A visual display method for industrial chain quality mapping

Abstract

The application discloses a kind of industrial chain quality atlas drawing visual display method, according to the execution order of each process of target industrial chain, each process node is visually processed, and available atlas is obtained.After that, the connection between adjacent two process nodes, the negative influence caused to intermediate product and other factors are quantified by multidimensional vector.Then, based on the model trained based on multidimensional vector, the yield is predicted based on the influence of target industrial chain on intermediate product.The prediction result is image processing, to provide conditions for subsequent visual display.The method provided by the application realizes the quantization of target industrial chain and the display of prediction result by electronic digital data processing technology means.

Description

Technical Field

[0001] This application relates to the field of electronic digital data processing technology, and in particular to a method for digital computing or data processing specifically applicable to a particular application, specifically a visualization method for drawing a quality map of the industrial chain. Background Technology

[0002] The industrial chain promotes economies of scale. By organically linking its various links, companies can achieve large-scale production and specialization, thereby reducing production costs and increasing efficiency. This economies of scale enable companies to reduce unit costs while ensuring product quality, thus enhancing their market competitiveness.

[0003] However, in reality, achieving synchronized and highly efficient operation across the entire supply chain is virtually impossible. Because the supply chain involves numerous technological steps, a problem in any one step can impact upstream and downstream processes, even affecting the quality of the final product. For example, in the bread processing supply chain, a problem in the baking process forces the bread dough produced upstream to have an extended storage time, potentially leading to over-fermentation and ultimately affecting the quality of the final bread. Therefore, monitoring and evaluating the quality of the supply chain is essential. Furthermore, providing users with the results of such evaluations in a visual manner would facilitate their decision-making.

[0004] For example, the publication (announcement) number: CN106845796B, patent title: "An online prediction method for product quality in hydrocracking process" (main classification number: G06F17 / 00), utilizes process variables of the entire hydrocracking process to enable the predicted quality to reflect the overall operating status.

[0005] On the one hand, this demonstrates that digital data processing technology has great potential in quality supervision and analysis; on the other hand, it also shows that there is still a broad prospect for technological expansion in this field. Summary of the Invention

[0006] This application provides a visualization method for drawing a quality map of the industrial chain, so as to at least partially solve the above-mentioned technical problems.

[0007] The embodiments of this application adopt the following technical solutions:

[0008] In a first aspect, embodiments of this application provide a method for visualizing and displaying a quality map of an industry chain, the method comprising:

[0009] The target industry chain is graphically processed to obtain a usable graph. The usable graph includes process nodes that represent the order of processes in which the product is produced in the target industry chain, and each process node corresponds one-to-one with a process. There is a multi-dimensional vector between two adjacent process nodes. The multi-dimensional vector is used to represent the transfer time between the processes corresponding to the two process nodes, the negative impact index of the transfer on intermediate products, the difficulty index of the connection between processes, the added value of intermediate products, and the replacement impact index of the downstream process node in the two adjacent process nodes when the target industry chain is operating normally.

[0010] When the restart or adjustment of the target supply chain is detected, the target input, constructed based on the attribute data and bias vector of the target product to be produced using the target supply chain, is input into a pre-trained target LSTM model, and the output is used as the predicted yield. The predicted yield represents the yield of the intermediate products at each of the process nodes. The target LSTM model is trained using historical vectors. The bias vector represents the difference between the current situation of the target supply chain and the multi-dimensional vector. The historical vector is a representation of the historical situation corresponding to the multi-dimensional vector.

[0011] Add the predicted yield to the hidden layer of the corresponding process node in the available graph;

[0012] When a user operation is detected, the selected process node is designated as the first target node.

[0013] Display the portion of the attribute data of the target product that corresponds to the first target node, and display the hidden layer corresponding to the first target node.

[0014] In an optional embodiment of this specification, the target LSTM model is obtained using the following steps:

[0015] Based on the attribute data collected during the production of historical products in the target industrial chain, a sample sequence and its corresponding sample tag set are constructed. Each sample sequence contains several sample nodes arranged in the order of the processes in which historical products were produced in the target industrial chain. Each sample node corresponds one-to-one with a process, and each sample node represents the attributes of the intermediate products processed by its corresponding process and its historical vector with downstream nodes. The sample tag set contains several sample tags, which correspond one-to-one with the sample nodes. Each sample tag represents the yield rate of the output of its corresponding sample node.

[0016] The target LSTM model is obtained by training the sample sequence and the sample label set.

[0017] In an optional embodiment of this specification, the method further includes:

[0018] The process node whose predicted yield is lower than the preset yield threshold is designated as the second target node;

[0019] Highlight the second target node.

[0020] In an optional embodiment of this specification, the method further includes:

[0021] For each process node, a corresponding target downstream node is determined; the target downstream node is the process node located downstream of the process node and that produces the intermediate product with the largest added value in the downstream process node.

[0022] When the process node is determined to be the second target node, information about its corresponding target downstream node is added to its hidden layer.

[0023] In an optional embodiment of this specification, the method further includes:

[0024] For each process node, a corresponding target upstream node is determined; the target upstream node is the process node located upstream of the process node and has the greatest impact on the yield of the intermediate products produced by the process node.

[0025] When the process node is determined to be the second target node, information about its corresponding target upstream node is added to its hidden layer.

[0026] In an optional embodiment of this specification, the method further includes:

[0027] When at least two adjacent process nodes are determined to be the second target node, the process node located upstream of and adjacent to it is also regarded as the target upstream node.

[0028] In an optional embodiment of this specification, the method further includes:

[0029] The yield threshold is negatively correlated with the value of the predicted yield corresponding to the target product.

[0030] In an optional embodiment of this specification, the method further includes:

[0031] For each of the aforementioned process nodes, the yield threshold corresponding to the intermediate product value-added of the adjacent downstream process node is negative, which is greater than the yield threshold corresponding to the intermediate product value-added of the adjacent downstream process node is non-negative.

[0032] In an optional embodiment of this specification, the method further includes:

[0033] In two adjacent process nodes, if the replacement impact index represented by their multidimensional vector is less than the preset replacement impact index threshold, and the predicted yield of the downstream process node is less than the average yield of the historical products of the target industry chain at that process node, then the downstream process node is also regarded as the second target node.

[0034] In an optional embodiment of this specification, the method further includes:

[0035] The attribute data collected during the production of historical products also includes event history attribute data; the event history attribute data is used to characterize historical events that occurred during the production of historical products, representing abnormal equipment status and abnormal operating conditions.

[0036] The target input was constructed using event attribute data of events that would cause the target industry chain to restart or be adjusted.

[0037] Secondly, embodiments of this application also provide a visualization display device for drawing a supply chain quality map, the device being used to implement the method steps in the first aspect.

[0038] Thirdly, embodiments of this application also provide an electronic device, including:

[0039] Processor; and

[0040] A memory configured to store computer-executable instructions, which, when executed, cause the processor to perform the steps of the method described in the first aspect.

[0041] Fourthly, embodiments of this application also provide a computer-readable storage medium storing one or more programs that, when executed by an electronic device including multiple applications, cause the electronic device to perform the steps of the method described in the first aspect.

[0042] The above-described technical solutions adopted in the embodiments of this application can achieve the following beneficial effects:

[0043] This application provides a visualization method for drawing a quality map of a supply chain. Based on the execution order of each process in the target supply chain, each process node is visualized to obtain a usable map. Then, multi-dimensional vectors are used to quantify factors such as the connection between adjacent process nodes and the negative impact on intermediate products. Next, a model trained based on multi-dimensional vectors is used to predict yield based on the impact of the target supply chain on intermediate products. This prediction result is then visualized to provide conditions for subsequent visualization. The method provided in this application utilizes electronic digital data processing technology to quantify the situation of the target supply chain and display the prediction results. Attached Figure Description

[0044] Figure 1 A schematic diagram illustrating the process of a visualization method for creating a supply chain quality map, as provided in the embodiments of this specification.

[0045] Figure 2 This is a schematic diagram of the structure of an electronic device in an embodiment of this specification. Detailed Implementation

[0046] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Similar elements in different embodiments are referred to by related similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the present application. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to the present application are not shown or described in the specification. This is to avoid obscuring the core parts of the present application with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0047] Furthermore, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for the clear description of a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed.

[0048] The serial numbers assigned to components in this document, such as "first" and "second," are used only to distinguish the described objects and have no sequential or technical meaning. The terms "connection" and "linkage" used in this application, unless otherwise specified, include both direct and indirect connections (linkages).

[0049] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0050] like Figure 1 As shown, the visualization method for drawing the supply chain quality map in this specification includes the following steps:

[0051] S100: Map the target industry chain to obtain a usable map.

[0052] The methods described in this manual are applicable to predictive data processing prior to supply chain restarts or adjustments (e.g., equipment in a specific process step undergoes maintenance, a previously suspended process step resumes operation, or equipment malfunctions due to suspension; or abnormal operating conditions, such as power outages). Supply chain restarts or adjustments often result in at least some equipment restarting, equipment not yet reaching stable operation, or changes in production plans, all of which can lead to instability.

[0053] The technical methods described in this specification can be applied to various industry chains and can be adaptively adjusted according to the different attributes of different industry chains. The industry chain to be examined in this specification is the target industry chain. The models in this specification correspond one-to-one with industry chains.

[0054] The available diagrams in this specification can be considered as a prototype for visual representation. The available diagrams contain process nodes (exemplarily represented by the origin) arranged in the order of the processes used to produce the product in the target supply chain, with each process node corresponding one-to-one with a specific process.

[0055] The relationship between two adjacent process nodes is represented by a multi-dimensional vector (for example, the connection between processes can be represented by lines, making the configuration of process nodes more adjustable; these lines can be used to graphically represent the multi-dimensional vector). This multi-dimensional vector characterizes the transfer time between the processes corresponding to the two process nodes under normal operation of the target industrial chain, and the negative impact index of the transfer on intermediate products (discontinuous production lines can have a negative impact on intermediate products, such as over-fermentation caused by the transfer of bread dough during bread making; this can be quantified based on expert experience). For example, the value ranges from 0 to 1, with 1 representing the greatest negative impact. The negative impact index is not only related to the situation of the industrial chain, but also to the loss-bearing capacity of the producer and the quality requirements of the target product. The greater the loss-bearing capacity, the smaller the value. The higher the quality requirements of the target product, the higher the value. The index also includes the difficulty index between processes (when processes are seamlessly connected, the difficulty index is low. If transshipment, intermediate processing, or intermediate storage is required, the difficulty index is high). For example, the value ranges from 0 to 1, with 1 representing the greatest difficulty. In the bread-making process, the divided dough is sent to the dough-forming equipment via an assembly line; this process has good continuity. The difficulty is relatively low. After the bread dough has fermented, it needs to be wrapped in a protective film to prevent dust contamination, and then transported manually to the baking workshop. This process is not continuous and is quite difficult. The added value of intermediate products (the value of intermediate products can be quantified through expert experience. Different process nodes produce different intermediate products. For example, the adjacent flour unpacking process node and flour sieving process node. Sifted flour tends to have a higher value than unpacked flour, i.e., positive, but the degree is limited. The adjacent flour sieving process node and dough kneading process node, because kneaded dough cannot be stored and has limited uses, tend to have a lower value, i.e., negative. In related technologies, used for...) The technical means for measuring value, where conditions permit, are applicable to this specification. The replacement impact index (where "impact" refers to a negative impact; this index can be determined based on expert experience. A larger value indicates a greater negative impact. For example, in adjacent fermentation and baking processes, the baking process is downstream. If a problem occurs in the baking process and it needs to be replaced, the fermented bread dough will be difficult to preserve, and longer intermediate transportation will also cause the bread dough to spoil; therefore, the index value will be larger. If there is no replaceable baking process node in the target supply chain, the index value will be even larger.)

[0056] In other words, a multidimensional vector represents the state of the target industrial chain during continuous and stable production. In this state, the intermediate products and the target product produced at each stage of the industrial chain meet expectations. The design of the industrial chain plays a more decisive role in determining the multidimensional vector.

[0057] S102: Based on the attribute data of the target product to be produced using the target industrial chain and the bias vector, construct the target input, input it into the pre-trained target LSTM model, and use the output as the predicted yield.

[0058] The attribute data of the target product is used to characterize the minimum requirements of the producer for the target product. For example, the amount of raw materials that can be used, the number of intermediate products that can be scrapped at each process node (which may be scrapped due to product quality errors), the parameters of the final target product (for bread, parameters may be aeration rate, moisture content, etc.), and the quantity of the target product obtained. The attribute data of other products (such as historical products) can also include the above dimensions.

[0059] The bias vector represents the difference between the current state of the target supply chain and the multidimensional vector. Taking bread as an example, the increased difficulty of transferring fermented bread dough will result in a higher difficulty index for the connection between processes in the bias vector compared to the multidimensional vector. In other words, the bias vector represents the difference between the restart or adjustment of the supply chain at this time and its historical stable production state. The bias vector and subsequent historical vectors focus more on the actual situation deviating from the design.

[0060] The predicted yield represents the predicted yield of intermediate products at each of the process nodes and the yield of the target product. If the intermediate product yield is low, and defective intermediate products cannot be scrapped in time, the target product yield will be low, and the production of the target product will consume more resources, resulting in greater losses.

[0061] The target LSTM (Long Short-Term Memory) model in this specification is trained using the aforementioned multidimensional vectors. In an optional embodiment of this specification, the target LSTM model is trained through the following steps.

[0062] Based on attribute data collected during the production of historical products in the target industrial chain (historical products may differ from target products in terms of output, raw materials, equipment status, etc.), and using the attribute data collected during production (the attribute data of historical products corresponds to the dimensions of the aforementioned attribute data of target products. In other optional embodiments, the attribute data may also include others, which will be explained later), a sample sequence and its corresponding sample label set are constructed. Each sample sequence contains several sample nodes arranged in the order of the process in which historical products were produced in the target industrial chain. Each sample node corresponds one-to-one with a process, and each sample node represents the attributes of the intermediate product processed by its corresponding process (e.g., fermentation process node, humidity, weight, allowable error, etc. of fermented dough) and its historical vector with downstream nodes. The sample label set contains several sample labels, which correspond one-to-one with the sample nodes, and each sample label represents the yield of the output of its corresponding sample node. An LSTM model is trained using the sample sequence and the sample label set to obtain the target LSTM model.

[0063] The historical vector represents the historical situation corresponding to the multidimensional vector. In some cases, the historical vector may be the same as the multidimensional vector. If there has been a deviation from the design of the industry chain in the past, this deviation can be reflected in the historical vector.

[0064] The LSTM model adopts the following natural laws:

[0065] 1. Forgetting Mechanism: LSTM simulates the forgetting process of human memory through the forgetting gate mechanism. The forgetting gate uses the sigmoid function to determine which information needs to be retained and which needs to be discarded, which is similar to the natural forgetting process of information in the human brain.

[0066] 2. Memory Mechanism: The cell states of LSTM act like an information conveyor belt, running throughout the entire time series. It maintains relatively stable information transmission across different time steps, enabling the network to remember information over long distances. This simulates the long-term storage and transmission of information in human memory.

[0067] 3. Learning Mechanism: The input and output gates control the writing of new information and the output of old information, simulating the process of humans learning new knowledge and recalling old knowledge. The input gate determines which new information needs to be added to the cell state, while the output gate determines which information needs to be output as the current hidden state.

[0068] The basic principles of the LSTM model include the following aspects:

[0069] 1. Cell state: The core of LSTM is the cell state, which runs through the entire time series and can maintain relatively stable information transmission at each time step of the sequence, thus enabling the network to remember information over long distances.

[0070] 2. Gating Mechanism: LSTM controls the updating and transmission of information through forget gates, input gates, and output gates. The forget gate determines which information needs to be retained and which needs to be discarded; the input gate controls how much information from the current input will be updated into the cell state; and the output gate determines which information from the cell state will be output as the hidden state at the current moment.

[0071] S104: Add the predicted yield to the hidden layer of the corresponding process node in the available graph.

[0072] In other words, each process node corresponds to its own hidden layer. A hidden layer is a layer that is not visible when not selected. Hiding information to a certain extent helps maintain the simplicity of the diagram.

[0073] S106: When a user operation is detected, the selected process node is taken as the first target node.

[0074] In related technologies, any techniques that enable selection based on user operation are applicable to this specification, where conditions permit. Furthermore, in related technologies, any techniques that enable deselection based on user operation are also applicable to this specification, where conditions permit.

[0075] S108: Display the portion of the attribute data of the target product corresponding to the first target node, and display the hidden layer corresponding to the first target node.

[0076] This application provides a visualization method for drawing a quality map of a supply chain. Based on the execution order of each process in the target supply chain, each process node is visualized to obtain a usable map. Then, multi-dimensional vectors are used to quantify factors such as the connection between adjacent process nodes and the negative impact on intermediate products. Next, a model trained based on multi-dimensional vectors is used to predict yield based on the impact of the target supply chain on intermediate products. This prediction result is then visualized to provide conditions for subsequent visualization. The method provided in this application utilizes electronic digital data processing technology to quantify the situation of the target supply chain and display the prediction results.

[0077] In a further optional embodiment of this specification, during model training, in addition to examining the deviation between the historical production process and the ideal state during supply chain design, unexpected events occurring during production are also included in the scope of examination. In this embodiment, the attribute data collected during the production of historical products also includes event history attribute data; the event history attribute data is used to characterize historical events that occurred during the production of historical products, representing abnormal equipment status (e.g., equipment failure, maintenance, etc.) and abnormal operating conditions (e.g., power outages, etc.). Event attribute data of events that led to the restart or adjustment of the target supply chain are also used when constructing the target input. Therefore, the model can learn the impact of unexpected events on the supply chain, and the output prediction results can be more objective and accurate.

[0078] Even after maintenance or adjustments, events that cause the target supply chain to restart or undergo adjustments may not be completely eliminated. Combined with the production and material conditions at the time, this could lead to low product yield. To avoid blindly resuming supply chain operations, in an optional embodiment of this specification, the process node with a predicted yield lower than a preset yield threshold is designated as a second target node. This second target node is highlighted to inform the user of potential risks at that process node, allowing the user to decide whether to further adjust the target supply chain or take other measures.

[0079] The yield threshold can be a value based on expert experience. In an optional embodiment of this specification, the yield threshold is negatively correlated with the predicted yield value corresponding to the target product, so as to increase the stringency of inspection of intermediate products when the yield of the target product is low, so as to detect problems as early as possible.

[0080] In some cases, the added value of intermediate products produced at a certain process stage may be negative. In this case, even if the yield of upstream output is not very high, the impact on the total value of the total output will not be significant. Therefore, in an optional embodiment of this specification, for each process node, the yield threshold corresponding to the case where the added value of intermediate products at the adjacent downstream process node is negative is greater than the yield threshold corresponding to the case where the added value of intermediate products at the adjacent downstream process node is non-negative.

[0081] Optionally, in two adjacent process nodes, if the multidimensional vector of the replacement impact index is less than the preset replacement impact index threshold (an empirical value that is positively correlated with the tolerability of increased production costs), and the predicted yield of the downstream process node is less than the average yield of the historical products of the target industry chain at that process node, then the downstream process node is also taken as the second target node, so as to discover as many potential quality-affecting risks in the industry chain as possible.

[0082] All technical means that can be used to achieve highlighting in related technologies are applicable to this specification, where conditions permit.

[0083] In a further optional embodiment of this specification, the prediction of the downstream transmission of potential hazards is also achieved through technical means. In this embodiment, for each process node, a corresponding target downstream node is determined; the target downstream node is the process node located downstream of the process node, and the process node whose intermediate products produced downstream of the process node have the largest added value compared to the output of the process node. When a process node is determined to be the second target node, information of its corresponding target downstream node is added to its hidden layer. The information of the target downstream node includes, but is not limited to, the predicted yield, and may also include information such as the expected output. If the yield of the second target node is too low, the qualified output of the downstream products will also decrease. When the product reaches the target downstream node, the reduction in the total value of its total output will be more significant, which will lead to increased costs and reduced profits, and it is necessary to inform the producer.

[0084] In addition to examining the downward transmission of potential problems, an optional embodiment of this specification also traces the source of the problems. Specifically, for each process node, its corresponding target upstream node is determined. The target upstream node is the process node located upstream of the process node and has the greatest impact on the yield of the intermediate products produced by the process node (e.g., in a bread-making supply chain. During the trial operation of the supply chain, it was found through experiments that the fermentation process has the greatest impact on the yield of the baking process, so the fermentation process node is the target upstream node of the baking process node. Since the first process node in the supply chain does not have an upstream node, it is not necessary to determine its upstream process node). When a process node is determined to be the second target node, information about its corresponding target upstream node is added to its hidden layer to indicate the possible source of the problem.

[0085] Optionally, when at least two adjacent process nodes are determined to be the second target node, the process node located upstream of and adjacent to it is also considered as the target upstream node to further indicate the possible source of the problem.

[0086] In a further optional embodiment of this specification, for each process node, a corresponding target upstream node is determined; the target upstream node is the process node located upstream of the process node and has the greatest impact on the yield of the intermediate products produced by the process node; when the corresponding target upstream node is determined to be the second target node, the information of the process node is displayed to express the transmission of potential problems.

[0087] Figure 2This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Please refer to it. Figure 2 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the electronic device may also include other hardware required for other business operations.

[0088] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 2 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0089] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0090] The processor reads the corresponding computer program from non-volatile memory into memory and then runs it, forming a visualization display device for drawing a supply chain quality map at the logical level. The processor executes the program stored in memory and is specifically used to execute any of the aforementioned visualization display methods for drawing a supply chain quality map.

[0091] The above is as stated in this application. Figure 1The visualization method for drawing a supply chain quality map disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0092] The electronic device can also perform Figure 1 A visualization method for drawing a quality map of the industrial chain is proposed, and its implementation is achieved. Figure 1 The functions of the embodiments shown are not described in detail here.

[0093] This application also proposes a computer-readable storage medium that stores one or more programs, the programs including instructions that, when executed by an electronic device including multiple applications, perform any of the aforementioned visualization methods for drawing supply chain quality maps.

[0094] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0095] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0096] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0097] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0098] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0099] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0100] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, 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, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0101] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0102] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0103] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An industrial chain quality mapping visual display method, characterized in that, The method includes: The target industry chain is graphically processed to obtain a usable graph. The usable graph includes process nodes that represent the order of processes in which the product is produced in the target industry chain, and each process node corresponds one-to-one with a process. Two adjacent process nodes are associated with each other through a multi-dimensional vector. The multi-dimensional vector is used to represent the transfer time between the processes corresponding to the two process nodes, the negative impact index of the transfer on intermediate products, the difficulty index of the connection between processes, the added value of intermediate products, and the replacement impact index of the downstream process node in the two adjacent process nodes when the target industry chain is operating normally. When the restart or adjustment of the target supply chain is detected, the target input, constructed based on the attribute data and bias vector of the target product to be produced using the target supply chain, is input into a pre-trained target LSTM model, and the output is used as the predicted yield. The predicted yield represents the yield of the intermediate products at each of the process nodes. The target LSTM model is trained using historical vectors. The bias vector represents the difference between the current situation of the target supply chain and the multi-dimensional vector. The historical vector is a representation of the historical situation corresponding to the multi-dimensional vector. Add the predicted yield to the hidden layer of the corresponding process node in the available graph; When a user operation is detected, the selected process node is designated as the first target node. Display the portion of the attribute data of the target product that corresponds to the first target node, and display the hidden layer corresponding to the first target node.

2. The method of claim 1, wherein, The target LSTM model is obtained using the following steps: Based on the attribute data collected during the production of historical products in the target industrial chain, a sample sequence and its corresponding sample tag set are constructed. Each sample sequence contains several sample nodes arranged in the order of the processes in which historical products were produced in the target industrial chain. Each sample node corresponds one-to-one with a process, and each sample node represents the attributes of the intermediate products processed by its corresponding process and its historical vector with downstream nodes. The sample tag set contains several sample tags, which correspond one-to-one with the sample nodes. Each sample tag represents the yield rate of the output of its corresponding sample node. The target LSTM model is obtained by training the sample sequence and the sample label set.

3. The method as described in claim 1, characterized in that, The method further includes: The process node whose predicted yield is lower than the preset yield threshold is designated as the second target node; Highlight the second target node.

4. The method as described in claim 3, characterized in that, The method further includes: For each process node, a corresponding target downstream node is determined; the target downstream node is the process node located downstream of the process node and that produces the intermediate product with the largest added value in the downstream process node. When the process node is determined to be the second target node, information about its corresponding target downstream node is added to its hidden layer.

5. The method as described in claim 3, characterized in that, The method further includes: For each process node, a corresponding target upstream node is determined; the target upstream node is the process node located upstream of the process node and has the greatest impact on the yield of the intermediate products produced by the process node. When the process node is determined to be the second target node, information about its corresponding target upstream node is added to its hidden layer.

6. The method as described in claim 5, characterized in that, The method further includes: When at least two adjacent process nodes are determined to be the second target node, the process node located upstream of and adjacent to it is also regarded as the target upstream node.

7. The method as described in claim 3, characterized in that, The method further includes: The yield threshold is negatively correlated with the value of the predicted yield corresponding to the target product.

8. The method as described in claim 3, characterized in that, The method further includes: For each of the aforementioned process nodes, the yield threshold corresponding to the intermediate product value-added of the adjacent downstream process node is negative, which is greater than the yield threshold corresponding to the intermediate product value-added of the adjacent downstream process node is non-negative.

9. The method as described in claim 3, characterized in that, The method further includes: In two adjacent process nodes, if the replacement impact index represented by their multidimensional vector is less than the preset replacement impact index threshold, and the predicted yield of the downstream process node is less than the average yield of the historical products of the target industry chain at that process node, then the downstream process node is also regarded as the second target node.

10. The method as described in claim 2, characterized in that, The method further includes: The attribute data collected during the production of historical products also includes event history attribute data; the event history attribute data is used to characterize historical events that occurred during the production of historical products, representing abnormal equipment status and abnormal operating conditions. The target input was constructed using event attribute data of events that would cause the target industry chain to restart or be adjusted.

Images