Intelligent detection method and system for grinding-in rate of automobile injection molding gear

Abstract

The invention provides an intelligent detection method and system for the grinding-in rate of an automobile injection molding gear, and relates to the technical field of material testing and analysis, and the method comprises the steps: obtaining the geometrical characteristic parameters of a meshing surface based on three-dimensional laser scanning; calculating a dynamic contact stress distribution index through a piezoelectric sensor array and a finite element model; measuring a tooth profile deviation fluctuation coefficient by using a high-speed camera system and an image processing algorithm; constructing a machine learning comprehensive evaluation model to output three-dimensional indexes; and a self-adaptive grinding scheme is generated according to the evaluation result, and closed-loop control machining is carried out. Through multi-source data fusion and intelligent optimization, the problems that a traditional method depends on manpower and is single in detection dimension are solved, accurate analysis of dynamic contact characteristics and real-time correction of machining parameters are achieved, and the detection precision and the production efficiency are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of material testing and analysis, and particularly to an intelligent detection method and system for the running-in rate of automotive injection-molded gears. Background Art

[0002] In the contact performance detection of automotive injection-molded gears, traditional methods mainly rely on the red lead powder contact method and manual visual assessment. During operation, red lead powder needs to be applied to the gear surface, and the contact quality is judged by the distribution of contact spots after meshing. However, this method can only reflect the static contact state and cannot capture the stress changes and tooth profile deviation fluctuations during the dynamic meshing process. Manual interpretation is highly subjective, with low quantization accuracy, and cannot identify microscopic surface defects, resulting in insufficient reliability of the detection results and difficulty in meeting the manufacturing requirements of high-precision gears.

[0003] Existing technologies have tried to introduce single-point stress sensors or two-dimensional image analysis methods to improve the objectivity of detection, but there are still significant limitations. Single-point sensors can only obtain local stress data and cannot comprehensively characterize the stress distribution characteristics of the contact area, resulting in a relatively high missed detection rate for abnormal stress concentration areas. Although two-dimensional image analysis can identify contact trajectory offsets, it lacks the depth information of three-dimensional topography and cannot quantify key parameters such as tooth surface waviness, making it difficult to accurately evaluate the contact stability under dynamic meshing.

[0004] Current detection systems generally adopt isolated data acquisition and analysis modules, and there is no effective coordination among geometric parameters, mechanical properties, and dynamic behavior data. The detection results are disjointed from the processing and correction links, and it is necessary to rely on manual experience to repeatedly try and adjust the grinding parameters, which is not only inefficient but also likely to cause a decrease in gear strength due to excessive modification. In addition, existing methods do not construct a real-time feedback closed-loop control mechanism, making it difficult to achieve intelligent optimization of detection-processing integration. Summary of the Invention

[0005] In order to solve the technical problems in the prior art, such as single detection means, reliance on manual experience, inability to analyze the contact characteristics of gears in real time and dynamically, and lack of intelligent closed-loop control, the present invention provides an intelligent detection method and system for the running-in rate of automotive injection-molded gears.

[0006] The technical solutions provided by the present invention are as follows:

[0007] First aspect:

[0008] An intelligent detection method for the running-in rate of automotive injection-molded gears provided by the present invention includes:

[0009] S1. Based on the three-dimensional topography data of the gear meshing surface collected by a three-dimensional laser scanner, obtain the geometric feature parameter set of the tooth surface contact area. The geometric feature parameter set at least includes the contact spot distribution density, the contact trajectory offset, and the tooth surface waviness;

[0010] S2. Construct a Dynamic Contact Stress Distribution Index (DCSDI). By using a piezoelectric stress sensor array installed on the gear test bench, collect the contact stress time-domain signal in real time, and calculate the dynamic contact stress distribution index in combination with the finite element contact analysis model;

[0011] S3. Establish a Profile Deviation Fluctuation Coefficient (PDFC). Use a high-speed imaging system to synchronously capture the image sequence of the contact trajectory during the meshing process of the gear pair, and obtain the time-series change characteristics of the tooth profile deviation through the image feature extraction algorithm;

[0012] S4. Input the dynamic contact stress distribution index and the tooth profile deviation fluctuation coefficient into the comprehensive evaluation model of the lapping rate. This model performs multi-dimensional matching analysis on the two characteristic parameters and the standard lapping rate database through machine learning algorithms, and outputs a three-dimensional evaluation index including the qualified rate of the contact area, the stress distribution uniformity, and the dynamic matching accuracy;

[0013] S5. Generate a contact optimization plan according to the three-dimensional evaluation index, adaptively grind the gear meshing surface through a numerical control profiling device, and real-time feedback and correct the processing parameters until the preset lapping rate threshold is reached.

[0014] Furthermore, the calculation method of the dynamic contact stress distribution index includes:

[0015]

[0016] where, σ max is the maximum stress value in the contact area, with the unit of MPa, σ avg is the average stress value in the contact area with the unit of MPa, N c is the stress concentration coefficient, determined according to the gear contact strength calculation method in ISO 6336 standard, ΔA is the area of the abnormal stress area with the unit of mm 2 , A0 is the theoretical contact area with the unit of mm 2 . The specific calculation process includes: perform spatial interpolation processing on the stress data collected by the piezoelectric sensor array, generate a contact stress nephogram, then use the region segmentation algorithm to identify the abnormal stress area, and perform area integral calculation in combination with the mesh division result of the finite element model.

[0017] Furthermore, the method for determining the fluctuation coefficient of the tooth profile deviation includes:

[0018]

[0019] where T is the complete meshing period, with the unit of s, and ΔF p (t) is the dynamic change amount of the pitch deviation, with the unit of μm, and F p0 is the reference value of the pitch tolerance, with the unit of μm, and Δf Hα (t) is the dynamic change amount of the tooth profile inclination deviation, with the unit of μm, and f Hα0 is the reference value of the tooth profile inclination tolerance, with the unit of μm. The calculation process includes: performing sub-pixel edge detection on the contact trajectory image obtained by the high-speed imaging system, eliminating the influence of motion blur by using the spatio-temporal domain filtering algorithm, and then inversely calculating the tooth profile deviation parameters through the gear meshing kinematic model.

[0020] Furthermore, the method for constructing the comprehensive evaluation model of the lapping rate in S4 includes:

[0021] Using the Improved Analytic Hierarchy Process (IAHP) to determine the weight distribution of each evaluation index, where an improved contact trajectory optimization algorithm is introduced to calculate the weight correction factor:

[0022]

[0023] where w i is the initial weight, α is the system stability coefficient, with the range of 0.1 - 0.3, β is the parameter sensitivity adjustment factor, with the range of 0.5 - 1.2, and C i is the confidence level of the characteristic parameter, which is calculated and determined according to the measurement uncertainty corresponding to the gear accuracy grade in ISO 1328 - 1:2013 standard.

[0024] Furthermore, the contact trajectory optimization algorithm adopts a hybrid optimization strategy of Genetic Algorithm (GA) and Particle Swarm Optimization (PSO):

[0025] where the fitness function is defined as:

[0026]

[0027] where k1, k2, and k3 are weighting coefficients, satisfying k1 + k2 + k3 = 1, and S contact is the ratio of the actual contact area to the theoretical contact area. The crossover probability P c and the mutation probability P mDynamically adjusted according to the population diversity index.

[0028] Furthermore, the finite element contact analysis model in S2 uses the Johnson-Cook dynamic constitutive model to describe the material properties: the expression is:

[0029]

[0030] where A is the yield stress of the material, in MPa, B is the strain hardening coefficient, n is the strain hardening index, C is the strain rate sensitivity coefficient, ε is the equivalent plastic strain, is the dimensionless strain rate, T is the relative temperature, and the model parameters are automatically matched and obtained through the material test database.

[0031] Furthermore, the image feature extraction algorithm in S3 uses an improved Canny edge detection algorithm, combines the gear meshing phase information to identify the contact trajectory, and the non-maximum suppression threshold is dynamically adjusted according to the image signal-to-noise ratio:

[0032]

[0033] T l = 0.4×T h

[0034] where σ n is the standard deviation of the image noise, I max 、I min are the maximum and minimum gray values of the image respectively, k is an empirical coefficient, with a range of 0.8 - 1.2, and the algorithm realizes real-time processing on the FPGA chip.

[0035] Second aspect:

[0036] An intelligent detection system for the lapping rate of an automotive injection-molded gear provided by the present invention includes:

[0037] A three-dimensional scanning module, equipped with a laser interferometer and a confocal displacement sensor, for obtaining the microscopic topography data of the gear meshing surface;

[0038] A stress analysis module, including a piezoelectric stress sensor array distributed in a ring and a high-speed data acquisition card, for real-time monitoring of the contact stress distribution;

[0039] An image processing module, integrated with a dual CCD high-speed camera and an LED pulse light source, for capturing dynamic images of the meshing contact trajectory;

[0040] A computing center, equipped with a GPU parallel computing unit, for running the finite element contact analysis model and the machine learning evaluation model;

[0041] The feedback control module, including a numerically controlled grinding machine and an adaptive PID controller, automatically adjusts the processing parameters according to the evaluation results.

[0042] Further, the piezoelectric stress sensor array of the stress analysis module is distributed in concentric circles, the distance between adjacent sensors is less than 1.5 mm, and it is connected to the calculation center through a fiber optic transmission protocol; the dual CCD high-speed cameras of the image processing module are arranged at a 90° orthogonal angle, and the frame rate is not less than 2000 fps.

[0043] Further, the numerically controlled grinding machine of the feedback control module includes a five-axis linkage robotic arm and a diamond grinding head, the grinding pressure control accuracy is ±0.1 N, and the position repeat positioning accuracy is ±2 μm; the adaptive PID controller exchanges data with the calculation center through a real-time Ethernet protocol.

[0044] The beneficial effects brought by the technical solution provided by the present invention at least include:

[0045] (1) In the present invention, by integrating three-dimensional topography scanning, dynamic stress analysis and tooth profile deviation dynamic monitoring technologies, a multi-dimensional data collaborative detection system is constructed, which significantly improves the comprehensiveness and accuracy of the lapping rate detection. The use of a high-precision sensor array and high-speed image processing technology realizes the real-time synchronous acquisition of the contact stress distribution and tooth profile deviation, overcoming the limitations of traditional single-point detection. Combined with a machine learning evaluation model, it can automatically identify abnormal areas under complex contact patterns, avoid the subjective errors of manual interpretation, greatly shorten the detection cycle, and ensure quality consistency in mass production.

[0046] (2) In the present invention, based on the feedback control mechanism of dynamic evaluation indicators, the detection results are intelligently associated with the processing parameters, and a targeted grinding plan is automatically generated. Through a multi-axis linkage precision processing device and an adaptive control algorithm, the grinding pressure, path and speed are dynamically adjusted, effectively eliminating local stress concentration and tooth profile deviation. This closed-loop control strategy can reduce the material waste caused by traditional trial-and-error processing, improve the contact uniformity and load-bearing capacity of the gear pair, extend the service life, and at the same time reduce the dependence on the experience of operators.

[0047] (3) In the present invention, the topography acquisition, stress analysis, image processing and numerically controlled machining modules are deeply integrated to construct a full-process intelligent detection - machining system. Through a unified data interface and a real-time communication protocol, efficient cooperation of each module is realized, avoiding data breaks caused by switching between multiple devices. The embedded expert knowledge base and optimization algorithm support the rapid adaptation of multiple types of gears, significantly improving the flexible production capacity of the production line and providing a complete solution for the digital upgrade of gear manufacturing. Description of the Drawings

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 A schematic diagram of a flow chart of an intelligent detection method for the grinding ratio of an automobile injection molded gear provided by an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of the structure of an intelligent detection system for the grinding rate of automotive injection molded gears provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of data flow and processing logic of an intelligent detection system for the grinding rate of automotive injection molded gears provided in an embodiment of the present invention.

[0052] In the figure: 301, three-dimensional scanning module; 302, stress analysis module; 303, image processing module; 304, computing center; 305, feedback control module. DETAILED DESCRIPTION

[0053] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0054] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0055] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0056] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are consistent.

[0057] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0058] Refer to the attached instruction manual Figure 1 , which shows a schematic flowchart of an intelligent detection method for the lapping rate of an automotive injection-molded gear provided by an embodiment of the present invention.

[0059] An embodiment of the present invention provides an intelligent detection method for the lapping rate of an automotive injection-molded gear. This method can be implemented by an intelligent detection device for the lapping rate of an automotive injection-molded gear. The intelligent detection device for the lapping rate of an automotive injection-molded gear can be a terminal or a server. The processing flow of the intelligent detection method for the lapping rate of an automotive injection-molded gear can include the following steps:

[0060] S1. Based on a 3D laser scanner, collect 3D topography data of the gear meshing surface, and obtain a set of geometric feature parameters of the tooth surface contact area. The set of geometric feature parameters at least includes the contact spot distribution density, the contact trajectory offset, and the tooth surface waviness.

[0061] In a possible implementation, a Keyence LJ-V7300 3D laser scanner is used, with a lateral resolution of 5μm and a longitudinal repeatability accuracy of 0.1μm. When scanning, the gear is installed on an air-bearing turntable (radial runout ≤ 2μm) and rotates at a constant speed of 10r / min. After the collected meshing point cloud data is denoised by Gaussian filtering (σ = 0.1mm), the following geometric feature parameters are extracted:

[0062] Contact spot distribution density: Identify contact spots through the Otsu adaptive threshold segmentation method. The calculation density formula is:

[0063]

[0064] where N spot is the number of spots, A contact is the contact area, and the qualified range is 150 - 400 pieces / cm 2 .

[0065] Contact trajectory offset: Generate a theoretical involute equation based on gear parameters, perform a cubic B-spline fitting on the actual trajectory, and calculate the root mean square deviation (RMS). The maximum allowable deviation is ≤ 15μm.

[0066] Tooth surface waviness: Use wavelet analysis to extract the surface components with wavelengths of 0.1 - 1mm, set the amplitude threshold to ±2μm, and mark the over-limit area as the grinding target.

[0067] S2. Construct a dynamic contact stress distribution index DCSDI. Real-time collect the contact stress time-domain signals through a piezoelectric stress sensor array installed on the gear test bench, and calculate the dynamic contact stress distribution index in combination with a finite element contact analysis model.

[0068] In a possible implementation, 16 Kistler 9313B piezoelectric stress sensors are installed on the gear test bench, distributed in a helix (pitch angle 15°), with a sampling frequency of 10 kHz. The signals are conditioned by the NI PXIe-4330 module. A gear contact model is constructed based on ANSYS Workbench, with a mesh size of 0.2 mm. The material model uses the Johnson-Cook dynamic constitutive equation:

[0069]

[0070] The parameters are set as: A = 80 MPa, B = 120 MPa, n = 0.35, C = 0.015, m = 1.2, and the reference strain rate is 1.0 s -1 , and the temperature parameter T* = (T - 25) / 235.

[0071] According to the formula

[0072]

[0073] where σ max and σ avg are obtained by interpolating the stress nephogram, and N c is calculated according to ISO 6336-2019. ΔA is the area of the abnormal stress region (by finite element mesh integration). Example: When σ max = 85 MPa, σ avg = 32 MPa, N c = 1.8, ΔA = 3.2 mm 2 and A0 = 25.6 mm 2 , DCSDI = 2.89. When it exceeds the threshold of 3.0, processing is triggered.

[0074] S3. Establish the profile deviation fluctuation coefficient PDFC. Use a high-speed camera system to synchronously capture the image sequence of the contact trajectory during the meshing process of the gear pair, and obtain the time-series change characteristics of the profile deviation through the image feature extraction algorithm.

[0075] In a possible implementation, a Photron FASTCAM Nova S12 high-speed camera (resolution 1280×1024, frame rate 5000 fps) is used, equipped with a 630 nm narrow-band filter and a pulsed LED light source (pulse width 50 μs), and a synchronous encoder is used to obtain the gear speed.

[0076] Calculate the blur length L = w × r × t according to the rotational speed ω (rad / s) exp , and use the Richardson-Lucy deconvolution to restore the image. Locate the edge based on the Zernike moment algorithm, with an accuracy of 0.1 pixel.

[0077] According to the formula

[0078]

[0079] Among them, T = 0.2 s (corresponding to 2000 r / min), ΔF p (t) fluctuates by ±12 μm (F p0 = 20 μm), Δf Hα (t) fluctuates by ±8 μm (f Hα0 = 15 μm), and the calculation result PDFC = 0.68, and the qualified threshold ≤ 0.7.

[0080] S4. Input the dynamic contact stress distribution index and the profile deviation fluctuation coefficient into the lapping rate comprehensive evaluation model. This model performs multi-dimensional matching analysis on the two characteristic parameters and the standard lapping rate database through machine learning algorithms, and outputs three-dimensional evaluation indicators including the contact area qualification rate, stress distribution uniformity, and dynamic matching accuracy.

[0081] In a possible implementation manner, the improved analytic hierarchy process (IAHP) is adopted to construct a judgment matrix (contact area qualification rate: stress distribution uniformity: dynamic matching accuracy = 1:1.5:2), calculate the initial weight w = [0.45, 0.35, 0.20], and through the correction factor formula:

[0082]

[0083] Adjust the parameters α = 0.2, β = 0.8, C_i = 0.92 (measured uncertainty according to ISO 1328-1), and the corrected weight is [0.53, 0.40, 0.22]. Combine the genetic algorithm (population size 200, crossover probability Pc = linearly decreasing from 0.8 to 0.2) and the particle swarm optimization (inertia weight ω = 0.9 - 0.4), and the fitness function is:

[0084]

[0085] Among them, S contact is the ratio of the actual contact area to the theoretical value, and the optimization goal is Fitness ≥ 0.85.

[0086] S5. Generate a contact optimization plan according to the three-dimensional evaluation indicators, perform adaptive grinding on the gear meshing surface through a numerical control profiling device, and provide real-time feedback to correct the processing parameters until the preset lapping rate threshold is reached.

[0087] In a possible implementation manner, the processing parameter mapping is as follows:

[0088] Qualified rate of contact area Abrasive pressure (N) Feed rate (mm / s) Path overlap rate <70％ 8.0±0.5 15 40％ 70％-85％ 6.5±0.3 20 30％ >85％ 5.0±0.2 25 20％

[0089] Adopt TX2-90L five-axis robotic arm (repeated positioning accuracy ±2μm), grinding pressure control accuracy ±0.1N, PID parameters are adjusted according to the rules:

[0090] K p = 2.5×(1 - S contact )

[0091] K i = 0.8 / K p

[0092] K d = 0.05×σ max

[0093] After processing every 3 tooth surfaces, re-detection is triggered until DCSDI < 3.0 and PDFC < 0.7 for three consecutive times.

[0094] The present invention also provides an intelligent detection system for the lapping rate of automotive injection-molded gears, which is applied to the intelligent detection method for the lapping rate of the above-mentioned automotive injection-molded gears, and includes:

[0095] As Figure 2 shown, this system is composed of a three-dimensional scanning module 301, a stress analysis module 302, an image processing module 303, a calculation center 304, and a feedback control module 305. The three-dimensional scanning module 301 includes a laser interferometer and a confocal displacement sensor, and transmits three-dimensional topography data to the calculation center 304 through a PCIe interface; the annular piezoelectric sensor array of the stress analysis module 302 is synchronized through the IEEE 1588 protocol and sends dynamic stress data to the calculation center 304; the double CCD cameras and FPGA chips of the image processing module 303 output tooth profile deviation data to the calculation center 304 in real time. The calculation center 304 integrates a GPU parallel computing unit, a machine learning evaluation model, and an expert knowledge base. After receiving all detection data, it generates processing instructions and sends them to the five-axis robotic arm and grinding head of the feedback control module 305 through the EtherCAT protocol. The feedback control module 305 transmits the processing status back to the calculation center 304 in real time to form a detection - processing closed loop.

[0096] The three-dimensional scanning module is configured with a laser interferometer and a confocal displacement sensor for obtaining the microscopic topography data of the gear meshing surface; the stress analysis module includes an annularly distributed piezoelectric stress sensor array and a high-speed data acquisition card for real-time monitoring of the contact stress distribution; the image processing module is integrated with double CCD high-speed cameras and an LED pulse light source for capturing dynamic images of the meshing contact trajectory; the calculation center is equipped with a GPU parallel computing unit for running a finite element contact analysis model and a machine learning evaluation model; the feedback control module includes a numerically controlled grinding machine and an adaptive PID controller for automatically adjusting processing parameters according to the evaluation results.

[0097] In a possible implementation, the piezoelectric stress sensor array of the stress analysis module is distributed in concentric circles, the spacing between adjacent sensors is less than 1.5 mm, and it is connected to the calculation center through the fiber optic transmission protocol; the dual CCD high-speed cameras of the image processing module are arranged at a 90° orthogonal angle, and the frame rate is not less than 2000 fps.

[0098] In a possible implementation, the numerically controlled grinding machine of the feedback control module includes a five-axis linkage robotic arm and a diamond grinding head, the grinding pressure control accuracy is ±0.1 N, and the position repeatability accuracy is ±2 μm; the adaptive PID controller and the calculation center perform data interaction through the real-time Ethernet protocol.

[0099] As Figure 3 shown, the data stream starts from the contact spot distribution density, trajectory offset, and tooth surface waviness output by the three-dimensional scanning module 301, and is input into the geometric feature fusion unit of the calculation center 304; the stress cloud map and DCSDI index provided by the stress analysis module 302 are input into the mechanical property analysis unit; the sub-pixel edge coordinates and PDFC coefficients of the image processing module 303 are input into the dynamic behavior evaluation unit. The calculation center 304 fuses multi-source data through the comprehensive evaluation model, calls the hybrid algorithm library of genetic algorithm and particle swarm optimization to generate the optimal processing parameters, and outputs the grinding pressure, path planning, and speed instructions to the feedback control module 305. After the feedback control module 305 executes the processing, it triggers the re-detection process of the three-dimensional scanning module 301, the stress analysis module 302, and the image processing module 303 to form a closed-loop verification. The data streams of each module progress unidirectionally, and finally return to the initial detection end through the verification of the processing effect.

[0100] In a possible implementation, the three-dimensional scanning module is Keyence LJ-V7300, which transmits data to the calculation center through the PCIe3.0x8 interface (bandwidth 6 GB / s). The stress analysis module is a Kistler 9313B sensor distributed in concentric circles (spacing 1.5 mm), and is synchronized through the IEEE 1588 protocol (time jitter < 1 μs). Image processing module: Xilinx Kintex-7 FPGA implements the Canny algorithm, and the dynamic threshold calculation is:

[0101]

[0102] The parameter k = 1.0, σ n = 2.3 - 3.5 (standard deviation of dark field noise), and the processing delay < 2 ms.

[0103] The computing center runs the PyTorch model for NVIDIA A100 GPU, and the database contains Johnson-Cook parameters and ISO failure thresholds for materials such as PA66 and PEEK. The feedback control module controls the five-axis grinder using the EtherCAT protocol (cycle 1ms), with a grinding head size of #2000 and a surface roughness of Ra≤0.1μm.

[0104] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0105] (1) In the present invention, by integrating three-dimensional morphology scanning, dynamic stress analysis and tooth profile deviation dynamic monitoring technology, a multi-dimensional data collaborative detection system is constructed to significantly improve the comprehensiveness and accuracy of grinding rate detection. High-precision sensor arrays and high-speed image processing technology are used to achieve real-time synchronous acquisition of contact stress distribution and tooth profile deviation, overcoming the limitations of traditional single-point detection. Combined with a machine learning evaluation model, abnormal areas under complex contact patterns can be automatically identified, avoiding subjective errors in manual interpretation, greatly shortening the detection cycle, and ensuring quality consistency in mass production.

[0106] (2) In the present invention, based on the feedback control mechanism of dynamic evaluation indicators, the detection results are intelligently associated with the processing parameters, and a targeted grinding plan is automatically generated. Through multi-axis linkage precision machining equipment and adaptive control algorithms, the grinding pressure, path and speed are dynamically adjusted to effectively eliminate local stress concentration and tooth profile deviation. This closed-loop control strategy can reduce the material waste caused by traditional trial and error processing, improve the contact uniformity and load-bearing capacity of the gear pair, extend the service life, and reduce the dependence on the operator's experience.

[0107] (3) In the present invention, the morphology acquisition, stress analysis, image processing and CNC machining modules are deeply integrated to construct a full-process intelligent detection-machining system. Through a unified data interface and real-time communication protocol, efficient collaboration of each module is achieved to avoid data faults caused by switching of multiple devices. The embedded expert knowledge base and optimization algorithm support the rapid adaptation of multiple models of gears, significantly improve the flexible production capacity of the production line, and provide a complete solution for the digital upgrade of gear manufacturing.

[0108] The above contents are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0109] There are a few points to note:

[0110] (1) The attached drawings of the embodiments of the present invention only relate to the structures involved in the embodiments of the present invention, and other structures can refer to the general design.

[0111] (2) For clarity, in the attached drawings used to describe the embodiments of the present invention, the thickness of layers or regions is enlarged or reduced, that is, these drawings are not drawn to actual scale. It can be understood that when an element such as a layer, film, region, or substrate is referred to as being "on" or "under" another element, the element can be "directly" on or under the other element or there can be intermediate elements.

[0112] (3) Without conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0113] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An intelligent detection method for the lapping rate of an automotive injection-molded gear, characterized in that, Including: S1. Based on the three-dimensional laser scanner to collect the three-dimensional topography data of the gear meshing surface, obtain the geometric feature parameter set of the tooth surface contact area, and the geometric feature parameter set at least includes the contact spot distribution density, the contact trajectory offset, and the tooth surface waviness; S2. Construct the dynamic contact stress distribution index DCSDI, collect the contact stress time-domain signal in real time through the piezoelectric stress sensor array installed on the gear test bench, and calculate the dynamic contact stress distribution index in combination with the finite element contact analysis model; S3. Establish the profile deviation fluctuation coefficient PDFC, use the high-speed camera system to synchronously capture the contact trajectory image sequence during the meshing process of the gear pair, and obtain the time-series change characteristics of the profile deviation through the image feature extraction algorithm; S4. Input the dynamic contact stress distribution index and the profile deviation fluctuation coefficient into the lapping rate comprehensive evaluation model, which conducts multi-dimensional matching analysis on the two characteristic parameters and the standard lapping rate database through the machine learning algorithm, and outputs a three-dimensional evaluation index including the contact area qualification rate, the stress distribution uniformity, and the dynamic matching accuracy; S5. Generate a contact optimization plan according to the three-dimensional evaluation index, adaptively grind the gear meshing surface through the numerical control profiling equipment, and real-time feedback and correct the processing parameters until the preset lapping rate threshold is reached.

2. The intelligent detection method for the lapping rate of an automotive injection-molded gear according to claim 1, wherein The calculation method of the dynamic contact stress distribution index includes: Among them, σ max is the maximum stress value in the contact area, with the unit of MPa, and σ avg is the average stress value in the contact area with the unit of MPa, N c is the stress concentration coefficient, which is determined according to the gear contact strength calculation method in ISO 6336 standard. ΔA is the area of the abnormal stress region with the unit of mm 2 , and A0 is the theoretical contact area with the unit of mm 2 . The specific calculation process includes: performing spatial interpolation processing on the stress data collected by the piezoelectric sensor array, generating a contact stress nephogram, then using a region segmentation algorithm to identify the abnormal stress region, and performing area integral calculation in combination with the mesh division result of the finite element model.

3. An intelligent detection method for the lapping rate of an automotive injection molded gear according to claim 1, characterized in that, The determination method of the profile deviation fluctuation coefficient includes: where T is the complete meshing period in s, and ΔF p (t) is the dynamic change of the pitch deviation in μm, and F p0 is the reference value of the pitch tolerance in μm, and Δf Hα (t) is the dynamic change of the profile inclination deviation in μm, and f Hα0 is the reference value of the profile inclination tolerance in μm. The calculation process includes: performing sub-pixel edge detection on the contact trajectory image obtained by the high-speed camera system, eliminating the influence of motion blur by using the spatio-temporal domain filtering algorithm, and then inversely calculating the profile deviation parameters through the gear meshing kinematic model.

4. The intelligent detection method for the lapping rate of an automotive injection molded gear according to claim 1, wherein, The construction method of the lapping rate comprehensive evaluation model described in S4 includes: Use the improved analytic hierarchy process IAHP to determine the weight distribution of each evaluation index, and introduce the contact trajectory optimization algorithm to calculate the weight correction factor: Among them, w i is the initial weight, α is the system stability coefficient, with a range of 0.1 - 0.3, β is the parameter sensitivity adjustment factor, with a range of 0.5 - 1.2, C i is the confidence level of the characteristic parameter, which is calculated and determined according to the measurement uncertainty corresponding to the gear accuracy grade in ISO 1328-1:2013 standard.

5. The intelligent detection method for the lapping rate of an automotive injection-molded gear according to claim 4, wherein The contact trajectory optimization algorithm adopts the hybrid optimization strategy of genetic algorithm GA and particle swarm optimization PSO: Where the fitness function is defined as: wherein, k1, k2, and k3 are weighting coefficients that satisfy k1 + k2 + k3 = 1, and S contact is the ratio of the actual contact area to the theoretical contact area, and the crossover probability P c and the mutation probability P m are dynamically adjusted according to the population diversity index.

6. The intelligent detection method for the lapping rate of an automotive injection-molded gear according to claim 1, wherein The finite element contact analysis model described in S2 uses the Johnson-Cook dynamic constitutive model to describe the material properties: the expression is: Wherein, A is the yield stress of the material, with the unit of MPa; B is the strain strengthening coefficient; n is the strain hardening index; C is the strain rate sensitivity coefficient; ε is the equivalent plastic strain, is the dimensionless strain rate; T is the relative temperature; and the model parameters are automatically obtained by matching with the material test database.

7. An intelligent detection method for the lapping rate of an automotive injection-molded gear according to claim 1, characterized in that The image feature extraction algorithm described in S3 uses the improved Canny edge detection algorithm, combines the gear meshing phase information to identify the contact trajectory, and the non-maximum suppression threshold is dynamically adjusted according to the image signal-to-noise ratio: T l = 0.4 × T h Among them, σ n is the standard deviation of image noise, I max , I min are the maximum and minimum gray values of the image respectively, k is an empirical coefficient with a range of 0.8 - 1.2, and the algorithm realizes real-time processing on the FPGA chip.

8. An intelligent detection system for the lapping rate of an automotive injection molded gear, characterized in that, Including: Three-dimensional scanning module (301), configured with a laser interferometer and a confocal displacement sensor, for obtaining the microscopic topography data of the gear meshing surface; Stress analysis module (302), including a piezoelectric stress sensor array with a circular distribution and a high-speed data acquisition card, for real-time monitoring of the contact stress distribution; Image processing module (303), integrated with a dual-CCD high-speed camera and an LED pulse light source, for capturing the dynamic image of the meshing contact trajectory; Computing center (304), equipped with a GPU parallel computing unit, for running the finite element contact analysis model and the machine learning evaluation model; Feedback control module (305), including a numerical control grinding machine and an adaptive PID controller, for automatically adjusting the processing parameters according to the evaluation results.

9. The intelligent detection system for the lapping rate of an automotive injection-molded gear according to claim 8, wherein, Including: The piezoelectric stress sensor array of the stress analysis module (302) is distributed in concentric circles, the distance between adjacent sensors is less than 1.5 mm, and it is connected to the calculation center (304) through a fiber optic transmission protocol; the dual CCD high-speed cameras of the image processing module (303) are arranged at a 90° orthogonal angle, and the frame rate is not less than 2000 fps.

10. The intelligent detection system for the lapping rate of an automotive injection-molded gear according to claim 8, characterized in that, including: The numerically controlled grinding machine of the feedback control module (305) includes a five-axis linkage robotic arm and a diamond grinding head, the grinding pressure control accuracy is ±0.1 N, and the position repeat positioning accuracy is ±2 μm; the adaptive PID controller and the calculation center (304) perform data interaction through a real-time Ethernet protocol.

Images