Assembly line performance change demand driven reverse variant network modeling method

Abstract

The invention aims to provide an assembly line performance change demand driven reverse variant network modeling method. The method comprises the following steps: constructing a five-dimensional association model matched with a to-be-changed assembly line; establishing a mapping and constraint relationship among the elements of each dimension; identifying a key assembly link and a change transmission mechanism, and constructing a dynamic feedback model among assembly process parameters, production line configuration parameters and assembly line performance; obtaining a topological index of any node in a network corresponding to the five-dimensional association model, and quantifying propagation risks from an initial node to a potential influence range; establishing an optimization function; solving the optimization function, and outputting a group of optimal solution sets; any solution in the optimal solution set represents an optimal change strategy which balances risks, time and cost. The production efficiency can be improved, the operation cost can be reduced, and the market competitiveness can be enhanced.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of industrial assembly line change requirements, and particularly relates to an assembly line performance change requirement driven reverse variant network modeling method. BACKGROUND

[0002] Currently, in the production practice of manufacturing enterprises, as the core production system, the performance of the assembly line directly determines the quality of the product, the production efficiency and the market competitiveness. However, with the rapid changes in market demand, the shortening of product life cycle and the growth of customer individualization demand, the assembly line often faces frequent performance change requirements. For example, in order to adapt to the introduction of new product models, improve the production rhythm, reduce energy consumption or improve product quality, the assembly line often needs to be adjusted and optimized locally or as a whole.

[0003] Traditionally, the performance change management of the assembly line usually relies on the experience of engineers, trial and error method or simple change notification process. This management mode has significant limitations: long change cycle, difficult to quickly respond to market changes; high change cost, due to lack of systematic evaluation, may lead to repeated investment or waste of resources; high change risk, especially in complex assembly lines, local changes may trigger unpredictable chain reactions, i.e. change propagation, and further affect other functions, structures, behaviors, executions and even the performance of the final product.

[0004] The existing technology lacks a deep understanding of the performance change driving mechanism of the assembly line when solving these problems, and fails to establish a cross-dimensional and dynamic correlation mechanism between functions, structures, behaviors, executions and performance. This means that it is impossible to accurately predict and control the scope of change. Moreover, for risk assessment and quantification of change propagation, and how to balance and optimize among multiple objectives such as risk, time and cost, the existing method also fails to provide a complete and efficient solution. Especially in the case of fine engineering problems involving product surface quality, assembly surface connection characteristics and complex three-dimensional geometric size and geometric tolerance change propagation mechanism, the existing technology is difficult to provide comprehensive analysis and optimization strategies, resulting in low efficiency of change management, and even hindering the efficient operation and sustainable development of manufacturing enterprises.

[0005] Therefore, based on this, the present disclosure provides a reverse variant network modeling method for fine engineering such as product surface quality, assembly surface connection characteristics and complex three-dimensional geometric size and geometric tolerance change propagation mechanism. SUMMARY

[0006] The present disclosure aims to provide an assembly line performance change requirement driven reverse variant network modeling method to solve at least one technical problem in the prior art.

[0007] The technical solution of the present disclosure is as follows:

[0008] An assembly line performance change requirement driven reverse variant network modeling method, comprising:

[0009] Obtaining performance change requirement information of an assembly line to be changed;

[0010] According to the performance change requirement information, a five-dimensional association model matched with the assembly line to be changed is constructed; the five-dimensional association model includes performance dimension, execution dimension, behavior dimension, structure dimension and function dimension, and mapping and constraint relationship between elements of each dimension are established, so as to convert scattered knowledge and data into structured five-dimensional association model;

[0011] Identifying key assembly links and change transmission mechanism, and constructing a dynamic feedback model between assembly process parameters, line configuration parameters and assembly line performance;

[0012] According to the five-dimensional association model, topological indexes of any node in the network corresponding to the five-dimensional association model are obtained, and propagation risks within the potential influence range from the initial node are quantified; and an optimization function based on change propagation risk, change propagation time and change propagation cost is established;

[0013] Solving the optimization function, and outputting a set of optimal solutions; any solution in the set of optimal solutions represents an optimal change strategy that balances risk, time and cost.

[0014] The mapping and constraint relationship between elements of each dimension are established, and the scattered knowledge and data are converted into structured five-dimensional association model, comprising:

[0015] Establishing cross-dimensional parameterized mapping, including performance-execution mapping, execution-behavior mapping, behavior-structure mapping and structure-function mapping;

[0016] Setting restriction conditions between elements within any dimension; and defining mutual dependence and logical restrictions between different dimensional elements;

[0017] Parameterized modeling and modular design of key components, modules and process steps in the assembly line are performed, and are abstracted into reusable and configurable element nodes;

[0018] According to the constraint relationship between dimensional elements, cross-dimensional parameterized mapping and elements within each dimension, a knowledge graph or relational database is formed;

[0019] The effectiveness of the established cross-dimensional parameterized mapping and constraint relationship between elements within each dimension is simulated and verified, and iterative optimization and adjustment are performed according to the simulation results, so as to ensure that the five-dimensional association model can accurately reflect the actual dynamic behavior of the assembly line.

[0020] The key assembly link is identified and the change transmission mechanism is changed, including:

[0021] The surface quality of the key assembly surface is quantitatively evaluated, and the surface lubrication condition is analyzed;

[0022] The connection characteristics of the key assembly surface in the assembly process are judged to identify the factors affecting the assembly accuracy;

[0023] The transmission mechanism of the three-dimensional geometric dimension and the tolerance change is judged, the cumulative effect of the size and the tolerance in the assembly process is identified, and the key assembly link affecting the accuracy and the performance is obtained;

[0024] According to the key assembly link, a dynamic feedback model between assembly process parameters, line configuration parameters and assembly line performance is constructed to simulate the dynamic influence and feedback mechanism of different process parameters and line configuration changes on assembly cycle, equipment utilization and product qualification rate.

[0025] The dynamic feedback model between the assembly process parameters, the line configuration parameters and the assembly line performance is constructed, including:

[0026] The input variables of the dynamic feedback model include assembly process parameters and line configuration parameters;

[0027] The output variables of the dynamic feedback model include assembly line performance indicators;

[0028] According to the causal chain and feedback loop between the input variables and the output variables, a mathematical expression of the dynamic feedback model is constructed;

[0029] The dynamic feedback model is parameterized and verified, and the model is trained and optimized using historical performance change data to ensure that the model can accurately reflect the dynamic response characteristics of the assembly line.

[0030] According to the five-dimensional correlation model, the topological index of any node in the network corresponding to the five-dimensional correlation model is obtained, and the propagation risk from the initial node to the potential influence range is quantified, including:

[0031] The key topological index of any node in the network is obtained; the key topological index includes node complexity, node betweenness, clustering coefficient and node set information degree, so as to quantify the change propagation risk from the initial node to the potential influence range;

[0032] The time cost required from the demand proposal to the final completion of the change is quantified, including analysis time, design time, implementation time and verification time;

[0033] The economic cost required from demand submission to final completion of the quantification change, including R&D investment, equipment modification cost, material consumption and production loss.

[0034] The optimization function based on change propagation risk, change propagation time and change propagation cost is established, including:

[0035] The optimization function is:

[0036] min(ω R ·Risk total +ω T ·Time total +ω C ·Cost total );

[0037] Wherein, Risk total is the total change propagation risk after quantification; Time total is the total change propagation time; Cost total is the total change propagation cost; ω R , ω T , ω C are weight coefficients respectively.

[0038] An assembly line performance change demand driven reverse variant network system based on the assembly line performance change demand driven reverse variant network modeling method, comprising:

[0039] A collection module acquires performance change demand information of an assembly line to be changed;

[0040] A first processing module interacts with the collection module to construct a five-dimensional association model matched with the assembly line to be changed according to the performance change demand information;

[0041] An identification module interacts with the first processing module to identify key assembly links and change transmission mechanisms, and construct a dynamic feedback model between assembly process parameters, production line configuration parameters and assembly line performance;

[0042] A second processing module interacts with the first processing module to obtain a topological index of any node in the five-dimensional association model corresponding network according to the five-dimensional association model, and to quantify the propagation risk from an initial node to a potential influence range; and to establish an optimization function based on change propagation risk, change propagation time and change propagation cost;

[0043] A third processing module interacts with the second processing module to solve the optimization function and output an optimal solution set;

[0044] The output module interacts with the third processing module for data output, and is configured to output any solution in the optimal solution set to obtain an optimal change strategy that balances risk, time and cost.

[0045] Advantages:

[0046] The advantages of the present disclosure at least include:

[0047] The method of the present application establishes a five-dimensional correlation model based on "function-structure-behavior-execution-performance", realizes comprehensive and in-depth understanding of the change driving mechanism of assembly line performance, thereby improving the scientificity and accuracy of change analysis; moreover, by quantifying network topology indexes such as node complexity, node betweenness, clustering coefficient and node set information degree, the propagation risk of change in different dimensions and links is accurately evaluated and controlled, and the uncertainty and potential loss in the change process are reduced; and a multi-objective optimization function based on change propagation risk, time and cost is constructed, and an intelligent optimization algorithm is used for solving, thereby planning the optimal change propagation path and strategy, shortening the change cycle and reducing the change cost; moreover, surface engineering analysis, connection property evaluation and size and geometric tolerance transmission analysis are integrated into the identification of key assembly links, and combined with system dynamics modeling, the change scheme is closer to the actual production demand, and the feasibility and implementation effect are improved. The present application provides a set of forward-looking, systematic and intelligent solution for manufacturing enterprises to cope with complex and changeable assembly line performance change requirements, which has important strategic significance and practical application value for improving production efficiency, reducing operating cost and enhancing market competitiveness. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 is a five-dimensional reverse variant network modeling diagram of an assembly line driven by performance change requirements;

[0049] Figure 2 is a flow chart of the reverse variant network modeling method of the assembly line performance change requirement driven by the present disclosure;

[0050] Figure 3 is a system block diagram of the system of the present disclosure. DETAILED DESCRIPTION

[0051] In the following, various embodiments of the present disclosure will be described more fully. The present disclosure can have various embodiments, and adjustments and changes can be made therein. However, it should be understood that there is no intention to limit various embodiments of the present disclosure to the specific embodiments disclosed herein, but the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives falling within the spirit and scope of various embodiments of the present disclosure.

[0052] Hereinafter, the term "include" or "may include" used in various embodiments of the disclosure indicates the presence of the disclosed functions, operations, or elements, and does not limit one or more functions, operations, or elements from being added. Also, as used in various embodiments of the disclosure, the terms "include", "have", and their conjugates merely indicate the presence of the mentioned features, numbers, steps, operations, elements, components, or combinations thereof, and do not preclude the presence or addition of one or more other features, numbers, steps, operations, elements, components, or combinations thereof.

[0053] In various embodiments of the disclosure, the expression "or" or "at least one of A or / and B" includes any combination of the listed terms or all the combinations thereof. For example, the expression "A or B" or "at least one of A or / and B" can include A, can include B, or can include both A and B.

[0054] The expressions (such as "first", "second", etc.) used in various embodiments of the disclosure can modify various constituent elements in various embodiments, but can not limit the corresponding constituent elements. For example, the above expressions do not limit the order and / or importance of the described elements. The above expressions are used only for the purpose of distinguishing one element from another element. For example, the first user device and the second user device indicate different user devices, although both are user devices. For example, the first element can be referred to as the second element, and likewise, the second element can be referred to as the first element, without departing from the scope of various embodiments of the disclosure.

[0055] It should be noted that if a description connects one constituent element to another constituent element, the first constituent element can be directly connected to the second constituent element, and a third constituent element can be "connected" between the first constituent element and the second constituent element. Conversely, when one constituent element is "directly connected" to another constituent element, it can be understood that there is no third constituent element between the first constituent element and the second constituent element.

[0056] The term "user" used in various embodiments of the disclosure can indicate a person using an electronic device or a device (for example, an artificial intelligence electronic device) using an electronic device.

[0057] The terms used in the various embodiments of the present disclosure are used only for the purpose of describing particular embodiments and are not intended to limit the various embodiments of the present disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly dictates otherwise. Unless defined otherwise, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the present disclosure belong. The terms (such as terms defined in a generally used dictionary) will be interpreted to have the same meaning as the contextual meaning in the relevant technical field and will not be interpreted to have idealized or overly formal meanings, unless clearly defined in the various embodiments of the present disclosure.

[0058] Term explanation:

[0059] FSBEP model: refers to a system model integrating five dimensions of Function, Structure, Behavior, Execution, and Performance, for describing an assembly line system.

[0060] Function: the task and role assumed by an assembly line or module, such as material transportation, part assembly, quality inspection, etc.

[0061] Structure: the physical composition and layout of an assembly line, such as the number of stations, equipment types, connection methods, etc.

[0062] Behavior: the running logic and action flow of an assembly line or module, such as processing cycle, operation sequence, human-machine interaction method, etc.

[0063] Execution: the specific running state and parameters of an assembly line or module, such as equipment speed, precision, temperature, force, etc.

[0064] Performance: the comprehensive indicators of efficiency, quality, cost, flexibility, etc. of an assembly line or module.

[0065] Parameter node: abstracting key components or modules in an assembly line as nodes in a network and describing their key features in a parameterized manner, such as size, model, performance parameters, etc.

[0066] Node complexity: measures the complexity of the interior of a node or its connections.

[0067] Node betweenness: represents the "traffic hub" status of a node in the network, i.e. how many shortest paths pass through the node. The higher the betweenness of a node, the greater the risk of change propagation.

[0068] Clustering coefficient: measures the tightness of connections between a node and its neighbors. Nodes with high clustering coefficients may have faster change propagation.

[0069] Node set information degree: measures the amount of information contained in a specific node set, used to assess the breadth of change impact. Through these indicators, combined with expert knowledge and historical data, the propagation risk of changes from the initial node to the potential impact range is quantified, and the corresponding risk model is established.

[0070] Reverse variant network: a network model that traces back and plans the change path and strategy of function, structure, behavior, execution, etc. from the final performance requirement.

[0071] Change propagation: refers to the chain reaction caused by a local change in a complex system, leading to unexpected or expected changes in other functions, structures, behaviors, executions or performance elements.

[0072] Change propagation path: refers to the path of design change transmission and diffusion in a system.

[0073] Change propagation risk: refers to the possibility and severity of technical uncertainties and negative chain effects that may be encountered during the implementation of design or process changes, such as deterioration of other performance indicators, project delays, cost overruns, etc. In this invention, it can be quantitatively evaluated based on network topology characteristics such as node complexity, betweenness, clustering coefficient and node attributes such as information degree, historical failure rate, etc.

[0074] Change propagation time: the time required for a change to complete after passing through a series of propagation paths from the source.

[0075] Change propagation cost: the various fees incurred during the propagation of changes, including design, verification, production adjustment, etc. Specific embodiment 1:

[0077] The present application provides a kind of assembly line performance change demand driven reverse variant network modeling method, to realize the accurate analysis of assembly line performance change, efficient planning, risk control and multi-objective optimization.This method mainly includes the following steps:

[0078] As Figure 1 And Figure 2 A kind of assembly line performance change demand driven reverse variant network modeling method, comprising:

[0079] S1: performance requirement analysis and change problem identification; this step is the starting point of change management. First, R&D personnel or production managers will propose specific assembly line performance change requirements based on market feedback, new product development requirements or existing product performance bottlenecks, etc.

[0080] For example, the capacity of an engine assembly line needs to be increased by 20%, or the assembly accuracy of a gearbox assembly needs to be improved by one level. These macro performance requirements will be decomposed into more specific, measurable technical problems. For example, capacity improvement may be decomposed into: shortening the tact time of key processes, optimizing material distribution processes, improving equipment utilization, etc.; assembly accuracy improvement may be decomposed into: reducing the fit clearance of key matching parts, controlling deformation during assembly, reducing surface roughness, etc.

[0081] In this process, the specific manifestations of these technical problems in the dimensions of product function, structure, assembly behavior, execution process, and final performance need to be identified. For example, improving sealing may require changing the material of the sealing ring, which in turn affects the pre-tightening force during assembly, requiring adjustment of the parameters of the tightening tool, and ultimately resulting in a reduction in leakage rate. The material of the sealing ring mentioned above can be regarded as structure; the pre-tightening force is regarded as behavior; improving sealing is regarded as function; adjusting the parameters of the tightening tool is regarded as execution; and the leakage rate is regarded as performance.

[0082] S2: "Performance-Execution-Behavior-Structure-Function" Five-Dimensional Internal Association Network Transformation:

[0083] A "performance-execution-behavior-structure-function" five-dimensional product and assembly line association model is constructed, as shown in Figure 1 The model is the basis for realizing reverse variant planning.

[0084] Performance Dimension: Represents the macro performance indicators of the product or assembly line, such as capacity, quality pass rate, energy consumption, noise, reliability, etc.

[0085] Execution Dimension: Represents specific execution actions or equipment parameters on the assembly line, such as robot trajectory, tightening torque, welding current, glue application speed, sensor data, etc.

[0086] Behavior Dimension: Represents the dynamic characteristics and physical behavior of components or systems during assembly, such as fit resistance, friction, vibration, thermal deformation, tolerance accumulation, etc.

[0087] Structure Dimension: Represents the geometric dimensions, form and position tolerances, material properties, and connection relationships between components of the product.

[0088] Function Dimension: Represents the specific functions to be achieved by the product or component, such as support, sealing, transmission, electrical conduction, etc.

[0089] Using parameterization, modularization, and knowledge technologies, combined with assembly models, historical cases, and artificial intelligence methods, mapping and constraint relationships between elements of each dimension are established, and dispersed knowledge and data are transformed into structured association networks.

[0090] The core of the reverse variant is to trace back the driving factors and influence paths from performance requirements to execution, behavior, structure, and function through these mapping relationships, forming a complex network of associations, thereby achieving analysis from "fruit" to "cause". This network not only includes the internal connections of each dimension, but also emphasizes the dependency between dimensions, such as execution-behavior mapping, behavior-structure mapping, structure-function mapping, and performance directly related to execution and function, etc.

[0091] Performance-Execution Mapping: Establish a parameterized mapping relationship between macro performance indicators such as capacity, quality pass rate, energy consumption, noise, and reliability, and specific execution actions or device parameters such as robot trajectory, tightening torque, welding current, glue application speed, and sensor data. Through historical data analysis, expert experience, or simulation, the impact of execution parameter changes on performance indicators is quantified or qualitatively described.

[0092] Execution-Behavior Mapping: Establish a parameterized mapping relationship between specific execution actions or device parameters on the assembly line and the dynamic characteristics and physical behavior exhibited by components or systems, such as the insertion resistance of mating parts, friction, vibration, thermal deformation, and tolerance accumulation. Through physical models, mechanical analysis, or sensor data acquisition, the impact of execution on behavior is quantified.

[0093] Behavior-Structure Mapping: Establish a parameterized mapping relationship between the physical behavior of components or systems and the structural elements of products or assembly lines, such as geometric dimensions, form and position tolerances, material properties, and connection relationships between components. Through structural mechanics analysis, finite element simulation, or experimental data, the impact of behavior on structural requirements is quantified.

[0094] Structure-Function Mapping: Establish a parameterized mapping relationship between the structural elements of products or assembly lines and the specific functions that products or components are intended to achieve, such as support, sealing, transmission, and electrical conduction. Through function analysis, structural design specifications, or knowledge base, define the supporting role of structure on function.

[0095] Define intra-dimension and cross-dimension constraints:

[0096] Intra-dimension constraints: Define the limiting conditions between elements within each dimension, for example, in the structure dimension, the dimensions and form and position tolerances of components must meet the assembly fit requirements; in the execution dimension, device operating parameters must comply with safety and process specifications.

[0097] Cross-dimension logical constraints: Define the mutual dependence and logical restrictions between different dimension elements, for example, performance goals such as improving product pass rate impose strict constraints on execution parameters such as detection accuracy and behavior characteristics such as tolerance accumulation; or functional requirements impose restrictions on structural design and material selection.

[0098] Structured transformation of knowledge and data:

[0099] Parametric modeling and modular design: Parametric modeling and modular design of key components, modules, and process steps in the assembly line, abstracted as reusable, configurable element nodes.

[0100] Knowledge engineering and data integration: Extract, integrate, and structure each dimension of elements and their mapping and constraint relationships from heterogeneous data sources, such as assembly models (CAD models, BOMs, historical case databases), and artificial intelligence methods (machine learning, data mining, natural language processing), forming a unified, queryable knowledge graph or relational database.

[0101] Dynamic feedback and iterative optimization: Use system dynamics models or other simulation tools to simulate and verify the effectiveness of the established mapping and constraint relationships, and perform iterative optimization and adjustment based on simulation results to ensure that the five-dimensional correlation model accurately reflects the actual dynamic behavior of the assembly line.

[0102] For example, when the sealing performance of a product is required to be improved, the system will query the execution parameters that may affect the sealing performance, such as the amount of glue, behavior such as the curing characteristics of the glue layer, structure such as the design of the sealing groove, and even functions such as the dustproof function, and identify their correlation.

[0103] S3. Key assembly links and change transmission mechanism identification:

[0104] Quantitative evaluation of the surface quality of the key assembly surface and analysis of the surface lubrication condition;

[0105] Through finite element analysis or contact mechanics model, analyze the connection characteristics of the combined assembly surface during the assembly process, such as the actual contact area, contact pressure distribution, and contact deformation, to identify the influence of these factors on the assembly accuracy;

[0106] Use tolerance chain analysis, statistical tolerance analysis, or Monte Carlo simulation to analyze the transmission mechanism of three-dimensional geometric dimension and position tolerance change driving, identify the cumulative effect of size and position tolerance in the assembly process, and accurately identify the key nodes that affect accuracy and performance under the background of performance change;

[0107] For the identified key assembly links, build a dynamic feedback model between assembly process parameters, line configuration parameters, and assembly line performance to simulate the dynamic influence and feedback mechanism of different process parameters and line configuration changes on assembly cycle, equipment utilization, product qualification rate, and other performance indicators.

[0108] Specifically, surface engineering and connection characteristic analysis: quantitatively evaluate the surface quality of key assembly surfaces, including measuring surface roughness (Ra, Rz) and surface waviness (Wa, Wz) using a profilometer, and analyzing surface lubrication conditions. At the same time, through finite element analysis or contact mechanics model, analyze the connection characteristics of the combined assembly surface during assembly, such as actual contact area, contact pressure distribution and contact deformation, to understand the influence of these factors on assembly accuracy. For example, in the assembly of precision bearings, the roughness of the mating surface of the shaft and the hole is too large, which will cause poor contact, affecting the assembly accuracy and bearing life; insufficient contact area may cause local high stress, leading to early failure.

[0109] Dimension and geometric tolerance transfer analysis: in-depth analysis of the transfer mechanism driven by three-dimensional geometric dimension and geometric tolerance changes. Using methods such as tolerance chain analysis, statistical tolerance analysis or Monte Carlo simulation, identify the cumulative effect of dimensions and geometric tolerances during assembly, so as to accurately identify which assembly links are critical to accuracy and performance under performance changes. For example, a multi-stage assembled box, the small deviation of the geometric tolerance of the internal parts may lead to the final axis parallelism out of tolerance after multi-layer superposition.

[0110] System dynamics modeling: for the identified key assembly links, use system dynamics methods to build a dynamic feedback model between assembly process parameters, production line configuration parameters and assembly line performance. Simulate the dynamic influence and feedback mechanism of different process parameters, such as tightening torque, welding current and production line configuration, such as workstation layout, equipment selection, on performance indicators such as assembly cycle time, equipment utilization, product yield, etc.

[0111] The dynamic feedback model is a supplement and deepening of the dynamic relationship between the "execution-behavior-performance" dimensions in the five-dimensional correlation model. It is not an independent model for simulating how changes in assembly process parameters and production line configuration parameters affect assembly line behavior and dynamically affect assembly line performance in the key assembly links identified by the five-dimensional correlation model, and provides the basis for predicting and optimizing change schemes.

[0112] Determine the input and output variables of the dynamic feedback model, the input variables include assembly process parameters such as tightening torque, welding current, glue application speed, and production line configuration parameters such as workstation layout, equipment selection, production cycle, and the output variables include assembly line performance indicators such as assembly cycle time, equipment utilization, product yield, energy consumption and noise;

[0113] Collect historical data of input and output variables, and through data preprocessing including data cleaning, missing value filling and outlier processing to improve data quality;

[0114] Based on the historical data, a system dynamics method is used to establish the causal chain and feedback loop between the input and output variables, and a mathematical expression of the dynamic feedback model is constructed, which can include differential equations, difference equations or state space equations to describe the dynamic influence of parameter changes on performance indicators and feedback mechanisms.

[0115] The dynamic feedback model is parameterized and verified, and the model is trained and optimized using historical performance change data to ensure that the model can accurately reflect the dynamic response characteristics of the assembly line.

[0116] Specific model example: dynamic feedback model of engine cylinder head bolt tightening

[0117] Suppose on an engine assembly line, in order to improve the "airtightness pass rate" (a performance indicator) of the product, the "key assembly link" is identified as the "cylinder head bolt tightening" process. A dynamic feedback model is constructed for this link.

[0118] 1). Determine the input and output variables of the dynamic feedback model

[0119] Input variables:

[0120] Assembly process parameters: tightening torque output by tightening tool.

[0121] Line configuration parameters: tightening speed of tightening equipment.

[0122] Output variables:

[0123] Assembly line performance indicators: product pass rate (measured by the inverse of the "sealing pressure compliance rate" or "leakage rate" of the cylinder head) and assembly cycle.

[0124] 2). Construct causal chain and feedback loop

[0125] According to physical principles and process experience, the causal relationship and feedback mechanism between input and output variables can be established:

[0126] Causal chain 1 (performance path): tightening torque (T) determines the axial pre-tightening force (F) of the bolt. The pre-tightening force directly affects the sealing pressure (P) of the cylinder head gasket. The sealing pressure must be within a suitable range, too low will cause leakage, too high may damage the gasket or cylinder body, ultimately determines the product pass rate (Q).

[0127] Causal chain 2 (efficiency path): tightening speed (S) directly determines the time to complete a single bolt tightening operation, thereby affecting the total assembly cycle (B).

[0128] Feedback loop: Excessive pre-tightening force can cause a small structural deformation (D) of the cylinder. This deformation in turn can cause uneven distribution of the sealing pressure, which negatively impacts the product yield. This is a typical negative feedback loop.

[0129] 3) Mathematical expression of the dynamic feedback model

[0130] Based on the above causal relationships, a simplified mathematical expression (e.g., difference equation or state-space equation) can be constructed to describe this dynamic system.

[0131] Bolt pre-tightening force model:

[0132] F(t) = k1 · T(t) - k2 · S(t)

[0133] Where F(t) is the bolt pre-tightening force at time t. k1 is the torque conversion coefficient. k2 is the speed influence coefficient, indicating that higher tightening speed can cause the pre-tightening force to drop slightly or become unstable due to dynamic effects.

[0134] Sealing pressure and structural deformation model:

[0135] P(t) = k3 · F(t) - k4 · D(t-1)

[0136]

[0137] P(t) is the sealing pressure. D(t) is the cumulative deformation of the cylinder.

[0138] The sealing pressure is generated by the pre-tightening force, but is negatively affected by the structural deformation at the previous time (k4).

[0139] The deformation amount accumulates with the action of the pre-tightening force (k5), while it has a certain stress release (k6).

[0140] Product yield model:

[0141] Q(t) = f(P(t))

[0142] Q(t) is the product yield. It is a function f of the sealing pressure, usually a non-linear function. When P is within the optimal interval [P_min, P_max], Q is highest; outside this interval, it drops rapidly.

[0143] Assembly rhythm model:

[0144]

[0145] B(t) is the assembly rhythm. T_base is the fixed operation time of the workstation, and N is the number of bolts.

[0146] 4) Dynamic feedback model calibration and validation

[0147] Parameter calibration: use historical production data (such as different batches of tightening torque set value, tightening speed, corresponding product leakage rate and production cycle record) to fit and calibrate the undetermined coefficients (k1, k2, k3, k4, k5, k6) in the model.

[0148] Model validation: use another set of independent production data to verify the accuracy of the model, and ensure that the model can accurately reflect the dynamic response characteristics of the assembly line.

[0149] Through this verified dynamic feedback model, simulation can be performed before change: for example, input different tightening torque and tightening speed combinations to predict their comprehensive impact on product qualification rate and assembly cycle, so as to find the optimal process parameters and achieve the goal of performance change.

[0150] S4. Multi-objective planning of change propagation risk, time and cost:

[0151] The change propagation risk is evaluated and quantified, specifically, the key topological indicators of the function, structure, behavior, execution, performance elements and the like corresponding to each node in the network are calculated, including node complexity, node betweenness, clustering coefficient and node set information degree, so as to quantify the change propagation risk from the initial node to the potential influence range;

[0152] Quantify the time cost required from the demand proposal to the final completion, including analysis time, design time, implementation time and verification time; quantify the economic cost required from the demand proposal to the final completion, including R&D investment, equipment modification cost, material consumption and production loss.

[0153] In this embodiment, based on the constructed reverse variant network, the method of network science is used to quantify the change propagation risk. For each node in the network, i.e. function, structure, behavior, execution, performance element, its key topological indicators are calculated, such as node complexity, node betweenness, clustering coefficient and node set information degree. Through these indicators, the propagation risk of change from the initial node to the potential influence range is quantified. The time cost required from the demand proposal to the final completion is quantified, including analysis, design, implementation, verification and other stages, and the economic cost, including R&D investment, equipment modification, material consumption, production loss, etc., to evaluate the change propagation time and cost.

[0154] At the same time, a multi-objective optimization function is established which comprehensively considers the change propagation risk, change propagation time and change propagation cost. The function aims to find the optimal change strategy, so that the above three objectives are minimized or optimally balanced under certain weights, for example: min(ωR • Risk total + ω T • Time total + ω C • Cost total , where Risk total is the quantified total change propagation risk, Time total is the total change propagation time, Cost total is the total change propagation cost, ω R , ω T , and ω C are weight coefficients that can be flexibly set according to the strategic preferences of the enterprise, for example, prioritizing risk control, prioritizing time reduction, or prioritizing cost reduction.

[0155] S5: Multi-objective optimization algorithm for change propagation path

[0156] Optimization algorithm selection and solution: intelligent optimization algorithms are used to solve the multi-objective optimization function constructed in S4, such as genetic algorithms, particle swarm optimization algorithms, ant colony algorithms, deep learning, or reinforcement learning, etc. These algorithms can search for optimal solution sets in large-scale, high-dimensional solution spaces, effectively handling complex multi-objective optimization problems.

[0157] Reverse variant network construction and decision-making: the output of the optimization algorithm will be a set of Pareto optimal solutions, each solution representing an optimal change strategy that balances risk, time, and cost. Ultimately, the invention constructs a performance change-oriented "function-structure-behavior-execution-performance" reverse variant network. This network not only clearly shows which functions, structures, behaviors, and execution elements need to be changed to achieve a specific performance improvement or adjustment, but more importantly, it provides optimal change paths, priorities, and specific operation guidance, making the change decision-making process transparent and intelligent. For example, when noise reduction is needed, the network can point out key structures that may affect noise, such as bearing types, behaviors such as assembly clearance, execution such as tightening torque, and even functions such as vibration reduction, and recommend an optimal change combination to minimize risk and cost.

[0158] The invention provides a forward-looking, systematic, and intelligent solution for manufacturing enterprises to respond to complex and variable assembly line performance change demands, which has important strategic significance and practical application value for improving production efficiency, reducing operating costs, and enhancing market competitiveness. Specific embodiment 2:

[0160] The present disclosure also provides an embodiment:

[0161] As Figure 3The application discloses an assembly line performance change demand driven reverse modification network system based on the assembly line performance change demand driven reverse modification network modeling method in the specific embodiment 1, and comprises a collection module 100, a first processing module 200, an identification module 300, a second processing module 400, a third processing module 500 and an output module 600. The collection module 100 is used for acquiring performance change demand information of an assembly line to be changed. The first processing module 200 interacts with the collection module 100, and is used for constructing a five-dimensional correlation model matched with the assembly line to be changed according to the performance change demand information. The identification module 300 interacts with the first processing module 200, and is used for identifying a key assembly link and a change transmission mechanism, and constructing a dynamic feedback model between assembly process parameters, line configuration parameters and assembly line performance. The second processing module 400 interacts with the first processing module 200, and is used for acquiring a topological index of any node in a network corresponding to the five-dimensional correlation model according to the five-dimensional correlation model, quantifying a propagation risk from an initial node to a potential influence range, and establishing an optimization function based on a change propagation risk, a change propagation time and a change propagation cost. The third processing module 500 interacts with the second processing module 400, and is used for solving the optimization function and outputting an optimal solution set. The output module 600 interacts with the third processing module 500, and is used for outputting any solution in the optimal solution set and acquiring an optimal change strategy balanced between risks, time and cost.

[0162] The above disclosure is only several specific implementation scenarios of the present disclosure, but the present disclosure is not limited thereto, and any changes that can be thought of by those skilled in the art should fall within the protection scope of the present disclosure. The above serial numbers of the present disclosure are only for description, and do not represent the advantages and disadvantages of the implementation scenarios.

Claims

1. A method for modeling inverse variant networks driven by assembly line performance change requirements, characterized in that, include: Obtain performance change requirements for the assembly line to be modified; Based on the performance change requirement information, a five-dimensional association model matching the assembly line to be changed is constructed; The five-dimensional association model includes: performance dimension, execution dimension, behavior dimension, structure dimension and function dimension, and establishes mapping and constraint relationships between elements of each dimension, transforming scattered knowledge and data into a structured five-dimensional association model; Identify key assembly steps and change transmission mechanisms, and construct a dynamic feedback model between assembly process parameters, production line configuration parameters, and assembly line performance. Based on the five-dimensional correlation model, the topological index of any node in the network corresponding to the five-dimensional correlation model is obtained, and the propagation risk from the initial node to the potential influence range is quantified; and an optimization function based on change propagation risk, change propagation time and change propagation cost is established. Solve the optimization function to output a set of optimal solutions; any solution in the optimal solution set represents an optimal change strategy that balances risk, time and cost.

2. The reverse variant network modeling method driven by assembly line performance change requirements according to claim 1, characterized in that, The process of establishing mapping and constraint relationships between elements of each dimension, transforming scattered knowledge and data into a structured five-dimensional relational model, includes: Establish cross-dimensional parameterized mappings, including: performance-execution mapping, execution-behavior mapping, behavior-structure mapping, and structure-function mapping; Set constraints between elements within any dimension; and define interdependencies and logical constraints between elements in different dimensions; The key components, modules, and process steps in the assembly line are parametrically modeled and modularly designed, and abstracted into reusable and configurable element nodes. Based on the elements in each dimension, cross-dimensional parameterized mapping, and the constraints between elements within each dimension, a knowledge graph or relational database is formed. The effectiveness of the established cross-dimensional parametric mapping and the constraint relationships of the internal elements of each dimension is simulated and verified. Based on the simulation results, iterative optimization and adjustment are carried out to ensure that the five-dimensional correlation model can accurately reflect the actual dynamic behavior of the assembly line.

3. The reverse variant network modeling method driven by assembly line performance change requirements according to claim 1, characterized in that, The mechanism for identifying key assembly processes and change transfer includes: The surface quality of key assembly surfaces is quantitatively assessed, and the surface lubrication status is analyzed. Determine the connection characteristics of the key assembly surfaces during the assembly process to identify factors that affect assembly accuracy; Determine the transmission mechanism driven by changes in three-dimensional geometric dimensions and geometric tolerances, identify the cumulative effect of dimensions and geometric tolerances in the assembly process, and identify key assembly links that affect accuracy and performance; Based on the key assembly steps, a dynamic feedback model is constructed between assembly process parameters, production line configuration parameters, and assembly line performance to simulate the dynamic impact and feedback mechanism of different process parameters and production line configuration changes on assembly cycle time, equipment utilization rate, and product qualification rate.

4. The reverse variant network modeling method driven by assembly line performance change requirements according to claim 1, characterized in that, The construction of a dynamic feedback model between assembly process parameters, production line configuration parameters, and assembly line performance includes: The input variables of the dynamic feedback model include: assembly process parameters and production line configuration parameters; The output variables of the dynamic feedback model include: assembly line performance indicators; Based on the causal chain and feedback loop between the input and output variables, construct the mathematical expression of the dynamic feedback model; The dynamic feedback model is calibrated and validated, and historical performance change data is used to train and optimize the model to ensure that the model can accurately reflect the dynamic response characteristics of the assembly line.

5. The reverse variant network modeling method driven by assembly line performance change requirements according to claim 1, characterized in that, The step of obtaining the topological index of any node in the network corresponding to the five-dimensional correlation model, based on the five-dimensional correlation model, and quantifying the propagation risk from the initial node to the potential impact range includes: For any node in the network, obtain its key topology indicators; the key topology indicators include: node complexity, node betweenness, clustering coefficient and node set information degree, thereby quantifying the risk of change propagation from the initial node to the potential impact range; The time cost required to quantify a change from the time the requirement is proposed to its final completion includes analysis time, design time, implementation time, and verification time. The economic costs required to quantify the change from the initial proposal to its final completion include R&D investment, equipment modification costs, material consumption, and production downtime losses.

6. The reverse variant network modeling method driven by assembly line performance change requirements according to claim 1, characterized in that, The establishment of the optimization function based on change propagation risk, change propagation time, and change propagation cost includes: The optimization function is: min(ω R ·Risk total +oh T ·Time total +oh C ·Cost total ); Among them, Risk total For the quantified total change propagation risk; Time total Total change propagation time; Cost total Total cost of change propagation; ω R ω T ω C These are the weighting coefficients.

7. A reverse variant network system driven by assembly line performance change requirements, based on the reverse variant network modeling method driven by assembly line performance change requirements according to any one of claims 1-6, characterized in that, include: The data acquisition module obtains performance change requirements for the assembly line to be modified. The first processing module interacts with the acquisition module to construct a five-dimensional association model that matches the assembly line to be changed based on the performance change requirement information. The identification module interacts with the first processing module to identify key assembly links and change transmission mechanisms, and to build a dynamic feedback model between assembly process parameters, production line configuration parameters and assembly line performance. The second processing module interacts with the first processing module to obtain the topological index of any node in the network corresponding to the five-dimensional association model, and quantifies the propagation risk from the initial node to the potential impact range based on the five-dimensional association model. And establish an optimization function based on change propagation risk, change propagation time, and change propagation cost; The third processing module interacts with the second processing module to solve the optimization function and output a set of optimal solutions. The output module interacts with the third processing module to output any solution in the optimal solution set, thereby obtaining an optimal change strategy that balances risk, time, and cost.

Images