GB/T 47472-2026Intelligent assembly platform architecture for complex products (English PDF)

6.1 Structural Subsystem

6.1.1 Overview The structural subsystem consists of five functional modules and three types of module interfaces. The functional modules mainly include the basic support module and the positioning clamp. The system includes a support module, a motion execution module, a logistics conveying module, and auxiliary modules. Module interfaces are divided into fixed connection interfaces and detachable connection interfaces. And motion connection interfaces. Each functional module is organically integrated through module interfaces to achieve complete assembly functionality from static support to dynamic execution. The specific components of the structural subsystem are shown in Figure 3. Figure

6.1.2 Functional Modules

6.1.2.1 Basic Support Module The basic support module is the foundation of the intelligent assembly platform for complex products, providing a stable and reliable basic framework and mounting base for the entire system. Accuracy. Other modules should be mounted directly or indirectly on top of this module to ensure that geometric accuracy and stability are maintained under static/dynamic loads. The design of this module must fully consider static/dynamic stiffness, natural frequency matching, and thermal stability, including the frame, base, housing, and moving platform. Qualitative properties and vibration reduction/isolation factors directly determine the upper limit of the platform's accuracy and long-term stability.

6.1.2.2 Positioning and clamping module The positioning and clamping module is responsible for determining and maintaining the precise spatial pose of components during the assembly process, providing consistent positioning for all assembly operations. Repeatable references. This includes flexible fixtures, fixed fixtures, and fixture families adapted to hold a variety of parts with varying specifications. This module requires... It can adapt to the production needs of multiple varieties and variable batches, and achieve flexibility in the assembly process through rapid model changeover.

6.1.2.3 Motion Execution Module The motion execution module provides the components needed to perform assembly actions, enabling precise execution of physical assembly actions according to predetermined paths and methods. This includes serial and parallel robots, running modules, slides, transmission mechanisms, and end effectors. The performance of this module directly determines the assembly speed. Precision and flexibility.

6.1.2.4 Logistics Conveying Module The logistics and conveying module performs the transfer, loading, and unloading of materials, components, and finished products between workstations and warehouses according to the assembly process, forming a continuous flow. The production flow includes conveyor lines, elevators, rotary tables, etc. This module typically integrates automatic identification technologies such as RFID and barcodes, and can communicate with upper-level management systems. Real-time interaction with the management system ensures the accuracy, efficiency, and traceability of material flow, which is key to forming a continuous and efficient production flow.

6.1.2.5 Auxiliary Supporting Modules Auxiliary modules should provide necessary support functions for the assembly process, such as energy transmission and safety protection, including cables and conduits. Safety fences, protective nets, safety gates, etc. The design of this module directly affects the reliability, ease of maintenance, and operational safety of the entire system.

6.1.3 Module Interface

6.1.3.1 Fixed connection interface Used for permanent or rigid connections between modules, requiring little or no disassembly between modules. Includes welding, bolted connections, and standard flanges. Connection methods such as linking.

6.1.3.2 Detachable connection interface Interface patterns used to achieve flexibility and modularity should ensure both rapid disassembly and high-precision reinstallation. This includes machine... Quick-change disc interface for end effectors, quick-change docking interface for tooling, modular slot positioning interface, etc.

6.1.3.3 Motion Connection Interface It provides precise relative motion while ensuring connectivity between functional modules. This includes precision linear guide pairs, precision ball screw pairs, and precision... Rotating shaft systems, etc.

6.2 Control Subsystem

6.2.1 Overview The control subsystem consists of three parts. hardware modules, software modules, and module interfaces. The hardware modules enable communication with the structural subsystem. Connection and control, including the main control computer, controller, driver, electrical components, motor, and data conversion module; software modules can realize the structure. The execution and coordination of the subsystem, data interaction with the sensing subsystem, the intelligent process decision-making and optimization subsystem, and the basic information subsystem, including the use of User management, multithreaded programming, control algorithms, human-computer interaction interfaces, etc.; module interfaces enable communication between the control subsystem and various subsystems and peripherals. Signal, data, and command transmission, including input interfaces, output interfaces, communication interfaces, debugging interfaces, special interfaces, and interface protocols. Control subsystem. The specific components of the system are shown in Figure 4. Figure

6.2.2 Hardware Module

6.2.2.1 Main Control Computer The main control computer is the brain of the entire system, providing interfaces for the control system hardware modules, processing various sensor data, and operating... Interface display screen and software operating environment.

6.2.2.2 Controller The controller is a hardware and software system with the main control computer as its core. By integrating it into a network, it can realize motion control and input/output. Signal control and sensor control, etc., and the control module can be reconfigured through the combination of different controllers. The controller interacts with various devices via communication protocols. sending precise position, speed, or torque commands to the actuators; and communicating via data... The digital/analog input/output module connects to control solenoid valves, relays, and other switching electrical components, or reads status signals from buttons, limit switches, etc. The system can control the switching or brightness adjustment of machine vision lighting; and collect feedback signals from force sensors, temperature sensors, displacement sensors, etc. in real time. The system performs data processing, logical judgment, and closed-loop control.

6.2.2.3 Driver Based on the signals sent by the controller, power amplification and feedback control are performed, enabling control of the motor's position, speed, and torque. It can monitor the status, faults, and abnormalities of equipment, and perform diagnosis and handling. It can also monitor the position, speed, load, and other parameters of actuators in real time. The system counts data and performs fault detection and anomaly handling based on set rules and algorithms.

6.2.2.4 Electrical Components Electrical components mainly include power supplies, switches, relays, capacitors, resistors, fuses, circuit breakers, connectors, and plugs.

6.2.2.5 Motor It serves as the power source for the motion execution module within the structural subsystem.

6.2.2.6 Data Conversion Module Tools used for converting between different data formats.

6.2.3 Software Modules

6.2.3.1 User Management Responsible for managing user permissions and authentication, including user registration, login, and access control functions.

6.2.3.2 Multithreaded Programming Parallel control of the structured subsystem is achieved through multithreaded programming, including thread management, thread synchronization, message passing, event handling, and data processing. Shared and shared memory, as well as error handling and exception handling, etc. This parallel framework is the core operating mechanism of the controller system software, which can efficiently coordinate and schedule multiple critical tasks.

a) Allocate high-priority real-time threads to motion control algorithms with extremely high real-time requirements to ensure precise timing control;

b) Create dedicated threads or thread pools for image processing algorithms to fully utilize multi-core processor resources and avoid blocking the main control loop;

c) Through a carefully designed thread synchronization and message passing mechanism, ensure smooth communication between motion control, image processing, and the main logic of the controller. It can exchange data securely, efficiently, and in real time and work collaboratively.

6.2.3.3 Motion Control Algorithm The control algorithms include path planning algorithms, multi-axis interpolation algorithms, and servo loop control algorithms. The control algorithm is the core of the control subsystem, which can transform target position, velocity profile, and process requirements into precise, coordinated, and real-time shaft systems. Motion commands. It comprises three levels. path planning algorithm, multi-axis interpolation algorithm, and servo control algorithm. Path planning algorithm. In a Cartesian coordinate system, it can generate paths based on the starting point, ending point, obstacle avoidance constraints, speed/acceleration limits, process requirements, etc. To achieve the desired spatial motion trajectory. This includes point-to-point movement, generation and optimization of continuous paths (such as straight lines, arcs, and spline curves), and speed... Proactive control is used to reduce impact and improve efficiency. Multi-axis interpolation algorithm. It can decompose the spatial trajectory and velocity profile generated by path planning into various motion axes according to the time axis at each control point. It specifies the precise position, speed, and acceleration settings to be achieved within the control cycle. It is responsible for coordinating the strict synchronization relationships between multiple axes (such as linear and circular axes). Arc and spiral motions ensure spatial accuracy and smoothness of complex trajectories. Servo loop control algorithm. It can receive the setpoint output from interpolation, combine it with actual position/velocity feedback from encoders, linear scales, etc., and then... The closed-loop control strategy calculates and outputs the final torque/current command to the driver in real time, accurately tracking the setpoint and overcoming load disturbances, friction, etc. Factors such as mechanical resonance are taken into account to ensure the final execution accuracy and dynamic response performance. These three levels work closely together to form a complete motion control solution, enabling the system to complete various tasks efficiently, accurately, and flexibly. A complex movement task.

6.2.3.4 Human-Computer Interaction Interface A software module with a user interface to allow operators to interact with and monitor the system. The user interface is accessed via a touchscreen display and control panel. Presented in the form of panels, graphical interfaces, etc., operators use the interface to control the intelligent assembly platform, set parameters, and handle faults. The interface design should include a graphical user interface, menus, buttons, text boxes, charts, etc. It considers user ease of operation and visual appeal, using layout... The layout, style, and visual elements are used to achieve a user-friendly interface design.

6.2.4 Module Interface

6.2.4.1 Input Interface As a physical functional interface, it is used to control the controller I/O circuits within the subsystem to obtain signals and data from peripherals, and can connect the structural subsystems. The hardware sensing signals in the system are safely and accurately sent to the controller, such as the data signals collected by the sensors.

6.2.4.2 Output Interface As a physical functional interface, it is used to send signals and data from the controller I/O circuit within the control subsystem to peripheral devices, enabling the controller to... The command signal is converted into a power signal to drive other subsystems, such as the control command sent to the actuator.

6.2.4.3 Communication Interface It has interfaces for communication with various modules in the structural subsystem and sensing subsystem, enabling data exchange and communication with other devices. Real-time performance, reliability, and accuracy are paramount. Communication interfaces may include fieldbus interfaces, Ethernet interfaces, serial communication interfaces, wireless communication interfaces, and dedicated interfaces. Use protocol gateways, etc.

6.2.4.4 Debugging Interface Used for interacting with devices based on specific protocols during the debugging of control subsystem hardware or software, enabling program downloading and online debugging. It includes functions such as testing and troubleshooting. Hardware debugging interfaces may include JTAG, SWD, etc.; software debugging interfaces may include MeterSphere. (An open-source continuous testing tool), Visual Studio (An integrated development environment), etc.

6.2.4.5 Special Interfaces Interfaces specifically designed for particular functional modules, such as temperature and humidity measurement module interfaces, safety I/O module interfaces, etc.

6.2.4.6 Interface Protocol An interface protocol is the language and rules governing data exchange between different modules. It should possess characteristics such as strong real-time performance, high reliability, and good anti-interference capabilities. Features. According to communication interface type, commonly used fieldbus interface protocols include PROFIBUS, Modbus, etc.; Ethernet interface protocols include... PROFINET, EtherCAT, etc.; serial communication interface protocols include UART, RS-232 (asynchronous transmission standard interface), CAN, etc.; wireless... Communication interface protocols include WAPI, Bluetooth, and 5G; proprietary protocols include IO-Link (a standardized sensor and actuator communication technology). OPCUA, etc.

6.3 Sensing Subsystem

6.3.1 Overview The perception subsystem consists of two parts. functional modules and module interfaces. The functional modules mainly include vision sensors and force/torque sensors. Sensors, including position/displacement sensors, environmental condition monitoring sensors, identification and tracking sensors, and safety protection sensors; the module interface mainly includes... This includes mechanical mounting interfaces, electrical and signal interfaces, and data communication interfaces. The functional modules of the sensing subsystem are fixed to the structure via the mechanical mounting interfaces. On the structural subsystem, it connects to the control subsystem via electrical and signal interfaces, and transmits the collected data to the basic information system via a data communication interface. The sensory subsystem is shown in Figure 5.

6.3.2 Functional Modules

6.3.2.1 Vision Sensor Acquiring image information of the assembly target and converting it into digital signals can be used for identification, positioning, and detection. It is mainly divided into 2D cameras and 3D cameras. Cameras fall into two main categories. 2D cameras analyze grayscale or color information to perform surface defect detection, character recognition, and precise two-dimensional positioning; 3D cameras... Acquire depth information to generate 3D point cloud data for 3D guidance, disordered grasping, high-precision dimensional measurement, and 3D defect analysis of complex objects. Analysis. Vision sensors are non-contact measurement devices and are the primary sensing means for achieving automated, intelligent quality control and flexible assembly. Their selection requires... Taking into account the effects of resolution, accuracy, speed, and ambient light, the image information acquired by the vision sensor undergoes image preprocessing and feature extraction. After extraction and discriminant analysis, precise motion adjustment commands are provided to the structural subsystem to achieve adaptive adjustment of the assembly position.

6.3.2.2 Force/Torque Sensor Measuring the interaction forces and torques between the end effector and the assembled components. This includes tension/compression sensors, torque sensors, and more. Force sensors, etc. Force/torque sensors enable intelligent assembly platforms to perform complex tasks requiring force feedback, such as shaft hole insertion and force-controlled machining. Grinding, precision pressing, drag teaching, and collision detection are the core components for achieving compliant assembly, ensuring product quality, and ensuring operational safety.

6.3.2.3 Position/Displacement Sensor Used to detect the presence, distance, angle, or minute displacement changes of objects. Includes photoelectric sensors, ultrasonic sensors, proximity switches, etc. Encoders, laser displacement sensors, etc. Position/displacement sensors provide data on the relative spatial relationship between the actuator and the target object. According to the data, it is a prerequisite for achieving precise action execution and process control.

6.3.2.4 Environmental Condition Monitoring Sensors Used to monitor the overall environmental status of the assembly unit, ensuring stable precision assembly process conditions and equipment health. Includes temperature sensors, humidity sensors, etc. Temperature sensors, vibration sensors, pressure sensors, etc.

6.3.2.5 Identification and Tracking Sensors Used to identify workpiece identity information, enabling material traceability. Includes RFID readers, barcode/QR code readers, etc. RFID uses radio waves to read and write information on carrier tags without direct contact. Its advantages include no need for direct viewing, batch reading capability, resistance to oil stains, and large data storage capacity. Large scale is the core of achieving end-to-end material tracking. Barcode readers are a cost-effective visual recognition solution, typically installed at workstations. On the inlet or conveyor line, the system automatically identifies the workpiece type, batch, and process parameters, and binds the physical materials with the digital information in the MES/ERP system. Determination is key to achieving flexible production and lean management.

6.3.2.6 Safety Protection Sensors Devices used to protect operator safety include safety light curtains/light screens and safety laser scanners.

6.3.3 Module Interface

6.3.3.1 Mechanical mounting interface Provide stable, accurate, and repeatable physical mounting and positioning for the sensor, ensuring that its observation reference is consistent with the coordinate system of the structural subsystem. This includes standard optical interfaces, mounting brackets and adjustment mechanisms, standard flanges, quick-change interfaces, etc.

6.3.3.2 Electrical and Signal Interfaces It powers the sensor and transmits the raw analog or digital signals it generates. This includes a power interface, an analog signal interface, and digital I/O interfaces. Mouth, etc.

6.3.3.3 Data Communication Interface This enables bidirectional, structured data exchange between sensors, data acquisition devices, and a host computer or main controller. The transmitted data is no longer the raw information. It's not a number, but an encapsulated data packet. Interface types include Industrial Ethernet protocol, fieldbus, point-to-point sensor bus, and wireless connection. Mouth, etc.

6.4 Intelligent Process Decision and Optimization Subsystem

6.4.1 Overview The intelligent process decision-making and optimization subsystem mainly includes a virtual and physical data layer, a virtual and physical interaction layer, a precise digital twin assembly model layer, and an assembly knowledge layer. The intelligent process decision-making and optimization subsystem comprises modules such as the model recognition layer, the intelligent decision-making layer for process parameters, and the process flow optimization layer. The intelligent process decision-making and optimization subsystem starts from the basic information subsystem. Data is acquired from the integrated sensing subsystem, with product performance and production efficiency as the main optimization objectives, to intelligently determine assembly process parameters. The strategy optimizes the assembly process, defining and outputting process parameters and the entire process flow for the control subsystem to use. The specific components of the energy process decision and optimization subsystem are shown in Figure 6.

6.4.2 Functional Modules

6.4.2.1 Virtual and Physical Data Layer This dataset contains complex product assembly information, comprising heterogeneous data from multiple sources in both physical and virtual spaces. The physical space data is not limited to Limited by theoretical design information, measured assembly parameters, module operating parameters, control signals, and assembly history data. The data in the virtual space includes... Simulation, prediction, and verification data for the twin assembly model. The data in the dataset undergoes preprocessing steps such as cleaning, fusion, dimensionality reduction, and transformation. After the holiday, data should be stored in the virtual and physical data layers in a hierarchical, categorized, or normalized manner.

6.4.2.2 Virtual-Real Interaction Layer This layer realizes the virtual-physical mapping of complex product assembly processes, transmitting twin data from the physical space to the twin assembly model layer, and utilizing... The decision-making information in the virtual space controls various activities in the actual assembly process.

6.4.2.3 Precise Digital Twin Assembly Model Layer This layer, based on the physical space information of the virtual and real data layers, utilizes analytical methods such as 3D modeling, finite element method, and intelligent mapping algorithms to perform analysis in the virtual... High-precision spatial mapping of complex product assembly processes integrates influencing factors from multiple disciplines, multiple physics fields, and multiple scales. Its core is to establish accurate data... The digital twin model should cover the actual geometric errors (including assembly contact surface distribution errors), dimensional errors, and actual assembly errors caused by the machining of parts. Mechanical states are used to create an intuitive, easy-to-use data carrier that can be easily integrated with 3D design models. Virtual simulation is used to generate assembly accuracy, stress, and other parameters. Predictive data on performance indicators and efficiency provide a basis for assembly process optimization, performance prediction, and parameter control, comprehensively reflecting the entire lifecycle of product assembly. cycle.

6.4.2.4 Assembling the Knowledge Model Layer The assembly knowledge model construction, guided by the top-level ontology architecture of the assembly process, involves establishing a basic knowledge model for product assembly and assembly data. Source knowledge model and assembly process knowledge model. The product assembly basic knowledge model covers the product's basic components, assembly relationships, and process characteristics, etc. The assembly resource knowledge model should describe the personnel, equipment, materials, and environment involved in the assembly process. The assembly process knowledge model... The process route and process parameters should be described.

6.4.2.5 Intelligent Decision-Making Layer for Process Parameters The intelligent decision-making for process parameters is based on the established precise digital twin model of the product assembly process. This is achieved by acquiring the empty... The system intelligently simulates assembly actions based on process attributes such as constraints, dimensions, mating relationships, assembly movements, and time, and provides assembly parameters under different machining error conditions. The assembly performance indicators corresponding to the process parameters are used to establish a nonlinear mapping relationship between assembly process parameters and performance indicators. Based on the computational data, knowledge reasoning and matching are performed in the assembly process design based on the assembly knowledge model to obtain the optimal assembly process. parameter.

6.4.2.6 Process Flow Optimization Layer The goal is to minimize assembly time by optimizing the process flow, thereby improving overall assembly efficiency. This is described in the assembly process knowledge model. The constraints between different process flows were identified. By incorporating these constraints into a time-series optimization algorithm, the optimal assembly can be obtained quickly and accurately. Based on the timing sequence, and then the process documents generated by the basic information subsystem, control instruction files are quickly generated.

6.4.3 Module Interface The intelligent process decision-making and optimization subsystem has an interface for interaction with the sensing subsystem, and can receive real-time assembly data collected by the sensing subsystem. Source data and assembly process data are used for process decision-making and optimization. The intelligent process decision-making and optimization subsystem has an interface with the basic information subsystem, and can receive basic product assembly data and assembly information. Source data, assembly process data, and quality and traceability data are used for process decision-making and optimization. Product assembly basics and assembly resources are also sent. Knowledge, assembly process knowledge, etc., are stored...