Electrical component assembly guiding method and system

Abstract

The invention relates to the technical field of electrical component assembly, and particularly discloses an electrical component assembly guiding method and system. Electrical component digital design model data and assembly site real-time environment perception data are obtained, an initial assembly sequence and an assembly path are generated according to the data, then an assembly environment adaptation factor is obtained by comparing a real-time environment with the assembly path, calculating an interference proportion and an unreachable path proportion and weighting, the path is corrected accordingly, and the assembly sequence and the assembly path are obtained. The method comprises the following steps: forming a corrected assembly guide path adaptive to a current environment, then quantifying appearance, mechanics and pose deviations and weighting through an assembly image, force / moment and position / attitude data to obtain a quality risk coefficient, and finally generating a cooperative control signal in combination with the corrected path and the quality risk coefficient to drive execution units such as a mechanical arm to act. Through real-time feedback fine tuning and step circulation, full-process assembly is completed, and the problems of poor adaptability to a complex electrical component assembly environment, incomplete quality control and insufficient operation precision are solved.

Description

Technical Field

[0001] This invention relates to the field of electrical component assembly technology, and in particular to an electrical component assembly guidance method and system. Background Technology

[0002] As core functional modules of high-end equipment, new energy systems, and power equipment, electrical components directly determine the operational performance, reliability, and safety of end products through their assembly quality, playing an irreplaceable role in key areas such as industrial production, energy supply, and intelligent manufacturing. With technological iteration and upgrading, the structure of electrical components is becoming increasingly complex, the number of sub-components is increasing significantly, and assembly constraints (such as geometric constraints, process constraints, and timing constraints) are becoming more diverse, continuously raising the requirements for assembly accuracy, process stability, and efficiency.

[0003] Current electrical component assembly methods are primarily manual, relying heavily on the operator's skill level and experience. When faced with assembly tasks involving multiple sub-components and complex constraints, manual assembly is prone to errors such as incorrect assembly, omissions, or deviations in assembly sequence. This is especially problematic when product models change frequently, leading to lengthy retraining cycles and high adaptation costs for operators, hindering rapid response to production demands. Furthermore, manual operation suffers from poor consistency in controlling key process parameters (such as bolt tightening torque, connector insertion / extraction force, and component alignment accuracy), resulting in significant quality dispersion and failing to meet the stringent process standards of high-reliability products. Therefore, a guided assembly method for electrical components is needed to address these issues. Summary of the Invention

[0004] The purpose of this invention is to provide a method for guiding the assembly of electrical components, comprising:

[0005] An electrical component assembly guidance method, comprising:

[0006] Acquire digital design model data of electrical components to be assembled and real-time environmental perception data of the assembly site;

[0007] The hierarchical assembly relationship of electrical components is obtained based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and the assembly path;

[0008] The assembly environment adaptation factor is obtained based on the real-time environmental perception data and the assembly path.

[0009] The assembly path is corrected based on the assembly environment adaptation factor and the initial assembly sequence to obtain a corrected assembly guidance path;

[0010] Acquire real-time process quality monitoring data during the assembly process, and obtain a quality risk coefficient based on the real-time process quality monitoring data and preset standard process parameter thresholds;

[0011] A coordinated control signal is generated based on the quality risk coefficient and the modified assembly guidance path, and the electrical components are controlled to perform assembly guidance according to the coordinated control signal until the assembly is completed.

[0012] Preferably, the step of obtaining the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes the steps of initial assembly sequence and assembly path, includes:

[0013] A topology analysis is performed on the digital design model data to obtain the assembly relationships of multiple sub-components, and an assembly relationship topology diagram is constructed based on the multiple assembly relationships.

[0014] The initial assembly sequence is obtained based on the assembly relationship topology diagram and preset assembly constraints;

[0015] The three-dimensional geometric information and target assembly pose coordinates of each sub-component are obtained according to the initial assembly sequence, and the preliminary spatial trajectories of multiple sub-components are obtained according to multiple sets of three-dimensional geometric information and multiple sets of target assembly pose coordinates.

[0016] Multiple smoothing coefficients between adjacent trajectories are obtained based on the multiple preliminary spatial trajectories;

[0017] The assembly path is obtained by splicing and smoothing the multiple preliminary spatial trajectories according to the multiple smoothing coefficients.

[0018] Preferably, the step of obtaining the assembly environment adaptation factor based on the real-time environment perception data and the assembly path includes:

[0019] The installation movement trajectories of multiple sub-components to be assembled are obtained based on the real-time environmental perception data. Overlapping position areas are obtained based on the multiple installation movement trajectories and the assembly path. Interference ratio is obtained based on the overlapping position areas and the preset total installation area. The interference ratio is used as the interference severity value.

[0020] Based on the assembly path, obtain the passable area information in the current environment, and based on the passable area information and the preset assembly space area, obtain the set of unreachable path points;

[0021] The total number of assembly path points is obtained based on the preset assembly space area, and the proportion of unreachable paths is obtained based on the set of unreachable path points and the total number of assembly path points.

[0022] The assembly environment adaptation factor is obtained by weighted fusion calculation based on the interference severity value and the proportion of unreachable paths.

[0023] Preferably, the step of correcting the assembly path based on the assembly environment adaptation factor and the initial assembly sequence to obtain a corrected assembly guidance path includes:

[0024] Obtain a preset adaptability level classification rule, wherein the classification rule includes multiple level intervals and the corresponding adaptation factor threshold range for each interval, and compare the assembly environment adaptation factor with multiple adaptation factor threshold ranges to obtain the adaptability level of the current assembly environment.

[0025] The set of assembly stages to be adjusted is obtained based on the adaptability level and the initial assembly sequence;

[0026] For each assembly stage in the assembly stage set, obtain the set of alternative assembly poses for the corresponding sub-component.

[0027] Based on the set of alternative assembly poses and the assembly path, path replanning is performed to obtain multiple alternative paths;

[0028] Obtain path adaptation references for multiple candidate paths, wherein the path adaptation references include path quality adaptation coefficients and assembly constraint completion coefficients, and obtain multiple comprehensive path adaptation coefficients based on the multiple path quality adaptation coefficients and the multiple assembly constraint completion coefficients.

[0029] Each comprehensive path adaptation coefficient is corrected sequentially according to the assembly environment adaptation factor to obtain multiple corrected comprehensive path adaptation coefficients;

[0030] The adaptation coefficients of multiple correction paths are sorted, and the candidate correction path with the highest adaptation coefficient is selected as the correction assembly guidance path.

[0031] Preferably, the step of acquiring real-time process quality monitoring data during the assembly process and obtaining a quality risk coefficient based on the real-time process quality monitoring data and a preset standard process parameter threshold includes:

[0032] Acquire multimodal process quality monitoring data collected in real time during the assembly process, wherein the multimodal process quality monitoring data includes assembly image data, assembly force / torque data, and assembly position / orientation data;

[0033] The visual appearance features of the assembly area are extracted based on the assembly image data, and the appearance quality deviation value is obtained based on the visual appearance features and the preset standard appearance feature threshold.

[0034] Based on the assembly force / torque data, extract the force / torque time sequence variation characteristics during the assembly process, and obtain the mechanical process deviation value based on the force / torque time sequence variation characteristics and the preset standard force / torque curve threshold.

[0035] Based on the assembly position / attitude data, extract the pose deviation features between the actual assembly pose and the target assembly pose, and obtain the pose deviation value based on the pose deviation features and the preset standard pose tolerance threshold.

[0036] The quality risk coefficient is obtained by weighted fusion calculation based on the appearance quality deviation value, the mechanical process deviation value, and the posture deviation value.

[0037] Preferably, the step of generating a coordinated control signal based on the quality risk coefficient and the modified assembly guidance path, and controlling the electrical components to perform assembly guidance according to the coordinated control signal until the assembly is completed, includes:

[0038] Based on the modified assembly guidance path, a basic control instruction sequence containing path points, timestamps, and state vectors is generated;

[0039] The basic control command sequence is adjusted based on the quality risk coefficient to obtain an adjusted control command sequence;

[0040] The adjustment control command sequence is encoded and time-synchronized to obtain a collaborative control signal, and the collaborative control signal is sent to a preset assembly execution unit, wherein the preset assembly execution unit includes a multi-degree-of-freedom robotic arm and a precision positioning platform.

[0041] Based on the path point sequence in the collaborative control signal, the desired angle and speed commands of each joint of the robotic arm are calculated and issued in real time, driving the end effector of the robotic arm to move along the corrected assembly guide path to the specified assembly pose.

[0042] During the assembly process, the feedback data from the force sensor, position sensor and vision sensor on the assembly execution unit are read in real time and compared in a closed loop with the process parameter instructions embedded in the collaborative control signal to generate a real-time execution deviation vector.

[0043] Based on the real-time execution deviation vector, the motion trajectory or force of the robotic arm end effector is finely adjusted in real time according to the preset impedance control model.

[0044] When the actual assembly contact force of the current assembly step reaches the preset target range, the assembly posture error converges to the allowable tolerance, and the visual inspection confirms that the part is in place, the current step is determined to be completed, and a step completion signal is sent to the control system.

[0045] After receiving the step completion signal, the system automatically loads the collaborative control signal corresponding to the next assembly step according to the initial assembly sequence, and repeats the above execution, monitoring and fine-tuning process.

[0046] The above steps are repeated until all steps in the initial assembly sequence have been completed and an assembly completion signal is output, thus ending the entire electrical component assembly guidance process.

[0047] This application also provides an electrical component assembly guidance system, comprising:

[0048] The data acquisition module is used to acquire digital design model data of electrical components to be assembled and real-time environmental perception data of the assembly site.

[0049] The assembly relationship parsing module is used to obtain the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and the assembly path;

[0050] An environment adaptation assessment module is used to obtain an assembly environment adaptation factor based on the real-time environment perception data and the assembly path.

[0051] The path dynamic correction module is used to correct the assembly path based on the assembly environment adaptation factor, the initial assembly sequence and the real-time environment perception data to obtain a corrected assembly guidance path.

[0052] The process quality monitoring and analysis module is used to acquire real-time process quality monitoring data during the assembly process, and to obtain the quality risk coefficient based on the real-time process quality monitoring data and preset standard process parameter thresholds.

[0053] The collaborative control and execution module is used to generate a collaborative control signal based on the quality risk coefficient and the modified assembly guidance path, and to control the electrical components to perform assembly guidance according to the collaborative control signal until the assembly is completed.

[0054] Preferably, the assembly relationship parsing module includes:

[0055] The structural analysis unit is used to perform topological analysis on the digital design model data, obtain the assembly relationship of multiple sub-components, and construct an assembly relationship topology diagram based on the multiple assembly relationships.

[0056] A sequence planning unit is used to obtain an initial assembly sequence based on the assembly relationship topology diagram and preset assembly constraints.

[0057] The trajectory generation unit is used to obtain the three-dimensional geometric information and target assembly pose of each sub-component according to the initial assembly sequence, and to obtain the preliminary spatial trajectory of multiple sub-components according to multiple three-dimensional geometric information and multiple target assembly poses.

[0058] A smoothness evaluation unit is used to obtain multiple smoothness coefficients between adjacent trajectories based on multiple preliminary spatial trajectories;

[0059] The path synthesis unit is used to splice and smooth multiple preliminary spatial trajectories according to multiple smoothing coefficients to obtain an assembly path.

[0060] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0061] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0062] The beneficial effects of this application are as follows: This invention acquires digital design model data of electrical components and real-time environmental perception data of the assembly site. Based on this, it analyzes hierarchical assembly relationships and generates an initial assembly sequence and assembly path. Next, by comparing the real-time environment and the assembly path, it calculates the interference ratio and the proportion of unreachable paths, and weights them to obtain an assembly environment adaptation factor. Based on this, it corrects the path, forming a modified assembly guidance path adapted to the current environment. Subsequently, it collects assembly images, force / torque, and position / attitude data through multimodal sensors, quantifies appearance, mechanical, and pose deviations, and weights them to obtain a quality risk coefficient. Finally, it combines the modified path and the quality risk coefficient to generate a collaborative control signal, driving the actions of robotic arms and other execution units. Through real-time feedback fine-tuning and step looping, it completes the entire assembly process. This method effectively solves the problems of poor adaptability to the assembly environment of complex electrical components, incomplete quality control, and insufficient operational precision. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of a method flow according to an embodiment of this application.

[0064] Figure 2 This is a schematic diagram of the system structure according to an embodiment of this application.

[0065] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0066] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0067] like Figure 1 As shown, this application provides a method for guiding the assembly of electrical components, including:

[0068] An electrical component assembly guidance method, comprising:

[0069] S1. Obtain digital design model data of the electrical components to be assembled and real-time environmental perception data of the assembly site;

[0070] S2. Obtain the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and the assembly path;

[0071] S3. Obtain the assembly environment adaptation factor based on the real-time environment perception data and the assembly path.

[0072] S4. The assembly path is modified according to the assembly environment adaptation factor and the initial assembly sequence to obtain the modified assembly guidance path;

[0073] S5. Obtain real-time process quality monitoring data during the assembly process, and obtain the quality risk coefficient based on the real-time process quality monitoring data and the preset standard process parameter threshold.

[0074] S6. Generate a collaborative control signal based on the quality risk coefficient and the modified assembly guidance path, and control the electrical components to perform assembly guidance based on the collaborative control signal until the assembly is completed.

[0075] As described in steps S1-S6 above, with the increasingly widespread application of electrical components in key fields such as high-end equipment and new energy systems, their structural complexity is constantly increasing, the number of sub-components is significantly increasing, and the geometric constraints, process constraints, and timing constraints in the assembly process are becoming increasingly diverse. This places higher demands on the accuracy, consistency, and adaptability of assembly operations. Traditional manual assembly methods are difficult to accurately handle assembly requirements under complex constraints, and they are highly dependent on the skill proficiency and experience of operators. When product models change, the adaptation cost is high and the response speed is slow. At the same time, the consistency of key process parameter control is poor, which cannot meet the stringent standards of high-reliability products. This invention first acquires the digital design model data of the electrical components to be assembled and the real-time environmental perception data of the assembly site. This is the basic data support for the entire assembly guidance process. The digital design model data originates from the electrical component design phase and includes core data such as the 3D geometry of sub-components, assembly constraints, and target assembly poses. Generated through computer-aided design (CAD) tools, it is imported into the system via a data interface. Real-time environmental perception data is collected from devices deployed at the assembly site, including vision sensors, LiDAR, and position sensors. This data covers obstacle distribution within the assembly space, the real-time position of sub-components to be assembled, and the surrounding environment of the assembly execution unit. Acquiring this data provides the initial basis for subsequent analysis of assembly relationships and assessment of environmental adaptability. For example, in the assembly of a new energy electrical control module, the digital design model data clearly defines the 3D dimensions, installation positions, and connection constraints of each capacitor, resistor, and chip to the circuit board. Meanwhile, the real-time environmental perception data uses vision sensors to capture the actual placement of each component on the assembly platform and the presence of obstacles such as tools or debris, ensuring that subsequent assembly path planning aligns with the actual site conditions.

[0076] The hierarchical assembly relationship of electrical components is obtained based on the digital design model data, including the initial assembly sequence and assembly path. The core of this step is to transform the design information in the digital model into an executable assembly operation framework. In specific implementation, the digital design model data is first analyzed for topology structure to sort out the connection relationships and dependencies between various sub-components, and then an assembly relationship topology diagram is constructed to clearly present the assembly sequence logic of the sub-components. Then, combined with preset assembly constraints, such as the installation priority of sub-components, spatial position interference constraints, and process requirement constraints, the initial assembly sequence is planned from the assembly relationship topology diagram to clarify which sub-components are assembled first and which are assembled later. Subsequently, based on the initial assembly sequence, the three-dimensional geometric information (such as size and shape) and target assembly pose coordinates (i.e., the spatial position and attitude that should be in after assembly) of each sub-component are extracted. Based on this information, a preliminary spatial trajectory of each sub-component from the initial position to the target position is generated. Next, the smoothing coefficient between adjacent preliminary spatial trajectories is calculated. This coefficient is used to evaluate the continuity at the junction of adjacent trajectories to avoid trajectory abrupt changes that cause instability in the motion of the assembly execution unit. Finally, all preliminary spatial trajectories are spliced ​​and smoothed according to the smoothing coefficient to form a continuous and stable assembly path. For example, in the assembly of a precision electrical sensor, topology analysis revealed that the outer casing sub-component must be assembled before the internal sensing element; otherwise, the sensing element cannot be accurately positioned. Based on this, the initial assembly sequence was determined to be "outer casing fixing - circuit board installation - sensing element assembly - cover plate encapsulation". Then, based on the three-dimensional dimensions and target pose of each component, the movement trajectory of each component was generated. After smoothing, it was ensured that the robotic arm could move the components smoothly without jamming, avoiding component collisions or positioning deviations caused by sudden changes in trajectory.

[0077] The assembly environment adaptation factor is obtained based on real-time environmental perception data and assembly path. The purpose is to evaluate the degree of matching between the preset assembly path and the actual assembly environment, and to provide a basis for subsequent path correction. The specific process is as follows: First, the installation movement trajectories of multiple sub-components to be assembled are extracted from real-time environmental perception data and compared with the assembly path generated in the second step. The overlapping areas of the two are identified, and the proportion of the overlapping area to the preset total installation area is calculated, i.e., the interference ratio. This ratio is used as the interference severity value to reflect the component interference risk that may occur during actual installation. Next, based on the assembly path, the passable area information in the current environment is analyzed to determine which path points on the assembly path are in impassable areas (such as obstacles or insufficient space), forming an unreachable path point set. Then, the total number of assembly path points within the preset assembly space area is counted, and the proportion of path points in the unreachable path point set to the total number of points is calculated, i.e., the unreachable path ratio. Finally, the interference severity value and the unreachable path ratio are weighted and fused to obtain the assembly environment adaptation factor. The value of this factor directly reflects the degree of adaptation between the assembly path and the current environment. The lower the value, the worse the adaptation, and path correction is required. For example, at an electrical control cabinet assembly site, real-time environmental perception data shows that due to the temporary placement of tools, a certain section of the preset assembly path is blocked. Calculations show that the interference rate is 15% and the unreachable path rate is 12%. After weighted fusion, the assembly environment adaptation factor is 0.78, indicating that the current assembly path has certain environmental incompatibility issues and needs to be adjusted accordingly.

[0078] The assembly path is corrected based on the assembly environment adaptation factor and the initial assembly sequence to obtain the corrected assembly guidance path. This step is crucial for responding to environmental changes and optimizing assembly execution. First, a preset adaptability level classification rule is obtained. This rule includes multiple level intervals and corresponding adaptation factor threshold ranges. The assembly environment adaptation factor obtained in the third step is compared with these threshold ranges to determine the adaptability level of the current assembly environment. A higher level indicates less need for existing path adjustments, while a lower level requires adjustments to more assembly stages. Next, based on the adaptability level and the initial assembly sequence, the assembly stages that need adjustment are selected, forming a set of assembly stages to be adjusted. For example, a lower adaptability level may require adjustments to the paths of multiple critical assembly stages, while a higher adaptability level only requires adjustments to a few stages. Then, for each assembly stage to be adjusted, a set of candidate assembly poses for the corresponding sub-component is generated by combining the 3D geometric information of the sub-component, assembly constraints, and real-time environmental perception data. That is, in addition to the initial target assembly pose, other suitable poses are selected. The system determines the spatial position and orientation required for assembly. Based on the set of alternative assembly poses and the original assembly path, path replanning is performed to generate multiple alternative paths. Then, the path adaptation reference for each alternative path is extracted, including the path quality adaptation coefficient (evaluating the smoothness and length rationality of the path) and the assembly constraint completion coefficient (evaluating whether the path meets all assembly constraint requirements). These two coefficients are weighted to obtain the comprehensive path adaptation coefficient for each alternative path. Next, the assembly environment adaptation factor is used to correct each comprehensive path adaptation coefficient, eliminating the influence of environmental adaptability on path evaluation, resulting in a corrected comprehensive path adaptation coefficient. Finally, all corrected comprehensive path adaptation coefficients are sorted, and the alternative corrected path with the highest value is selected as the corrected assembly guidance path, ensuring that the corrected path adapts to the current environment while meeting assembly quality and constraint requirements. For example, in the assembly of a motor stator and rotor, the initial assembly path had a high proportion of unreachable paths and a low adaptation factor due to the slight tilt of the workbench. By classifying the adaptability level, the path that needs to be adjusted in the rotor installation stage was determined, generating three sets of alternative assembly poses, corresponding to three alternative paths. After calculating the comprehensive path adaptation coefficient of each path and combining it with the environmental adaptation factor for correction, the path with the highest correction coefficient was finally selected as the corrected assembly guide path. This path avoids the influence of the tilted area and ensures that the rotor can be assembled with the stator smoothly and accurately.

[0079] Real-time process quality monitoring data is acquired during the assembly process, and a quality risk coefficient is obtained based on this data and preset standard process parameter thresholds. Its function is to assess the quality status during assembly in real time, providing a basis for subsequent control adjustments. First, real-time process quality monitoring data is acquired through multimodal sensors deployed on the assembly execution unit. This data includes assembly image data (collected by vision sensors, reflecting the appearance of the assembly area), assembly force / torque data (collected by force sensors, recording the force and torque changes during assembly), and assembly position / posture data (collected by position sensors, reflecting the actual assembly posture of sub-components). Next, the assembly image data is processed to extract visual appearance features of the assembly area, such as whether there are scratches on the component surface, whether the connection gaps are uniform, and whether the markings are aligned. These features are compared with preset standard appearance feature thresholds to calculate the appearance quality deviation value. The assembly force / torque data is then further processed... Time-series analysis is performed to extract the temporal variation characteristics of force / torque, such as the rate of torque increase during bolt tightening and the torque fluctuation range during the steady-state phase. These characteristics are compared with preset standard force / torque curve thresholds to obtain mechanical process deviation values. Assembly position / posture data is analyzed to extract posture deviation characteristics between the actual assembly posture and the target assembly posture, such as spatial position deviation and posture angle deviation. Combined with preset standard posture tolerance thresholds, posture deviation values ​​are calculated. Finally, the appearance quality deviation value, mechanical process deviation value, and posture deviation value are weighted and fused to obtain a quality risk coefficient. The larger the coefficient, the higher the risk of quality problems in the current assembly process, and the more timely the adjustment of control strategies is required. For example, in the assembly of an electrical connector, real-time acquired assembly image data shows that there are slight scratches at the connector insertion and removal points, with a corresponding appearance quality deviation value of 0.12. Force sensor data shows that the fluctuation range of insertion and removal torque exceeds the standard curve threshold, with a mechanical process deviation value of 0.18. Position sensor data shows that the actual insertion depth deviates little from the target depth, with a pose deviation value of 0.05. After weighted fusion calculation, the quality risk coefficient is 0.13, indicating that there is a certain quality risk in the current assembly, and the insertion and removal force and speed need to be fine-tuned.

[0080] Based on the quality risk coefficient and the revised assembly guidance path, a collaborative control signal is generated, and the electrical components are controlled according to the signal to guide the assembly until completion. This is the execution link that transforms the planning and monitoring results into actual assembly operations. First, based on the corrected assembly guide path, a basic control command sequence is generated, containing path points (the spatial positions that the assembly execution unit should reach at each moment), timestamps (the execution time corresponding to each path point), and state vectors (the motion state parameters of the assembly execution unit). Then, the basic control command sequence is adjusted according to the quality risk coefficient. If the quality risk coefficient is high, parameters such as motion speed and force are appropriately adjusted to reduce the risk of quality fluctuations during assembly. If the quality risk coefficient is low, the stability of the parameters is maintained, resulting in an adjusted control command sequence. Next, the adjusted control command sequence is encoded to ensure the accuracy of signal transmission and is time-synchronized, matching the execution time of each command with the assembly rhythm to form a collaborative control signal. This signal is then sent to the preset assembly execution unit, which typically includes a multi-degree-of-freedom robotic arm and a precision positioning platform, responsible for the actual assembly operation. After receiving the collaborative control signal, the assembly execution unit calculates the desired angle and speed commands for each joint of the robotic arm in real time based on the path point sequence in the signal, driving the end effector of the robotic arm to carry the sub-component along the corrected assembly guide path to the designated assembly pose. During the assembly process, feedback data from force sensors, position sensors, and vision sensors on the assembly execution unit are read in real time. This feedback data is compared in a closed loop with the process parameter commands embedded in the collaborative control signal to generate a real-time execution deviation vector. This vector reflects the difference between the actual execution state and the preset command. Based on the preset impedance control model, the motion trajectory or force of the robotic arm end effector is finely adjusted in real time according to the real-time execution deviation vector to ensure the accuracy of the assembly action. For example, when the deviation vector shows a slight deviation between the actual pose and the target pose, the trajectory is corrected by finely adjusting the joint angle of the robotic arm. When the actual assembly contact force of the current assembly step reaches the preset target range, the assembly pose error converges to the allowable tolerance, and the visual inspection confirms that the part is in place, the current assembly step is determined to be completed, and a step completion signal is sent to the control system. After receiving the signal, the control system automatically loads the collaborative control signal corresponding to the next assembly step according to the initial assembly sequence and repeats the above execution, monitoring, and fine-tuning process. This cycle continues until all steps in the initial assembly sequence are completed, and an assembly completion signal is output, ending the entire assembly guidance process.

[0081] In summary, this invention provides comprehensive support for subsequent steps by integrating design and field data; the hierarchical assembly relationship analysis step transforms design data into an executable assembly framework, ensuring the rationality of the assembly logic; the environmental adaptation assessment and dynamic path correction steps enable the assembly path to flexibly respond to changes in the field environment, reducing interference and unreachability risks; the process quality monitoring step comprehensively captures assembly deviations through multimodal data, providing a reliable basis for quality risk assessment; and the collaborative control execution step transforms the optimized path and quality monitoring results into precise assembly operations, ensuring assembly accuracy and stability through real-time closed-loop adjustments. The entire technical solution specifically addresses the problems of poor environmental adaptability, incomplete quality control, and insufficient operational precision in the assembly of complex electrical components.

[0082] In one embodiment, step S2, which involves obtaining the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes an initial assembly sequence and an assembly path, includes:

[0083] S21. Perform topological analysis on the digital design model data to obtain the assembly relationships of multiple sub-components, and construct an assembly relationship topology diagram based on the multiple assembly relationships;

[0084] S22. Obtain the initial assembly sequence based on the assembly relationship topology diagram and preset assembly constraints;

[0085] S23. Obtain the three-dimensional geometric information and target assembly pose coordinates of each sub-component according to the initial assembly sequence, and obtain the preliminary spatial trajectory of multiple sub-components according to multiple sets of three-dimensional geometric information and multiple sets of target assembly pose coordinates;

[0086] S24. Obtain multiple smoothing coefficients between adjacent trajectories based on the multiple preliminary spatial trajectories;

[0087] S25. The multiple preliminary spatial trajectories are spliced ​​and smoothed according to the multiple smoothing coefficients to obtain the assembly path.

[0088] As described in steps S21-S25 above, this invention performs topological analysis on the digital design model data to obtain the assembly relationships of multiple sub-components and construct an assembly relationship topology diagram. The digital design model data originates from the design phase of electrical components and is generated through modeling using professional design tools such as CAD. It contains core information such as the three-dimensional geometric parameters, connection methods, and constraints of the sub-components. After being imported into the system via a data interface, the model data is deconstructed using a topological analysis algorithm. This algorithm identifies the connection nodes, contact relationships, and dependencies between sub-components (e.g., a sub-component depends on another component as an installation reference), identifies the assembly relationships of all sub-components, and then constructs an assembly relationship topology diagram using nodes to represent sub-components and edges to represent assembly relationships. For example, the digital design model of an electrical control box includes sub-components such as a housing, circuit board, interface module, and fixing bolts. Through topology analysis, it is found that the circuit board needs to be connected to the mounting posts on the housing first, the interface module needs to be installed based on the reserved interface on the circuit board, and the fixing bolts are used to fasten the housing and the cover plate. The topology diagram built based on these relationships can clearly show the assembly dependency logic of each component, avoiding the problem of reversed order in subsequent sequence planning. The core value of this step is to transform the scattered design data into a visualized and logical assembly relationship network, providing a foundation for subsequent sequence planning.

[0089] The initial assembly sequence is obtained based on the assembly relationship topology diagram and preset assembly constraints. These preset assembly constraints are conditions set in accordance with industry process standards, product design requirements, and the characteristics of the assembly execution unit. They include geometric constraints (e.g., no spatial interference between sub-components), process constraints (e.g., welded components must be assembled before packaged components), and timing constraints (e.g., fixing must be completed before wiring). In practical implementation, based on the dependencies in the assembly relationship topology diagram, a constraint satisfaction algorithm is used to filter assembly sequences that meet all preset assembly constraints. Priority is given to planning the assembly sequence of sub-components with no or few dependencies, and then sub-components that depend on other components are gradually included, ultimately forming the initial assembly sequence. For example, in the assembly of a certain motor component, the preset assembly constraint clearly states that "the stator must be fixed to the frame first before the rotor can be assembled with the stator." Combining the dependencies between the stator and frame, and between the rotor and stator in the assembly relationship topology diagram, an initial assembly sequence of "frame positioning - stator installation - rotor assembly - end cover fixing" is planned. This sequence strictly follows geometric and process constraints, ensuring the logical rationality of the assembly process and avoiding problems such as components being unable to be installed or unable to be disassembled and adjusted after installation due to incorrect sequence.

[0090] Based on the initial assembly sequence, the 3D geometric information and target assembly pose coordinates of each sub-component are obtained, and preliminary spatial trajectories of multiple sub-components are derived accordingly. The 3D geometric information of the sub-components (such as dimensions, shape, and mounting interface dimensions) is directly extracted from the digital design model data. The target assembly pose coordinates refer to the spatial position (X, Y, Z coordinates) and attitude (pitch, yaw, roll angle) that each sub-component should be in after assembly. These coordinates are determined by analyzing the assembly reference planes and positioning hole positions of each sub-component in the digital design model. After obtaining the above data, a trajectory planning algorithm is used, starting from the initial placement position of the sub-component (determined by visual positioning information from real-time environmental perception data at the assembly site) and ending at the target assembly pose coordinates, to generate a preliminary spatial trajectory for each sub-component. This trajectory is a discrete set of spatial points, clearly defining the approximate movement path of the sub-component from its initial position to its target position. For example, the initial placement position of a capacitor component is in the material area of ​​the assembly platform, and the target assembly pose coordinates are the specified mounting hole position on the circuit board. The preliminary spatial trajectory generated by the trajectory planning algorithm clarifies the movement path of the capacitor component, which needs to be lifted vertically to the preset height, then moved horizontally to the top of the circuit board, and finally lowered vertically to insert into the mounting hole. This step provides a clear spatial movement direction for each sub-component and is the basis for the generation of the assembly path.

[0091] Multiple smoothing coefficients are obtained between adjacent trajectories based on several preliminary spatial trajectories. Here, "adjacent" refers to the spatial trajectories of two connected sub-components in the initial assembly sequence, or the connection relationship between adjacent path segments within the same sub-component trajectory. The smoothing coefficient is used to evaluate the continuity at the trajectory connection point. Its calculation is based on parameters such as trajectory curvature change, path point spacing, and rate of change of motion direction, and is implemented using a smoothness evaluation algorithm. Specifically, the algorithm calculates the tangent angle and path point density change of adjacent trajectories at the connection point. The smaller the angle and the smoother the path point density change, the closer the smoothing coefficient is to 1, indicating a smoother connection. If the angle is too large or the path point density changes abruptly, the smoothing coefficient is less than 1, indicating a risk of abrupt change in the connection. For example, the initial trajectory endpoint of the circuit board assembly and the initial trajectory starting point of the interface module assembly have a tangent angle of 15° at the connection point, and the path point spacing remains consistent. The calculated smoothness coefficient is 0.92, indicating that the two trajectories are connected smoothly. However, in the trajectory of a certain sub-component, the tangent angle between adjacent path segments is 60°, and the path point spacing abruptly changes from 5mm to 2mm, with a smoothness coefficient of 0.35, indicating a risk of abrupt change at this connection point, requiring further processing. The core function of this step is to identify connection problems in the initial spatial trajectory, providing a quantitative basis for subsequent path smoothing processing and avoiding instability in the motion of the assembly execution unit due to abrupt trajectory changes.

[0092] The assembly path is obtained by splicing and smoothing multiple preliminary spatial trajectories based on multiple smoothing coefficients. Specifically, a path smoothing algorithm is used. For junctions with low smoothing coefficients, transition path points are inserted and the path curvature is adjusted to optimize the trajectory, ensuring a continuous transition between adjacent trajectories. For abrupt changes in the trajectory of the same sub-component, the distribution of path points is adjusted through curve fitting (such as Bézier curve fitting) to gradually change the trajectory curvature. Simultaneously, the algorithm also considers the motion characteristics of the assembly execution unit (such as the joint range of motion and maximum speed of the robotic arm) to optimize the spliced ​​trajectory as a whole, ensuring that the path length is within a reasonable range and that all path points are within the reachable space of the assembly execution unit. For example, for an abrupt change in the trajectory with a smoothing coefficient of 0.35, three transition path points are inserted, decomposing the original 60° tangent angle into three gradually changing 15° angles. Then, the distribution of path points is adjusted through Bézier curve fitting, gradually changing the curvature of this trajectory from 0.2mm⁻¹ to 0.05mm⁻¹, resulting in a continuous assembly path without abrupt changes after splicing. This step optimizes the initial trajectory to form a continuous path that conforms to the motion characteristics of the assembly execution unit, reducing the risk of vibration and collision during the assembly process and improving motion stability.

[0093] In one embodiment, step S3, which involves obtaining the assembly environment adaptation factor based on the real-time environment perception data and the assembly path, includes:

[0094] S31. Obtain the installation movement trajectory of multiple sub-components to be assembled based on the real-time environmental perception data, obtain the overlapping position area based on the multiple installation movement trajectories and the assembly path, obtain the interference ratio based on the overlapping position area and the preset total installation area, and use the interference ratio as the interference severity value.

[0095] S32. Obtain passable area information in the current environment based on the assembly path, and obtain a set of unreachable path points based on the passable area information and the preset assembly space area.

[0096] S33. Obtain the total number of assembly path points according to the preset assembly space area, and obtain the proportion of unreachable paths according to the set of unreachable path points and the total number of assembly path points.

[0097] S34. Perform a weighted fusion calculation based on the interference severity value and the proportion of unreachable paths to obtain the assembly environment adaptation factor.

[0098] As described in steps S31-S34 above, this invention first acquires the installation movement trajectories of multiple sub-components to be assembled based on real-time environmental perception data, and calculates the interference ratio in conjunction with the assembly path, using this ratio as the interference severity value. The real-time environmental perception data is collected in real-time by devices such as visual sensors, lidar, and position sensors deployed at the assembly site, covering information such as the real-time position, movement state, and distribution of surrounding obstacles of the sub-components to be assembled. A data fusion algorithm processes the multi-sensor data to extract the installation movement trajectory of each sub-component to be assembled. This trajectory reflects the actual movement path of the sub-component from its initial position to its target position in the actual environment. The assembly path is a preset ideal trajectory obtained based on the analysis of the digital design model data. The two are compared spatially, and a spatial overlap region detection algorithm identifies the overlapping areas in space between the installation movement trajectory and the assembly path. This overlapping area indicates a potential risk of interference between the sub-component and other objects or the path itself during actual movement. Subsequently, the ratio of the spatial volume of this overlapping area to the spatial volume of the preset total installation area is calculated to obtain the interference ratio, which is used as the interference severity value. The larger this value, the higher the risk of interference collisions during actual assembly. For example, at a precision electrical sensor assembly site, real-time environmental perception data showed that because the tool was temporarily placed at the edge of the assembly platform, the installation trajectory of the sensor element to be assembled overlapped with the preset assembly path in the area near the edge of the platform. Calculations showed that the volume of the overlapping area was 0.02 m³, the preset total installation area volume was 0.2 m³, and the interference ratio was 10%, meaning the severity of the interference was 10%. This result clearly quantified the interference risk in the current environment, providing a clear basis for whether the path needs to be adjusted to avoid the interference area. The core value of this step lies in quantifying the interference risk, thus avoiding the omission of potential interference hazards caused by relying solely on subjective judgment.

[0099] The assembly path information is obtained by acquiring passable area information in the current environment and combining it with a preset assembly space area to obtain a set of inaccessible path points. The passable area information of the assembly path is obtained through spatial mapping analysis of obstacle information in real-time environmental perception data and the assembly path. Specifically, using the 3D coordinates and dimensions of obstacles in the environmental perception data, an obstacle model of the site environment is constructed. Then, spatial collision detection is performed on all path points of the assembly path and the obstacle model. If a path point is inside an obstacle or too close to an obstacle (less than a preset safe distance threshold), it is determined to be an inaccessible path point. The preset assembly space area is a reasonable range of assembly activities set according to product design requirements and the actual space dimensions of the assembly site, clearly defining the spatial boundaries where the assembly path should be located. By comparing all path points of the assembly path with the preset assembly space area, path points that exceed this area or are located in inaccessible areas are filtered out, forming a set of inaccessible path points. For example, the pre-set assembly space for an electrical control cabinet is a cuboid space with a length of 1.2m, a width of 0.8m, and a height of 1.0m. By analyzing the path points of the assembly path, it was found that 5 path points are outside the space. Moreover, the areas corresponding to these 5 path points are blocked by pipelines. Collision detection determined that they were impassable and included them in the set of inaccessible path points. This step, by accurately identifying inaccessible path points, lays the foundation for the subsequent calculation of the proportion of inaccessible paths.

[0100] The total number of points on the assembly path is obtained based on the preset assembly space area, and the percentage of inaccessible paths is calculated by combining this with the set of inaccessible path points. The total number of points on the assembly path refers to the number of all discrete path points included in the preset assembly path. This number is set by the accuracy requirements during path planning; generally, higher accuracy requirements result in more path points. For example, the path point spacing in a high-precision assembly path might be set to 0.1mm, corresponding to a relatively large total number of points. The ratio of the number of path points in the inaccessible path point set to the total number of points on the assembly path is the percentage of inaccessible paths. This ratio directly reflects the proportion of paths in the preset assembly path that cannot be actually executed. For example, if an assembly path has a total of 1000 points and the inaccessible path point set contains 40 path points, the calculated percentage of inaccessible paths is 4%. This value indicates that 4% of the preset path cannot be traversed normally in the current environment and needs to be avoided through path adjustments. This step, by quantifying the percentage of inaccessible paths, provides an intuitive quantitative indicator for evaluating the overall accessibility of the path.

[0101] The assembly environment adaptability factor is obtained by weighted fusion calculation of the interference severity value and the proportion of unreachable paths. The weighted fusion calculation adopts a linear weighted model, where the weight coefficients are set based on the degree of impact of the two risks on the assembly process. This is calibrated through extensive experimental data and industry experience. For example, in the assembly of precision electronic components, interference collisions may cause permanent damage to parts, resulting in a higher impact; therefore, the weight coefficient for the interference severity value can be set to 0.6, and the weight coefficient for the proportion of unreachable paths can be set to 0.4. In the assembly of large electrical equipment, path accessibility has a greater impact on assembly efficiency, and the weight coefficients can be appropriately adjusted to 0.4 for the interference severity value and 0.6 for the proportion of unreachable paths. In specific calculations, the interference severity value and the proportion of unreachable paths are multiplied by their respective weight coefficients, and then the two results are added together to obtain the assembly environment adaptability factor. The value of this factor ranges from 0 to 1. The closer the value is to 1, the better the adaptability of the assembly path to the current environment, and the less significant the adjustment is required; the closer the value is to 0, the worse the adaptability, and the more important the path needs to be corrected. For example, in the assembly of a certain electrical component, the interference severity value is 8%, the proportion of unreachable paths is 5%, and the weighting coefficients are set to 0.6 and 0.4 respectively. The calculated assembly environment adaptability factor is 0.6×0.08+0.4×0.05=0.068. This value indicates that the current path has poor adaptability to the environment and path replanning is required in subsequent steps. This step achieves a comprehensive quantification of the adaptability of the assembly environment by integrating the evaluation results of the two core dimensions, avoiding the limitations of single-dimensional evaluation.

[0102] In one embodiment, step S4, which modifies the assembly path based on the assembly environment adaptation factor and the initial assembly sequence to obtain a modified assembly guidance path, includes:

[0103] S41. Obtain a preset adaptability level division rule, wherein the division rule includes multiple level intervals and the corresponding adaptation factor threshold range for each interval, and compare the assembly environment adaptation factor with multiple adaptation factor threshold ranges to obtain the adaptability level of the current assembly environment.

[0104] S42. Obtain the set of assembly stages to be adjusted based on the adaptability level and the initial assembly sequence;

[0105] S43. For each assembly stage in the assembly stage set, obtain the set of alternative assembly poses for the corresponding sub-component.

[0106] S44. Perform path replanning based on the set of alternative assembly poses and the assembly path to obtain multiple alternative paths;

[0107] S45. Obtain path adaptation references for multiple candidate paths, wherein the path adaptation references include path quality adaptation coefficients and assembly constraint completion coefficients, and obtain multiple comprehensive path adaptation coefficients based on the multiple path quality adaptation coefficients and the multiple assembly constraint completion coefficients.

[0108] S46. Each comprehensive path adaptation coefficient is corrected sequentially according to the assembly environment adaptation factor to obtain multiple corrected comprehensive path adaptation coefficients.

[0109] S47. Sort the adaptation coefficients of multiple correction synthesis paths and select the candidate correction path with the highest adaptation coefficient as the correction assembly guidance path.

[0110] As described in steps S41-S47 above, this invention obtains a preset adaptability level classification rule, compares the assembly environment adaptation factor with the adaptation factor threshold range, and obtains the adaptability level of the current assembly environment. The preset adaptability level classification rule is based on a large amount of assembly scenario experimental data and industry process experience, and includes multiple level intervals and corresponding adaptation factor threshold ranges. For example, it can be divided into four intervals: high adaptability level (adaptation factor 0.8-1.0), medium adaptability level (0.5-0.8), low adaptability level (0.3-0.5), and very low adaptability level (0-0.3). Each interval corresponds to different path adjustment requirements. This rule is pre-stored in the system's parameter configuration module and can be fine-tuned according to the assembly characteristics of different electrical components. The assembly environment adaptation factor is calculated through steps S31 to S34 above, reflecting the degree of matching between the current environment and the preset path. By substituting this adaptation factor into the level classification rule and comparing it with the threshold range of each interval, the adaptability level of the current assembly environment can be determined. For example, the assembly environment compatibility factor of a certain electrical connector assembly is 0.45, which belongs to the low compatibility level according to the classification rules. This indicates that there are obvious environmental incompatibility problems in the preset path, and the path needs to be adjusted for many key assembly stages. The core value of this step is to clarify the scope and priority of path adjustment through level quantification, and avoid the waste of efficiency caused by blind adjustment.

[0111] The set of assembly stages to be adjusted is obtained based on the adaptability level and the initial assembly sequence. The initial assembly sequence is derived from the topology and preset assembly constraints of the digital design model, clarifying the assembly order and corresponding stages of each sub-component. Different adaptability levels correspond to different adjustment ranges. At high adaptability levels, only a few critical path points need to be adjusted, without changing the entire assembly stage; at medium adaptability levels, some core assembly stages need to be adjusted; at low and very low adaptability levels, most or even all assembly stages require path replanning. By establishing a mapping relationship between adaptability levels and the adjustment range of assembly stages, the assembly stages that need adjustment are selected from the initial assembly sequence, forming the set of assembly stages to be adjusted. For example, the initial assembly sequence of a motor stator assembly includes four stages: "stator positioning - coil installation - lead wire welding - insulation encapsulation". The current adaptability level is low. According to the mapping relationship, the two core stages "stator positioning" and "coil installation" need to be adjusted. These two stages are included in the set of assembly stages to be adjusted. This step, by combining the logical correlation of the assembly sequence, ensures that the path adjustment does not disrupt the rationality of the overall assembly process.

[0112] For each assembly stage in the set of assembly stages to be adjusted, a set of candidate assembly poses for the corresponding sub-component is obtained. This set of candidate assembly poses for the sub-component is generated based on digital design model data, preset assembly constraints, and real-time environmental perception data. The digital design model data provides the sub-component's 3D geometric information and core assembly benchmark requirements; the preset assembly constraints clarify the geometric and technological conditions that the pose must meet; and the real-time environmental perception data reflects the spatial limitations of the current environment. Using a pose optimization algorithm, under the premise of satisfying all constraints, all possible spatial positions (X, Y, Z coordinates) and attitudes (pitch angle, yaw angle, roll angle) of the sub-component are searched to form the set of candidate assembly poses. For example, the assembly stage to be adjusted for a certain capacitor component is "capacitor welding and positioning". Its digital design model requires that the capacitor pins be precisely aligned with the circuit board pads. The preset assembly constraints stipulate that the deviation of the welding posture shall not exceed ±0.1mm. Combined with the spatial range of the circuit board without obstacles displayed by the real-time environmental perception data, three candidate assembly postures that meet the requirements are searched through the posture optimization algorithm, forming a set of candidate assembly postures. This step provides a variety of posture options for subsequent path replanning and ensures the flexibility of path adjustment.

[0113] Based on the set of candidate assembly poses and the assembly path, path replanning is performed to obtain multiple candidate paths. Path replanning adopts a sampling-based path planning algorithm (such as the RRT algorithm). Starting from the initial placement position of the sub-component (determined by visual positioning information from real-time environmental perception data) and ending at each pose in the set of candidate assembly poses, it combines the passable area information of the current environment (extracted from real-time environmental perception data) to avoid obstacles and interference areas, generating multiple continuous and smooth candidate paths. For example, the set of candidate assembly poses for an interface module contains two target poses. Starting from the initial position of the module on the assembly platform, two candidate paths are planned using the RRT algorithm. Path 1 avoids the tool obstacle on the left side of the platform, and path 2 avoids the pipeline obstruction on the right side of the platform. Both paths meet the motion smoothness requirements. This step, through multi-target pose path planning, provides sufficient alternatives for subsequent selection of the optimal path.

[0114] The system obtains path adaptation references for multiple candidate paths, including path quality adaptation coefficients and assembly constraint completion coefficients, and then obtains multiple comprehensive path adaptation coefficients. The path quality adaptation coefficient is used to evaluate the motion characteristics of the candidate paths. Its calculation is based on parameters such as path length, smoothness, and motion time, obtained through weighted summation. For example, the shorter the path length, the higher the smoothness, and the more reasonable the motion time, the closer the coefficient is to 1. The weighting of this parameter is combined with the motion characteristics of the assembly execution unit. For example, the motion energy consumption of the robotic arm is positively correlated with the path length; therefore, the path length weight can be set to 0.3, the smoothness weight to 0.5, and the motion time weight to 0.2. The assembly constraint completion coefficient is used to evaluate whether the candidate paths meet all preset assembly constraints. Through a constraint satisfaction detection algorithm, it verifies one by one whether the candidate paths meet geometric constraints, process constraints, and other requirements. If they are fully satisfied, the coefficient is 1; if they are partially satisfied, the corresponding value is calculated according to the degree of satisfaction (e.g., if 80% of the constraints are satisfied, the coefficient is 0.8). The comprehensive path fit coefficient is calculated by linearly weighting the path quality fit coefficient and the assembly constraint completion coefficient. The weighting coefficients are set according to the assembly requirements. For example, in high-precision assembly scenarios, the weight of the assembly constraint completion coefficient can be set to 0.6, and the weight of the path quality fit coefficient to 0.4. In high-efficiency assembly scenarios, the weighting coefficients can be adjusted to 0.6 and 0.4. For instance, if the path quality fit coefficient of a candidate path is 0.92 and the assembly constraint completion coefficient is 0.95, and both weighting coefficients are set to 0.5, then the comprehensive path fit coefficient is (0.92 + 0.95) × 0.5 = 0.935. This step comprehensively considers the quality and constraint satisfaction of the candidate path through multi-dimensional evaluation, avoiding path defects caused by single-dimensional evaluation.

[0115] Each comprehensive path adaptation coefficient is sequentially corrected based on the assembly environment adaptation factor, resulting in multiple corrected comprehensive path adaptation coefficients. The correction process employs a multiplicative correction model, multiplying the comprehensive path adaptation coefficient of each candidate path by the assembly environment adaptation factor to achieve a quantitative correction of the path's comprehensive adaptability based on environmental adaptability. This is because the comprehensive path adaptation coefficient primarily assesses the quality and constraint satisfaction of the path itself, without fully considering the impact of environmental adaptability. The assembly environment adaptation factor, however, directly reflects the degree of support the current environment provides for path execution. By multiplying the two, the corrected coefficients can more comprehensively reflect the path's overall performance in the current environment. For example, if the overall path fit coefficient of a candidate path is 0.935 and the current assembly environment fit factor is 0.45, the corrected overall path fit coefficient is 0.935 × 0.45 ≈ 0.421. If the overall path fit coefficient of another candidate path is 0.89, the corrected coefficient is 0.89 × 0.45 ≈ 0.401. This step, by correcting the environment fit factor, makes the path evaluation results more consistent with the actual execution scenario, avoiding the selection of paths that perform well in ideal conditions but are not suitable for the current environment.

[0116] Multiple modified integrated path adaptation coefficients are sorted, and the candidate modified path with the highest adaptation coefficient is selected as the modified assembly guidance path. The sorting is in descending order; a higher adaptation coefficient indicates better overall adaptability of the candidate path in the current environment, satisfying assembly constraints and path quality requirements while adapting to the actual environment. For example, in an assembly scenario with three candidate paths and adaptation coefficients of 0.421, 0.401, and 0.385 respectively, after descending sorting, the candidate path with the highest coefficient of 0.421 is selected as the modified assembly guidance path. This path avoids interference and unreachability issues in the environment while ensuring assembly accuracy and motion stability. This sorting and filtering step ensures that the final modified path is the optimal choice under the current environment.

[0117] In one embodiment, step S5, which involves acquiring real-time process quality monitoring data during the assembly process and obtaining a quality risk coefficient based on the real-time process quality monitoring data and a preset standard process parameter threshold, includes:

[0118] S51. Acquire multimodal process quality monitoring data collected in real time during the assembly process, wherein the multimodal process quality monitoring data includes assembly image data, assembly force / torque data, and assembly position / orientation data.

[0119] S52. Extract the visual appearance features of the assembly area based on the assembly image data, and obtain the appearance quality deviation value based on the visual appearance features and the preset standard appearance feature threshold.

[0120] S53. Extract the force / torque time sequence change characteristics during the assembly process based on the assembly force / torque data, and obtain the mechanical process deviation value based on the force / torque time sequence change characteristics and the preset standard force / torque curve threshold.

[0121] S54. Extract the pose deviation features between the actual assembly pose and the target assembly pose based on the assembly position / pose data, and obtain the pose deviation value based on the pose deviation features and the preset standard pose tolerance threshold.

[0122] S55. A quality risk coefficient is obtained by weighted fusion calculation based on the appearance quality deviation value, the mechanical process deviation value, and the posture deviation value.

[0123] As described in steps S51-S55 above, this invention acquires multimodal process quality monitoring data collected in real time during the assembly process, including assembly image data, assembly force / torque data, and assembly position / attitude data. This data is collected in real time by various types of sensors deployed in the assembly execution unit and assembly site. Assembly image data is acquired by a high-definition vision sensor (such as an industrial camera), which is installed above the assembly platform or at the end of the robotic arm, clearly capturing real-time images of the assembly area and recording the appearance, alignment, and other characteristics of the components. Assembly force / torque data is acquired by force sensors installed on the end effector of the robotic arm or assembly fixture, recording in real time the force and torque changes generated by component contact, tightening, and other actions during the assembly process. Assembly position / attitude data is acquired by position sensors (such as laser positioning sensors) and attitude sensors, accurately obtaining the actual spatial coordinates and attitude angles of the sub-components during the assembly process. For example, in the assembly process of an electrical connector, a vision sensor collects 15 frames of images of the assembly area per second, a force sensor records the changes in force during the insertion and removal of the connector in real time, and a position / attitude sensor accurately captures the insertion depth and tilt angle of the connector. This multimodal data comprehensively reflects the real-time status of the assembly process from different dimensions, providing rich raw data support for subsequent quality analysis.

[0124] S52 extracts visual appearance features of the assembly area from assembly image data and obtains appearance quality deviation values ​​by combining them with preset standard appearance feature thresholds. After preprocessing (such as noise reduction and image enhancement), the assembly image data is processed using image feature extraction algorithms (such as feature extraction algorithms based on convolutional neural networks, which contain 3 convolutional layers and 2 pooling layers with kernel sizes of 3×3, 5×5, and 3×3, and a stride of 1, respectively, to extract deep features such as edges, textures, and colors through convolution operations) to extract visual appearance features of the assembly area, including whether there are scratches, stains, or deformations on the component surface, whether the installation joints are uniform, and whether the markings are aligned. The preset standard appearance feature thresholds are based on product design requirements and industry quality standards and include thresholds for scratch length, joint width, and marking alignment deviation, such as stipulating that the scratch length on the component surface should not exceed 0.5mm and the joint width deviation should not exceed 0.1mm. The extracted visual appearance features are compared with the preset standard appearance feature thresholds, and the degree of difference between the two is calculated using a deviation quantification algorithm to obtain the appearance quality deviation value. The larger the value, the greater the deviation between the appearance quality and the standard requirements. For example, after feature extraction of the image data of a capacitor component assembly, a scratch with a length of 0.3 mm was found on its surface, and the joint width deviation was 0.08 mm. According to the deviation quantification algorithm, the appearance quality deviation value was calculated to be 0.25. This step achieves an objective assessment of the assembly appearance quality by accurately extracting appearance features and quantifying deviation.

[0125] S53 extracts the temporal variation characteristics of force / torque based on assembly force / torque data, and obtains the mechanical process deviation value by combining it with a preset standard force / torque curve threshold. The assembly force / torque data is continuous temporal data. A temporal feature extraction algorithm (such as the sliding window method, with a window size of 50ms and a step size of 10ms, calculating parameters such as the maximum, minimum, average, and rate of change within the window) is used to extract the temporal variation characteristics of force / torque, including the peak value of the applied force, the rate of increase, the fluctuation range of the stable phase, and the holding time of the torque. The preset standard force / torque curve threshold is obtained based on statistical analysis of force / torque data from a large number of qualified assembly samples. For example, in the bolt tightening process, the standard force / torque curve clearly defines the slope of the torque increase phase, the peak torque range, and the allowable torque fluctuation value in the stable phase. The extracted temporal variation characteristics of force / torque are compared with the preset standard force / torque curve threshold, and the degree of deviation is calculated using a curve similarity analysis algorithm to obtain the mechanical process deviation value. For example, in a bolt fastening assembly, the peak torque extracted from the force / torque data is 8 N·m, while the preset standard peak torque range is 7-7.5 N·m. The rate of increase is 15% faster than the standard curve. The mechanical process deviation value is calculated to be 0.32 by curve similarity analysis. This step effectively evaluates the rationality of the assembly mechanical process by capturing the temporal variation law of force / torque and quantifying the deviation, thus avoiding component damage or loose connection caused by improper force / torque.

[0126] S54 extracts pose deviation features from assembly position / attitude data and obtains the pose deviation value by combining it with a preset standard pose tolerance threshold. Assembly position / attitude data includes the actual spatial coordinates (X, Y, Z) and attitude angles (pitch, yaw, roll) of the sub-component. Using a pose deviation feature extraction algorithm, the difference between the actual assembly pose and the target assembly pose (analyzed from the digital design model data) is calculated to obtain pose deviation features, including position deviation amounts (e.g., 0.1mm deviation in the X-axis direction, 0.05mm deviation in the Y-axis direction) and attitude deviation angles (e.g., 0.5° pitch deviation). The preset standard pose tolerance threshold is established based on the product design's precision requirements, clearly defining the allowable deviation range for position and attitude. For example, it specifies that the total position deviation must not exceed 0.2mm, and the attitude deviation angle must not exceed 1°. The extracted pose deviation features are compared with the preset standard pose tolerance threshold, and the total deviation is calculated using a deviation accumulation algorithm to obtain the pose deviation value. For example, the actual positional deviation of a certain circuit board assembly is 0.12mm on the X-axis, 0.08mm on the Y-axis, and 0.05mm on the Z-axis. The attitude deviation angles are pitch angle 0.3° and yaw angle 0.2°. The positional deviation value is calculated to be 0.18 by the deviation accumulation algorithm. This step accurately quantifies the degree of deviation of the assembly posture and provides a direct basis for evaluating the assembly accuracy.

[0127] S55 calculates the quality risk coefficient by weighted fusion of appearance quality deviation, mechanical process deviation, and pose deviation. The weighted fusion calculation uses a linear weighted model, with the weight coefficients determined based on the degree of influence of different deviation types on assembly quality. This is calibrated through extensive experimental data and engineering experience. For example, in the assembly of precision electronic components, pose deviation has the greatest impact on subsequent component installation and product performance, and its weight coefficient can be set to 0.4; mechanical process deviation is second, with a weight of 0.35; and appearance quality deviation has a weight of 0.25. In the assembly of electrical equipment with high appearance requirements, the weights for appearance quality deviation, pose deviation, and mechanical process deviation can be adjusted to 0.4, 0.35, and 0.25 respectively. In the specific calculation, the three deviation values ​​are multiplied by their corresponding weight coefficients, and the products are summed to obtain the quality risk coefficient. This coefficient ranges from 0 to 1. The closer the value is to 1, the higher the assembly quality risk, and the more timely the adjustment of the assembly operation is required; the closer the value is to 0, the better the assembly quality meets the standard requirements. For example, the appearance quality deviation of an electrical component assembly is 0.25, the mechanical process deviation is 0.32, and the positional deviation is 0.18, with weighting coefficients of 0.25, 0.35, and 0.4, respectively. The calculated quality risk coefficient is 0.25×0.25+0.32×0.35+0.18×0.4=0.0625+0.112+0.072=0.2465. This value indicates that there is a certain risk in the current assembly quality, but it is within a controllable range. The risk can be reduced by fine-tuning the assembly parameters. This step, by integrating the results of multi-dimensional deviation assessment, achieves a comprehensive quantification of assembly quality risk, avoiding the limitations of single-dimensional assessment.

[0128] In one embodiment, step S6, which generates a coordinated control signal based on the quality risk coefficient and the modified assembly guidance path, and controls the electrical components to perform assembly guidance according to the coordinated control signal until the assembly is completed, includes:

[0129] S61. Based on the modified assembly guidance path, generate a basic control instruction sequence containing path points, timestamps, and state vectors;

[0130] S62. Adjust the basic control command sequence according to the quality risk coefficient to obtain an adjusted control command sequence;

[0131] S63. Encode and synchronize the adjustment control command sequence to obtain a collaborative control signal, and send the collaborative control signal to a preset assembly execution unit, wherein the preset assembly execution unit includes a multi-degree-of-freedom robotic arm and a precision positioning platform.

[0132] S64. Based on the path point sequence in the cooperative control signal, calculate and issue the desired angle and speed commands for each joint of the robotic arm in real time, and drive the end effector of the robotic arm to move along the corrected assembly guide path to the specified assembly posture.

[0133] S65. During the assembly process, the feedback data from the force sensor, position sensor and vision sensor on the assembly execution unit are read in real time and compared in a closed loop with the process parameter instructions embedded in the collaborative control signal to generate a real-time execution deviation vector.

[0134] S66. Based on the real-time execution deviation vector, finely adjust the motion trajectory or force of the robotic arm end in real time according to the preset impedance control model.

[0135] S67. When the actual assembly contact force of the current assembly step reaches the preset target range, the assembly posture error converges to the allowable tolerance, and the visual inspection confirms that the part is in place, the current step is determined to be completed, and a step completion signal is sent to the control system.

[0136] S68. After receiving the step completion signal, the system automatically loads the collaborative control signal corresponding to the next assembly step according to the initial assembly sequence, and repeats the above execution, monitoring and fine-tuning process.

[0137] S69. Repeat the above steps until all steps in the initial assembly sequence have been completed and an assembly completion signal is output, thus ending the entire electrical component assembly guidance process.

[0138] As described in steps S61-S69 above, the present invention first generates a basic control instruction sequence containing path points, timestamps, and state vectors based on the modified assembly guidance path. The optimal path, adapted to the current assembly environment, includes the coordinates of all spatial path points in the sub-component assembly, trajectory smoothing parameters, and other information. When generating the basic control instruction sequence, an instruction encoding algorithm is used to bind each path point in the modified assembly guide path with a corresponding timestamp. The timestamp is calculated based on the movement speed of the assembly execution unit (set in combination with the maximum movement speed of the robotic arm and the assembly accuracy requirements, such as 0.1m / s for precision assembly), ensuring that the execution unit moves according to the preset rhythm. The state vector contains parameters such as the motion mode of the execution unit (e.g., uniform speed, variable speed) and the attitude of the end effector (e.g., clamping force level). For example, in the assembly of a resistor element, the modified assembly guide path contains 100 path points from the material area to the circuit board mounting position. The instruction encoding algorithm assigns a corresponding timestamp to each path point (the time interval between adjacent path points is 0.02s) and sets the state vector to "uniform speed movement, clamping force level 3". The generated basic control instruction sequence provides a clear motion reference for the execution unit. The core of this step is to transform the optimized path into an executable instruction set and build the basic framework for control execution.

[0139] The basic control command sequence is adjusted based on the quality risk coefficient to obtain the adjusted control command sequence. The quality risk coefficient is calculated through steps S51 to S55 above and reflects the degree of quality deviation and risk level in the current assembly process. Different risk levels correspond to different command adjustment strategies. The adjustment process uses a parameter correction algorithm. If the quality risk coefficient is high (e.g., greater than 0.3), it indicates that there are significant appearance, mechanical, or pose deviations. Parameters in the control commands, such as motion speed (reduced by 10%-20%) and end force (adjusted according to the type of deviation, such as reducing clamping force or assembly force if the mechanical deviation is too large), need to be adjusted to reduce the accumulation of deviations. If the quality risk coefficient is low (e.g., less than 0.1), the basic command parameters remain unchanged, and only the timing synchronization accuracy is fine-tuned. For example, the quality risk coefficient of a certain capacitor assembly is 0.35, mainly due to mechanical process deviation (excessive bolt tightening torque). The parameter correction algorithm adjusts the torque command in the bolt tightening stage of the basic control command from 8 N·m to 7 N·m, and at the same time reduces the movement speed in this stage from 0.08 m / s to 0.06 m / s. By adjusting the command, the quality risk problem is directly and specifically solved. This step realizes the dynamic binding of quality risk and control command, making the control action more in line with quality requirements.

[0140] The adjustment control command sequence is encoded and time-synchronized to obtain a coordinated control signal, which is then sent to the preset assembly execution unit. The encoding process uses an industry-standard communication encoding protocol (such as Profinet) to convert information such as path points, timestamps, and state vectors in the adjustment control command sequence into digital signals recognizable by the execution unit, ensuring the accuracy and anti-interference capabilities of signal transmission. Time synchronization is achieved through a clock synchronization algorithm, using the system's unified clock as a reference to calibrate the execution time of each control command, avoiding assembly deviations caused by asynchronous movements of multiple execution units (such as a robotic arm and a precision positioning platform). The preset assembly execution unit includes a multi-degree-of-freedom robotic arm and a precision positioning platform. The robotic arm is responsible for the gripping, movement, and assembly of sub-components, while the precision positioning platform is used to fix the base or position auxiliary components. The two achieve coordinated movement through coordinated control signals. For example, in the assembly of a certain electrical module, the robotic arm performs the actions of picking up and installing the circuit board, while the precision positioning platform adjusts the position of the base synchronously. The timing synchronization algorithm ensures that when the robotic arm places the circuit board, the base has been accurately positioned to the target position. After the collaborative control signal is encoded and transmitted, the action error between the two is controlled within ±0.01s. This step ensures the effective transmission of control signals and the coordinated linkage of multiple execution units, providing signal support for precise assembly.

[0141] Based on the path point sequence in the collaborative control signal, the desired angle and speed commands for each joint of the robotic arm are calculated and issued in real time, driving the end effector of the robotic arm to move along the corrected assembly guide path to the designated assembly pose. The path point sequence contains all the spatial coordinate information of the corrected assembly guide path. Using a forward kinematics algorithm, based on the structural parameters of the robotic arm (such as the number of joints and link lengths, which are preset in the actuator control system), the spatial coordinates of each path point are converted into the desired angles of each joint (such as the rotation angles of the 6 joints of a 6-DOF robotic arm). At the same time, the desired speed of each joint is calculated in combination with the timestamp (to ensure that the end effector arrives at the path point at the preset time). For example, when assembling an interface module for a 6-DOF robotic arm, the coordinates of a certain path point in the collaborative control signal are (X=100mm, Y=50mm, Z=80mm). The forward kinematics algorithm calculates the expected angles of each joint as 30°, 45°, 60°, etc., based on the length of the robotic arm links (link 1 is 30mm long, link 2 is 40mm long, etc.). At the same time, the joint velocity is calculated as 5° / s based on the timestamp. After the command is issued, the end effector of the robotic arm is driven to move precisely along the path. This step transforms the spatial path into joint motion commands for the robotic arm, realizing the physical execution of the path.

[0142] During the assembly process, feedback data from force sensors, position sensors, and vision sensors on the assembly execution unit are read in real time and compared in a closed loop with the process parameter instructions embedded in the collaborative control signal to generate a real-time execution deviation vector. Force sensors collect data such as contact force and clamping force during the assembly process, position sensors collect the actual position coordinates of the robotic arm end and components, and vision sensors collect real-time image data of the assembly area. This feedback data is transmitted to the control system in real time through sensor interfaces. The process parameter instructions are preset standard parameters (such as target assembly force, target position coordinates, and standard appearance features) embedded in the collaborative control signals. The closed-loop comparison uses a deviation calculation algorithm to compare the differences between the feedback data and the process parameter instructions one by one, such as the difference between the actual assembly force and the target assembly force, the three-dimensional difference between the actual position coordinates and the target coordinates, and the pixel difference between the actual appearance features and the standard features. These differences are integrated into a real-time execution deviation vector (such as [force deviation 0.5N, X-axis position deviation 0.03mm, Y-axis position deviation 0.02mm, appearance deviation 0.01 pixels]). This vector comprehensively reflects the deviation between the actual execution state and the preset instructions, providing a precise basis for subsequent fine-tuning.

[0143] Based on the real-time execution deviation vector, the motion trajectory or force of the robotic arm's end effector is fine-tuned in real time according to a preset impedance control model. The preset impedance control model is constructed based on the robotic arm's dynamic characteristics and assembly process requirements, and includes stiffness, damping, and inertia parameters. These parameters are determined through experimental calibration (e.g., in precision assembly scenarios, the stiffness parameter is set to 50 N / m, and the damping parameter to 10 N·s / m). The core function of the model is to calculate the required fine-tuning amount based on the magnitude and direction of the deviation vector. For example, if the real-time execution deviation vector shows an X-axis position deviation of 0.05 mm (actual position slightly to the right) and a force deviation of 0.8 N (actual force too large), the impedance control model calculates that the robotic arm's end effector trajectory needs to be fine-tuned to the left by 0.05 mm based on the stiffness parameter, and that the end effector force needs to be reduced by 0.6 N based on the damping parameter. The fine-tuning command is sent to the robotic arm joint controller in real time. By adjusting the joint angle and speed, precise correction of the trajectory and force is achieved. This step realizes dynamic closed-loop fine-tuning of the execution process, effectively compensating for real-time deviations and improving assembly accuracy.

[0144] When the actual assembly contact force of the current assembly step reaches the preset target range, the assembly pose error converges to the allowable tolerance, and visual inspection confirms that the component is in place, the current step is determined to be complete, and a step completion signal is sent. The preset target range (e.g., assembly contact force 3-5N) and allowable tolerance (e.g., pose error ±0.02mm) are both set based on product design requirements and process standards and stored in the control system. The actual assembly contact force is extracted from the force sensor feedback data, the assembly pose error is calculated from the position sensor feedback data and the target pose, and visual inspection confirms whether the component is completely in place (e.g., whether the circuit board is completely attached to the mounting post without any warping) by analyzing the images collected by the visual sensor. For example, in a resistor assembly step, when the force sensor feedback contact force is 4.2N (within the 3-5N target range), the pose error is 0.01mm (less than the ±0.02mm allowable tolerance), and the visual image shows that the resistor pin is fully inserted into the pad, the control system determines that the step is complete and sends a step completion signal to the system. This step provides clear and multi-dimensional judgment criteria for the completion of the assembly steps, ensuring the assembly quality of each step.

[0145] After receiving the step completion signal, the system automatically loads the corresponding collaborative control signal for the next assembly step according to the initial assembly sequence, and repeats the above execution, monitoring, and fine-tuning process. The initial assembly sequence clearly defines the order of all assembly steps. The control system has a built-in step management module that stores the collaborative control signal corresponding to each step. For example, if the initial assembly sequence is "resistor assembly - capacitor assembly - interface module assembly", when the resistor assembly step completion signal is received, the step management module automatically calls the collaborative control signal corresponding to the capacitor assembly, repeats steps S64-S67, and drives the robotic arm to perform the capacitor assembly action. This step achieves automatic connection of assembly steps without manual intervention, improving assembly efficiency and process continuity.

[0146] S69 executes the above steps repeatedly until all steps in the initial assembly sequence are completed, outputting an assembly completion signal and ending the assembly guidance process. The cyclic control is implemented through the system's main controller, which monitors the execution progress of the initial assembly sequence in real time. When the completion signal of the last assembly step is received, it confirms that all sub-components have been assembled as required and outputs an assembly completion signal (including assembly time, quality risk coefficient statistics, and deviation data for each step), prompting the operator to perform subsequent inspection or unloading operations. For example, the assembly of an electrical control box involves 8 steps. After approximately 15 minutes of cyclic execution, if all steps meet the quality requirements, the system outputs an assembly completion signal. During the assembly process, the average pose error of each step is controlled within ±0.015mm, the mechanical process deviation is less than 0.1, and the overall quality risk coefficient remains stable below 0.1. This step marks the completion of the entire assembly guidance process, ensuring the integrity and closed-loop control of the assembly process.

[0147] This application also provides an electrical component assembly guidance system, comprising:

[0148] The data acquisition module is used to acquire digital design model data of electrical components to be assembled and real-time environmental perception data of the assembly site.

[0149] The assembly relationship parsing module is used to obtain the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and the assembly path;

[0150] An environment adaptation assessment module is used to obtain an assembly environment adaptation factor based on the real-time environment perception data and the assembly path.

[0151] The path dynamic correction module is used to correct the assembly path based on the assembly environment adaptation factor, the initial assembly sequence and the real-time environment perception data to obtain a corrected assembly guidance path.

[0152] The process quality monitoring and analysis module is used to acquire real-time process quality monitoring data during the assembly process, and to obtain the quality risk coefficient based on the real-time process quality monitoring data and preset standard process parameter thresholds.

[0153] The collaborative control and execution module is used to generate a collaborative control signal based on the quality risk coefficient and the modified assembly guidance path, and to control the electrical components to perform assembly guidance according to the collaborative control signal until the assembly is completed.

[0154] In one embodiment, the assembly relationship parsing module includes:

[0155] The structural analysis unit is used to perform topological analysis on the digital design model data, obtain the assembly relationship of multiple sub-components, and construct an assembly relationship topology diagram based on the multiple assembly relationships.

[0156] A sequence planning unit is used to obtain an initial assembly sequence based on the assembly relationship topology diagram and preset assembly constraints.

[0157] The trajectory generation unit is used to obtain the three-dimensional geometric information and target assembly pose of each sub-component according to the initial assembly sequence, and to obtain the preliminary spatial trajectory of multiple sub-components according to multiple three-dimensional geometric information and multiple target assembly poses.

[0158] A smoothness evaluation unit is used to obtain multiple smoothness coefficients between adjacent trajectories based on multiple preliminary spatial trajectories;

[0159] The path synthesis unit is used to splice and smooth multiple preliminary spatial trajectories according to multiple smoothing coefficients to obtain an assembly path.

[0160] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0161] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0162] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in this application and in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0163] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, apparatus, article, or method. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, apparatus, article, or method that includes that element.

[0164] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for guiding the assembly of electrical components, characterized in that, include: Acquire digital design model data of electrical components to be assembled and real-time environmental perception data of the assembly site; The hierarchical assembly relationship of electrical components is obtained based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and the assembly path; The assembly environment adaptation factor is obtained based on the real-time environmental perception data and the assembly path. The assembly path is corrected based on the assembly environment adaptation factor and the initial assembly sequence to obtain a corrected assembly guidance path; Acquire real-time process quality monitoring data during the assembly process, and obtain a quality risk coefficient based on the real-time process quality monitoring data and preset standard process parameter thresholds; A coordinated control signal is generated based on the quality risk coefficient and the modified assembly guidance path, and the electrical components are controlled to perform assembly guidance according to the coordinated control signal until the assembly is completed.

2. The electrical component assembly guidance method according to claim 1, characterized in that, The step of obtaining the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and assembly path, includes: A topology analysis is performed on the digital design model data to obtain the assembly relationships of multiple sub-components, and an assembly relationship topology diagram is constructed based on the multiple assembly relationships. The initial assembly sequence is obtained based on the assembly relationship topology diagram and preset assembly constraints; The three-dimensional geometric information and target assembly pose coordinates of each sub-component are obtained according to the initial assembly sequence, and the preliminary spatial trajectories of multiple sub-components are obtained according to multiple sets of three-dimensional geometric information and multiple sets of target assembly pose coordinates. Multiple smoothing coefficients between adjacent trajectories are obtained based on the multiple preliminary spatial trajectories; The assembly path is obtained by splicing and smoothing the multiple preliminary spatial trajectories according to the multiple smoothing coefficients.

3. The electrical component assembly guidance method according to claim 1, characterized in that, The step of obtaining the assembly environment adaptation factor based on the real-time environment perception data and the assembly path includes: The installation movement trajectories of multiple sub-components to be assembled are obtained based on the real-time environmental perception data. Overlapping position areas are obtained based on the multiple installation movement trajectories and the assembly path. Interference ratio is obtained based on the overlapping position areas and the preset total installation area. The interference ratio is used as the interference severity value. Based on the assembly path, obtain the passable area information in the current environment, and based on the passable area information and the preset assembly space area, obtain the set of unreachable path points; The total number of assembly path points is obtained based on the preset assembly space area, and the proportion of unreachable paths is obtained based on the set of unreachable path points and the total number of assembly path points. The assembly environment adaptation factor is obtained by weighted fusion calculation based on the interference severity value and the proportion of unreachable paths.

4. The electrical component assembly guidance method according to claim 1, characterized in that, The step of correcting the assembly path based on the assembly environment adaptation factor and the initial assembly sequence to obtain a corrected assembly guidance path includes: Obtain a preset adaptability level classification rule, wherein the classification rule includes multiple level intervals and the corresponding adaptation factor threshold range for each interval, and compare the assembly environment adaptation factor with multiple adaptation factor threshold ranges to obtain the adaptability level of the current assembly environment. The set of assembly stages to be adjusted is obtained based on the adaptability level and the initial assembly sequence; For each assembly stage in the assembly stage set, obtain the set of alternative assembly poses for the corresponding sub-component. Based on the set of alternative assembly poses and the assembly path, path replanning is performed to obtain multiple alternative paths; Obtain path adaptation references for multiple candidate paths, wherein the path adaptation references include path quality adaptation coefficients and assembly constraint completion coefficients, and obtain multiple comprehensive path adaptation coefficients based on the multiple path quality adaptation coefficients and the multiple assembly constraint completion coefficients. Each comprehensive path adaptation coefficient is corrected sequentially according to the assembly environment adaptation factor to obtain multiple corrected comprehensive path adaptation coefficients; The adaptation coefficients of multiple correction paths are sorted, and the candidate correction path with the highest adaptation coefficient is selected as the correction assembly guidance path.

5. The electrical component assembly guidance method according to claim 1, characterized in that, The step of acquiring real-time process quality monitoring data during the assembly process and obtaining a quality risk coefficient based on the real-time process quality monitoring data and preset standard process parameter thresholds includes: Acquire multimodal process quality monitoring data collected in real time during the assembly process, wherein the multimodal process quality monitoring data includes assembly image data, assembly force / torque data, and assembly position / orientation data; The visual appearance features of the assembly area are extracted based on the assembly image data, and the appearance quality deviation value is obtained based on the visual appearance features and the preset standard appearance feature threshold. Based on the assembly force / torque data, extract the force / torque time sequence variation characteristics during the assembly process, and obtain the mechanical process deviation value based on the force / torque time sequence variation characteristics and the preset standard force / torque curve threshold. Based on the assembly position / attitude data, extract the pose deviation features between the actual assembly pose and the target assembly pose, and obtain the pose deviation value based on the pose deviation features and the preset standard pose tolerance threshold. The quality risk coefficient is obtained by weighted fusion calculation based on the appearance quality deviation value, the mechanical process deviation value, and the posture deviation value.

6. The electrical component assembly guidance method according to claim 1, characterized in that, The step of generating a coordinated control signal based on the quality risk coefficient and the modified assembly guidance path, and controlling the electrical components to perform assembly guidance according to the coordinated control signal until the assembly is completed, includes: Based on the modified assembly guidance path, a basic control instruction sequence containing path points, timestamps, and state vectors is generated; The basic control command sequence is adjusted based on the quality risk coefficient to obtain an adjusted control command sequence; The adjustment control command sequence is encoded and time-synchronized to obtain a collaborative control signal, and the collaborative control signal is sent to a preset assembly execution unit, wherein the preset assembly execution unit includes a multi-degree-of-freedom robotic arm and a precision positioning platform. Based on the path point sequence in the collaborative control signal, the desired angle and speed commands of each joint of the robotic arm are calculated and issued in real time, driving the end effector of the robotic arm to move along the corrected assembly guide path to the specified assembly pose. During the assembly process, the feedback data from the force sensor, position sensor and vision sensor on the assembly execution unit are read in real time and compared in a closed loop with the process parameter instructions embedded in the collaborative control signal to generate a real-time execution deviation vector. Based on the real-time execution deviation vector, the motion trajectory or force of the robotic arm end effector is finely adjusted in real time according to the preset impedance control model. When the actual assembly contact force of the current assembly step reaches the preset target range, the assembly posture error converges to the allowable tolerance, and the visual inspection confirms that the part is in place, the current step is determined to be completed, and a step completion signal is sent to the control system. After receiving the step completion signal, the system automatically loads the collaborative control signal corresponding to the next assembly step according to the initial assembly sequence, and repeats the above execution, monitoring and fine-tuning process. The above steps are repeated until all steps in the initial assembly sequence have been completed and an assembly completion signal is output, thus ending the entire electrical component assembly guidance process.

7. An electrical component assembly guidance system, characterized in that, include: The data acquisition module is used to acquire digital design model data of electrical components to be assembled and real-time environmental perception data of the assembly site. The assembly relationship parsing module is used to obtain the hierarchical assembly relationship of electrical components based on the digital design model data, wherein the hierarchical assembly relationship includes the initial assembly sequence and the assembly path; An environment adaptation assessment module is used to obtain an assembly environment adaptation factor based on the real-time environment perception data and the assembly path. The path dynamic correction module is used to correct the assembly path based on the assembly environment adaptation factor, the initial assembly sequence and the real-time environment perception data to obtain a corrected assembly guidance path. The process quality monitoring and analysis module is used to acquire real-time process quality monitoring data during the assembly process, and to obtain the quality risk coefficient based on the real-time process quality monitoring data and preset standard process parameter thresholds. The collaborative control and execution module is used to generate a collaborative control signal based on the quality risk coefficient and the modified assembly guidance path, and to control the electrical components to perform assembly guidance according to the collaborative control signal until the assembly is completed.

8. An electrical component assembly guidance system according to claim 7, characterized in that, The assembly relationship parsing module includes: The structural analysis unit is used to perform topological analysis on the digital design model data, obtain the assembly relationship of multiple sub-components, and construct an assembly relationship topology diagram based on the multiple assembly relationships. A sequence planning unit is used to obtain an initial assembly sequence based on the assembly relationship topology diagram and preset assembly constraints. The trajectory generation unit is used to obtain the three-dimensional geometric information and target assembly pose of each sub-component according to the initial assembly sequence, and to obtain the preliminary spatial trajectory of multiple sub-components according to multiple three-dimensional geometric information and multiple target assembly poses. A smoothness evaluation unit is used to obtain multiple smoothness coefficients between adjacent trajectories based on multiple preliminary spatial trajectories; The path synthesis unit is used to splice and smooth multiple preliminary spatial trajectories according to multiple smoothing coefficients to obtain an assembly path.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images