A method and system for assembling and controlling automotive seat sensors

Abstract

This invention relates to the field of machine vision-controlled sensor assembly technology, specifically a method and system for assembling automotive seat sensors. The invention includes: visually inspecting and rejecting products with appearance and pin defects at two stages; adjusting the forming pressure based on the coplanarity of incoming materials to measure the actual coplanarity; matching welding power based on the local coplanarity of the pins; calculating residual strain using a thermo-coupling model; and implementing slow cooling if the strain exceeds the standard; determining the potting volume based on the pit morphology and coplanarity; adjusting the curing temperature based on the residual strain and collecting point clouds on the back side; fitting a zero-point output quadratic model to reject products with nonlinear errors exceeding the standard; registering two types of point clouds to calculate the vertical gap; setting the reference glue amount based on zero-point drift and compensating in real time; and controlling the glue dispensing from the screw pump; correcting the pressure based on residual strain and drift; and using impedance control and PID force closed-loop control; completing the mounting if the standard is met, and immediately alarming if abnormalities occur. This invention achieves adaptive, traceable, and precise assembly of automotive seat sensor assemblies.

Description

Technical Field

[0001] This invention relates to the field of machine vision control sensor assembly technology, specifically to a method and system for assembling and controlling automotive seat sensors. Background Technology

[0002] Automotive seat sensors are core sensing components for smart cockpits, airbag triggering, seat posture adjustment, and occupant recognition. Their assembly precision and fit directly determine the overall vehicle safety performance and intelligent interactive experience. However, sensors themselves are mostly flexible films, sheets, or small electronic devices, which are easily damaged by improper force, and the accuracy of the installation position directly affects the sensor performance and the comfort of drivers and passengers.

[0003] Currently, most mainstream automotive seat sensor assembly processes employ a semi-automatic assembly mode assisted by manual labor. While automation continues to advance, existing technologies still have shortcomings: dimensional deviations during the manufacturing stage of automotive seat sensors (such as surface unevenness after potting) cannot be adaptively compensated during the assembly and integration stage, resulting in partial suspension or overpressure of the sensor after mounting, affecting the linearity of pressure detection; the superposition of welding residual strain and potting shrinkage stress during the integration curing thermal cycle causes the sensor zero-point drift to exceed the design range, easily leading to sensor failure due to excessive compression, or improper assembly resulting in later detachment and signal abnormalities.

[0004] Furthermore, traditional calibration often uses linear models, ignoring nonlinear errors, and does not set a threshold for rejecting defective products. This allows defective products with large zero-point drift to enter the integration stage, resulting in unstable finished product performance. Existing adhesive application does not adjust the amount of adhesive in real time according to the sensor drift coefficient and the gap between the foam and the sensor, which easily leads to localized insufficient adhesive or excessively thick adhesive layers. At the same time, the mounting pressure is fixed and lacks compliant force control, which can easily damage the sensor or cause improper mounting. Different car models have different seat patterns and sensor mounting points. Existing assembly control solutions have poor universality and cannot adapt to the needs of multi-category, flexible mass production.

[0005] Therefore, a method and system for assembling and controlling automotive seat sensors is needed. Summary of the Invention

[0006] This invention provides a method and system for assembling and controlling automotive seat sensors.

[0007] The technical solution of this invention is as follows: A method for assembling and controlling automotive seat sensors includes the following steps: S1. Acquire a top-view image of the sensor chip, extract its outline and marking features, complete appearance defect detection and orientation correction, and remove defective workpieces from the production line, completing the first rejection process; perform pin trimming and bending on the remaining chips after rejection, then obtain pin quality data, and remove sensor chips with pin length and end coplanarity outside the set range from the production line, completing the second rejection process; adjust the forming pressure based on the infeed coplanarity and target coplanarity, and obtain the final measured coplanarity through the coplanarity response model; S2. Fix the formed chip on the platform, obtain the pin position and pin height through the vision inspection unit, calculate the pin local coplanarity, weld the pin to the flexible circuit board, and spatially modulate the welding power density according to the pin local coplanarity. At the same time, the welding residual strain is obtained by combining the temperature field in the thermo-mechanical coupling model. If the welding residual strain exceeds the threshold, the cooling rate is adjusted. S3 encapsulates the sensitive area of ​​the sensor with epoxy resin, and the curing temperature is adjusted according to the obtained welding residual strain. After encapsulation and curing, the point cloud on the back of the sensor is obtained through the vision inspection unit. S4. Measure the zero-point output value of the sensor at multiple temperatures and under no load, and fit a quadratic model to obtain the zero-point temperature drift coefficient and nonlinear error. If the nonlinear error is greater than the threshold, it is discarded. S5. Obtain the point cloud of the seat foam installation area, align it with the point cloud on the back of the sensor to the same coordinate system, calculate the compensation height of each point as the vertical gap, determine the cross-sectional area of ​​the reference glue line according to the zero-point temperature drift coefficient of the sensor, and then compensate the amount of glue in real time in combination with the compensation height. S6. Mount the sensor onto the foam coated with adhesive. Correct the mounting pressure based on the welding residual strain and zero-point drift coefficient. Use impedance control and PID force closed-loop to correct the position in real time. If the force and position deviation conditions are met simultaneously, the sensor is determined to be mounted in place and the assembly process ends; otherwise, an alarm is triggered.

[0008] Furthermore, regarding the final measured coplanarity described in S1 The process is as follows: in, Indicates the characteristic pressure of the material; Indicates random error; Indicates the coplanarity of the incoming materials; This indicates the forming pressure.

[0009] Furthermore, regarding the pin local coplanarity described in S2, the first vision detection unit scans all pin tips line by line, records the position coordinates and height of each pin, and simultaneously measures the height of the reference plane to quantify the pin local coplanarity at the position coordinates.

[0010] Furthermore, the calculation process for the welding residual strain mentioned in S2 is as follows: in, This indicates the total time from when the welding heat source begins to act on the weld joint area until the heat source is removed, and the weld joint completes melting and begins to solidify. This indicates the difference in thermal expansion coefficients; It represents the temperature field.

[0011] Furthermore, point clouds are generated on the back of the sensor described in S3. The results are obtained through processing by the second vision detection unit. ,in, This represents the point cloud on the back of the sensor; This indicates the height of the sampling point, which is the height relative to the original substrate surface before potting; Indicates the planar coordinates of the sampling point; This indicates the sampling point number in the point cloud.

[0012] Furthermore, point clouds are generated in the seat foam installation area described in S5. The results are obtained through processing by the second vision detection unit. ,in, Represents the planar coordinates of the bubble point cloud; This represents the height coordinates of the kth sampling point in the point cloud of the car seat foam installation area.

[0013] Furthermore, in step S5, the point cloud on the back of the sensor is aligned to the same coordinate system, and the compensation height of each point is obtained as the vertical gap. The specific process is as follows: Then, by using an iterative nearest-point algorithm, the point cloud on the back of the sensor is transformed into the coordinate system of the car seat foam mounting area to obtain the transformed point cloud. Calculate the compensation height for each corresponding point, i.e., the vertical clearance: To obtain the maximum unevenness .

[0014] Furthermore, in the local adhesive line cross-sectional area described in S5 The process is as follows: in, This represents the unevenness compensation coefficient. The coefficient for converting to adhesive volume adjustment is determined experimentally. Indicates the cross-sectional area of ​​the reference adhesive line; This represents the zero-point temperature drift coefficient.

[0015] Furthermore, regarding the mounting pressure described in S6 The process is as follows: in, Indicates the reference mounting force, when the welding residual strain And zero-point temperature drift coefficient Standard mounting force at that time; Indicates the influence coefficient of welding residual strain; This indicates that the drift coefficient influence coefficient is a positive number; This represents the zero-point temperature drift coefficient.

[0016] An automotive seat sensor assembly control system includes the following: The feeding and forming module acquires a top-view image through the first vision detection unit during feeding, extracts features to complete appearance defect detection and orientation correction, and completes the first rejection process; it acquires pin quality data, compares the pins after cutting and bending to reject unqualified products, and completes the second rejection process; it adjusts the forming pressure according to the coplanarity of the incoming material and the target coplanarity, and obtains the final measured coplanarity through the coplanarity response model; The welding temperature control module obtains the pin position and pin height to get the pin local coplanarity. At the same time, it modulates the welding power density spatially based on the pin local coplanarity. In the thermo-mechanical coupling model, it combines the temperature field to get the welding residual strain. If the welding residual strain exceeds the threshold, the cooling fan speed is reduced; otherwise, the slow cooling protection is not activated. The potting scanning module determines the potting volume for sensor potting protection based on the surface pit volume and pin coplanarity. It adjusts the curing temperature based on the obtained welding residual strain. After potting and curing, the point cloud on the back of the sensor is obtained through the second vision detection unit. The calibration rejection module measures the zero-point output value of the sensor when it is unloaded and fits a quadratic model to obtain the zero-point temperature drift coefficient and nonlinear error. If the nonlinear error is greater than the threshold, it is judged as unqualified and the rejection process is completed three times. The standard adhesive application module acquires the point cloud of the seat foam installation area and aligns it with the point cloud on the back of the sensor to the same coordinate system. The compensation height of each point is obtained as the vertical gap. During adhesive application, the cross-sectional area of ​​the reference adhesive line is determined based on the zero-point temperature drift coefficient of the sensor. The amount of adhesive is then compensated in real time by combining the compensation height. At the same time, the volumetric flow rate of adhesive is controlled by adjusting the speed of the screw pump. The mounting control module mounts the sensor onto the glued foam. It corrects the mounting pressure based on the welding residual strain and zero-point drift coefficient. It uses impedance control and PID force closed-loop to correct the position in real time. If both force and position deviation conditions are met, the sensor is determined to be mounted in place and the assembly process ends; otherwise, an alarm is triggered.

[0017] The beneficial effects of this invention are as follows: 1. This invention transmits individual characteristics from the component manufacturing stage, such as pin coplanarity, welding residual strain, potting profile and shrinkage rate, zero-point drift coefficient, etc., to the seat integration stage in real time, enabling key parameters such as adhesive application amount, mounting force, and curing temperature to achieve single-piece-level adaptive adjustment, thereby completely eliminating process fragmentation and effectively suppressing assembly failures caused by individual differences.

[0018] 2. This invention dynamically adjusts the forming pressure based on the coplanarity of the incoming materials, modulates the welding power according to the local coplanarity of the pins, and adjusts the potting and curing temperature according to the residual welding strain, so that the process parameters of each step are adapted to the initial state of each chip, effectively reducing the performance differences between finished products and ensuring the consistency of mass production.

[0019] 3. This invention utilizes a thermo-coupling model to calculate welding residual strain in real time. When the strain exceeds a threshold, it actively reduces the cooling fan speed to slow down the cooling process. At the same time, it reduces the potting and curing temperature based on the residual strain, thereby reducing thermal stress damage to sensitive areas and preventing excessive zero-point drift, weld fatigue cracking, or potting layer peeling from the sensor during use, thus extending the product's service life.

[0020] 4. This invention precisely aligns the point cloud of the seat foam with the point cloud on the back of the sensor to obtain the vertical gap at each position. It then adaptively adjusts the reference adhesive amount based on the zero-point drift coefficient and compensates for local adhesive amounts in real time according to the gap. This compensates for unevenness between the foam and the sensor, optimizing the adhesive layer thickness distribution, ensuring no bubbles or peeling after installation, and improving bonding strength.

[0021] 5. This invention corrects the desired mounting pressure based on welding residual strain and zero-point drift coefficient, employs impedance control to achieve soft contact, and prevents impact damage; it also superimposes a PID force closed loop to eliminate steady-state errors, and only determines the mounting position when both force deviation and position deviation conditions are met; otherwise, an alarm is triggered. This balances sensor protection and assembly accuracy, reduces scrap rates, and improves the stability of automated assembly. Detailed Implementation

[0022] To better understand the above technical solutions, specific implementation methods will be described in detail below. It should also be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0023] This embodiment provides a method for assembling and controlling automotive seat sensors, including the following steps: S1. Acquire a top-view image of the sensor chip, extract its outline and marking features, complete appearance defect detection and orientation correction, and remove defective workpieces from the production line to complete the first rejection process; cut and bend the pins of the remaining chips after rejection, then obtain the pin quality data, and remove sensor chips whose pin length and end coplanarity are not within the set range from the production line to complete the second rejection process; adjust the forming pressure by the inlet coplanarity and the target coplanarity, and obtain the final measured coplanarity by the coplanarity response model.

[0024] In this embodiment of the invention, during the feeding stage, the conveying mechanism transports disordered sensor chips to the positioning station, collects top view images of the chips, extracts their outlines and marking features, completes appearance defect detection and orientation correction, removes workpieces with appearance defects, and corrects workpieces with incorrect orientation.

[0025] First, the randomly stacked sensor chips are transported to a linear feeding track in a scattered manner by a vibratory feeder or flexible feeder. Then, a positioning mechanism, such as a pneumatic stop or a negative pressure adsorption position, guides the chips one by one to the vision positioning station to achieve steady-state fixation of individual chips.

[0026] Next, the vision inspection unit, which includes a high-resolution industrial camera, a coaxial light source or a ring light source, is activated and distributed at different positions on the production line to acquire a clear top-down image from directly above the chip. The acquired image is then transmitted and processed in real time.

[0027] Furthermore, identify appearance defects and orientation errors of sensor chips, discard sensor chips with appearance defects, correct chips with orientation errors, and perform pressure positioning on sensor chips that are not discarded. Specifically, a top-view image of the sensor chip is acquired, and the chip's outline, surface texture, and marking features are extracted. The extracted features are then compared with a preset standard chip image. If the area of ​​a missing corner, surface crack, or stain is detected to exceed the preset pixel threshold, it is determined to be an appearance defect and removed from the production line, which is a rejection. If the position, number, or arrangement of the positioning markers is found to be inconsistent with the standard template, it is determined to be an incorrect orientation and the orientation is adjusted. When it becomes a qualified chip, the remaining chips after one rejection are completed, and the cutting and bending process continues.

[0028] Furthermore, for qualified chips, the material barcode of the current batch of chips is read, and then the sensor chip pins are cut and bent. Specifically, the sensor chip pins are bent to the target angle, and then the pin quality data, including pin length, bending angle, coplanarity, etc., are obtained through the vision inspection unit. Sensor chips with pin length and coplanarity within the set range are considered as initially qualified chips, and sensor chips that are not within the set range are removed from the production line. This means that sensor chips with pin length and end coplanarity that are not within the set range are removed from the production line, which is the second rejection.

[0029] In this embodiment of the invention, the coplanarity of the incoming materials is defined as follows: , representing the actual measured value, indicating the original unevenness of all pin tips before any bending or shaping of the chip; the target coplanarity is . Let be the pre-designed target value, representing the degree of unevenness that the tips of all pins will exhibit after the chip is pressed and bent by the forming mold. Then, the forming pressure is... The specific process is as follows: in, This indicates the reference pressure, which is the pressure used when the coplanarity of the incoming material is exactly equal to the target value; This represents the proportional coefficient in proportional control, which converts length deviation into pressure compensation. In this process, the pressure is adjusted based on the coplanarity of the incoming material. If the coplanarity of the incoming material pins is greater than the target value (i.e., more uneven), the forming pressure is increased to ensure sufficient deformation; if the incoming material is already of good quality, the pressure is reduced to avoid excessive bending.

[0030] The coplanarity of the pins after forming is then measured. A coplanarity response model must satisfy the following relationship to demonstrate how pressure affects the final measured coplanarity: in, Indicates the characteristic pressure of the material; This represents random errors, including mold gaps, material fluctuations, etc.

[0031] At the same time, a unique file is created for each sensor chip, recording its ID and initial incoming material parameters.

[0032] S2. Fix the formed chip on the platform, obtain the pin position and pin height through the vision inspection unit, calculate the pin local coplanarity, weld the pin to the flexible circuit board, and spatially modulate the welding power density according to the pin local coplanarity. At the same time, the welding residual strain is obtained by combining the temperature field in the thermo-mechanical coupling model. If the welding residual strain exceeds the threshold, the cooling rate is adjusted.

[0033] The formed sensor chip is fixed on a precision platform, and the first vision inspection unit scans all pin tips line by line, recording the position coordinates of each pin. That is, the first The planar coordinates and pin height of each pin tip in the chip's local coordinate system That is, the first The height coordinates of each pin tip, relative to a reference plane, are measured simultaneously, along with the height of the reference plane. The calculation yields: ;in, To represent a plane, This represents the indices of different pins on the chip; that is, it outputs a discrete two-dimensional array. This corresponds to the pin coordinates. Furthermore, for the area between pins, i.e., the location without pins, we can... Defined as 0 or obtained through interpolation to achieve a continuous distribution, but in actual soldering, only the pin positions are considered. That's it; then you can obtain the local coplanarity of different pins. That is, quantification in location The height difference between the tip of the pin and a certain reference plane.

[0034] In this embodiment of the invention, the chip is soldered onto a flexible circuit board, and the pins are connected to the pads using laser reflow soldering or laser soldering, with a soldering power density distribution. Based on the local coplanarity of the pins Spatial modulation is performed, and the specific process is as follows: in, Indicates the reference power density; This represents the power compensation factor, which is typically between 0.2 and 0.5.

[0035] Residual deformation generated during welding was calculated using a thermo-coupling model. The specific process is as follows: in, It represents the total time from when the welding heat source (such as a laser beam) begins to act on the weld joint area until the heat source is removed, the weld joint has completed melting and begins to solidify; This represents the difference in thermal expansion coefficients, specifically the difference between the thermal expansion coefficient of the pin material and the thermal expansion and water absorption coefficient of the pad material. In practical applications... It is considered as a comprehensive equivalent coefficient and is calibrated experimentally: the actual welding residual strain under different welding temperature curves is measured, and the coefficient is derived by reverse calculation. The temperature field is represented by a spatiotemporal function, indicating the temperature at each location on the sensor chip surface during the welding process. At every moment The temperature value is determined by the welding power density distribution. The results were calculated using parameters such as the thermal properties of the material, welding speed, and time.

[0036] If actual measurement It will automatically reduce the cooling fan speed. To slow down cooling, among which, This is the maximum permissible residual weld strain threshold, an upper limit calibrated through experiments or simulations. It refers to the residual weld strain generated during welding. Exceed At that time, it was believed that the solder joint posed a significant risk of thermal stress, which could lead to problems such as excessive zero-point drift, solder joint fatigue cracking, or potting layer peeling during sensor use. If If the temperature is within a controllable range, it means that the thermal stress during welding is within acceptable limits and slow cooling protection is not required. Adjusting the cooling rate should be done towards a lower value.

[0037] S3. The sensitive area of ​​the sensor is protected by epoxy resin potting. The curing temperature is adjusted according to the obtained welding residual strain. After potting and curing, the point cloud on the back of the sensor is obtained through the vision inspection unit.

[0038] Specifically, the sensitive area of ​​the sensor is protected by epoxy resin potting. The curing temperature is adjusted according to the obtained welding residual strain. After potting and curing, the point cloud on the back of the sensor is obtained through a vision inspection unit. The potting shrinkage rate is estimated and recorded based on the adjusted curing temperature and curing time.

[0039] In this embodiment of the invention, the sensitive area of ​​the sensor is protected by epoxy resin potting. The sensitive area refers to the core functional area that is directly responsible for sensing pressure changes and converting them into electrical signals. For example, the transducer, the circular or square thin film area in the center of the sensor, or the bare die of the signal processing chip (ASIC / IC). The extremely weak electrical signals (microvolts / millivolts) generated by the sensing unit need to be amplified, filtered, and digitized by the micro-integrated circuit chip adjacent to the sensing area.

[0040] During the sealing and curing process, the curing temperature is determined based on the welding residual strain. The adjustment process is as follows: in, This indicates the actual cured temperature after adjustment; This indicates the baseline curing temperature, which is usually a fixed value. This indicates the maximum permissible threshold for residual welding strain; This represents the stress suppression coefficient, which controls the extent of temperature reduction; it is typically set to 0-0.3. The larger, the same The lower the temperature, the more significant the reduction. In this embodiment of the invention, based on the measured residual welding strain... The final set temperature after lowering the reference temperature, due to and The value inside the parentheses is less than 1, therefore In other words, the greater the residual welding strain, the lower the potting and curing temperature.

[0041] After potting and curing, a laser profilometer is used to scan the back of the sensor (the surface to be bonded to the seat foam). The scan is then processed by a vision inspection unit to obtain a point cloud of the back of the sensor. The specific process is as follows: in, This represents the point cloud on the back of the sensor; This indicates the height of the sampling point, which is the height relative to the original substrate surface before potting; This represents the planar coordinates of the sampling point, with the origin being the geometric center of the sensor. This indicates the sampling point number in the point cloud.

[0042] In this embodiment of the invention, the sensor sealing shrinkage rate is recorded simultaneously. (Can be used to adjust the integrated curing temperature), adjust according to the actual curing temperature. and curing time The estimation process is as follows: in, The shrinkage coefficient of the material is determined experimentally. Indicates ambient temperature; Indicates the reference temperature.

[0043] S4. Measure the zero-point output value of the sensor at multiple temperatures and under no load, and fit a quadratic model to obtain the zero-point temperature drift coefficient and nonlinear error. If the nonlinear error is greater than the threshold, it is discarded.

[0044] During the electrical calibration process in the sensor manufacturing stage, the sensor is placed in a temperature chamber, and three or more temperature points are set sequentially (e.g., -20℃, 23℃, 85℃). After each temperature point stabilizes, the zero-point output value under no-load conditions is measured. Now, we fit a quadratic model, as shown below: in, It is 23℃; Indicates reference temperature The zero-point output value is a constant term. This represents the zero-point temperature drift coefficient, indicating the temperature deviation from zero by a given value. At 1℃, the zero-point output changes approximately; This represents the second-order nonlinear coefficient of the zero-point temperature drift.

[0045] Specifically, the zero-point temperature drift coefficient is the sensor that will falsely report a pressure signal when the ambient temperature changes, even if no one is sitting in the seat.

[0046] Calculate nonlinear error after calibration The process is as follows: in, This represents the calibrated sensor output value, the measured value after zero-point drift and sensitivity temperature compensation, under ideal conditions. ; This indicates the applied standard load.

[0047] if If the sensor fails to meet the standard, it will be deemed unqualified. Substandard sensors will be rejected three times to prevent them from entering the expensive seat integration stage, thus preventing them from proceeding to the next step.

[0048] S5. Obtain the point cloud of the seat foam installation area, align it with the point cloud on the back of the sensor to the same coordinate system, calculate the compensation height of each point as the vertical gap, determine the cross-sectional area of ​​the reference adhesive line based on the zero-point temperature drift coefficient of the sensor, and then compensate the amount of adhesive in real time in combination with the compensation height.

[0049] After the sensors are calibrated, the seat integration stage begins. The seat foam installation area is then scanned and processed by the second vision detection unit to obtain a point cloud of the seat foam installation area. The specific process is as follows: in, The planar coordinates of the foam point cloud are represented by the geometric center of the foam installation area; This represents the height coordinates of the k-th sampling point in the point cloud of the car seat foam installation area. Then, through ICP registration (i.e., iterative nearest point algorithm), the... Transform the coordinates to the coordinate system of the car seat foam installation area to obtain the transformed point cloud. Calculate the compensation height for each corresponding point, i.e., the vertical clearance: To obtain the maximum unevenness This reflects the most severe misfit between the foam mounting surface of the car seat and the back of the car sensor.

[0050] In this embodiment of the invention, during adhesive application, the adhesive applicator moves along a preset trajectory (e.g., a circular path around the sensor contour), and at each position on the trajectory, there is a corresponding planar coordinate. , For the local unevenness at this location, interpolation is used to obtain the discrete... The continuous values ​​obtained from it.

[0051] Then based on the compensation height The amount of adhesive is adjusted together with the zero-point temperature drift coefficient of the sensor to achieve adaptive control of the adhesive application. The specific process is as follows: The larger the drift of the sensor, the thicker the adhesive layer needs to buffer thermal stress, i.e., the cross-sectional area of ​​the reference adhesive line. Zero-point temperature drift coefficient of the sensor Related, therefore, ;in, Represents a reference constant, when the sensor has no drift ( The standard cross-sectional area of ​​the adhesive line at that time; This represents the sensitivity coefficient.

[0052] Furthermore, during actual adhesive application, the local cross-sectional area of ​​the adhesive line... According to the compensation height Real-time compensation is performed as follows: in, This represents the unevenness compensation coefficient, which accounts for the height difference. The coefficient for adjusting the glue volume is determined experimentally. Simultaneously, the glue dispensing volumetric flow rate... The specific process is as follows: This is achieved by adjusting the screw pump speed, as shown in the following process: , in, Indicates the glue application speed; This indicates the displacement constant of the glue pump; The coefficient of compressibility is a correction factor for the slight change in volume of glue when it is compressed or extruded in a screw pump; it is typically 0.95-1.05. This indicates the rotational speed of the screw pump.

[0053] Ultimately, by adjusting the screw pump speed... This ensures that the actual glue output matches the required local glue line cross-sectional area. This simultaneously compensates for individual sensor differences (such as thicker adhesive layer for large drift) and local bonding gaps (such as applying more adhesive to recessed areas).

[0054] S6. Mount the sensor onto the foam coated with adhesive. Correct the mounting pressure based on the welding residual strain and zero-point drift coefficient. Use impedance control and PID force closed-loop to correct the position in real time. If the force and position deviation conditions are met simultaneously, the sensor is determined to be mounted in place and the assembly process ends; otherwise, an alarm is triggered.

[0055] In this embodiment of the invention, the sensor is mounted onto a foam coated with adhesive, and the mounting pressure is based on the residual welding strain. and drift coefficient Make corrections; Expected mounting power The calculation process is as follows: in, Indicates the reference mounting force, when the welding residual strain And zero-point drift Standard mounting force at that time; This represents the influence coefficient of welding residual strain, and is a positive number. This indicates that the drift coefficient is a positive number; the greater the residual welding strain and the greater the drift, the smaller the allowable mounting force should be to prevent damage.

[0056] During the mounting process, impedance control and PID force closed-loop are used together to correct the position in real time. The specific process is as follows: in, Represents the position correction vector; The inertia matrix is ​​a positive definite diagonal matrix, tuned by the installation engineer according to actual requirements. yes The inverse matrix; Represents the damping matrix; This represents the desired force vector, which contains the desired forces in each direction, where Direction is In other directions, the value is 0; Represents the current velocity vector; Indicates the control period, the sampling interval of the control system; express The force actually applied at any given moment; Indicates the expected mounting force; Indicates proportional gain; Indicates integral gain; To eliminate steady-state error, the interval [0, Tz] is divided into countless tiny time intervals, and in each tiny time interval... Within, take the error value at that moment. ,take Then add them all up; Represents differential gain; This indicates the rate of change of force error.

[0057] At the moment the sensor first comes into contact with the foam, impedance control makes the installation robotic arm act like a soft spring, avoiding impact and protecting the sensor. Once the sensor has made stable contact, the PID force closed loop adds a correction value to precisely adjust the contact force to the desired value, eliminating steady-state error.

[0058] Next, the placement accuracy is checked, as shown in the following process: in, Indicates the allowable force deviation; This represents the actual measured position vector; Represents the target position vector; This indicates the allowable positional deviation.

[0059] When both of the above conditions are met simultaneously, the sensor is determined to have been correctly attached to the seat foam, and the assembly process ends. If the conditions cannot be met within the specified maximum attachment time, the system will issue an alarm, and the workpiece will be marked as an attachment failure, requiring manual intervention.

[0060] This embodiment provides an automotive seat sensor assembly and control system, including the following: The feeding and forming module acquires a top-view image through the first vision detection unit during feeding, extracts features to complete appearance defect detection and orientation correction, and completes the first rejection process; it acquires pin quality data, compares the pins after cutting and bending to reject unqualified products, and completes the second rejection process; it adjusts the forming pressure according to the coplanarity of the incoming material and the target coplanarity, and obtains the final measured coplanarity through the coplanarity response model; The welding temperature control module obtains the pin position and pin height to get the pin local coplanarity. At the same time, it modulates the welding power density spatially based on the pin local coplanarity. In the thermo-mechanical coupling model, it combines the temperature field to get the welding residual strain. If the welding residual strain exceeds the threshold, the cooling fan speed is reduced; otherwise, the slow cooling protection is not activated. The potting scanning module determines the potting volume for sensor potting protection based on the surface pit volume and pin coplanarity. It adjusts the curing temperature based on the obtained welding residual strain. After potting and curing, the point cloud on the back of the sensor is obtained through the second vision detection unit. The calibration rejection module measures the zero-point output value of the sensor when it is unloaded and fits a quadratic model to obtain the zero-point temperature drift coefficient and nonlinear error. If the nonlinear error is greater than the threshold, it is judged as unqualified and the rejection process is completed three times. The standard adhesive application module acquires the point cloud of the seat foam installation area and aligns it with the point cloud on the back of the sensor to the same coordinate system. The compensation height of each point is obtained as the vertical gap. During adhesive application, the cross-sectional area of ​​the reference adhesive line is determined based on the zero-point temperature drift coefficient of the sensor. The amount of adhesive is then compensated in real time by combining the compensation height. At the same time, the volumetric flow rate of adhesive is controlled by adjusting the speed of the screw pump. The mounting control module mounts the sensor onto the glued foam. It corrects the mounting pressure based on the welding residual strain and zero-point drift coefficient. It uses impedance control and PID force closed-loop to correct the position in real time. If both force and position deviation conditions are met, the sensor is determined to be mounted in place and the assembly process ends; otherwise, an alarm is triggered.

Claims

1. A method for assembling and controlling automotive seat sensors, characterized in that, Includes the following steps: S1. Acquire a top-view image of the sensor chip, extract its outline and marking features, complete appearance defect detection and orientation correction, and remove defective workpieces from the production line, completing the first rejection process; trim and bend the pins of the remaining chips after rejection, then obtain pin quality data, and remove sensor chips with pin length and end coplanarity outside the set range from the production line, completing the second rejection process; adjust the forming pressure based on the infeed coplanarity and target coplanarity, and obtain the final measured coplanarity through the coplanarity response model; S2. Fix the formed chip on the platform, obtain the pin position and pin height through the vision inspection unit, calculate the pin local coplanarity, weld the pin to the flexible circuit board, and spatially modulate the welding power density according to the pin local coplanarity. At the same time, the welding residual strain is obtained by combining the temperature field in the thermo-mechanical coupling model. If the welding residual strain exceeds the threshold, the cooling rate is adjusted. S3 encapsulates the sensitive area of ​​the sensor with epoxy resin. The curing temperature is adjusted according to the obtained welding residual strain. After encapsulation and curing, the point cloud on the back of the sensor is obtained through the vision inspection unit. S4. Measure the zero-point output value of the sensor at multiple temperatures and under no load, and fit a quadratic model to obtain the zero-point temperature drift coefficient and nonlinear error. If the nonlinear error is greater than the threshold, it is discarded. S5. Obtain the point cloud of the seat foam installation area, align it with the point cloud on the back of the sensor to the same coordinate system, calculate the compensation height of each point as the vertical gap, determine the cross-sectional area of ​​the reference glue line according to the zero-point temperature drift coefficient of the sensor, and then compensate the amount of glue in real time in combination with the compensation height. S6. Mount the sensor onto the foam coated with adhesive. Correct the mounting pressure based on the welding residual strain and zero-point drift coefficient. Use impedance control and PID force closed-loop to correct the position in real time. If the force and position deviation conditions are met simultaneously, the sensor is determined to be mounted in place and the assembly process ends; otherwise, an alarm is triggered.

2. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, The final measured coplanarity described in S1 The process is as follows: in, Indicates the characteristic pressure of the material; Indicates random error; Indicates the coplanarity of the incoming materials; This indicates the forming pressure.

3. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, In the pin local coplanarity described in S2, the first vision detection unit scans all pin tips line by line, records the position coordinates and pin height of each pin, and measures the height of the reference plane to quantify the pin local coplanarity at the position coordinates.

4. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, The calculation process for the welding residual strain mentioned in S2 is as follows: in, This indicates the total time from when the welding heat source begins to act on the weld joint area until the heat source is removed, and the weld joint completes melting and begins to solidify. This indicates the difference in the coefficients of thermal expansion. It represents the temperature field.

5. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, Point cloud on the back of the sensor described in S3 The results are obtained through processing by the second vision detection unit. ,in, This represents the point cloud on the back of the sensor; This indicates the height of the sampling point, which is the height relative to the original substrate surface before potting; Indicates the planar coordinates of the sampling point; This indicates the sampling point number in the point cloud.

6. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, Point cloud in the seat foam installation area described in S5 The results are obtained through processing by the second vision detection unit. ,in, Represents the planar coordinates of the bubble point cloud; This represents the height coordinates of the kth sampling point in the point cloud of the car seat foam installation area.

7. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, In step S5, the point cloud on the back of the sensor is aligned to the same coordinate system, and the compensation height of each point is obtained as the vertical gap. The specific process is as follows: Then, by using an iterative nearest-point algorithm, the point cloud on the back of the sensor is transformed into the coordinate system of the car seat foam mounting area to obtain the transformed point cloud. Calculate the compensation height for each corresponding point, i.e., the vertical clearance: To obtain the maximum unevenness .

8. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, In the local adhesive line cross-sectional area described in S5 The process is as follows: in, This represents the unevenness compensation coefficient. The coefficient for converting to adhesive volume adjustment is determined experimentally. Indicates the cross-sectional area of ​​the reference adhesive line; This represents the zero-point temperature drift coefficient.

9. The method for assembling and controlling an automotive seat sensor according to claim 1, characterized in that, The mounting pressure described in S6 The process is as follows: in, Indicates the reference mounting force, when the welding residual strain And zero-point temperature drift coefficient Standard mounting force at that time; Indicates the influence coefficient of welding residual strain; This indicates that the drift coefficient influence coefficient is a positive number; This represents the zero-point temperature drift coefficient.

10. A vehicle seat sensor assembly control system, applied to the vehicle seat sensor assembly control method according to claim 1, characterized in that, Includes the following: The feeding and forming module acquires a top-view image through the first vision detection unit during feeding, extracts features to complete appearance defect detection and orientation correction, and completes the first rejection process; it acquires pin quality data, compares the pins after cutting and bending to reject unqualified products, and completes the second rejection process; it adjusts the forming pressure according to the coplanarity of the incoming material and the target coplanarity, and obtains the final measured coplanarity through the coplanarity response model; The welding temperature control module obtains the pin position and pin height to get the pin local coplanarity. At the same time, it modulates the welding power density spatially based on the pin local coplanarity. In the thermo-mechanical coupling model, it combines the temperature field to get the welding residual strain. If the welding residual strain exceeds the threshold, the cooling fan speed is reduced; otherwise, the slow cooling protection is not activated. The potting scanning module determines the potting volume for sensor potting protection based on the surface pit volume and pin coplanarity. It adjusts the curing temperature based on the obtained welding residual strain. After potting and curing, the point cloud on the back of the sensor is obtained through the second vision detection unit. The calibration rejection module measures the zero-point output value of the sensor when it is unloaded and fits a quadratic model to obtain the zero-point temperature drift coefficient and nonlinear error. If the nonlinear error is greater than the threshold, it is judged as unqualified and the rejection process is completed three times. The standard adhesive application module acquires the point cloud of the seat foam installation area and aligns it with the point cloud on the back of the sensor to the same coordinate system. The compensation height of each point is obtained as the vertical gap. During adhesive application, the cross-sectional area of ​​the reference adhesive line is determined based on the zero-point temperature drift coefficient of the sensor. The amount of adhesive is then compensated in real time by combining the compensation height. At the same time, the volumetric flow rate of adhesive is controlled by adjusting the speed of the screw pump. The mounting control module mounts the sensor onto the glued foam. It corrects the mounting pressure based on the welding residual strain and zero-point drift coefficient. It uses impedance control and PID force closed-loop to correct the position in real time. If both force and position deviation conditions are met, the sensor is determined to be mounted in place and the assembly process ends; otherwise, an alarm is triggered.

Images