A precise welding and assembly method for an intelligent robot

Abstract

The present invention discloses a precision welding and assembly method for an intelligent manipulator, which relates to the field of product welding and assembly, wherein the method comprises: obtaining G product parts of a target manufactured product; based on an intelligent assembly submodule, making assembly decisions on the G product parts according to a product design plan to obtain a parts assembly decision; based on the parts assembly decision, assembling the G product parts according to an assembly manipulator to obtain a product parts assembly; based on an intelligent welding submodule, making welding decisions on the product parts assembly according to a product design plan to generate a product welding decision; based on the product welding decision, welding the product parts assembly according to a welding manipulator. The method solves the technical problem of low welding and assembly accuracy of products in the prior art, which leads to poor welding and assembly quality of products. The method achieves the technical effect of improving the welding and assembly accuracy of products and enhancing the welding and assembly quality of products.

Description

Technical Field

[0001] The present invention relates to the field of product welding and assembly, and in particular to a precise welding and assembly method of an intelligent robot arm. Background Art

[0002] Welding assembly plays a crucial role in product production. It's a crucial method for joining materials, providing a stable structure and improving product strength and durability. In industries such as aerospace, automotive, and shipbuilding, welding assembly is widely used to connect key components and structural parts to ensure product safety and reliability. However, existing technologies often suffer from low welding assembly accuracy, resulting in poor quality. Summary of the Invention

[0003] This application provides a precision welding and assembly method for an intelligent robot. This method addresses the existing technical problem of low welding and assembly accuracy, which results in poor welding and assembly quality. This method achieves the technical effect of improving welding and assembly accuracy and enhancing the welding and assembly quality of products.

[0004] In view of the above problems, the present application provides a precise welding and assembly method for an intelligent robot.

[0005] In a first aspect, the present application provides a precision welding and assembly method for an intelligent robot, wherein the method is applied to a precision welding and assembly system for an intelligent robot, and the method includes: obtaining a product design scheme for a target manufactured product; obtaining G product parts of the target manufactured product, wherein G is a positive integer greater than 1; based on an intelligent assembly submodule, making assembly decisions on the G product parts according to the product design scheme to obtain a parts assembly decision; based on the parts assembly decision, assembling the G product parts according to an assembly robot to obtain a product parts assembly; based on an intelligent welding submodule, making welding decisions on the product parts assembly according to the product design scheme to generate a product welding decision; based on the product welding decision, welding the product parts assembly according to a welding robot.

[0006] In a second aspect, the present application also provides a precision welding and assembly system for an intelligent robot, wherein the system includes: a product design scheme acquisition module, the product design scheme acquisition module is used to obtain the product design scheme of the target manufactured product; a product part acquisition module, the product part acquisition module is used to obtain G product parts of the target manufactured product, wherein G is a positive integer greater than 1; an assembly decision module, the assembly decision module is used to make assembly decisions on the G product parts according to the product design scheme based on the intelligent assembly sub-module, and obtain a part assembly decision; a part assembly module, the part assembly module is used to assemble the G product parts according to the assembly robot based on the part assembly decision, and obtain a product part assembly; a welding decision module, the welding decision module is used to make welding decisions on the product part assembly according to the product design scheme based on the intelligent welding sub-module, and generate a product welding decision; and a welding module is used to weld the product part assembly according to the welding robot based on the product welding decision.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The intelligent assembly submodule makes assembly decisions for the G product parts of the target product, obtaining a parts assembly decision. The assembly robot assembles the G product parts according to the parts assembly decision, obtaining a product parts assembly. The intelligent welding submodule makes welding decisions for the product parts assembly, generating a product welding decision. The welding robot welds the product parts assembly according to the product welding decision. This achieves the technical effect of improving the accuracy and quality of product welding assembly.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings of the embodiments of the present invention. Obviously, the drawings described below only relate to some embodiments of the present invention and are not intended to limit the present invention.

[0011] Figure 1 A schematic diagram of a process for precise welding and assembly of an intelligent robot arm is provided for this application;

[0012] Figure 2 This is a structural diagram of a precision welding and assembly system for an intelligent robot arm for this application. DETAILED DESCRIPTION

[0013] This application provides a precise welding and assembly method using an intelligent robot. This method addresses the technical problem of low welding and assembly accuracy and poor quality of products in the prior art. This method achieves the technical effect of improving the welding and assembly accuracy and quality of products.

[0014] Example 1

[0015] Please see the attached Figure 1 The present application provides a method for precise welding and assembly of an intelligent manipulator, wherein the method is applied to a precise welding and assembly system of an intelligent manipulator, and the method specifically comprises the following steps:

[0016] Obtain product design solutions for target manufacturing products;

[0017] Obtain G product parts of the target manufactured product, where G is a positive integer greater than 1;

[0018] Connect to the intelligent robotic arm precision welding and assembly system described in this application to read the product design plan of the target manufactured product. The target manufactured product can be any part to be processed by precision welding and assembly using the intelligent robotic arm precision welding and assembly system. The target manufactured product includes G product parts. G is a positive integer greater than 1. The product design plan includes product type information and design drawings corresponding to the target manufactured product, as well as part structure design information corresponding to each product part of the target manufactured product.

[0019] Based on the intelligent assembly submodule, make assembly decisions for the G product parts according to the product design plan to obtain part assembly decisions;

[0020] Among them, obtaining part assembly decisions includes:

[0021] Obtaining a control variable set of the assembly robot;

[0022] Based on the control variable set, construct the intelligent assembly submodule;

[0023] Wherein, constructing the intelligent assembly submodule includes:

[0024] Obtaining backtracking constraints based on the G part structure data and the product design plan;

[0025] Based on the control variable set, obtaining a backtracking target;

[0026] Perform part assembly control backtracking based on the backtracking constraint and the backtracking target to obtain a part assembly control record set;

[0027] Training the BP neural network according to the parts assembly control record set, obtaining an output precision operator after each training P times, where P is a preset number of training times;

[0028] If the output precision operator satisfies the output precision operator constraint, the intelligent assembly submodule is generated.

[0029] A precision welding and assembly system connected to an intelligent robot described in this application reads the assembly robot's control variable set. An assembly robot is a robot that can automatically perform assembly tasks. The assembly robot includes components such as the robot body, control system, gripper, and auxiliary equipment. The control variable set includes multiple control variables for the assembly robot. These control variables include the assembly robot's operating force, movement direction, movement distance, movement speed, and control angle.

[0030] Furthermore, G part structure data and product design plans are set as backtracking constraints. A control variable set is set as a backtracking target. Part assembly control records are read from the precision welding and assembly system of the intelligent manipulator based on the backtracking constraints and the backtracking target to obtain a part assembly control record set. The part assembly control record set includes multiple part assembly control records. Each part assembly control record includes historical part structure data, historical product design plans, and historical part assembly decisions. The historical part assembly decisions include multiple historical control parameters for multiple control variables of the assembly manipulator corresponding to the historical part structure data and historical product design plans. Subsequently, a BP neural network is trained based on the part assembly control record set. After each P training round, an output precision operator is obtained, and it is determined whether the output precision operator satisfies the output precision operator constraint. If the output precision operator satisfies the output precision operator constraint, an intelligent assembly submodule is generated. The BP neural network is a multi-layer feedforward neural network trained using an error back propagation algorithm. P is a preset number of training rounds, which is pre-set and determined by the precision welding and assembly system of the intelligent manipulator. The output accuracy operator is the output accuracy of the BP neural network when trained P times based on the part assembly control record set. The output accuracy operator constraints include an output accuracy range pre-set and determined by the precision welding and assembly system of the intelligent manipulator. The intelligent assembly submodule is a BP neural network trained based on the part assembly control record set, where the output accuracy operator satisfies the output accuracy operator constraints. The intelligent assembly submodule includes an input layer, a hidden layer, and an output layer.

[0031] Collect basic information of the G product parts to obtain G part data sets;

[0032] Extract structural parameters based on the G part data sets to obtain G part structural data;

[0033] Based on the G part structure data and the product design plan, the part assembly decision is generated according to the intelligent assembly submodule.

[0034] Based on the parts assembly decision, assemble the G product parts using an assembly robot to obtain a product parts assembly;

[0035] A precision welding and assembly system connected to the intelligent robot retrieves basic information from G product parts to obtain G part data sets. Each part data set includes a part image and part structure data corresponding to each product part. The part image is the image data information corresponding to the product part. The part structure data includes the dimensional and structural parameter information corresponding to the product part. Subsequently, part structure data is extracted from the G part data sets to obtain G part structure data. The G part structure data and the product design plan are input into the intelligent assembly submodule to obtain a part assembly decision. Then, the assembly robot assembles the G product parts according to the part assembly decision to obtain a product part assembly. The part assembly decision includes multiple control variable parameters of the assembly robot corresponding to the G part structure data and the product design plan. The multiple control variable parameters include operating force parameters, movement direction parameters, movement distance parameters, movement speed parameters, control angle parameters, etc. corresponding to multiple control variables in the control variable set. The product part assembly includes the G product parts assembled after the assembly robot assembles the G product parts.

[0036] The intelligent assembly sub-module is used to make assembly decisions for G product parts, and the assembly robot is used to assemble G product parts, thereby improving the accuracy and intelligence of product part assembly.

[0037] After obtaining G parts data sets, it includes:

[0038] Based on the G product parts, obtain the g-th product part, where g is a positive integer and belongs to G;

[0039] Matching the g-th part dataset corresponding to the g-th product part according to the G part datasets, wherein the g-th part dataset includes the g-th part image and the g-th part structure data;

[0040] Performing surface defect recognition based on the g-th part image to obtain a surface defect degree of the g-th part;

[0041] Determining whether the surface defect degree of the g-th part meets a preset surface defect degree;

[0042] If the surface defect degree of the g-th part meets the preset surface defect degree, a g-th part defect warning signal is generated.

[0043] G product parts are sequentially extracted to obtain a g-th product part. The g-th product part is each product part in the G product parts. Where g is a positive integer, g belongs to G. Next, a g-th part dataset corresponding to the g-th product part is extracted from the G part datasets. The g-th part dataset includes a g-th part image and g-th part structure data corresponding to the g-th product part.

[0044] Further, surface defect recognition is performed on the g-th part image to obtain the surface defect degree of the g-th part, and it is determined whether the surface defect degree of the g-th part meets the preset surface defect degree. If the surface defect degree of the g-th part meets the preset surface defect degree, a g-th part defect warning signal is generated. The surface defect degree of the g-th part is data information used to characterize the degree of surface defects of the g-th product part. The greater the surface defect degree of the g-th product part, the higher the corresponding surface defect degree of the g-th part. The preset surface defect degree includes a part surface defect degree range pre-set and determined by the precision welding and assembly system of the intelligent manipulator. The g-th part defect warning signal is data information used to characterize that the surface defect degree of the g-th part meets the preset surface defect degree and the surface defect degree of the g-th product part is high.

[0045] Preferably, when surface defects are identified on the g-th part image, a precision welding and assembly system connected to the intelligent manipulator reads multiple part surface defect identification records. Each part surface defect identification record includes a historical part image and a historical part surface defect degree. Then, the multiple part surface defect identification records are continuously self-trained and learned to a convergence state according to the convolutional neural network to generate a surface defect recognition network. The g-th part image is input into the surface defect recognition network to obtain the surface defect degree of the g-th part. The convolutional neural network is a type of feedforward neural network that includes convolution calculations and has a deep structure. The surface defect recognition network includes an input layer, a hidden layer, and an output layer. The surface defect recognition network has the function of matching the surface defect degree of the part on the input g-th part image.

[0046] By identifying surface defects of G product parts and combining them with preset surface defect degrees, part defect warning signals are adaptively generated, and surface defect verification of G product parts is achieved before welding and assembly, thereby improving the welding and assembly quality of the products.

[0047] extracting the g-th part structural design information corresponding to the g-th product part based on the product design solution;

[0048] Comparing the g-th part structural design information with the g-th part structural data to obtain the g-th part structural standardization;

[0049] Determining whether the structural standard degree of the g-th part meets a preset structural standard degree;

[0050] If the structural standard of the g-th part does not meet the preset structural standard, a g-th part structure warning signal is generated.

[0051] Extract the g-th part structural design information corresponding to the g-th product part from the product design plan. Then, compare the g-th part structural design information with the g-th part structural data to obtain the g-th part structural standard, and determine whether the g-th part structural standard meets the preset structural standard. If the g-th part structural standard does not meet the preset structural standard, generate a g-th part structural warning signal. The g-th part structural standard is data information used to characterize the consistency between the g-th part structural data and the g-th part structural design information. The stronger the consistency between the g-th part structural data and the g-th part structural design information, the higher the corresponding g-th part structural standard. The preset structural standard includes a part structural standard range pre-set and determined by the precision welding and assembly system of the intelligent manipulator. The g-th part structural warning signal is a warning prompt information used to characterize that the g-th part structural standard does not meet the preset structural standard and the structural standard of the g-th product part is poor.

[0052] By identifying the part structure standards of G product parts and combining them with the preset structure standards, adaptively generating part structure warning signals, it is possible to perform part structure verification on G product parts before welding and assembly, thereby improving the welding and assembly reliability of the product.

[0053] Based on the intelligent welding submodule, a welding decision is made on the product parts assembly according to the product design plan to generate a product welding decision;

[0054] Among them, generating product welding decisions includes:

[0055] activating a welding decision space within the intelligent welding submodule;

[0056] Based on the product design solution, obtaining a target product type;

[0057] Based on the target product type and the welding decision space, a matching welding decision domain is obtained;

[0058] Based on the matching welding decision domain, obtaining a first welding decision particle;

[0059] After the assembly robot completes the assembly of G product parts according to the parts assembly decision, the precision welding and assembly system of the intelligent robot automatically activates the welding decision space within the intelligent welding submodule. The intelligent welding submodule includes a welding decision space. The welding decision space includes multiple welding decision domains corresponding to multiple product types. Each welding decision domain includes multiple welding decision records corresponding to each product type. Each welding decision record includes the historical motion control parameters of the welding robot (the historical movement speed, historical acceleration, historical motion stroke, etc. of each axis of the welding robot) and the historical welding process parameters (historical welding current, historical arc voltage, historical welding speed, historical welding distance, historical shielding gas flow, etc.) corresponding to the historical product parts assembly of each product type.

[0060] Next, the product type information within the product design solution is recorded as the target product type. The target product type is matched with multiple welding decision domains within the welding decision space. The welding decision domain corresponding to the target product type is set as the matching welding decision domain. Multiple welding decision records within the matching welding decision domain are randomly selected to obtain a first welding decision particle. The first welding decision particle can be a random welding decision record within the matching welding decision domain.

[0061] Performing welding fitness analysis based on the product part assembly and the first welding decision particle to obtain first particle welding fitness;

[0062] Wherein, obtaining the first particle welding adaptability includes:

[0063] activating the welding-digital twin submodule, modeling the product part assembly according to the welding-digital twin submodule, and generating a product part assembly model;

[0064] Based on the welding-digital twin module, simulate welding the product parts assembly model according to the first welding decision particles to obtain a welded product parts assembly model;

[0065] Welding quality evaluation is performed based on the welding product parts assembly model to generate the first particle welding fitness.

[0066] If the welding adaptability of the first particle meets the preset welding adaptability, the product welding decision is generated according to the first welding decision particle.

[0067] Based on the product welding decision, the product parts assembly is welded using a welding manipulator.

[0068] The welding-digital twin module within the precision welding and assembly system of the intelligent manipulator is activated, and the welding-digital twin module is used to model the product part assembly to generate a product part assembly model. The welding-digital twin module includes a conventional digital twin platform. A digital twin platform is based on digital twin technology. The digital twin platform can be used to simulate and predict product welding processes. Digital twin technology integrates the physical world with digital models. Digital twin technology collects various data to construct digital models of entities, and uses various simulation analysis tools to simulate and predict entities. The product part assembly model includes a simulation model corresponding to the product part assembly. Subsequently, the first welding decision particle is uploaded to the welding-digital twin module. The welding-digital twin module simulates welding on the product part assembly model according to the first welding decision particle to obtain a welded product part assembly model. The welded product part assembly model includes a product part assembly model that has been simulated and welded according to the first welding decision particle. Furthermore, the weld quality of the welded product part assembly model is evaluated to generate the welding fitness of the first particle. The first particle welding fitness is data used to characterize the welding quality of a welding product part assembly model. The higher the welding quality of the welding product part assembly model, the greater the corresponding first particle welding fitness.

[0069] Furthermore, a determination is made as to whether the first particle welding fitness satisfies a preset welding fitness. If the first particle welding fitness satisfies the preset welding fitness, the first welding decision particle is output as a product welding decision. The welding robot welds the product part assembly according to the product welding decision, thereby improving the degree of welding automation and welding quality of the product. If the first particle welding fitness does not satisfy the preset welding fitness, an iterative optimization search is continued for multiple welding decision records within the matching welding decision domain until a product welding decision that satisfies the preset welding fitness is obtained. The preset welding fitness comprises a particle welding fitness range pre-determined by the precision welding and assembly system of the intelligent robot.

[0070] In summary, the precise welding and assembly method of an intelligent robot provided in this application has the following technical effects:

[0071] 1. The intelligent assembly submodule makes assembly decisions for the target product's G parts, generating a part assembly decision. The assembly robot assembles the G parts according to the part assembly decision, generating a product part assembly. The intelligent welding submodule makes welding decisions for the product part assembly, generating a product welding decision. The welding robot welds the product part assembly according to the product welding decision. This achieves the technical effect of improving the accuracy and quality of product welding assembly.

[0072] 2. By identifying surface defects of G product parts and combining them with preset surface defect degrees, part defect warning signals are adaptively generated. This allows surface defect verification of G product parts before welding and assembly, thereby improving the welding and assembly quality of the product.

[0073] 3. By identifying the part structure standards of G product parts and combining them with the preset structure standards, a part structure early warning signal is adaptively generated, and the part structure of G product parts can be verified before welding and assembly, thereby improving the welding and assembly reliability of the product.

[0074] Example 2

[0075] Based on the same invention concept as the precise welding and assembly method of an intelligent manipulator in the aforementioned embodiment, the present invention also provides a precise welding and assembly system for an intelligent manipulator, see the attached Figure 2 , the system comprising:

[0076] A product design scheme obtaining module, wherein the product design scheme obtaining module is used to obtain a product design scheme of a target manufactured product;

[0077] a product parts acquisition module, the product parts acquisition module being used to obtain G product parts of the target manufactured product, wherein G is a positive integer greater than 1;

[0078] An assembly decision module, wherein the assembly decision module is used to make assembly decisions for the G product parts according to the product design plan based on the intelligent assembly submodule to obtain a part assembly decision;

[0079] A parts assembly module, configured to assemble the G product parts according to the parts assembly decision using an assembly robot to obtain a product parts assembly;

[0080] A welding decision module, wherein the welding decision module is used to make a welding decision on the product part assembly according to the product design scheme based on the intelligent welding submodule to generate a product welding decision;

[0081] A welding module is used to weld the product parts assembly according to the welding manipulator based on the product welding decision.

[0082] Furthermore, the system also includes a parts assembly decision generation module to perform the following steps:

[0083] Obtaining a control variable set of the assembly robot;

[0084] Based on the control variable set, construct the intelligent assembly submodule;

[0085] Collect basic information of the G product parts to obtain G part data sets;

[0086] Extract structural parameters based on the G part data sets to obtain G part structural data;

[0087] Based on the G part structure data and the product design plan, the part assembly decision is generated according to the intelligent assembly submodule.

[0088] Furthermore, the system further includes a construction module to perform the following operation steps:

[0089] Obtaining backtracking constraints based on the G part structure data and the product design plan;

[0090] Based on the control variable set, obtaining a backtracking target;

[0091] Perform part assembly control backtracking based on the backtracking constraint and the backtracking target to obtain a part assembly control record set;

[0092] Training the BP neural network according to the parts assembly control record set, obtaining an output precision operator after each training P times, where P is a preset number of training times;

[0093] If the output precision operator satisfies the output precision operator constraint, the intelligent assembly submodule is generated.

[0094] Furthermore, the system also includes a parts defect warning module to perform the following steps:

[0095] Based on the G product parts, obtain the g-th product part, where g is a positive integer and belongs to G;

[0096] Matching the g-th part dataset corresponding to the g-th product part according to the G part datasets, wherein the g-th part dataset includes the g-th part image and the g-th part structure data;

[0097] Performing surface defect recognition based on the g-th part image to obtain a surface defect degree of the g-th part;

[0098] Determining whether the surface defect degree of the g-th part meets a preset surface defect degree;

[0099] If the surface defect degree of the g-th part meets the preset surface defect degree, a g-th part defect warning signal is generated.

[0100] Furthermore, the system also includes a part structure early warning module to perform the following steps:

[0101] extracting the g-th part structural design information corresponding to the g-th product part based on the product design solution;

[0102] Comparing the structural design information of the g-th part with the structural data of the g-th part to obtain the structural standardization of the g-th part;

[0103] Determining whether the structural standard degree of the g-th part meets a preset structural standard degree;

[0104] If the structural standard of the g-th part does not meet the preset structural standard, a g-th part structure warning signal is generated.

[0105] Furthermore, the system also includes a product welding decision generation module to perform the following steps:

[0106] activating a welding decision space within the intelligent welding submodule;

[0107] Based on the product design solution, obtaining a target product type;

[0108] Based on the target product type and the welding decision space, a matching welding decision domain is obtained;

[0109] Based on the matching welding decision domain, obtaining a first welding decision particle;

[0110] Performing welding fitness analysis based on the product part assembly and the first welding decision particle to obtain first particle welding fitness;

[0111] If the welding adaptability of the first particle meets the preset welding adaptability, the product welding decision is generated according to the first welding decision particle.

[0112] Furthermore, the system further includes a welding fitness analysis module to perform the following steps:

[0113] activating the welding-digital twin submodule, modeling the product part assembly according to the welding-digital twin submodule, and generating a product part assembly model;

[0114] Based on the welding-digital twin module, simulate welding the product parts assembly model according to the first welding decision particles to obtain a welded product parts assembly model;

[0115] Welding quality evaluation is performed based on the welding product parts assembly model to generate the first particle welding fitness.

[0116] The precise welding and assembly system of an intelligent robot provided by an embodiment of the present invention can execute the precise welding and assembly method of an intelligent robot provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0117] The modules included are divided only according to functional logic, but are not limited to the above division, as long as they can achieve the corresponding functions; in addition, the specific names of the functional modules are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0118] This application provides a precision welding and assembly method for an intelligent manipulator, wherein the method is applied to a precision welding and assembly system for an intelligent manipulator. The method comprises: using an intelligent assembly submodule to make assembly decisions for G product parts of a target manufactured product to obtain a part assembly decision; an assembly manipulator assembling the G product parts according to the part assembly decision to obtain a product part assembly; using an intelligent welding submodule to make welding decisions for the product part assembly to generate a product welding decision; and a welding manipulator welding the product part assembly according to the product welding decision. This method achieves the technical effect of improving the welding and assembly accuracy and quality of the product.

[0119] Although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments. Without departing from the concept of the present invention, it can also include more other equivalent embodiments, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A precise welding and assembly method for an intelligent manipulator, characterized in that: The method comprises: Obtain product design solutions for target manufacturing products; Obtain G product parts of the target manufactured product, where G is a positive integer greater than 1; Based on the intelligent assembly submodule, make assembly decisions for the G product parts according to the product design plan to obtain part assembly decisions; Based on the parts assembly decision, assemble the G product parts using an assembly robot to obtain a product parts assembly; Based on the intelligent welding submodule, a welding decision is made on the product parts assembly according to the product design plan to generate a product welding decision; Based on the product welding decision, welding the product parts assembly according to the welding manipulator; The intelligent assembly submodule is used to make assembly decisions for the G product parts according to the product design solution to obtain part assembly decisions, including: Obtaining a control variable set of the assembly robot; Based on the control variable set, construct the intelligent assembly submodule; Collect basic information of the G product parts to obtain G part data sets; Extract structural parameters based on the G part data sets to obtain G part structural data; Based on the G part structure data and the product design plan, the part assembly decision is generated according to the intelligent assembly submodule.

2. The method according to claim 1, wherein Based on the control variable set, the intelligent assembly submodule is constructed, including: Obtaining backtracking constraints based on the G part structure data and the product design plan; Based on the control variable set, obtaining a backtracking target; Perform part assembly control backtracking based on the backtracking constraint and the backtracking target to obtain a part assembly control record set; Training the BP neural network according to the parts assembly control record set, and obtaining an output precision operator after each training P times, where P is a preset number of training times; If the output precision operator satisfies the output precision operator constraint, the intelligent assembly submodule is generated.

3. The method according to claim 1, wherein After obtaining G parts data sets, including: Based on the G product parts, obtain the g-th product part, where g is a positive integer; Matching the g-th part dataset corresponding to the g-th product part according to the G part datasets, wherein the g-th part dataset includes the g-th part image and the g-th part structure data; Performing surface defect recognition based on the g-th part image to obtain a surface defect degree of the g-th part; Determining whether the surface defect degree of the g-th part meets a preset surface defect degree; If the surface defect degree of the g-th part meets the preset surface defect degree, a g-th part defect warning signal is generated.

4. The method according to claim 3, wherein The method comprises: extracting the g-th part structural design information corresponding to the g-th product part based on the product design solution; Comparing the structural design information of the g-th part with the structural data of the g-th part to obtain the structural standardization of the g-th part; Determining whether the structural standard degree of the g-th part meets a preset structural standard degree; If the structural standard of the g-th part does not meet the preset structural standard, a g-th part structure warning signal is generated.

5. The method according to claim 1, wherein Based on the intelligent welding submodule, a welding decision is made for the product parts assembly according to the product design scheme, and a product welding decision is generated, including: activating a welding decision space within the intelligent welding submodule; Based on the product design solution, obtaining a target product type; Based on the target product type and the welding decision space, a matching welding decision domain is obtained; Based on the matching welding decision domain, obtaining a first welding decision particle; Performing welding fitness analysis based on the product part assembly and the first welding decision particle to obtain first particle welding fitness; If the welding adaptability of the first particle meets the preset welding adaptability, the product welding decision is generated according to the first welding decision particle.

6. The method according to claim 5, wherein Performing welding fitness analysis based on the product part assembly and the first welding decision particle to obtain first particle welding fitness includes: activating the welding-digital twin submodule, modeling the product part assembly according to the welding-digital twin submodule, and generating a product part assembly model; Based on the welding-digital twin module, simulate welding the product parts assembly model according to the first welding decision particles to obtain a welded product parts assembly model; Welding quality evaluation is performed based on the welding product parts assembly model to generate the first particle welding fitness.

7. A precise welding and assembly system for an intelligent manipulator, characterized in that: The system is used to perform the method according to any one of claims 1 to 6, and the system comprises: A product design scheme obtaining module, wherein the product design scheme obtaining module is used to obtain a product design scheme of a target manufactured product; A product parts acquisition module, the product parts acquisition module is used to obtain G product parts of the target manufactured product, where G is a positive integer greater than 1; An assembly decision module, wherein the assembly decision module is used to make assembly decisions for the G product parts according to the product design plan based on the intelligent assembly submodule to obtain a part assembly decision; A parts assembly module, configured to assemble the G product parts according to the parts assembly decision using an assembly robot to obtain a product parts assembly; A welding decision module, wherein the welding decision module is used to make a welding decision on the product part assembly according to the product design scheme based on the intelligent welding submodule to generate a product welding decision; A welding module is used to weld the product parts assembly according to the welding manipulator based on the product welding decision.

Images