A data fusion method for household appliance production quality safety traceability

Abstract

The present application relates to the technical field of data traceability management, and particularly relates to a data fusion method for household appliance production quality safety traceability. The method comprises the following steps: in the household appliance assembly process, periodically and synchronously collecting hardware component installation state data, assembly equipment operating parameters and assembly material attribute parameters, and fusing them into real-time assembly data groups; calculating the deviation between the real-time assembly data groups and standard assembly data groups, determining the assembly deviation amount, and marking the assembly deviation amount as a quality characteristic parameter; adjusting the assembly equipment operating state according to the quality characteristic parameter, and synchronously reading the assembly equipment feedback signals in the adjustment process. The present application forms a quality monitoring and abnormality traceability system covering the whole household appliance assembly process through data fusion technology and data correlation traceability technology, so as to improve the accuracy, stability and traceability of household appliance production quality safety management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data traceability management, and particularly relates to a data fusion method for quality and safety traceability of household appliance production. BACKGROUND

[0002] The multi-component integration characteristics of household appliances increase the complexity of assembly, and problems such as looseness and misplacement may occur in the assembly process, affecting the performance of the device. However, assembly quality problems are difficult to be completely discovered by automatic detection equipment, and need to be checked and tested manually. In large-scale production, the efficiency and accuracy of manual inspection are limited, and it is difficult to ensure the assembly quality of each device; once the device fails, it is difficult to trace to the specific assembly link. In addition, in the face of the complex assembly scene of multi-component integration of household appliances, the existing technology does not establish a synchronous collection and fusion mechanism for hardware component installation state data, assembly equipment operating parameters and assembly material attribute parameters, and cannot form real-time assembly data groups covering the whole process of multi-component assembly. This leads to the fact that problems such as looseness and misplacement occurring in the assembly process cannot be captured by automatic detection equipment through data correlation analysis, and can only rely on manual inspection and testing, which directly limits the automatic identification ability of assembly quality problems. SUMMARY

[0003] Therefore, it is necessary to provide a data fusion method for quality and safety traceability of household appliance production to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a data fusion method for quality and safety traceability of household appliance production is applied to an assembly link in the production process of household appliances, the household appliances are composed of multiple hardware components, the assembly link involves multiple assembly stations, and the method comprises the following steps:

[0005] Step S1: In the assembly process of household appliances, periodically synchronously collect hardware component installation state data, assembly equipment operating parameters and assembly material attribute parameters, and fuse them into real-time assembly data groups;

[0006] Step S2: Calculate the deviation of the real-time assembly data groups and the standard assembly data groups, determine the assembly deviation amount, and mark the assembly deviation amount as a quality characteristic parameter;

[0007] Step S3: Adjust the operating state of the assembly equipment according to the quality characteristic parameter, and synchronously read the feedback signal of the assembly equipment in the adjustment process; determine the equipment operating deviation amount according to the difference between the feedback signal and the parameters in the real-time assembly data groups;

[0008] Step S4: Compare the operating deviation amount with a preset operating deviation threshold, and record the comparison result in real time; if the comparison result is within the preset assembly range, it is determined that the assembly process is in a normal state, and the assembly condition is maintained until the end of the production cycle.

[0009] Step S5: If the assembly process is outside the preset range, it is determined that the assembly process is in an abnormal state. The assembly process is stopped immediately and the abnormal information is recorded. The abnormal information is associated with the corresponding hardware components and assembly stations to generate a traceability record for the production of home appliances.

[0010] The beneficial effects of this invention are as follows:

[0011] 1) This method breaks through the traditional single data acquisition mode by employing multi-source synchronous data acquisition technology to organically integrate hardware component installation status data, assembly equipment operating parameters, and assembly material attribute parameters to form a real-time assembly data set. This innovative data fusion strategy ensures deep correlation between data from different dimensions during the assembly process, providing a comprehensive and accurate data foundation for subsequent quality analysis. By constructing a deviation calculation model between the real-time assembly data set and the standard assembly data set, the assembly deviation is accurately marked as a quality characteristic parameter, realizing the quantitative expression of assembly quality differences. This provides innovative quantitative indicators for quality control and improves the scientific rigor and accuracy of quality analysis.

[0012] 2) Based on quantified quality characteristic parameters, this method innovatively employs adaptive equipment control technology to dynamically adjust the operating status of assembly equipment. During the adjustment process, real-time feedback signal analysis technology is used to accurately determine the equipment operating deviation by comparing the feedback signal from the assembly equipment with parameters in the real-time assembly data set, thus achieving closed-loop control of the equipment adjustment process. This intelligent equipment control mechanism significantly improves the accuracy and real-time performance of equipment operating status adjustment. Simultaneously, by comparing the operating deviation with a preset operating deviation threshold in real time, an intelligent monitoring system for the assembly process status is constructed. When the comparison result is within the preset assembly range, the system automatically maintains stable assembly operation, ensuring the efficiency and stability of the assembly process and improving the stability and controllability of the production process.

[0013] 3) When the operational deviation exceeds the preset assembly range, this method immediately activates the anomaly response mechanism. It innovatively utilizes data association and traceability technology to precisely link the anomaly information with the corresponding hardware components and assembly stations. By constructing a traceability record system for home appliance production, it achieves full-chain traceability of the abnormal assembly process, providing a clear data path for root cause analysis of quality problems. This innovative anomaly traceability model not only quickly locates the point where quality problems arise but also provides strong data support for subsequent quality improvement and preventative measures, enhancing the accuracy and effectiveness of home appliance production quality and safety management. Attached Figure Description

[0014] Fig. 1 A flowchart illustrating the steps of a data fusion method for tracing the quality and safety of home appliance production.

[0015] Fig. 2 Diagram of equipment and apparatus for the production and assembly of home appliances;

[0016] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0018] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0019] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0020] To achieve the above objectives, please refer to Figs. 1-2 A data fusion method for quality and safety traceability in the production of home appliances is proposed and applied to the assembly stage of the home appliance production process. Home appliances consist of multiple hardware components, and the assembly stage involves multiple assembly stations. The method includes the following steps:

[0021] Step S1: During the assembly of home appliances, periodically and synchronously collect hardware component installation status data, assembly equipment operating parameters, and assembly material attribute parameters, and merge them into a real-time assembly data group.

[0022] Preferably, the specific steps for collecting hardware component installation status data in step S1 are as follows:

[0023] Identify plate-shaped components, column-shaped components, and frame-type components that are assembled hardware components;

[0024] When identified as a plate-shaped component, the data acquisition operation is performed in the following order: first, the alignment data of the mounting holes on its edge is acquired, and then the insertion status data of its surface interface is acquired.

[0025] When a columnar component is identified, the data acquisition operation is performed in the following order: first, its axial installation depth data is acquired, and then its radial circumferential surface fit data is acquired.

[0026] When a component is identified as a frame-type component, the data collection operation is performed in the following order: first, the gap width data of its splicing parts is collected, and then the horizontal height difference data of its four corners is collected.

[0027] For each data collection point of each hardware component, multiple consecutive data collections are performed, and the average value of the multiple data collections is taken as the valid data of that data collection point to determine the installation status data of the hardware component.

[0028] In this embodiment of the invention, when a hardware component enters the assembly station, a visual recognition device is used to identify the component type, distinguishing between plate-shaped components, column-shaped components, and frame-type components. The visual recognition device has a shooting range covering the entire component, a sampling frequency of 1 time / second, and an identification accuracy controlled within ±2mm to ensure accurate component type identification.

[0029] In one operation step of this embodiment of the invention, when the component is identified as a plate-shaped assembly (e.g., a control panel or motherboard of a home appliance), the following operations are performed:

[0030] 1) Acquiring alignment data for edge mounting holes: Deploy a laser alignment sensor directly above the hole, maintaining a vertical distance of 50cm between the sensor and the hole. Each time data is collected, a laser beam is emitted through the hole, and the coordinate deviation of the laser reflection point is recorded. This process is repeated five times, with a 0.5-second interval between each acquisition. For example, if the deviation values ​​for a certain hole are 0.1mm, 0.2mm, 0.1mm, 0.3mm, and 0.2mm respectively, the average value of 0.18mm is taken as the valid alignment data for that hole.

[0031] 2) Acquire surface interface mating status data: Install an industrial camera on the side of the interface to capture side images of the interface after mating. The image resolution is 1280×720 pixels, and three consecutive shots are taken with a 1-second interval between each shot. Analyze the overlap area ratio between the plug and socket in the images, and take the average of the three ratios as the valid mating status data. For example, if the overlap area ratios in the three shots are 98%, 97%, and 99%, the average of 98% is the valid data for that interface.

[0032] In one operation step of this embodiment of the invention, when the component is identified as a columnar assembly (such as a support column or connecting shaft of a home appliance), the following operations are performed:

[0033] 1) Acquiring Axial Installation Depth Data: Install a displacement sensor directly above the component along its axis. Set the initial distance between the sensor probe and the top of the component to 100mm. Record the real-time distance from the sensor to the top of the component during each acquisition, and continuously acquire data 4 times with a 0.5-second interval between each acquisition. For example, if the four distance values ​​are 85.2mm, 85.3mm, 85.1mm, and 85.4mm, the average value of 85.25mm is the valid axial installation depth data (installation depth = initial distance - real-time distance, i.e., 14.75mm).

[0034] 2) Acquiring radial circumferential surface fit data: Three pressure sensors are evenly arranged on the radial outer side of the component. The contact pressure between the sensors and the component surface is maintained at 5N. The pressure feedback value of each sensor is collected four times consecutively, with a 1-second interval between each collection. The average of the four feedback values ​​for each sensor is calculated, and the arithmetic mean of the three average values ​​is taken as the valid fit data. For example, if the average pressure values ​​of the three sensors are 4.8N, 5.0N, and 5.2N, the total average of 5.0N is the valid fit data.

[0035] In one operation step of this embodiment of the invention, when a component is identified as a frame-type component (such as the outer shell frame or internal support frame of a home appliance), the following operations are performed:

[0036] 1) Collect gap width data at the splicing joint: Install laser width gauges on both sides of the splicing joint, maintaining a horizontal distance of 30cm between the gauges and the joint. Select 5 measurement points evenly along the joint, taking 3 consecutive measurements at each point, with a 0.5-second interval between each measurement. Take the average of the 3 measurements at each point as the gap width for that point. Then, sum the average values ​​of the 5 points to obtain the valid data for the splicing joint. For example, if the average widths of the 5 points at a splicing joint are 0.2mm, 0.3mm, 0.2mm, 0.4mm, and 0.3mm respectively, the total average of 0.28mm is the valid gap width data.

[0037] 2) Collect horizontal height difference data at the four corners: Place a height sensor at each corner, ensuring the sensor's measurement reference surface is level with the assembly platform. Collect three consecutive height values ​​for each corner, with a one-second interval between each measurement. Calculate the height difference between each corner and the reference surface, and take the average of the three differences as the valid data for that corner. Then, calculate the difference between the maximum and minimum valid data for each of the four corners; this difference will be the final valid horizontal height difference data. For example, if the average height differences for the four corners are 0.1mm, 0.2mm, 0.1mm, and 0.3mm, the maximum difference of 0.2mm is the valid horizontal height difference data.

[0038] It should be noted that if, in multiple data collections at a certain collection point, the deviation of a single data point from the average value exceeds ±0.3 mm (numerical type) or ±5% (percentage / pressure type), then that data point will be discarded and recollected until the number of consecutive valid data points meets the requirement.

[0039] Preferably, the specific steps for collecting the assembly equipment operating parameters in step S1 are as follows:

[0040] The equipment type is determined based on the main operation type it undertakes in the current assembly process. When the main operation type is to provide driving force to move the parts, the assembly equipment is determined to be a power supply type; when the main operation type is to directly perform part gripping, positioning or fastening operations, the assembly equipment is determined to be a mechanical execution type.

[0041] If the equipment is identified as a power supply device, after the equipment is started, the voltage value is collected at the peak position of each power supply cycle and the current value is collected at the trough position; the collection time point of adjacent cycles is adjusted in a proportional increment manner, and the temperature gradient data of the power supply line is recorded synchronously.

[0042] If the device is classified as a mechanical actuator, when performing linear motion, speed values ​​are collected at the 1 / 4, 1 / 2, and 3 / 4 positions of the stroke; when performing rotational motion, torque values ​​are collected at π / 4 radian intervals according to the rotation angle; after each action cycle, an equal calibration force is applied during measurement, and the angular deviation value of the robotic arm joint is collected.

[0043] In this embodiment of the invention, before the assembly process starts, the work list of the current assembly process is retrieved through the process information system deployed on the assembly line. The list contains the preset work descriptions of the assembly equipment required for the process (such as "driving the conveyor belt to transport components" and "grabbing the motherboard and positioning it for installation").

[0044] Please see Fig. 2 If the main function of the equipment described in the work list is "providing driving force" or "outputting power to move components," and the equipment outputs power through components such as motors and hydraulic pumps to indirectly drive the assembly components (such as conveyor belts and conveyor rollers) to move, without any direct contact with the components to be assembled, then it is classified as power supply equipment 101. If the main function of the equipment described in the work list is "grabbing components," "positioning and installing," or "tightening screws," and the equipment directly contacts the components to be assembled through actuators such as robotic arms, grippers, and screwdrivers to complete operations such as grabbing, aligning the installation position, or fastening, then it is classified as mechanical execution equipment 102. If the equipment has both types of operations (such as a conveyor with a grabbing function), then the "main operation type" marked in the work list shall prevail.

[0045] In one operational step of this embodiment of the invention, after the device is started, it is connected to the power supply line via a voltage sensor and a current sensor. The power supply cycle is set to 20 milliseconds (corresponding to a power frequency of 50Hz). Within each cycle, when the voltage waveform reaches its peak (e.g., the 90° phase point of a sine wave), the voltage sensor collects a voltage value with an accuracy of 0.1V; when the current waveform reaches its trough (e.g., the 270° phase point of a sine wave), the current sensor collects a current value with an accuracy of 0.01A. The collection time points of adjacent cycles are adjusted in a proportionally increasing manner, with the increment ratio set to 1.2.

[0046] It should be noted that the peak acquisition time for the first cycle is 10 milliseconds, the peak acquisition time for the second cycle is 10 × 1.2 = 12 milliseconds, the peak acquisition time for the third cycle is 12 × 1.2 = 14.4 milliseconds, and so on; the current acquisition time points are adjusted proportionally. Temperature sensors are used synchronously to collect temperature gradient data of the power supply line. One temperature sensor is placed at the input end, the middle section, and the output end of the power supply line. The sampling frequency is consistent with the voltage and current acquisition frequency. After each acquisition, the temperature difference between two adjacent points (such as the temperature difference between the input end and the middle section, and the temperature difference between the middle section and the output end) is calculated as the temperature gradient data, with an accuracy of 0.1℃.

[0047] In one operational step of this embodiment of the invention, when the device performs linear motion (such as a pusher cylinder pushing a component to move along a guide rail), the motion stroke is determined by a displacement sensor. Trigger points are set at 1 / 4, 1 / 2, and 3 / 4 of the stroke. When the device moves to a trigger point, a speed sensor collects an instantaneous speed value, with an accuracy of 0.1 m / s. For example, when the total stroke is 200 mm, speeds are collected at 50 mm, 100 mm, and 150 mm. When the device performs rotational motion (such as the rotation of a screw tightening mechanism), the rotation angle is monitored by an angle encoder. Every π / 4 radian (i.e., 45°) of rotation, a torque sensor collects a torque value, with an accuracy of 0.1 N·m. For example, starting from 0°, torque values ​​are collected sequentially at 45°, 90°, 135°, etc.

[0048] It should be noted that after each action cycle is completed (such as when the robotic arm completes one grasp-place cycle), an equivalent calibration force (such as a force of 5N) is applied to the joint of the robotic arm using a calibration device. The actual angle of the joint is collected by the angle sensor, compared with the theoretical standard angle, and the angle deviation value (actual angle - theoretical angle) is calculated. The accuracy is retained to 0.1°.

[0049] Preferably, the specific steps for collecting assembly material attribute parameters in step S1 are as follows:

[0050] The type of assembly material is determined by identifying the arrangement pattern of surface feature points. When the lines connecting adjacent feature points show at least three different angles and there is no obvious row and column alignment feature, it is determined to be a dot matrix distribution material. When the lines connecting adjacent feature points show only two mutually perpendicular directions and form a regular grid, it is determined to be a checkerboard distribution material.

[0051] If the material is determined to be distributed in a dot matrix pattern, the attribute parameters are collected sequentially from the feature points at the edge of the material to the feature points at the center.

[0052] If the material is determined to be distributed in a chessboard pattern, collect the attribute parameters of the feature points in the odd-numbered rows along the horizontal direction of the chessboard, and then collect the attribute parameters of the feature points in the even-numbered columns along the vertical direction of the chessboard.

[0053] During the collection process of each row or column of the checkerboard-shaped material distribution, an interval skipping method is adopted. The interval skipping method is to skip the second feature point after collecting the first feature point, skip the fourth feature point after collecting the third feature point, and so on until the collection of that row or column is completed.

[0054] In this embodiment of the invention, when the assembly material enters the inspection station, an industrial camera is used to photograph the surface of the material to obtain an image of the distribution of surface feature points. The image resolution is set to 2048×1536 pixels, and the shooting distance is maintained at 30cm to ensure that the feature points are fully presented. The image is transmitted to an image analysis device to identify and mark the coordinate positions of all feature points (with the lower left corner of the material as the origin, and the unit is mm).

[0055] For any two adjacent feature points, calculate the angle between the line connecting them and the horizontal direction (range 0°~180°), accurate to 1°, using the coordinate difference. If at least three different angles (e.g., 30°, 60°, 90°) are found, and all feature points show no obvious row or column alignment (the number of feature points in any row or column does not exceed 30% of the total number), then the material is classified as a dot matrix distribution. If the angles of the lines connecting feature points on the surface of a material include 25°, 50°, 75°, and 100°, and no more than three consecutive feature points are arranged in a straight line, then this category applies. If the lines connecting adjacent feature points only show two mutually perpendicular directions, 0° (horizontal) and 90° (vertical), and the feature points form a regular grid (the spacing deviation between feature points in each row and column does not exceed 0.5mm), then the material is classified as a checkerboard distribution. If the feature points are arranged with a horizontal spacing of 10mm and a vertical spacing of 8mm, and each row and column is neatly aligned, then this category applies.

[0056] In one operation step of this embodiment of the invention, firstly, the material edge feature points (feature points ≤ 5mm away from the material edge) are identified and sequentially marked as P1, P2, P3... in a clockwise direction; then, the center feature point (the point with the smallest average distance to all edge feature points) is determined and marked as Pc; the acquisition sequence is P1→P2→...→Pc, and the attribute parameters (such as material hardness and surface roughness) of each feature point are acquired 3 times, with an interval of 1 second between each acquisition.

[0057] In one implementation of this invention, the row numbers are marked from left to right (row 1, row 2, ...), and odd-numbered rows such as row 1, row 3, row 5, etc. are selected; each row adopts an interval skipping sampling method: the first feature point is sampled, the second one is skipped, the third one is sampled, the fourth one is skipped, until the end of the row.

[0058] In one implementation of this invention, column numbers are labeled sequentially from front to back (column 1, column 2, ...), and even-numbered columns such as 2, 4, and 6 are selected. Each column also uses an interval sampling method, with the same rules as row sampling. Attribute parameters for each feature point are sampled twice, with a 0.5-second interval between each sampling.

[0059] It should be noted that the adjacency of feature points is determined by a distance of ≤10mm; feature points outside this range are not considered adjacent. When calculating angles, if the difference between two angles is ≤5°, they are considered the same angle. During the data acquisition process, if the data of a feature point deviates from that of its adjacent points by more than 10%, the data must be acquired again.

[0060] Step S2: Calculate the deviation between the real-time assembly data set and the standard assembly data set, determine the assembly deviation, and mark the assembly deviation as a quality characteristic parameter.

[0061] Preferably, step S2 includes the following steps:

[0062] Step S21: Divide the real-time assembly data group into installation status data subgroup, assembly equipment operating parameter subgroup and assembly material attribute parameter subgroup according to data type, and at the same time, divide the standard assembly data group into three standard subgroups accordingly.

[0063] Step S22: For the installation status data subgroup and the corresponding standard subgroup, calculate the deviations in the order of hardware component assembly. First, calculate the independent installation deviation of each component, and then calculate the cooperative installation deviation of adjacent components.

[0064] Step S23: For the subgroup of operating parameters of the assembly equipment and the corresponding standard subgroup, calculate the deviation in segments according to the time nodes of equipment operation, and calculate the parameter fluctuation deviation in the equipment start-up stage, stable operation stage and pre-shutdown stage respectively;

[0065] Step S24: For the assembly material attribute parameter subgroup and the corresponding standard subgroup, calculate the deviation according to the assembly and use order of the material. First calculate the initial attribute deviation when not assembled, and then calculate the dynamic attribute deviation during the assembly process.

[0066] Step S25: Integrate the above three types of deviation calculation results, and weight them according to the proportions of 40% for installation status data deviation, 30% for assembly equipment operating parameter deviation, and 30% for assembly material attribute parameter deviation to obtain the assembly deviation amount, and mark the assembly deviation amount as a quality characteristic parameter.

[0067] In one operational step of this embodiment of the invention, the installation status data subgroup includes the installation position coordinates of the hardware components (accurate to 0.1 mm), the tightening torque value (accurate to 0.1 N·m), and the interface insertion depth (accurate to 0.1 mm), etc.; for example, the installation status data of a motherboard is "coordinates (100.2, 50.5) mm, screw torque 3.2 N·m, interface insertion depth 8.0 mm". The assembly equipment operating parameter subgroup includes the equipment operating speed (e.g., conveyor belt 2.5 m / min), working pressure (e.g., pneumatic tool 0.6 MPa), operating temperature (e.g., motor 45℃), etc. The assembly material property parameter subgroup includes the material size specifications (e.g., plastic part length 50.0 mm), material hardness (e.g., metal part HRC30), surface roughness (e.g., Ra 1.6 μm), etc. The standard assembly data group is divided into three standard subgroups according to the above classification, where the standard values ​​are set according to the home appliance assembly process documents, for example, the standard installation torque is 3.0 ± 0.2 N·m.

[0068] In another embodiment of the invention, the hardware components are assembled sequentially according to their order (e.g., first the base plate, then the bracket, and finally the panel). The independent installation deviation of each component is calculated: for example, the real-time installation coordinates of the base plate are (200.3, 100.1) mm, and the standard coordinates are (200.0, 100.0) mm, with a deviation of "+0.3 mm in the X direction and +0.1 mm in the Y direction"; the real-time screw torque is 3.4 N·m, and the standard torque is 3.0 N·m, with a deviation of +0.4 N·m. The coordinated installation deviation of adjacent components is calculated: for example, the real-time gap between the bracket and the base plate is 0.3 mm, and the standard gap is 0.2 mm, with a deviation of +0.1 mm; the real-time parallelism measurement of the panel and the bracket is 0.5 mm / m, and the standard value is 0.3 mm / m, with a deviation of +0.2 mm / m.

[0069] In another operating mode of this invention, during the equipment startup phase (0-30 seconds after startup): parameters are collected every 5 seconds, and the deviation between the real-time value and the standard startup curve is calculated. For example, during the 10th second of the conveyor belt startup phase, the real-time speed is 1.2 m / min, the standard speed is 1.0 m / min, and the deviation is +0.2 m / min. During the stable operation phase (30 seconds after startup to 30 seconds before shutdown): parameters are collected every 10 seconds, and the deviation between the real-time value and the standard stable value is calculated. For example, during stable operation of the welding equipment, the real-time current is 150 A, the standard current is 145 A, and the deviation is +5 A. During the pre-shutdown phase (30 seconds before shutdown): parameters are collected every 5 seconds, and the deviation between the real-time value and the standard shutdown curve is calculated. For example, during the 10th second before the robotic arm stops, the real-time speed is 0.3 m / s, the standard speed is 0.2 m / s, and the deviation is +0.1 m / s. The parameter fluctuation deviation is taken as the average of the absolute values ​​of all deviation values ​​within each phase.

[0070] In another embodiment of the invention, the materials are processed according to the assembly and usage sequence (e.g., using the outer shell first, then the internal support, and finally the connector). The initial property deviation before assembly is calculated: for example, the real-time length of the outer shell before assembly is 120.2 mm, the standard length is 120.0 mm, and the deviation is +0.2 mm. The dynamic property deviation during assembly is calculated: for example, the real-time hardness of the support during assembly is HRC29, the standard hardness is HRC30, and the deviation is -1 HRC; the real-time roughness Ra of the connector after assembly is 1.8 μm, the standard roughness Ra is 1.6 μm, and the deviation is +0.2 μm.

[0071] Step S3: Adjust the operating status of the assembly equipment according to the quality characteristic parameters, and read the feedback signal of the assembly equipment synchronously during the adjustment process; determine the equipment operating deviation by the difference between the feedback signal and the parameters in the real-time assembly data set;

[0072] Preferably, in step S3, adjusting the operating status of the assembly equipment based on quality characteristic parameters, and synchronously reading the feedback signal from the assembly equipment during the adjustment process, includes:

[0073] Identify the hardware component types associated with the quality characteristic parameters. When associated with plate-shaped components, adjust the operation status of the assembly equipment to the transmission components related to the assembly of plate-shaped components; when associated with column-shaped components, adjust the orientation to the positioning components related to the assembly of column-shaped components; when associated with frame-type components, adjust the orientation to the splicing components related to the assembly of frame-type components.

[0074] For the transmission components associated with the plate-shaped assembly, adjustments are made in the order of first reducing the conveyor belt speed and then reducing the spacing between the clamping devices. After each adjustment, the current state is maintained until the conveying cycle of one plate-shaped assembly is completed.

[0075] For the positioning components associated with the columnar assembly, adjustments are made in the order of first increasing the rotational speed of the rotating platform and then increasing the pressure of the pressing device, with a pause after each adjustment until the rotating platform completes half a revolution;

[0076] For the splicing parts associated with the frame-type components, the adjustments are performed in the order of first slowing down the splicing speed and then adjusting the splicing alignment accuracy. After each adjustment, the current state is maintained until the splicing operation of one frame edge is completed.

[0077] Read the initial feedback signal of the assembly equipment before starting the adjustment;

[0078] When adjusting the transmission components, a feedback signal is read at the conveyor belt turning node; when adjusting the positioning components, a feedback signal is read at the moment the pressing device contacts the columnar component; and when adjusting the splicing components, a feedback signal is read at the overlapping point of the frame edge.

[0079] In this embodiment of the invention, the associated hardware component type identifier is extracted from the stored entries of quality characteristic parameters (such as "deviation X-timestamp-workstation-plate component"). The system presets three types of identifier rules: when the entry contains the keyword "plate", it corresponds to a plate component (such as a circuit board or panel); when it contains the keyword "column", it corresponds to a column component (such as a support column or shaft-like part); when it contains the keyword "frame", it corresponds to a frame-like component (such as a shell frame or bracket frame). If the identifier is ambiguous, the determination is assisted by calling the installation status data subgroup associated with the parameter (such as the hole alignment data of the plate component or the gap data of the frame-like component).

[0080] In another implementation of this invention, the first step is to reduce the conveyor belt speed. If the original speed is 2 m / min, adjust it to 1.5 m / min (75% of the original speed). This setting is achieved via the speed adjustment knob on the equipment control panel, with an accuracy of 0.1 m / min. The second step is to reduce the clamping device spacing. If the original spacing is 100 mm, adjust it to 80 mm (a 20% reduction). This is achieved by rotating the adjusting screw of the clamping device. Each rotation corresponds to a 2 mm change in spacing, requiring 10 rotations. After each adjustment, maintain the current speed and spacing unchanged until one conveying cycle of the plate-shaped component is completed (the time from the component entering the conveyor belt inlet to leaving the outlet, e.g., 20 seconds).

[0081] In another implementation of this invention, the first step is to increase the rotational speed of the rotating platform. If the original speed was 30 rpm, it is adjusted to 40 rpm (an increase of 33%). This is achieved by adjusting the speed controller, with each level corresponding to 5 rpm, requiring an increase of two levels. The second step is to increase the pressure of the pressing device. If the original pressure was 5 N, it is adjusted to 7 N (an increase of 40%). This is achieved by adjusting the pressure via a pneumatic valve. Each 1 / 4 turn of the valve corresponds to a pressure change of 0.5 N, requiring one full rotation. After each adjustment, the platform is paused and allowed to complete half a revolution (originally, the second half of the revolution took 1 second, after adjustment it takes 0.75 seconds). An angle sensor on the platform triggers a pause end signal.

[0082] In another implementation of this invention, the first step is to slow down the splicing speed. If the original speed is 10 mm / s, adjust it to 8 mm / s (a 20% reduction). This is achieved by setting the speed controller of the propulsion motor; each division on the controller corresponds to 1 mm / s, requiring a reduction of 2 divisions. The second step is to adjust the splicing alignment accuracy. If the original allowable alignment error range is ±0.5 mm, adjust it to ±0.3 mm (a 40% reduction). This is achieved by operating the dial of the fine-tuning mechanism; each division corresponds to 0.05 mm, requiring an inward adjustment of 4 divisions. After each adjustment, maintain the current speed and accuracy parameters until the splicing operation of one frame edge is completed (e.g., a single frame edge is 200 mm long, taking 25 seconds at a speed of 8 mm / s).

[0083] It should be noted that the initial feedback signals of the assembly equipment are read through the equipment signal interface, including the current speed of the conveyor belt (e.g., 2.0 m / min), the rotation speed of the rotating platform (e.g., 30 rpm), and the pressure of the splicing device (e.g., 5 N), and recorded as initial state values. Position sensors are installed at conveyor belt turning points (e.g., at 90° corners). When the plate-shaped component triggers the sensor, the conveyor belt operating current (e.g., 1.2 A) and the clamping device pressure (e.g., 3 N) are read as feedback signals. At the instant the pressing device contacts the columnar component (when the pressure sensor value jumps from 0 to ≥1 N), the real-time rotation speed of the rotating platform (e.g., 40 rpm) and the displacement of the pressing device (e.g., 5 mm) are read as feedback signals. At the point where the frame edges align (when the position sensor detects a distance of ≤0.1 mm between the two side frames), the operating voltage of the propulsion motor (e.g., 220 V) and the alignment error value (e.g., 0.3 mm) are read as feedback signals.

[0084] Preferably, in step S3, determining the equipment operating deviation by the difference between the feedback signal and the parameters in the real-time assembly data set includes:

[0085] The feedback signals are divided into start-up phase signal segments, running phase signal segments, and pause phase signal segments according to the time nodes of the assembly process. At the same time, the parameters in the real-time assembly data group are divided into the three phases accordingly.

[0086] For the startup phase, first calculate the difference between the feedback signal segment and the corresponding installation status data, then calculate the difference between the signal segment and the corresponding equipment operating parameters. Both calculations are performed by comparing the peak points of the signal fluctuations in sequence.

[0087] For the operation phase, according to the assembly sequence of hardware components, the difference between the feedback signal segment and the real-time assembly data parameter at the time of assembly of each component is calculated in turn. First, the difference in material attribute parameters is compared, and then the difference in installation status data is compared.

[0088] For the pause phase, the difference between the feedback signal segment and the last set of real-time assembly data parameters before the pause is calculated. When comparing, the stable value part of the signal is extracted first, and then the static value part of the data parameters is extracted.

[0089] The difference calculation results of the three stages are combined according to the proportion of 20% for the start-up stage, 60% for the operation stage, and 20% for the pause stage to obtain the equipment operation deviation.

[0090] In one operation of this embodiment of the invention, the process begins from the moment the device is powered on and receives a start command, until the first hardware component enters the assembly station. For example, after pressing the start button, the conveyor belt starts running until the first plate-shaped component is transported to the positioning point; this process lasts 15 seconds and is recorded as the start phase. From the moment the first component enters the assembly station, the process ends before the device receives a pause command. For example, the process of the components sequentially completing operations such as positioning, fastening, and testing lasts 180 seconds and is recorded as the running phase. From the moment the device receives a pause command, the process ends until all moving parts come to a complete stop. For example, after pressing the pause button, the conveyor belt gradually decelerates to a stop, lasting 10 seconds and is recorded as the pause phase.

[0091] In another operation of this embodiment of the invention, waveform analysis is performed on the start-up signal segment (such as a current signal) to identify the peak points of signal fluctuations (e.g., the three peak points of the current during startup, occurring at the 3rd, 7th, and 12th seconds, with values ​​of 5A, 4.8A, and 4.5A, respectively). Installation status data corresponding to the start-up data segment (e.g., the three speed peaks of the conveyor belt: 1.2m / s, 1.1m / s, and 1.0m / s) and equipment operating parameters (e.g., the three voltage peaks of the motor: 220V, 215V, and 210V) are extracted from the start-up data segment to ensure that the number of peak points is consistent with the signal segment.

[0092] In another operation of this embodiment of the invention, the difference between the feedback signal segment and the installation status data is compared sequentially according to the peak point order: the difference at the first peak point = 5A - 1.2m / s (Note: the units here need to be uniformly converted to dimensionless values ​​before calculation, for example, all calculated as a percentage of the standard value), the difference at the second peak point = 4.8A - 1.1m / s, and the difference at the third peak point = 4.5A - 1.0m / s. The average of the absolute values ​​of the three differences is taken as the first difference in the startup phase. The difference between the feedback signal segment and the equipment operating parameters is also compared sequentially according to the peak point order: the difference at the first peak point = 5A - 220V (same as above, calculated after conversion to dimensionless), the difference at the second peak point = 4.8A - 215V, and the difference at the third peak point = 4.5A - 210V. The average of the absolute values ​​is taken as the second difference in the startup phase. The total difference in the startup phase = (first difference + second difference) / 2.

[0093] In another operation of this embodiment of the invention, for the "motherboard" assembly: extract the signal value (e.g., vibration frequency 20Hz) during the motherboard assembly period in the running signal segment, corresponding to the material property parameters (motherboard thickness 2mm) and installation status data (installation position deviation 0.1mm) in the real-time assembly data. First, calculate the difference between the signal value and the material property parameter (20Hz - 2mm, converted to dimensionless), then calculate the difference between the signal value and the installation status data (20Hz - 0.1mm, converted to dimensionless), and take the average of the absolute values ​​of the two differences as the difference in motherboard assembly. 1) For the "heat sink" assembly: Similarly, extract the signal value (vibration frequency 18Hz), calculate the difference between the signal value and the heat sink material property (area 10cm²)... 2 The difference between the motherboard assembly and the installation state (95% fit) is averaged to obtain the difference in heatsink assembly. 2) For "casing" assembly: Repeat the above steps to calculate the difference in casing assembly. Total difference during operation = (motherboard assembly difference + heatsink assembly difference + casing assembly difference) / 3.

[0094] It should be noted that the stable value portion is extracted from the pause signal segment, i.e., the signal value (e.g., current 0.5A) after the equipment has come to a complete stop (e.g., 8-10 seconds after the pause command is issued); and the static value portion is extracted from the last set of real-time assembly data before the pause (e.g., the last recorded conveyor belt speed 0 m / s, torque 0 N·m). The pause stage difference = stable signal value - static data value (calculated after conversion to dimensionless, e.g., 0.5A - 0 m / s).

[0095] Step S4: Compare the running deviation with the preset running deviation threshold and record the comparison result in real time; if the comparison result is within the preset assembly range, the assembly process is determined to be in a normal state, and the assembly condition is maintained until the end of the production cycle.

[0096] Preferably, step S4 includes the following steps:

[0097] Step S41: The running deviation is divided into plate-shaped component-related deviation group, column-shaped component-related deviation group and frame-type component-related deviation group according to the hardware component type. At the same time, the preset running deviation threshold is divided into three groups of thresholds accordingly.

[0098] Step S42: For the plate-shaped component associated deviation group, compare the edge mounting hole alignment deviation and surface interface insertion status deviation with the corresponding group threshold in sequence. First compare the single deviation value, and then compare the cumulative value of three consecutive deviations.

[0099] Step S43: For the columnar component associated deviation group, compare it with the corresponding group threshold in the order of its axial installation depth deviation and radial circumferential surface fit deviation. When comparing, first extract the peak part of the deviation, and then extract the stable part of the deviation.

[0100] Step S44: For the associated deviation group of frame-type components, compare it with the corresponding group threshold in the order of the gap width deviation of its splicing part and the horizontal height difference deviation of the four corners.

[0101] Step S45: If the comparison results of all groups are within the preset assembly range, lock the operating parameter settings of the current assembly equipment, fix the material supply conveying rhythm, and keep the hardware component grasping and positioning path unchanged.

[0102] One of the separation rules in this embodiment of the invention is as follows: 1) All deviation values ​​generated during the assembly of plate-shaped components (such as edge mounting hole alignment deviation and surface interface insertion state deviation) account for 40% of the total operational deviation; 2) All deviation values ​​generated during the assembly of column-shaped components (such as axial installation depth deviation and radial circumferential surface fit deviation) account for 30% of the total operational deviation; 3) All deviation values ​​generated during the assembly of frame-type components (such as joint gap width deviation and four corner horizontal height difference deviation) account for 30% of the total operational deviation.

[0103] For example, the preset operating deviation threshold is divided into three groups of thresholds according to the above proportion. Assuming the total threshold is 0.5mm, the threshold for the plate component group is 0.2mm, the threshold for the column component group is 0.15mm, and the threshold for the frame component group is 0.15mm.

[0104] In another operation of this embodiment of the invention, a single deviation value is obtained by reading the real-time alignment deviation of the mounting holes at the edge of the plate-shaped component (e.g., 0.12 mm) and comparing it with the hole position threshold (e.g., 0.1 mm) in the plate-shaped component group threshold. If the deviation exceeds the threshold, it is marked as abnormal. The cumulative deviation value is calculated by summing the absolute values ​​of the alignment deviations collected three times consecutively (e.g., 0.12 mm, 0.13 mm, 0.11 mm) (0.36 mm) and comparing it with the cumulative threshold (e.g., 0.3 mm). If the deviation exceeds the threshold, it is marked as abnormal. The peak value of the deviation is extracted by taking the maximum value (e.g., 0.18 mm) from the axial mounting depth deviation curve and comparing it with the axial peak value threshold (e.g., 0.2 mm) in the columnar component group threshold. The stable value of the deviation is extracted by taking the stable value (e.g., 0.1 mm) within 5 consecutive seconds from the deviation curve and comparing it with the axial stability threshold (e.g., 0.15 mm). The gap width deviation of the splicing parts is compared by reading the real-time gap width deviation of the splicing parts of the frame components (e.g., 0.12mm) and comparing it with the gap threshold (e.g., 0.15mm) in the frame component group threshold; the horizontal height difference deviation of the four corners is compared by reading the horizontal height difference deviation of the four corners (e.g., 0.08mm) and comparing it with the height difference threshold (e.g., 0.1mm) in the frame component group threshold.

[0105] In another operation of this embodiment of the invention, the operating parameters of the assembly equipment are locked. The current operating parameters (e.g., conveyor belt speed 2m / min, clamping force 5N) are saved as fixed values ​​through the equipment control panel, preventing unauthorized modification. The material supply and conveying rhythm is fixed, with a material conveying interval set to 10 seconds per piece. The start and stop times of the conveyor motor are controlled by a timer. The hardware component's gripping and positioning path is maintained. The current gripping coordinates (e.g., X=100mm, Y=50mm, Z=30mm) and movement trajectory are fixed in the robotic arm control system, and subsequent operations directly call upon these path parameters.

[0106] It should be noted that once the parameters are locked, any adjustments require unlocking with administrator privileges; the fixed delivery rhythm and positioning path must be verified by ensuring the current settings are maintained after the device is restarted.

[0107] Step S5: If the assembly process is outside the preset range, it is determined that the assembly process is in an abnormal state. The assembly process is stopped immediately and the abnormal information is recorded. The abnormal information is associated with the corresponding hardware components and assembly stations to generate a traceability record for the production of home appliances.

[0108] Preferably, step S5 includes the following steps:

[0109] Step S51: When the assembly process is determined to be in an abnormal state, a deceleration signal is sent to the assembly equipment; when the equipment running speed drops to one-third of the initial speed, a stop signal is sent, and after the stop signal is sent, the current position of the equipment's executing component remains unchanged;

[0110] Step S52: Extract the real-time assembly data before the anomaly occurred in reverse order of the assembly process. First, extract the installation status data of the hardware components of the last complete assembly, then extract the operating parameters of the corresponding assembly equipment, and finally extract the attribute parameters of the relevant assembly materials.

[0111] In this embodiment of the invention, when the assembly process is determined to be in an abnormal state through the aforementioned steps, the system sends a deceleration signal to the control unit of the assembly equipment. The deceleration signal is a low-level signal lasting for 5 seconds, which is transmitted to the drive module through the signal transmission line inside the equipment.

[0112] In another operation of this embodiment of the invention, if the initial operating speed of the assembly equipment is 3 meters per minute, the drive module, after receiving the signal, gradually reduces the motor power supply frequency to decelerate the equipment. During the deceleration process, a speed sensor installed at the driven wheel of the conveyor belt monitors the operating speed in real time and feeds back the speed value to the control unit every 0.5 seconds. When the speed is detected to drop to 1 meter per minute (i.e., one-third of the initial speed), the control unit immediately sends a stop signal. The stop signal is a high-level signal lasting for 2 seconds, triggering the electromagnetic braking device of the equipment to stop the movement of the conveyor belt, robotic arm, and other actuators. After the stop signal is issued, the braking device remains locked to ensure that the spatial position of the actuators remains unchanged (e.g., the end of the robotic arm stops at a height of 50 mm from the assembly platform, and the deviation of a point on the conveyor belt surface from the positioning reference line does not exceed ±1 mm).

[0113] In another operation of this embodiment of the invention, real-time assembly data before the anomaly occurs is extracted according to the reverse order of the assembly process (e.g., the reverse order of "outer shell fastening → wiring connection → motherboard installation" is "motherboard installation → wiring connection → outer shell fastening").

[0114] 1) Extract the hardware component installation status data from the last complete assembly. For example, if the anomaly occurs during the casing fastening process, retrieve the final status data of the motherboard installation, including the torque values ​​of the four fixing screws of the motherboard (e.g., 3.2 N·m, 3.1 N·m, 3.3 N·m, 3.2 N·m) and the gap value between the motherboard and the chassis (e.g., 0.12 mm). The data comes from the stored records of the torque sensor and laser displacement sensor at the installation station.

[0115] 2) Extract the operating parameters of the corresponding assembly equipment. Taking the motherboard installation process as an example, extract the operating parameters of the motherboard installation robot arm, including the gripping force (e.g., 8N), moving speed (e.g., 0.5 m / s), and rotation angle (e.g., 90 degrees). The data comes from the real-time log of the robot arm control system.

[0116] 3) Extract the attribute parameters of relevant assembly materials. For example, the material type (e.g., FR-4), thickness (e.g., 1.6mm), and batch number (e.g., B20240518) of the motherboard. The data comes from material barcode scanning records and archived information from thickness detection sensors.

[0117] Preferably, step S5 further includes the following steps:

[0118] Step S53: Record the position of the equipment execution components under abnormal conditions. First, record the spatial coordinate association information of the mechanical execution unit, then record the output status association information of the power supply unit, and mark the time difference from the issuance of the stop signal to the complete stop of the equipment.

[0119] Step S54: The extracted real-time assembly data, marked abnormal locations and status information are integrated into an abnormality log file in chronological order, and the abnormality log file is associated with the unique identifier of the corresponding home appliance and stored as abnormal information.

[0120] Step S55: Associate the abnormal information with the corresponding hardware components and assembly stations to generate a traceability record for the production of home appliances.

[0121] In one implementation of this invention, the spatial coordinate association information of the mechanical execution unit is recorded: the spatial coordinates at the time of abnormal stop are read using an angle encoder installed at the joint of the robotic arm and a coordinate sensor at the end effector. For example, the three-dimensional coordinates of the end effector of the robotic arm are (X = 200mm, Y = 150mm, Z = 80mm), and the joint rotation angles are (θ1 = 30 degrees, θ2 = 60 degrees, θ3 = 45 degrees). These coordinates are associated with the equipment's reference coordinate system (with the lower left corner of the equipment as the origin) and then stored in the record. The output status association information of the power supply unit is recorded: the output parameters of power components such as motors and cylinders are read, such as the real-time current (e.g., 2.5A) and voltage (e.g., 220V) of the conveyor belt drive motor, and the working pressure (e.g., 0.6MPa) of the pressing cylinder. The data comes from the equipment's power monitoring module and pressure sensor. The time difference between the issuance of the stop signal and the complete cessation of the equipment is recorded by the timer of the control unit. The time difference is calculated to be 2.5 seconds, accurate to 0.001 seconds.

[0122] In another operation of this embodiment of the invention, the file content sequentially includes: the time of the anomaly (e.g., 2024-06-10 10:25:35), data of each process before the anomaly (installation status, equipment parameters, and material properties arranged in reverse order), coordinates of the mechanical execution unit, output status of the power supply unit, and the difference in stop time. The file format is text, with a timestamp (accurate to the second) preceding each piece of information.

[0123] It should be noted that the unique identifier obtained by scanning the barcode of the home appliance at the assembly line entrance (such as product serial number SN20240610001) is used to create an anomaly log file named "SN20240610001_anomaly log.txt". This file is stored in a designated path on the production data server (such as " / production records / anomaly data / 20240610 / "), and an association index between the identifier and the file path is established in the database. The traceability record clearly indicates that the anomaly occurred at workstation WS05, involving component PCB20240518003, along with the associated anomaly log file path and product serial number. The traceability record is presented in tabular form, containing fields such as "anomaly date, workstation number, component number, product serial number, anomaly type (marked here as assembly process anomaly), and record path". It is stored in the traceability management system's database and can be queried by product serial number or component number.

[0124] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0125] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A data fusion method for quality and safety traceability in the production of home appliances, characterized in that, An assembly process applied in the manufacturing of home appliances, where the appliances consist of multiple hardware components and the assembly process involves multiple assembly stations, the method includes the following steps: Step S1: During the assembly of home appliances, periodically and synchronously collect hardware component installation status data, assembly equipment operating parameters, and assembly material attribute parameters, and merge them into a real-time assembly data group. Step S2: Calculate the deviation between the real-time assembly data set and the standard assembly data set, determine the assembly deviation, and mark the assembly deviation as a quality characteristic parameter. Step S3: Adjust the operating status of the assembly equipment according to the quality characteristic parameters, and read the feedback signal of the assembly equipment synchronously during the adjustment process; determine the equipment operating deviation by the difference between the feedback signal and the parameters in the real-time assembly data set; In step S3, adjusting the operating status of the assembly equipment based on quality characteristic parameters, and synchronously reading the feedback signals from the assembly equipment during the adjustment process, includes: Identify the hardware component types associated with the quality characteristic parameters. When associated with plate-shaped components, adjust the operation status of the assembly equipment to the transmission components related to the assembly of plate-shaped components; when associated with column-shaped components, adjust the orientation to the positioning components related to the assembly of column-shaped components; when associated with frame-type components, adjust the orientation to the splicing components related to the assembly of frame-type components. For the transmission components associated with the plate-shaped assembly, adjustments are made in the order of first reducing the conveyor belt speed and then reducing the spacing between the clamping devices. After each adjustment, the current state is maintained until the conveying cycle of one plate-shaped assembly is completed. For the positioning components associated with the columnar assembly, adjustments are made in the order of first increasing the rotational speed of the rotating platform and then increasing the pressure of the pressing device, with a pause after each adjustment until the rotating platform completes half a revolution; For the splicing parts associated with the frame-type components, the adjustments are performed in the order of first slowing down the splicing speed and then adjusting the splicing alignment accuracy. After each adjustment, the current state is maintained until the splicing operation of one frame edge is completed. Read the initial feedback signal of the assembly equipment before starting the adjustment; When adjusting the transmission components, a feedback signal is read once at the conveyor belt turning node; when adjusting the positioning components, a feedback signal is read once at the moment the pressing device contacts the columnar component; when adjusting the splicing components, a feedback signal is read once at the overlapping point of the frame edge. In step S3, determining the equipment operating deviation by the difference between the feedback signal and the parameters in the real-time assembly data set includes: The feedback signals are divided into start-up phase signal segments, running phase signal segments, and pause phase signal segments according to the time nodes of the assembly process. At the same time, the parameters in the real-time assembly data group are divided into the three phases accordingly. For the startup phase, first calculate the difference between the feedback signal segment and the corresponding installation status data, then calculate the difference between the signal segment and the corresponding equipment operating parameters. Both calculations are performed by comparing the peak points of the signal fluctuations in sequence. For the operation phase, according to the assembly sequence of hardware components, the difference between the feedback signal segment and the real-time assembly data parameter at the time of assembly of each component is calculated in turn. First, the difference in material attribute parameters is compared, and then the difference in installation status data is compared. For the pause phase, the difference between the feedback signal segment and the last set of real-time assembly data parameters before the pause is calculated. When comparing, the stable value part of the signal is extracted first, and then the static value part of the data parameters is extracted. The difference calculation results of the three stages are combined according to the proportion of 20% for the start-up stage, 60% for the operation stage, and 20% for the pause stage to obtain the equipment operation deviation. Step S4: Compare the running deviation with the preset running deviation threshold and record the comparison result in real time; if the comparison result is within the preset assembly range, the assembly process is determined to be in a normal state, and the assembly condition is maintained until the end of the production cycle. Step S5: If the assembly process is outside the preset range, it is determined that the assembly process is in an abnormal state. The assembly process is stopped immediately and the abnormal information is recorded. The abnormal information is associated with the corresponding hardware components and assembly stations to generate a traceability record for the production of home appliances.

2. The data fusion method for traceability of production quality and safety of home appliances according to claim 1, characterized in that, The specific steps for collecting hardware component installation status data in step S1 are as follows: Identify plate-shaped components, column-shaped components, and frame-type components that are assembled hardware components; When identified as a plate-shaped component, the data acquisition operation is performed in the following order: first, the alignment data of the mounting holes on its edge is acquired, and then the insertion status data of its surface interface is acquired. When a columnar component is identified, the data acquisition operation is performed in the following order: first, its axial installation depth data is acquired, and then its radial circumferential surface fit data is acquired. When a component is identified as a frame-type component, the data collection operation is performed in the following order: first, the gap width data of its splicing parts is collected, and then the horizontal height difference data of its four corners is collected. For each data collection point of each hardware component, multiple consecutive data collections are performed, and the average value of the multiple data collections is taken as the valid data of that data collection point to determine the installation status data of the hardware component.

3. The data fusion method for traceability of production quality and safety of home appliances according to claim 1, characterized in that, The specific steps for collecting the assembly equipment operating parameters in step S1 are as follows: The equipment type is determined based on the main operation type it undertakes in the current assembly process. When the main operation type is to provide driving force to move the parts, the assembly equipment is determined to be a power supply type; when the main operation type is to directly perform part gripping, positioning or fastening operations, the assembly equipment is determined to be a mechanical execution type. If the equipment is identified as a power supply device, after the equipment is started, the voltage value is collected at the peak position of each power supply cycle and the current value is collected at the trough position; the collection time point of adjacent cycles is adjusted in a proportional increment manner, and the temperature gradient data of the power supply line is recorded synchronously. If the device is classified as a mechanical actuator, when performing linear motion, speed values ​​are collected at the 1 / 4, 1 / 2, and 3 / 4 positions of the stroke; when performing rotary motion, torque values ​​are collected at π / 4 radian intervals according to the rotation angle. After each action cycle, an equal calibration force is applied during measurement, and the angular deviation value of the robotic arm joint is collected.

4. The data fusion method for traceability of production quality and safety of home appliances according to claim 1, characterized in that, The specific steps for collecting assembly material attribute parameters in step S1 are as follows: The type of assembly material is determined by identifying the arrangement pattern of surface feature points. When the lines connecting adjacent feature points show at least three different angles and there is no obvious row and column alignment feature, it is determined to be a dot matrix distribution material. When the lines connecting adjacent feature points show only two mutually perpendicular directions and form a regular grid, it is determined to be a checkerboard distribution material. If the material is determined to be distributed in a dot matrix pattern, the attribute parameters are collected sequentially from the feature points at the edge of the material to the feature points at the center. If the material is determined to be distributed in a chessboard pattern, collect the attribute parameters of the feature points in the odd-numbered rows along the horizontal direction of the chessboard, and then collect the attribute parameters of the feature points in the even-numbered columns along the vertical direction of the chessboard. During the collection process of each row or column of the checkerboard-shaped material distribution, an interval skipping method is adopted. The interval skipping method is to skip the second feature point after collecting the first feature point, skip the fourth feature point after collecting the third feature point, and so on until the collection of that row or column is completed.

5. The data fusion method for traceability of production quality and safety of home appliances according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: Divide the real-time assembly data group into installation status data subgroup, assembly equipment operating parameter subgroup and assembly material attribute parameter subgroup according to data type, and at the same time, divide the standard assembly data group into three standard subgroups accordingly. Step S22: For the installation status data subgroup and the corresponding standard subgroup, calculate the deviations in the order of hardware component assembly. First, calculate the independent installation deviation of each component, and then calculate the cooperative installation deviation of adjacent components. Step S23: For the subgroup of operating parameters of the assembly equipment and the corresponding standard subgroup, calculate the deviation in segments according to the time nodes of equipment operation, and calculate the parameter fluctuation deviation in the equipment start-up stage, stable operation stage and pre-shutdown stage respectively; Step S24: For the assembly material attribute parameter subgroup and the corresponding standard subgroup, calculate the deviation according to the assembly and use order of the material. First calculate the initial attribute deviation when not assembled, and then calculate the dynamic attribute deviation during the assembly process. Step S25: Integrate the above three types of deviation calculation results, and weight them according to the proportions of 40% for installation status data deviation, 30% for assembly equipment operating parameter deviation, and 30% for assembly material attribute parameter deviation to obtain the assembly deviation amount, and mark the assembly deviation amount as a quality characteristic parameter.

6. The data fusion method for traceability of production quality and safety of home appliances according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: The running deviation is divided into plate-shaped component-related deviation group, column-shaped component-related deviation group and frame-type component-related deviation group according to the hardware component type. At the same time, the preset running deviation threshold is divided into three groups of thresholds accordingly. Step S42: For the plate-shaped component associated deviation group, compare the edge mounting hole alignment deviation and surface interface insertion status deviation with the corresponding group threshold in sequence. First compare the single deviation value, and then compare the cumulative value of three consecutive deviations. Step S43: For the columnar component associated deviation group, compare it with the corresponding group threshold in the order of its axial installation depth deviation and radial circumferential surface fit deviation. When comparing, first extract the peak part of the deviation, and then extract the stable part of the deviation. Step S44: For the associated deviation group of frame-type components, compare it with the corresponding group threshold in the order of the gap width deviation of its splicing part and the horizontal height difference deviation of the four corners. Step S45: If the comparison results of all groups are within the preset assembly range, lock the operating parameter settings of the current assembly equipment, fix the material supply conveying rhythm, and keep the hardware component grasping and positioning path unchanged.

7. The data fusion method for traceability of production quality and safety of home appliances according to claim 6, characterized in that, Step S5 includes the following steps: Step S51: When the assembly process is determined to be in an abnormal state, a deceleration signal is sent to the assembly equipment; when the equipment running speed drops to one-third of the initial speed, a stop signal is sent, and after the stop signal is sent, the current position of the equipment's executing component remains unchanged; Step S52: Extract the real-time assembly data before the anomaly occurred in reverse order of the assembly process. First, extract the installation status data of the hardware components of the last complete assembly, then extract the operating parameters of the corresponding assembly equipment, and finally extract the attribute parameters of the relevant assembly materials.

8. The data fusion method for traceability of production quality and safety of home appliances according to claim 7, characterized in that, Step S5 also includes the following steps: Step S53: Record the position of the equipment execution components under abnormal conditions. First, record the spatial coordinate association information of the mechanical execution unit, then record the output status association information of the power supply unit, and mark the time difference from the issuance of the stop signal to the complete stop of the equipment. Step S54: The extracted real-time assembly data, marked abnormal locations and status information are integrated into an abnormality log file in chronological order, and the abnormality log file is associated with the unique identifier of the corresponding home appliance and stored as abnormal information. Step S55: Associate the abnormal information with the corresponding hardware components and assembly stations to generate a traceability record for the production of home appliances.

Images