specification, and do not have to include all the components shown in the figure.

[0091] The electronic device in the above embodiments is used to implement the corresponding method 700 in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be elaborated herein.

[0092] Based on the same inventive concept, corresponding to the method of any of the above embodiments, the present application further provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the method 700 described in any of the foregoing embodiments.

[0093] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible by a computing device.

[0094] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the method 700 described in any of the foregoing embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be elaborated herein.

[0095] Those of ordinary skill in the art should understand that the discussion of any of the above embodiments is only exemplary and is not intended to imply that the scope of the present application (including the claims) is limited to these examples; under the concept of the present application, the technical features in the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations in different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of brevity.

[0096] In addition, for simplicity of explanation and discussion, and so as not to make the embodiments of the present application difficult to understand, well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Further, the devices may be shown in block diagram form in order to avoid making the embodiments of the present application difficult to understand, and this also takes into account the fact that details of the implementation of these block diagram devices are highly dependent on the platform on which the embodiments of the present application are to be implemented (i.e., these details should be fully within the understanding of those skilled in the art). In cases where specific details (such as circuits) are set forth to describe exemplary embodiments of the present application, it will be apparent to those skilled in the art that the embodiments of the present application may be practiced without these specific details or with variations of these specific details. Accordingly, these descriptions are to be regarded as illustrative rather than restrictive.

[0097] Although the present application has been described in connection with specific embodiments thereof, many alternatives, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art in light of the foregoing description. For example, other memory architectures (such as dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0098] Embodiments of the present application are intended to cover all such alternatives, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the embodiments of the present application shall be included within the protection scope of the present application.

Claims

1. A quality prediction method, comprising: Obtain the test data of each process of the target product during the manufacturing process; According to the detection data, the quality of the target product is predicted based on the quality prediction model to obtain a prediction result, which includes a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product.

2. The method of claim 1, wherein: The quality prediction model is constructed by the following method: Mapping each process in the manufacturing process into a neural network model; The transfer relationship between the plurality of neural network models corresponding to each process in the manufacturing process is connected to obtain the quality prediction model.

3. The method of claim 2, wherein: The transfer relationship includes a serial process, the processes in the manufacturing process include a first process and a second process, the plurality of neural network models include a first neural network model corresponding to the first process and a second neural network model corresponding to the second process, and the transfer relationship between the plurality of neural network models corresponding to each process in the manufacturing process is connected to obtain the quality prediction model further includes: In response to the transfer relationship between the first process and the second process being the serial process, the output layer of the first neural network model is connected with the quality characteristic nodes in the input layer of the second neural network model to obtain the quality prediction model.

4. The method of claim 2, wherein: The transfer relationship includes coupled parallel processes, the processes in the manufacturing process include a third process and a fourth process, the multiple neural network models include a third neural network model corresponding to the third process and a fourth neural network model corresponding to the fourth process, and the transfer relationship between the multiple neural network models corresponding to each process in the manufacturing process is connected to obtain the quality prediction model further includes: In response to the transfer relationship between the third process and the fourth process being the coupled parallel process, aligning the dimensions of an output layer of the third neural network model and an output layer of the fourth neural network model and then multiplying them to obtain a first target neural network model; The quality prediction model is obtained based on the first target neural network model.

5. The method of claim 2, wherein: The transfer relationship includes superimposed parallel processes, the processes in the manufacturing process include a fifth process and a sixth process, the multiple neural network models include a fifth neural network model corresponding to the fifth process and a sixth neural network model corresponding to the sixth process, and the transfer relationship between the multiple neural network models corresponding to each process in the manufacturing process is connected to obtain the quality prediction model further includes: In response to the transfer relationship between the fifth process and the sixth process being the superposition parallel process, aligning the dimensions of the output layer of the fifth neural network model and the output layer of the sixth neural network model and then adding them to obtain a second target neural network model; The quality prediction model is obtained based on the second target neural network model.

6. The method according to claim 4 or 5, wherein: The processes in the manufacturing process further include a seventh process, the plurality of neural network models further include a seventh neural network model corresponding to the seventh process, and the transfer relationship between the plurality of neural network models corresponding to each process in the manufacturing process is connected to obtain the quality prediction model further includes: In response to the third process and the fourth process being connected in parallel and having a transmission relationship with the seventh process being a serial process, connecting the output layer of the first target neural network model with the quality characteristic nodes in the input layer of the seventh neural network model to obtain the quality prediction model; In response to the fact that the fifth process and the sixth process are connected in parallel and their transmission relationship with the seventh process is a serial process, the output layer of the second target neural network model is connected to the quality characteristic nodes in the input layer of the seventh neural network model to obtain the quality prediction model.

7. The method of claim 2, wherein: Each process in the manufacturing process includes an input of the manufacturing process, a mechanism of the manufacturing process, a control of the manufacturing process and an output of the manufacturing process, the neural network model includes an input layer, a hidden layer and an output layer, and mapping each process in the manufacturing process to the neural network model further includes: Mapping the input of the manufacturing process to the input layer of the neural network model; and / or mapping the mechanism of the manufacturing process and the control of the manufacturing process into hidden layers of the neural network model; and / or The output of the manufacturing process is mapped to the output layer of the neural network model.

8. A quality prediction device, comprising: An acquisition module is configured to acquire detection data of each process of the target product in the manufacturing process; The prediction module is configured to predict the quality of the target product based on the quality prediction model according to the detection data to obtain a prediction result, wherein the prediction result includes a first quality prediction result corresponding to each process in the manufacturing process and a second quality prediction result of the target product.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.

Images