A method for quality consistency testing in the assembly process of electromechanical products

Abstract

This invention discloses a method for quality consistency detection in the assembly process of electromechanical products, specifically relating to the field of assembly quality inspection of electromechanical products. The method includes: S1. Pre-assembly three-dimensional coordinate benchmark modeling of components, establishing an assembly benchmark coordinate system, acquiring the three-dimensional model and extracting key feature points to generate a benchmark coordinate library; S2. Component production deviation detection and accessory optimization, acquiring actual coordinates, calculating deviations, scoring according to a multi-dimensional weighted model, and deciding on accessory replacement based on the score; S3. Initial assembly of components, installing accessories sequentially according to the assembly work instructions; S4. Second inspection, verifying the assembly sequence and detecting spatial position deviations of components and torque and angles of connectors; S5. Comprehensive quality consistency scoring and judgment, scoring according to the model and classifying quality levels; S6. Assembly deviation correction and verification, tracing the causes of deviations and making targeted corrections until the score meets the standard. This method improves detection accuracy and problem identification rate, reduces costs, and ensures assembly quality consistency.

Description

Technical Field

[0001] This invention relates to the field of electromechanical product assembly quality inspection technology, and more specifically, to a method for quality consistency inspection during the assembly process of electromechanical products. Background Technology

[0002] In the manufacturing process of electromechanical products, the assembly stage is a key process that determines the final performance, stability, and service life of the product, and the quality consistency inspection of the assembly process is the core means to ensure assembly quality. As downstream industries continue to increase their requirements for the precision of electromechanical products, the industry's technical demand for assembly quality inspection has gradually evolved from basic dimensional verification to a full-process, high-precision, and traceable approach.

[0003] Currently, the mainstream method for inspecting the assembly quality of electromechanical products in the industry mainly relies on manual visual inspection combined with two-dimensional dimensional measuring tools, supplemented by a binary judgment standard of pass / fail. The specific process is usually as follows: first, the basic linear dimensions of each component to be assembled are measured to confirm that the component dimensions meet the design drawing requirements, and then the component is put into assembly; after assembly, the key appearance dimensions of the assembly are checked again using two-dimensional measuring tools, while the experience of the inspectors is relied upon to judge whether there are obvious defects in the assembly appearance, and finally, the assembly quality is judged based on a comprehensive assessment of the measurement results and experience.

[0004] In addition, some production scenarios with a certain level of automation will introduce simple inspection equipment such as ordinary optical inspection instruments to assist in dimensional measurement, while recording basic assembly process information through the Manufacturing Execution System (MES).

[0005] However, in practical use, it still has some shortcomings. For example, existing technologies mostly use two-dimensional dimensional measurement or manual visual inspection, which cannot accurately capture the three-dimensional spatial position deviation of the parts. They can only provide a qualitative description of linear dimensions and cannot reflect the degree of deviation through quantitative data, resulting in a deviation identification omission rate of over 25%, making it difficult to detect hidden dimensional problems. Existing technologies only use a binary judgment of qualified / unqualified to determine whether parts need to be replaced, without establishing a multi-dimensional quantitative scoring system. This can easily lead to the over-replacement of repairable parts or the omission of unqualified parts, increasing assembly costs and wasting parts resources. The secondary inspection of existing technologies often repeats basic dimensional inspections and does not focus on the core assembly links—the compliance of the assembly sequence and the reliability of the torque and tightening angle of the connectors. The identification rate of hidden problems such as incorrect assembly sequence and excessive torque deviation is low, which can easily lead to assembly with defects and create hidden risks of product operation failure. Existing technologies do not establish a deviation traceability mechanism. After assembly deviations occur, they often adopt a blind rework method of disassembly and reassembly, which cannot locate the root cause of the deviation, resulting in low rework efficiency and extended production cycle. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a method for quality consistency detection in the assembly process of electromechanical products, which solves the problems mentioned in the background art through the following solution.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for quality consistency detection in the assembly process of electromechanical products, comprising: S1: three-dimensional coordinate reference modeling of components before assembly: establishing a unified reference coordinate system to provide a reference for the subsequent quantitative calculation of component deviations;

[0008] S2: Component production deviation detection and parts optimization: Fix the component on the reference fixture, re-collect the actual coordinates of key feature points, calculate the production deviation, score the parts based on the production deviation using a multi-dimensional weighted model, and make parts replacement decisions based on the scores.

[0009] S3: Preliminary assembly of components: According to the assembly work instructions, first install the base and align it with the reference coordinate system, then install the core functional components in sequence, and finally pre-assemble the connecting parts.

[0010] S4: Second inspection: The assembly sequence compliance is checked through the MES system. A portable coordinate measuring machine is used to collect the coordinates of key feature points of the components after initial assembly and to detect the spatial position deviation of the components and the torque reliability of the connectors.

[0011] S5: Comprehensive Quality Consistency Scoring and Judgment: Based on secondary inspection deviation, torque, average component score, and sequential compliance, a multi-dimensional weighted model is used to score the assembly quality. If the score is unqualified, the assembly must be disassembled and retested.

[0012] S6: Assembly Deviation Correction and Verification: Based on the data from the first two inspections, trace the cause of the deviation, take targeted corrective measures, and re-inspect and score the assembly deviation until the score is ≥80.

[0013] Preferably, the three-dimensional coordinate reference modeling method is as follows:

[0014] Based on the product design drawings, an assembly reference coordinate system is established on the testing platform with the mounting surface of the electromechanical product base as the XY plane and the center of the positioning hole as the origin O.

[0015] For all components to be assembled, a laser 3D scanner is used to acquire 3D models and extract key feature points of each component:

[0016] The key feature points include:

[0017] Mechanical components: center of positioning hole, edge point of mounting surface, centerline of shaft parts;

[0018] Connector: head center point, threaded section axis;

[0019] Import the coordinates of all key feature points of the components into the detection system to generate a component reference coordinate library.

[0020] Preferably, the method for acquiring the actual coordinates of the key feature points is as follows:

[0021] The components to be assembled are fixed one by one to the reference fixture of the testing platform, with a repeatability of ≤0.002mm. The actual X coordinates of the key feature points of each component are then re-acquired using a laser 3D scanner. 实 Y 实 Z 实 ;

[0022] The production deviations include dimensional deviations, geometric tolerance deviations, surface fit, and marking consistency.

[0023] The method for calculating the dimensional deviation is as follows:

[0024] ;

[0025] The calculation method for the accessory rating is as follows:

[0026] Construct a parts quality scoring model and score according to the following weights and rules:

[0027] Dimensional deviation weighting 40%: ΔL≤0.01mm earns 40 points, deducting 10 points for every 0.005mm exceeding the limit;

[0028] Geometric tolerances are weighted at 30%: flatness ≤ 0.005mm earns 30 points, and points are deducted proportionally for deviations.

[0029] Surface compatibility weight 20%: 20 points are awarded for mounting surface roughness Ra≤1.6μm, and 5-15 points are deducted proportionally for deviations outside the tolerance;

[0030] Identification consistency weight 10%: 10 points for consistency between model, QR code and BOM, otherwise 0 points.

[0031] Preferably, the compliance check of the assembly sequence is as follows:

[0032] By retrieving the process records of the assembly process through the MES system, it can be verified whether the component installation sequence conforms to the assembly work instructions;

[0033] If there is an error in the assembly sequence of core components, the assembly is directly judged as unqualified, and subsequent inspections are skipped, triggering a rework process.

[0034] Preferably, the spatial position deviation of the component is detected as follows:

[0035] A portable coordinate measuring machine was used to re-acquire the actual coordinates (X, Y, X) of key feature points of each component after initial assembly. 装 Y 装 Z装 );

[0036] Calculate the position deviation degree, including:

[0037] Linear position deviation:

[0038] ;

[0039] Three-dimensional angle deviation: Calculate the included angle between the normal vectors of the actual installation surface of the component and the reference surface.

[0040] Preferably, the torque and reliability detection of the connecting piece are as follows:

[0041] Use an intelligent torque wrench to detect the actual torque values (M 实 ) and tightening angles (θ 实 ) of all screws / bolts one by one;

[0042] Compare with the design standard values (M0, θ0), and calculate the deviation: (For example, for an M8 screw, M0 = 20 N·m, and ΔM ≤ 1 N·m is qualified), , and Δθ ≤ 5° is qualified.

[0043] Preferably, the weights and scoring rules of each dimension of the multi-dimensional weighted model are as follows:

[0044] Weight of the secondary detection position deviation degree 40%: ΔL 装 ≤ 0.02 mm and the angle deviation ≤ 0.05° get 40 points, and points are deducted proportionally for exceeding the tolerance;

[0045] Weights of the torque and angle of the connecting piece 30%: ΔM ≤ 0.5 N·m and Δθ ≤ 3° get 30 points, and 5 - 30 points are deducted for exceeding the tolerance;

[0046] Weight of the average score of the first detection of accessories 15%: If the average score of all accessories ≥ 90 points, get 15 points, and 3 points are deducted for every 5 - point decrease;

[0047] Weight of the compliance of the assembly sequence 15%: Get 15 points for complete compliance, and 0 points for incorrect sequence;

[0048] If the score ≥ 90 points, it is judged as excellent and directly enter the next process; if the score is 80 - 89 points, it is judged as qualified, record the deviation data for subsequent optimization, and no rework is required; if the score is 60 - 79 points, it is judged as to be corrected, and the S6 correction process needs to be executed; if the score < 60 points, it is judged as unqualified, disassemble the assembly and re - execute the S2 - S5 process.

[0049] Preferably, the method for tracing the source of the deviation reason is as follows:

[0050] [[ID=​

[0051] If the positional deviation is mainly due to the component itself: replace or repair the component, and repeat S2-S4;

[0052] If the positional deviation is due to the assembly operation: adjust the reference fixture, reassemble, and then perform S4 inspection;

[0053] If the torque deviation originates from the tool: calibrate the torque tool, retighten the connection, and check.

[0054] Preferably, the corrective measures include:

[0055] Position deviation correction: The micro-adjustment method is adopted, which adjusts the component installation gap by using precision shims, or corrects the spatial angle by using an assembly robot;

[0056] Torque deviation correction: Retighten the connector using the torque-angle method, first tightening the torque to 80% of M0, then tightening it to the rated value at angle θ0;

[0057] The corrected assembly is re-inspected using S4 and scored using S5 until the score is ≥80, ensuring that the quality consistency meets the standards.

[0058] The technical effects and advantages of this invention are as follows:

[0059] This invention establishes a unified assembly reference coordinate system, uses a laser 3D scanner to collect the 3D model of the component and extract key feature points, and calculates the 3D linear deviation by combining formulas. At the same time, it covers dimensions such as form and position tolerance and surface adaptability, transforming the deviation from a qualitative description to quantitative data and reducing the omission of deviation identification.

[0060] This invention uses a weighted scoring-based component decision-making system to avoid invalid replacements and reduce assembly costs.

[0061] This invention improves the identification rate of hidden problems and eliminates assembly with defects by focusing on key aspects such as sequence, position, and torque through secondary detection.

[0062] This invention uses a deviation tracing matrix to locate the cause, and the corrected result improves the pass rate compared to traditional blind rework assembly. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the process of the present invention;

[0064] Figure 2 This is a device connection diagram for the present invention. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] As attached Figure 1-2 The method shown is for quality consistency testing in the assembly process of electromechanical products, comprising:

[0067] S1: 3D coordinate datum modeling of components before assembly: Establish a unified datum coordinate system to provide a reference for the subsequent quantitative calculation of component deviations;

[0068] Specifically, the three-dimensional coordinate reference modeling method is as follows:

[0069] Based on the product design drawings (CAD model), an assembly reference coordinate system (O-XYZ) is established on the testing platform (such as a marble precision platform with a flatness of ≤0.005mm / m²), with the mounting surface of the electromechanical product base as the XY plane and the center of the positioning hole as the origin O.

[0070] For all components to be assembled (such as motors, reducers, brackets, connectors, etc.), a laser 3D scanner (scanning accuracy 0.003mm, point cloud density 1000 points / mm²) was used to acquire 3D models and extract key feature points of each component.

[0071] The key feature points include:

[0072] Mechanical components: center of positioning hole (A1, A2...), edge point of mounting surface (B1, B2...), axis of shaft parts (C1-C2);

[0073] Connectors (screws, bolts): head center point (D1, D2...), threaded section axis (E1-E2);

[0074] Import the coordinates of all key feature points of the components into the detection system to generate a component reference coordinate library, which will serve as the reference for subsequent deviation calculations (e.g., the reference coordinates of motor mounting hole A1 are (X0, Y0, Z0)).

[0075] S2: Component production deviation detection and parts optimization: Fix the component on the reference fixture, re-collect the actual coordinates of key feature points, calculate the production deviation, score the parts based on the production deviation using a multi-dimensional weighted model, and make parts replacement decisions based on the scores.

[0076] Specifically, the method for acquiring the actual coordinates of the key feature points is as follows:

[0077] The components to be assembled are fixed one by one to the reference fixture of the testing platform, with a repeatability of ≤0.002mm. The actual X coordinates of the key feature points of each component are then re-acquired using a laser 3D scanner. 实 Y 实 Z 实 ;

[0078] The production deviations include dimensional deviations, geometric tolerance deviations, surface fit, and marking consistency.

[0079] The method for calculating the dimensional deviation is as follows:

[0080] ;

[0081] The calculation method for the accessory rating is as follows:

[0082] Construct a parts quality scoring model (out of 100 points), and score according to the following weights and rules:

[0083] Dimensional deviation (weight 40%): 40 points for ΔL≤0.01mm, deduct 10 points for every 0.005mm exceeding the tolerance;

[0084] Geometric tolerance deviation (weight 30%): Flatness ≤ 0.005mm gets 30 points, exceeding the tolerance will result in a deduction of points proportionally;

[0085] Surface compatibility (weight 20%): 20 points are awarded for mounting surface roughness Ra≤1.6μm, and 5-15 points are deducted proportionally for deviations outside the tolerance.

[0086] Identification consistency (weight 10%): 10 points for model number, QR code and BOM consistency, otherwise 0 points.

[0087] The component replacement decision-making matching method is as follows:

[0088] Parts with a score of ≥85 are considered compatible and proceed directly to the assembly stage; parts with a score of 60 ≤ score <85 are considered repairable and, after being repaired by precision grinding and laser correction, undergo the above three-step process again. Only parts that meet the standards can be assembled; parts with a score <85 are considered unusable and are replaced with new parts, and this process is repeated until the part score is ≥85.

[0089] Taking the motor mounting holes of an industrial robot joint module as an example:

[0090] Design reference coordinates (X0, Y0, Z0): (50.000mm, 50.000mm, 100.000mm) (extracted from CAD model);

[0091] Actual detection coordinates (X) 实 ,Y 实 Z 实(50.008mm, 50.005mm, 100.003mm) (acquired by laser scanner);

[0092] Deviation calculation:

[0093] ΔL= ≈0.010mm, which meets the excellent standard of ΔL≤0.01mm, and scores 40 points in the dimension of size deviation in the accessory evaluation.

[0094] S3: Preliminary assembly of components: According to the assembly work instructions, first install the base and align it with the reference coordinate system, then install the core functional components in sequence, and finally pre-assemble the connecting parts.

[0095] S4: Second inspection: The assembly sequence compliance is checked through the MES system. A portable coordinate measuring machine is used to collect the coordinates of key feature points of the components after initial assembly and to detect the spatial position deviation of the components and the torque reliability of the connectors.

[0096] Specifically, the compliance check for the assembly sequence is as follows:

[0097] By retrieving process records (such as barcode scanning records and sensor trigger times) from the MES system, we can verify whether the component installation sequence conforms to the assembly work instructions (such as whether the installation of the bracket first and then the motor is performed).

[0098] If there is an error in the assembly sequence of core components (such as installing the motor directly without the locating pin), it will be directly judged as an assembly failure and trigger the rework process (skipping subsequent inspections).

[0099] The spatial position deviation of the component is detected as follows:

[0100] A portable coordinate measuring machine (with a measurement accuracy of 0.005 mm) was used to re-acquire the actual coordinates (X, Y, X) of key feature points of each component after initial assembly. 装 Y 装 Z 装 );

[0101] Calculating the position deviation includes:

[0102] Linear position deviation:

[0103] (For example, a linear deviation of the motor shaft centerline ≤ 0.03mm is considered acceptable).

[0104] Three-dimensional angular deviation: Calculate the angle between the normal vector of the actual mounting surface of the component and the reference plane (e.g., the angular deviation between the mounting surface of the reducer and the XY plane of the base is ≤0.1° to be considered acceptable).

[0105] The torque and reliability of the connectors are tested as follows:

[0106] Use a smart torque wrench (measurement accuracy ±1%, sampling rate 10 Hz) to detect the actual torque values (M 实 ), and the tightening angle (θ 实 ) of all screws / bolts one by one;

[0107] Compare with the design standard values (M0, θ0), and calculate the deviation: (For example, for an M8 screw, M0 = 20 N·m, and ΔM ≤ 1 N·m is qualified), (Δθ ≤ 5° is qualified);

[0108] Take the M8 motor fixing bolt as an example:

[0109] Design standard torque (M0): 20 N·m (determined according to the bolt material (8.8 - grade steel) and the connecting components (motor housing + aluminum alloy bracket));

[0110] Actual tightening torque (M 实 ): 19.5 N·m (stable value collected by the smart torque wrench);

[0111] Deviation calculation: ΔM = |19.5 - 20| = 0.5 N·m, meeting the excellent standard of ΔM ≤ 0.5 N·m, corresponding to 30 points in the torque and angle dimensions of the comprehensive score;

[0112] Design standard tightening angle (θ0): 45° (According to the bolt length of 20 mm, pitch of 1.25 mm, set the tightening depth of 1.5 mm, corresponding to a rotation angle of 432°, and the actual process is optimized to 45°, that is, after the torque reaches 16 N·m, rotate another 45° to 20 N·m);

[0113] Actual tightening angle (θactual): 43°;

[0114] Deviation calculation: Δθ = |43 - 45| = 2°, meeting the excellent standard of Δθ ≤ 3°, jointly supporting a full score in the torque and angle dimensions with the torque deviation (0.5 N·m);

[0115] For the connecting parts of key positions (such as motor fixing bolts), additionally detect the coincidence degree of the torque - angle curve, compare with the standard curve, and a coincidence degree ≥ 90% is qualified.

[0116] S5: Comprehensive quality consistency scoring and determination: Based on the secondary detection deviation, torque, average score of accessories, and sequence compliance, score according to the multi - dimensional weighted model, and determine the assembly quality according to the score. If unqualified, disassemble and retest;

[0117] The weights and scoring rules of each dimension of the multi - dimensional weighted model are as follows:

[0118] Degree of deviation of the secondary detection position (weight 40%): ΔL 装A score of 40 is awarded for a tolerance of ≤0.02mm and an angle deviation of ≤0.05°; points will be deducted proportionally for deviations exceeding the tolerance.

[0119] Connector torque and angle (weight 30%): 30 points for ΔM≤0.5N・m and Δθ≤3°, deduct 5-30 points for exceeding the tolerance;

[0120] First inspection average score of all parts (weight 15%): 15 points are awarded if the average score of all parts is ≥90 points, and 3 points are deducted for every 5 points decrease;

[0121] Assembly sequence compliance (weight 15%): 15 points for full compliance, 0 points for incorrect sequence.

[0122] It should be further explained that a score of ≥90 is considered excellent and proceeds directly to the next process; a score of 80-89 is considered acceptable, and the deviation data is recorded for subsequent optimization without rework; a score of 60-79 is considered to need correction and requires the execution of the S6 correction process; a score <60 is considered unacceptable, and the assembly is disassembled and the S2-S5 process is re-executed.

[0123] S6: Assembly Deviation Correction and Verification: Based on the data from the first two inspections, trace the cause of the deviation, take targeted corrective measures, and re-inspect and score the assembly deviation until it reaches ≥80 points.

[0124] The method for tracing the cause of the deviation is as follows:

[0125] By combining the detection data from S2 and S4, the cause can be located using the deviation tracing matrix:

[0126] If the positional deviation is mainly due to the component itself (e.g., the component score in S2 is <90 points): replace or repair the component, and repeat S2-S4;

[0127] If the positional deviation is caused by the assembly operation (such as inaccurate fixture positioning): adjust the reference fixture (such as replacing the positioning pin), reassemble, and then perform the S4 test.

[0128] If the torque deviation is due to a tool (such as a smart wrench calibration failure): calibrate the torque tool (to restore accuracy to ±0.5%), retighten the connector and test;

[0129] The corrective measures include:

[0130] Position deviation correction: The micro-adjustment method is adopted, which adjusts the component installation gap by using precision shims (thickness accuracy 0.001mm), or corrects the spatial angle by using an assembly robot (repeat positioning accuracy 0.002mm);

[0131] Torque deviation correction: Retighten the connector using the torque-angle method, first tightening the torque to 80% of M0, then tightening it to the rated value at angle θ0;

[0132] The corrected assembly is re-performed with S4 inspection and S5 scoring until the score is ≥80 points to ensure that the quality consistency meets the standards.

[0133] To implement the above methods, a consistency testing system for the assembly quality of electromechanical products is constructed, including: a testing and data acquisition layer, a data processing and decision-making layer, and an execution and correction layer;

[0134] The detection and data acquisition layer includes:

[0135] 3D Inspection Module: Laser 3D scanner, portable coordinate measuring machine, reference fixture, responsible for acquiring the 3D coordinates of parts and assemblies;

[0136] Torque detection module: intelligent torque wrench (with data storage function) and torque-angle curve analyzer, responsible for detecting the torque of the connector and the tightening process;

[0137] The data processing and decision-making layer includes:

[0138] Data processing module: Industrial computer (equipped with point cloud processing software such as Geomagic ControlX) performs deviation calculation and scoring model operation;

[0139] Decision and Traceability Module: The MES system interface records inspection data, scoring results, and corrective measures, enabling full-process traceability from component to assembly to inspection and correction.

[0140] The execution and correction layer includes:

[0141] Correction execution module: Precision gasket library, assembly robot, tool calibrator, responsible for deviation correction operations.

[0142] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0143] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for quality consistency detection of an electromechanical product assembly process, characterized in that, Comprise: S1: Assemble the front part of the three-dimensional coordinate reference modeling: Establish a unified reference coordinate system, provide a reference for subsequent part deviation quantification calculation; S2: Part production deviation detection and accessory optimization: Fix the part on the reference fixture, reacquire the actual coordinates of the key feature points, calculate the production deviation, score the accessories based on the production deviation according to the multi-dimensional weighting model, and match the accessory replacement decision according to the score; S3: Part initial assembly: According to the assembly instruction, first assemble the base and align the reference coordinate system, then assemble the core function piece one by one, and finally preassemble the connecting piece; S4: Second detection: Through the MES system, the assembly sequence compliance detection is carried out, the portable three-coordinate measuring machine is used to collect the key feature point coordinates of the parts after initial assembly and carry out part space position deviation detection and connecting piece torque reliability detection; The assembly sequence compliance detection is as follows: Through the MES system, the process record of the assembly process is called, and whether the part installation sequence meets the assembly instruction is verified; If there is an error in the core part assembly sequence, it is directly determined as unqualified assembly, and the subsequent detection is skipped to trigger the rework process; The connecting piece torque reliability detection is as follows: Using intelligent torque wrench, detecting actual torque value M of all screws / bolts one by one 实 With tightening angle θ 实 ; The deviation is calculated by comparing the design standard value (M0, θ0): , , Δθ≤5° is qualified; S5: Comprehensive quality consistency score and judgment: Based on the second detection deviation, torque, accessory score average, sequence compliance, score according to the multi-dimensional weighting model, judge the assembly quality according to the score, and disassemble and retest if unqualified; The weight and score rule of each dimension of the multi-dimensional weighting model is as follows: Secondary detection position deviation weight 40%: AL 装 ≤0.02mm and angle deviation ≤0.05° 40 points, out of tolerance according to the proportion of points; Connecting piece torque and angle weight 30%: ΔM≤0.5N・m and Δθ≤3° get 30 points, exceed the tolerance by 5-30 points; The first detection accessory score average weight is 15%: all accessory score average ≥90 points get 15 points, decrease by 3 points for every 5 points; Assembly sequence compliance weight 15%: completely compliant get 15 points, sequence error get 0 points; Score ≥90 points is judged as excellent, directly enter the next process; 80-89 points is judged as qualified, record the deviation data for subsequent optimization, no need to rework; 60-79 points is judged as to be corrected, S6 correction process needs to be executed; <60 points is judged as unqualified, disassemble the assembly body and execute S2-S5 process again; S6: Assembly deviation correction and verification: Based on the deviation reason of the previous two detection data, take corrective measures, re-detect the score after correction, until ≥80 points; The deviation reason tracing method is as follows: Combine the detection data of S2 and S4, locate the reason through the deviation tracing matrix: If the position deviation is mainly caused by the part itself: replace or repair the part, repeat S2-S4; If the position deviation is caused by assembly operation: adjust the reference fixture, re-initialize the assembly and execute S4 detection; If the torque deviation is caused by the tool: calibrate the torque tool, re-tighten the connecting piece and detect.

2. The method for quality consistency detection in the assembly process of electromechanical products according to claim 1, characterized in that: The three-dimensional coordinate reference modeling method is as follows: Based on product design drawings, take the base mounting surface of the electromechanical product as the XY plane on the detection platform, and establish the assembly reference coordinate system with the hole center as the origin O; For all parts to be assembled, use a laser three-dimensional scanner to collect three-dimensional models and extract key feature points of each part: The key feature points include: Mechanical parts: positioning hole center, mounting surface edge point, shaft center line; Connecting parts: head center point, thread segment axis; Import all the key feature point coordinates of the parts into the detection system to generate a part reference coordinate library.

3. The method for quality consistency detection in the assembly process of electromechanical products according to claim 1, characterized in that: The method for collecting the actual coordinates of the key feature points is as follows: The to-be-assembled components are fixed one by one on the reference fixture of the detection platform, the positioning accuracy is ≤0.002mm, and the actual coordinates X 实 , Y 实 , and Z 实 of key feature points of each component are re-acquired through a laser three-dimensional scanner. The production deviation includes size deviation, geometric tolerance deviation, surface adaptability and identification consistency; The calculation method of the size deviation is as follows: ； The calculation method of the fitting score of the fitting part is as follows: A fitting part quality score model is constructed, and the score is calculated according to the following weights and rules: Size deviation weight 40%: ΔL≤0.01mm gets 40 points, and each 0.005mm exceeds 10 points; Geometric tolerance deviation weight 30%: flatness ≤0.005mm gets 30 points, and the out-of-tolerance is deducted according to the proportion; Surface adaptability weight 20%: mounting surface roughness Ra≤1.6μm gets 20 points, and the out-of-tolerance is deducted according to the proportion; Identification consistency weight 10%: model, two-dimensional code and BOM are consistent to get 10 points, otherwise 0 points.

4. The method for quality consistency detection in the assembly process of electromechanical products according to claim 1, characterized in that: The part space position deviation detection is as follows: The portable three-coordinate measuring machine is used to re-collect the actual coordinates (X 装 , Y 装 , Z 装 ) of the key feature points of each component after the initial assembly. Calculate the position deviation degree, including: Linear position deviation: ； Three-dimensional angle deviation: calculate the normal vector angle between the actual mounting surface of the part and the reference surface.

5. The method for quality consistency detection for the assembly process of electromechanical products according to claim 1, characterized in that: The correction measures include: Position deviation correction: adopt micro-adjustment method, adjust the part mounting gap through precision gasket, or correct the space angle through assembly robot; Torque deviation correction: according to the torque-angle method, retighten the connecting part, first tighten the torque to 80% of M0, and then tighten to the rated value according to θ0 angle; Re-execute S4 detection and S5 scoring on the corrected assembly body until the score is ≥80 points to ensure that the quality consistency meets the standard.

Images