A method for generating the BOM of a circuit breaker product

Abstract

The present invention discloses a method for generating a BOM of a circuit breaker product, which includes the steps of: defining and constructing a circuit breaker product library, a circuit breaker component library, and a circuit breaker device library; associating keywords and constructing text data; respectively constructing a circuit breaker product component recommendation model and a circuit breaker product device recommendation model based on the FastText algorithm, and obtaining a trained component recommendation model and a device recommendation model after training; inputting the required component and device keywords into the model to intelligently recommend multiple relevant components and devices, and then selecting appropriate components and devices; updating the text data; updating the circuit breaker product library, the component library, and the device library. The present invention not only makes full use of the BOM of existing model circuit breakers to quickly and accurately construct a new circuit breaker product model or a BOM with complex components, but also can continuously expand and establish the product model library, component library, and period library required for the BOM, effectively reducing the production cost caused by inaccurate or incorrect manual creation of the BOM.

Description

Technical Field

[0001] The present invention belongs to the technical fields of industrial engineering and production management, and particularly relates to a method for generating a BOM of a circuit breaker product. Background Art

[0002] Driven by the production paradigm of the new generation of intelligent manufacturing, using intelligent means to assist or even replace manual labor is a new model to improve manufacturing efficiency and reduce the number of misoperations. Circuit breaker products have many product models and complex components. Different models of circuit breakers require a variety of components and devices, and there are also differences in the specifications and qualities of the same device. Manually formulating the BOM of a circuit breaker according to an order is slow and error-prone. Especially when encountering a circuit breaker involving non-standard and devices of the same specification but different qualities, the inaccuracy of the BOM will lead to an increase in production costs. Summary of the Invention

[0003] In order to overcome the disadvantages of time-consuming and error-prone in the process of manually generating a BOM, the present invention provides a method for generating a BOM of a circuit breaker product. When an order contains a circuit breaker with a new model or complex components, after determining the category and model of the circuit breaker, multiple suitable components and devices can be recommended from the component library and the device library successively by means of searching and matching component and device keywords. The devices are used to assemble the circuit breaker components, and the components are finally assembled into finished products. And when generating the BOM of the circuit breaker product, the product library, the product component library, and the product device library in the system are continuously enriched, so as to more conveniently generate the product BOM.

[0004] The technical solutions adopted by the present invention to solve the above technical problems are as follows:

[0005] A method for generating a BOM of a circuit breaker product, comprising the following steps:

[0006] Step 1: Define and construct a circuit breaker product library, a circuit breaker component library, and a circuit breaker device library;

[0007] Step 2: Associate each component in the component library with multiple component keywords, and store the data with the component keywords as attribute data and the component names as label data as component text data;

[0008] Associate each device in the device library with multiple device keywords, and store the data with the device keywords as attribute data and the device names as label data as device text data;

[0009] Step 3: Respectively construct a circuit breaker product component recommendation model and a circuit breaker product device recommendation model based on the FastText algorithm;

[0010] Tokenize the attribute data and add the __label__ prefix to the label data to obtain the processed text data. Use the text data to train the model to obtain a trained component recommendation model and a device recommendation model;

[0011] Step 4: Obtain the circuit breaker product information to be produced from the order, extract the main attribute parameters and secondary attribute parameters of the product, select or create the corresponding product category from the circuit breaker product library and then select the matching product model. This process is to sequentially search for the main attribute parameters and secondary attribute parameters in the product library until the parameter value of the circuit breaker model that appears last is found;

[0012] Based on the selected product model, determine the component information of the circuit breaker product. If it is necessary to update / supplement the components in the selected product, manually input the keywords of the components to be updated based on experience. After tokenizing the keywords, the component recommendation model recommends the matching components from the component library, and then manually select the most suitable one from them; if the component recommendation model fails to recommend, new components need to be created manually;

[0013] Based on the updated / supplemented components, determine the device information of the circuit breaker product. If it is necessary to update / supplement the devices in the selected product, manually input the keywords of the devices to be updated based on experience. After tokenizing the keywords, the device recommendation model recommends the matching devices from the device library, and then manually select the most suitable one from them; if the device recommendation model fails to recommend, select from the device library manually;

[0014] Finally, the components and devices form the three - layer circuit breaker product BOM;

[0015] Step 5: Save the input keywords and the selected components or devices to the corresponding text data and update the text data;

[0016] Step 6: Supplement the newly created product categories, components and devices to the corresponding libraries respectively, and update the circuit breaker product library, the circuit breaker component library and the circuit breaker device library.

[0017] The beneficial effects of a method for generating a circuit breaker product BOM proposed by the present invention are as follows:

[0018] A method for generating a BOM of a circuit breaker product according to the present invention divides the BOM into three layers: product model, component, and device in sequence, and constructs a circuit breaker model library, a circuit breaker component library, and a circuit breaker model library. When making the product BOM, it is associated with the BOM of the historical circuit breaker product model by means of sequential search of the main attribute and the secondary attribute. A text recommendation model is constructed by the FastText algorithm. After inputting the keywords of the required components and devices into the model, multiple relevant components and devices are intelligently recommended, and then appropriate components and devices are selected. This ensures the efficiency and accuracy when making the BOM. The present invention not only makes full use of the BOM of the existing model circuit breakers to quickly and accurately construct the BOM of the new circuit breaker product model or the BOM with complex components, but also can continuously expand the product model library, component library, and period library required for establishing the BOM, effectively reducing the production cost caused by inaccurate or incorrect manual creation of the BOM. Brief Description of the Drawings

[0019] Figure 1 It is a flowchart of a method for generating a BOM of a circuit breaker product according to the present invention;

[0020] Figure 2 It is a flowchart for generating a BOM of a circuit breaker product;

[0021] Figure 3 It is a three-layer BOM architecture diagram of the product, components, and devices of the present invention. Detailed Embodiments

[0022] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0023] The present invention discloses a method for generating a BOM of a circuit breaker product, which mainly includes the definition and association of a circuit breaker product library, a circuit breaker component library, and a circuit breaker device library, and the formulation process of a three-layer architecture BOM. The steps for generating the BOM are as follows: First, search for a suitable product model in the product library according to the parameters of the main / secondary attributes of the product or create a new product; secondly, use a text recommendation model to update the components in the product according to the component keywords. If the recommendation is unsuccessful, reconstruct / supplement the components in the product; finally, use the text recommendation model to update the devices in the product according to the device keywords. If the recommendation is unsuccessful, manually select from the device library. The present invention not only makes full use of the BOM of existing model circuit breakers, integrates an intelligent model into the BOM generation process, and quickly and accurately constructs a new circuit breaker product model or a BOM with complex components, but also can continuously expand the product model library, component library, and period library required for establishing the BOM, effectively reducing the production cost caused by inaccurate or incorrect manual creation of the BOM.

[0024] In one embodiment, as Figure 1 shown, the present invention provides a method for generating a BOM of a circuit breaker product, which mainly includes the definition and association of a circuit breaker product library, a circuit breaker component library, and a circuit breaker device library, and the formulation process of a three-layer architecture BOM. The specific implementation steps are as follows:

[0025] Step 1: Define and construct a circuit breaker product library, a circuit breaker component library, and a circuit breaker device library.

[0026] Define and construct a circuit breaker product library: The definition of the circuit breaker product library includes defining the circuit breaker category, circuit breaker name, main attribute of the circuit breaker, and secondary attribute of the circuit breaker. The circuit breaker category is composed of an enterprise code, a product code, and a design code. The name of the circuit breaker is further expanded on the basis of the circuit breaker category. The main attribute of the circuit breaker is the core parameter, which is the parameter that can realize the core function of the circuit breaker; the secondary attribute is the secondary parameter of the circuit breaker, and the purpose is to ensure the integrity of the circuit breaker. Finally, the defined circuit breaker category is stored in the circuit breaker product library, and each circuit breaker category contains multiple circuit breaker models. The circuit breaker in the customer order is to determine the specific category from the circuit breaker product library and then select the specific model of the circuit breaker.

[0027] The definition of each different circuit breaker model in the circuit breaker category is to combine the above-set main attribute and secondary attribute in a full permutation manner to form different models of circuit breakers. The maximum number of circuit breaker models under each category is Q, and the solution formula is shown in Equation (1), where P represents the number of possible main attributes and A represents the number of possible secondary attributes. These circuit breaker models are saved in units of categories.

[0028] Q = P! * A! (1)

[0029] The product models in the circuit breaker product library are stored in categories. For example, for the XXW1 series intelligent universal circuit breakers, XXW1 is the circuit breaker category, XX is the enterprise code, W is the product code, and 1 is the design code; the circuit breaker models under this category are further extended with the category number. For example, the main attribute is the frame current, and the parameter values are 1000A and 2000A. The first secondary attribute is the intelligent controller, and the parameters are type M and type H. The second secondary attribute is the installation method, and the parameters are draw-out type and fixed type. Eight models of circuit breakers are constructed in a full permutation manner according to the above attributes, as shown in Table 1.

[0030] Table 1 Example of circuit breaker models under the circuit breaker product category

[0031] Number Model Name 1 XXW1-1000A-M Type Intelligent Controller - Drawer Mounting Method 2 XXW1-2000A-M Type Intelligent Controller - Drawer Mounting Method 3 XXW1-1000A-H Type Intelligent Controller - Drawer Mounting Method 4 XXW1-2000A-H Type Intelligent Controller - Drawer Mounting Method 5 XXW1-1000A-M Type Intelligent Controller - Fixed Mounting Method 6 XXW1-2000A-M Type Intelligent Controller - Fixed Mounting Method 7 XXW1-1000A-H Type Intelligent Controller - Fixed Mounting Method 8 XXW1-2000A-H Type Intelligent Controller - Fixed Mounting Method

[0032] Define and construct the circuit breaker component library: Store it in units of components. The specific information of each component in the circuit breaker component library includes the number, name, and all the components that make it up. The component number is the component creation date and the creation number within that date. For example, 20221210-001, 20221210 is the component serial number, and 001 is the first component created within that date. The name is the regular naming of the component by the company, such as the dual power supply transfer switch. The components that make it up are the components included in each component when making the circuit breaker BOM. If the component is purchased, there is no corresponding device.

[0033] Define and construct the circuit breaker device library: Store it in units of devices. The information of each device in the circuit breaker device library includes the number, name, storage location, and supplier manufacturer. The rule for the number is the warehousing date and the part batch serial number. One number allows multiple devices to correspond to it, and these multiple devices are devices of the same specification. For example, 20221210-1000517122, 20221210 is the warehousing date, and 1000517122 is the batch number of multiple devices. The name is the product name provided by the device supplier, such as power switch\ESC-240-27\AC220V\DC27V\8A, where power switch is the device name, and ESC-240-27\AC220V\DC27V\8A is the device specification. The storage location is in the raw material placement area of the circuit breaker production workshop, such as A-10-3, indicating the 3rd grid of the 10th shelf in storage area A. The supplier manufacturer is the name of the manufacturer that provides the above devices, such as XXXX Company.

[0034] Step 2: Association between component keywords and components and association between device keywords and devices.

[0035] After the above circuit breaker product library, circuit breaker component library, and circuit breaker device library are defined and constructed, each component in the component library is associated with multiple component keywords, and each device in the device library is associated with multiple device keywords. For example, the keywords corresponding to the device power switch \ESC-240-27\AC220V\DC27V\8A are power switch, 8A, and quality grade A.

[0036] The data with component keywords as attribute data and component names as label data is stored as component text data, and the data with device keywords as attribute data and device names as label data is stored as device text data.

[0037] Step 3: Construct a circuit breaker product component recommendation model and a circuit breaker product device recommendation model based on the FastText algorithm respectively.

[0038] Perform word segmentation on the attribute data one by one. The word segmentation tool is the lcut word segmentation function in the jieba library, a third-party library of the Python language; add a __label__ prefix to the label data, and save the processed text data in the format of a.txt file. The text data set is divided into a training set, a test set, and a validation set in a ratio of 6:2:2. The training set and the test set are used to train the circuit breaker product component recommendation model and the circuit breaker product device recommendation model (collectively referred to as the text recommendation model), and the validation set is used to verify the performance of the text recommendation model. The number of devices or components recommended by the text recommendation model is set to 5 or other quantities. The construction of the text recommendation model is to train two independent multi-classification models, one is a component recommendation model, and the other is a device recommendation model. The two models are trained using the FastText algorithm. Before inputting the text data into the algorithm, word segmentation is first performed. The data used to train the model is the text data composed of historical components and their keywords and the text data composed of devices and their keywords. The segmented keywords and the processed component / device names are input into the FastText algorithm, and the optimal parameters of the algorithm for training the two models are searched respectively using the grid search method, and then the component / device recommendation model is trained. Manually input the keywords of components and devices to recommend multiple matching components and devices in the component library and the device library respectively.

[0039] The trained component recommendation model is embedded in the process of determining product components in the BOM generation method, and the trained device recommendation model is embedded in the process of determining product devices in the BOM generation method.

[0040] Step 4: Establish a three-layer BOM for circuit breaker products, circuit breaker components, and circuit breaker devices. The formulation process of the three-layer architecture BOM is to formulate a BOM that sequentially includes all information of circuit breaker models, components, and devices. The formulation rule is to change or update the components and devices of the product on the basis of the original BOM.

[0041] After constructing and associating the circuit breaker product library, circuit breaker component library, circuit breaker device library, and completing the training of the component / device text recommendation model, the BOM generation process of the circuit breaker product is carried out. The flowchart of the circuit breaker product BOM generation is as Figure 2 shown.

[0042] After obtaining the products to be produced from the order, obtain the information of the circuit breaker products to be produced, extract the main attribute parameters and secondary attribute parameters of the products, and select or create the corresponding product categories from the circuit breaker product library and then select the matching product models. This process is to search for the main attribute parameters and secondary attribute parameters from the product library one by one until the parameter value of the circuit breaker model that appears last is searched.

[0043] According to the selected product model, determine the component information of the circuit breaker product. If it is necessary to update / supplement the components in the selected product, manually input the keywords of the components to be updated based on experience. After segmenting the keywords, recommend 5 or fewer matching components from the component library, and manually select the most suitable one from them. If the recommendation is unsuccessful, it is necessary to manually create new components.

[0044] According to the updated / supplemented components, determine the device information of the circuit breaker product. If it is necessary to update / supplement the devices in the selected product, manually input the keywords of the devices to be updated based on experience. After segmenting the keywords, recommend 5 or fewer matching devices from the device library, and manually select the most suitable one from them. If the recommendation is unsuccessful, it is necessary to select from the device library manually.

[0045] Finally, the components and devices form the three-layer circuit breaker product BOM. The BOM structure is as Figure 3 shown, containing all the information of the product model, components, and devices.

[0046] Step Five: Update the keyword and component / device matching text data.

[0047] Because when inputting component / device keywords to judge whether suitable components / devices can be recommended is based on manual experience, different people may use different keywords when using the BOM generation method. Therefore, the same component / device will correspond to different keywords. Therefore, each time the BOM is generated, the input keywords and the selected components / devices are saved to the corresponding component / device text data.

[0048] Step Six: Update the circuit breaker product library, circuit breaker component library, and circuit breaker device library.

[0049] During the process of generating the product BOM, the newly created product categories, product components, and product devices are respectively supplemented into the corresponding libraries, enriching the types and quantities of BOMs included in the system, so that the entire BOM generation method has the advantages of iterative optimization and continuous feedback.

[0050] Further, in Steps 4 to 6, when the customer's order arrives, the circuit breaker models included in the order are manually disassembled. If there is a circuit breaker model that meets the requirements in the system, the circuit breaker model is directly applied, or the components and devices included therein are changed based on this model and changed to another new circuit breaker model under this category. After the change, a new number is automatically formed and saved separately. When determining whether there is a similar type of circuit breaker model, it is searched according to the parameters in the circuit breaker in the order, and the keyword search is performed in the order of the main and secondary attributes until the parameter value of the circuit breaker model that appears last is searched.

[0051] If a suitable circuit breaker model is not searched in the first input of the main attribute as described above, after determining the main and secondary attributes of the circuit breaker in the order, a new circuit breaker product category is created, saved in the product library in units of categories, and then a suitable circuit breaker model is selected from this category.

[0052] When a suitable circuit breaker model is selected from the circuit breaker model library as described above, if components need to be changed, enter the keywords of the components to be changed. These keywords will be automatically input into the trained component recommendation model after the automatic word segmentation operation in the system, and then the system recommends multiple components, and the most suitable component is manually selected from them.

[0053] If the component recommendation model recommends a suitable component, directly use the selected component or change the corresponding device based on the selected component. The changed component is stored according to the storage rules of the component library, and the changed component and its keywords are stored in the text data for training the component recommendation model.

[0054] If the component recommendation model does not recommend a suitable component, a new component is created, and the required devices are added according to the requirements. The new component is stored in the circuit breaker component library. The newly created component and its keywords are stored in the text data for training the component recommendation model.

[0055] For the change and addition of the above devices, keywords are also input according to the characteristics of the products in the customer order, and then these keywords are input into the trained device recommendation model after the word segmentation operation. The devices recommended by the model are manually selected. If the recommended devices are not suitable, they are manually re-screened from the device library, or new devices are created. The devices and their keywords selected manually or newly created are stored in the text data for training the device recommendation model.

[0056] The keywords input for recommending components and devices using the recommendation model are determined manually based on experience and knowledge according to the product model used and the characteristics of the components. After recommending the matching components and devices using the input keywords, the keywords are stored as attribute data and the component names are stored as label data. Therefore, the keywords corresponding to the same component or device may vary depending on the experience of different people.

[0057] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0058] The above-described embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the invention patent should be subject to the appended claims.

Claims

1. A method for generating a BOM of a circuit breaker product, characterized in that, it includes the following steps: Step 1: Define and construct a circuit breaker product library, a circuit breaker component library, and a circuit breaker device library; Step 2: Associate each component in the component library with multiple component keywords, and store the data with component keywords as attribute data and component names as label data as component text data; Associate each device in the device library with multiple device keywords, and store the data with device keywords as attribute data and device names as label data as device text data; Step 3: Based on the FastText algorithm, construct a circuit breaker product component recommendation model and a circuit breaker product device recommendation model respectively; Segment the attribute data and add the __label__ prefix to the label data to obtain the processed text data, and use the text data to train the model to obtain the trained component recommendation model and device recommendation model; Step 4: Obtain the circuit breaker product information to be produced from the order, and extract the main attribute parameters and secondary attribute parameters of the product. Select or create the corresponding product category from the circuit breaker product library and then select the matching product model. This process is to search for the main attribute parameters and secondary attribute parameters successively from the product library until the parameter value of the circuit breaker model that appears last is searched; According to the selected product model, determine the component information of the circuit breaker product. If it is necessary to update / supplement the components in the selected product, manually input the keywords of the components to be updated based on experience. After segmenting the keywords, the component recommendation model recommends the matching components from the component library, and manually select the most suitable one from them; if the component recommendation model fails to recommend successfully, it is necessary to manually create new components; According to the updated / supplemented components, determine the device information of the circuit breaker product. If it is necessary to update / supplement the devices in the selected product, manually input the keywords of the devices to be updated based on experience. After segmenting the keywords, the device recommendation model recommends the matching devices from the device library, and manually select the most suitable one from them; if the device recommendation model fails to recommend successfully, it is necessary to select from the device library manually; Finally, the components and devices form a three-layer BOM of the circuit breaker product; Step 5: Save the input keywords and the selected components or devices into the corresponding text data and update the text data; Step 6: Supplement the newly created product categories, components, and devices into the corresponding libraries respectively, and update the circuit breaker product library, the circuit breaker component library, and the circuit breaker device library.

2. The method for generating a BOM of a circuit breaker product according to claim 1, characterized in that, when segmenting the keywords, use the lcut segmentation function in the jieba library of the third-party library of the python language for segmentation.

3. The method for generating a BOM of a circuit breaker product according to claim 1, characterized in that, The product models in the circuit breaker product library are stored in units of categories. It is defined that the circuit breaker product library includes the definition of circuit breaker categories, circuit breaker names, main attributes of the circuit breaker, and secondary attributes of the circuit breaker. Among them, the circuit breaker category consists of an enterprise code, a product code, and a design code. The circuit breaker name is further extended based on the circuit breaker category. The main attribute of the circuit breaker is the core parameter, which is the parameter that can realize the core function of the circuit breaker. The secondary attribute is the secondary parameter of the circuit breaker, and the purpose is to ensure the integrity of the circuit breaker. Finally, the defined circuit breaker categories are stored in the circuit breaker product library, and each circuit breaker category contains multiple circuit breaker models.

4. A method for generating a BOM of a circuit breaker product according to claim 1, characterized in that, the specific information of each component in the circuit breaker component library includes a number, a name, and all the components it consists of. The number is the component creation date and the number created within that date. The name is a regular naming of the component. The components it consists of are the components included in each component when making the circuit breaker BOM.

5. A method for generating a BOM of a circuit breaker product according to claim 1, characterized in that, the information of each device in the circuit breaker device library includes a number, a name, a storage location, and a supplier manufacturer. The rule of the number is the warehousing date and the part batch serial number. One number allows multiple devices to correspond to it, and these multiple devices are devices of the same specification. The name is the product name provided by the device supplier. The storage location is in the raw material placement area of the circuit breaker production workshop. There are multiple shelves in the placement area, and the storage location is the specific shelf number where the parts are stored. The supplier manufacturer is the name of the manufacturer that provides the device.

Images