Intelligent spraying device for motor cast aluminum rotor end ring and core

Abstract

This invention relates to the field of intelligent spraying technology and discloses an intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor. The device includes: a support frame, a turntable rotatably mounted on a conical-bottomed workbench of the support frame, a U-shaped spraying frame, a protective shell, a fan and purification components, and control components. The control components integrate six functional modules: multi-source data acquisition and status perception, spraying object identification and process parameter library, adaptive adjustment of spraying process parameters, waste gas treatment and environmental protection linkage control, online evaluation and closed-loop feedback of spraying quality, and multi-station collaboration and safety interlock control. This invention achieves intelligent management and control of the entire rotor spraying process, enabling automatic rotor model identification, adaptive optimization of spraying parameters, graded energy-saving control of waste gas, online detection and closed-loop compensation of coating quality. This significantly improves spraying quality, paint utilization, and production efficiency, reduces energy consumption, meets the needs of flexible production, and achieves full lifecycle traceability of spraying quality.

Description

Technical Field

[0001] This invention relates to the field of intelligent spraying technology, specifically to an intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor. Background Technology

[0002] In the field of motor manufacturing, the end rings and core surfaces of cast aluminum rotors typically require a protective coating to improve their corrosion resistance, insulation performance, and appearance. Currently, conventional coating operations are mostly completed manually or with semi-automated equipment. Manual coating is not only inefficient and labor-intensive, but the coating quality also heavily relies on the operator's skill level, making it difficult to guarantee consistent coating thickness, coverage, and consistency. Existing semi-automated coating equipment usually consists of a simple rotating worktable and a fixed spray gun. While it improves efficiency to some extent by allowing operators to manually load and unload materials and start / stop the spraying process, it still has many shortcomings in practical applications.

[0003] First, existing spraying equipment lacks the ability to automatically identify rotor model, position and attitude; different models of rotors have large differences in geometric features such as end ring width, core length and number of slots, and the spraying path and process parameters need to be manually adjusted and set, which takes a long time to change and debug and is difficult to adapt to the flexible production needs of multiple varieties and small batches.

[0004] Secondly, the method of adjusting the spraying parameters is relatively crude. The flow rate and pressure of the spray gun are usually set once before production and kept constant during the spraying process. However, in actual production, the viscosity of the paint will change with temperature and batch, and the requirements for paint flow rate and spray pattern also vary in different spraying areas.

[0005] Secondly, the environmental protection treatment capacity is insufficient. The spraying process generates a large amount of waste gas and paint mist containing volatile organic compounds (VOCs), which poses a threat to the environment and the health of operators. Most existing equipment is only equipped with simple exhaust fans and lacks an efficient waste gas collection and purification system.

[0006] In addition, the quality inspection process is lagging behind and relies on manual labor. After the coating is completed, operators need to visually inspect or use simple instruments to spot check the coating quality. The inspection efficiency is low, the rate of missed inspection is high, and the quality data cannot be fed back to the coating process parameters in real time for closed-loop correction, making it difficult to trace and prevent quality problems.

[0007] In summary, existing coating devices for cast aluminum rotor end rings and iron cores of electric motors have significant shortcomings in terms of intelligence, flexibility, environmental protection, and quality control. There is an urgent need to develop an intelligent coating device that can achieve automatic model identification, adaptive adjustment of process parameters, environmentally friendly treatment of waste gas, online quality feedback, and multi-station collaborative control. Summary of the Invention

[0008] To address the technical problems in the prior art, this application provides an intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor.

[0009] The intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor provided in this application specifically includes:

[0010] A support frame is provided, and a conical bottom worktable is installed on the top of the support frame. The surface of the conical bottom worktable is provided with evenly distributed through holes, and a drain valve is provided at the bottom of the conical bottom worktable.

[0011] A turntable is rotatably connected to the surface of the conical-bottom worktable, and the surface of the turntable is rotatably connected with evenly distributed placement pins for placing the rotor to be sprayed.

[0012] The U-shaped spray frame is installed on the rear side of the surface of the conical bottom workbench, and multiple spray heads are installed between the U-shaped spray frames;

[0013] A protective housing is installed above the support frame. A control valve is installed on the rear side of the protective housing. The outlet of the control valve is connected to the spray head through multiple spray connection pipes. An exhaust fan is connected through the top of the protective housing.

[0014] A fan is installed at the bottom of the support frame. The fan is connected to the exhaust head through an exhaust pipe. The air outlet of the fan is connected to a purification component through an exhaust pipe.

[0015] A control element, which is mounted on the front side of the protective housing.

[0016] Preferably, the control components include a multi-source data acquisition and status sensing module, a spraying object identification and process parameter library module, a spraying process parameter adaptive adjustment module, a waste gas treatment and environmental protection linkage control module, a spraying quality online assessment and closed-loop feedback module, and a multi-station collaboration and safety interlock control module.

[0017] The multi-source data acquisition and state perception module is used to acquire multi-dimensional sensor data throughout the entire spraying process, perform timestamp synchronization, filtering and noise reduction, outlier removal and dimensional standardization, generate a standardized spraying state vector and send it synchronously to the corresponding module.

[0018] The spraying object identification and process parameter library module is used to intelligently identify the rotor model based on the acquired rotor image, match the corresponding spraying process parameter template, and perform structured storage, full life cycle management and intelligent recommendation of process parameters.

[0019] The adaptive adjustment module for spraying process parameters is used to dynamically adjust the spraying flow rate, spraying pressure, spraying trajectory, and multi-axis linkage parameters based on the spraying state vector and the reference process parameter template, so as to achieve adaptive optimization and precise control of the spraying process parameters.

[0020] The exhaust gas treatment and environmental protection linkage control module is used to dynamically adjust the fan operating parameters based on the VOCs concentration and equipment operating status during the spraying process, and to perform full life cycle management of the purification components and emergency graded response for exhaust gas treatment.

[0021] The online coating quality assessment and closed-loop feedback module is used for visual inspection and comprehensive evaluation of the coating quality after spraying, and performs differentiated closed-loop compensation and process parameter self-optimization based on the quality evaluation results.

[0022] The multi-station collaboration and safety interlock control module is used to coordinate the timing of various actuators and the multi-station collaborative scheduling, while also building an equipment health status assessment model and a preventive maintenance mechanism.

[0023] In summary, this application includes at least one of the following beneficial technical effects:

[0024] 1. This invention acquires rotor images through industrial vision sensors in the multi-source data acquisition and state perception module, and uses a recognition model built based on deep learning convolutional neural networks to automatically identify rotor models, contour boundaries, end face angles, and placement deviations. The spraying object recognition and process parameter library module automatically matches the corresponding spraying process parameter templates based on the recognition results, supporting structured storage and quick retrieval of no less than 50 commonly used rotor models. For new rotor models, only 10 to 20 sample images need to be collected to complete model iteration through transfer learning, effectively solving the problems of lack of automatic recognition capability, long changeover time, and difficulty in adapting to flexible production in the prior art.

[0025] 2. This invention utilizes an adaptive adjustment module for spraying process parameters. Based on the real-time feedback of the spraying distance from a laser displacement sensor, a fuzzy PID control algorithm is employed to correct the flow rate of each spraying head online, achieving a control accuracy within ±2%. Simultaneously, a mathematical model is established for spraying pressure, flow rate, and paint viscosity. The recursive least squares method is used to identify the real-time viscosity of the paint online. When the viscosity deviates from the preset range, the rotation speed of the feeding device is automatically adjusted to correct the spraying pressure. This solves the problem in existing technologies where spraying parameters are constant and cannot be adaptively adjusted according to changes in distance, paint viscosity fluctuations, and characteristics of different regions, thereby improving coating quality and paint utilization.

[0026] 3. This invention uses a waste gas treatment and environmental protection linkage control module, employing a segmented PID control algorithm to achieve dynamic hierarchical closed-loop control of the fan speed. The fan speed is automatically adjusted according to the VOCs concentration range (standby, adjustment, emergency). The module also has a three-level emergency response mechanism, which executes warning, suspension of spraying, and emergency stop operations in stages when the VOCs concentration exceeds the limit or the fan malfunctions. This overcomes the shortcomings of existing equipment that only has a simple exhaust fan and lacks dynamic adjustment and purification management.

[0027] 4. This invention utilizes an online coating quality assessment and closed-loop feedback module, employing an improved YOLOv8 deep learning model to detect coating defects on the rotor after coating. The minimum identifiable defect size is ≤0.3mm, and the coating thickness is indirectly measured through coating grayscale and thickness calibration curves. The module extracts four quality characteristics: coverage, uniformity, edge clarity, and defect area, constructs a comprehensive evaluation model, and generates a quality score of 0-100. Based on the score, the quality is divided into three levels: excellent, qualified, and requiring recoating. This solves the problems of lagging quality detection, reliance on manual labor, lack of closed-loop feedback, and difficulty in traceability in existing technologies. Attached Figure Description

[0028] Figure 1 This is a three-dimensional structural diagram of the present invention;

[0029] Figure 2 This is a rear-view stereoscopic structural diagram of the present invention;

[0030] Figure 3 This is a schematic diagram of the structure of the fan and purification components of the present invention;

[0031] Figure 4 This is a schematic diagram of the surface structure of the cone-bottom worktable of the present invention;

[0032] Figure 5 This is a schematic diagram of the control component architecture of the present invention.

[0033] Explanation of reference numerals in the attached drawings: 1. Support frame; 101. Conical bottom workbench; 102. Protective shell; 2. Turntable; 201. Placement pin; 3. U-shaped spray frame; 301. Spray head; 302. Control valve; 303. Spray connection pipe; 4. Fan; 401. Exhaust fan head; 402. Exhaust duct; 403. Exhaust duct; 404. Purification component; 5. Control component. Detailed Implementation

[0034] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. In addition, the forms of the various structures described in the following embodiments are merely illustrative. The intelligent spraying device for the end ring and core of the cast aluminum rotor of the motor involved in the present invention is not limited to the structures described in the following embodiments. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Please see Figures 1 to 4 As shown, this embodiment provides an intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor, including a support frame 1, a turntable 2, a U-shaped spraying frame 3, a protective shell 102, a fan 4, and a control component 5;

[0036] A conical bottom workbench 101 is installed above the support frame 1. The surface of the conical bottom workbench 101 has evenly distributed through holes for allowing waste liquid or droplets generated during the spraying process to enter the interior of the conical bottom workbench 101 through the through holes. A drain valve is provided at the bottom of the conical bottom workbench 101 for periodically draining the collected waste liquid. It should be noted that the specific form of the drain valve, such as a manual ball valve or a solenoid valve, can be selected according to actual needs. It is a conventional setting in this field and will not be described in detail here.

[0037] A turntable 2 is rotatably connected to the surface of the conical bottom worktable 101, and evenly distributed placement pins 201 are rotatably connected to the surface of the turntable 2 for placing the rotor to be sprayed. In this embodiment, the turntable 2 can be driven to rotate by a motor to realize automatic rotor repositioning or rotary spraying. It should be noted that the motor and its transmission method, such as belt drive, gear drive or direct drive, are conventional technical means in the field, and their specific models and installation positions can be adjusted according to the actual working conditions. This application does not make specific limitations on this.

[0038] A U-shaped spray frame 3 is installed on the rear side of the surface of the conical bottom worktable 101, and multiple spray heads 301 are installed between the U-shaped spray frames 3. The arrangement of the spray heads 301 can be adjusted according to the size of the rotor and the spraying requirements, for example, they can be spaced apart along the vertical or horizontal direction of the U-shaped spray frame 3. The type of spray head 301, such as an air spray gun, a high-pressure airless spray gun or an electrostatic spray gun, is a conventional choice in the art, and this application does not limit it.

[0039] A protective housing 102 is installed on the top of the support frame 1. A control valve 302 is installed on the rear side of the protective housing 102. The inlet of the control valve 302 is used to connect to an external spray material supply device. The outlet of the control valve 302 is connected to the spray head 301 through multiple spray connection pipes 303 to control the on / off state and flow rate of the spray material. It should be noted that the specific form of the external spray material supply device, such as a pressure tank, diaphragm pump or metering pump and its connection method with the control valve, are conventional technical means in this field, and this application does not make specific limitations on them.

[0040] A blower head 401 is connected through the top of the protective housing 102 to collect exhaust gas, mist droplets and dust generated during the spraying process; a fan 4 is installed at the bottom of the support frame 1, and the fan 4 is connected to the blower head 401 through a blower duct 402. The air outlet of the fan 4 is connected to a purification component 404 through an air outlet duct 403; the purification component 404 is used to filter, adsorb or purify the exhaust gas to meet environmental emission requirements; it should be noted that the specific structure of the purification component 404, such as an activated carbon filter layer, a HEPA filter or a water washing tower, can be selected according to the actual application scenario, which is a well-known technology in the field and will not be elaborated in detail here.

[0041] The control component 5 is installed on the front side of the protective housing 102. The software system of the control component 5 adopts a modular and layered architecture, such as... Figure 5 As shown, it includes: a multi-source data acquisition and status perception module, a spraying object identification and process parameter library module, a spraying process parameter adaptive adjustment module, a waste gas treatment and environmental protection linkage control module, a spraying quality online assessment and closed-loop feedback module, and a multi-station collaboration and safety interlock control module. Each module interacts with other modules through standardized data interfaces, forming a closed-loop control chain of "perception—analysis—decision—execution—feedback".

[0042] The multi-source data acquisition and status perception module is responsible for acquiring multi-dimensional sensor data throughout the entire spraying process and preprocessing it to provide a unified and reliable status input for subsequent modules. Specifically, this includes:

[0043] An industrial vision sensor installed inside the protective housing 102 is used to collect images of the turntable 2 and the rotor to be painted on the placement pin 201. The sensor obtains the rotor model, the outline boundary of the end ring and the iron core, the end face angle and the placement position deviation through the image recognition algorithm. The visual sampling frequency is not less than 30 frames / second and the recognition accuracy is ≤0.5mm.

[0044] The laser displacement sensor installed on one side of the U-shaped spray frame 3 is used to measure the straight distance between the spray head 301 and the rotor end face in real time. The measurement range is 50mm to 300mm, the accuracy is ±0.2mm, and the data is fed back to the spray height control circuit.

[0045] The miniature flow sensor installed between the control valve 302 and each spraying connection pipe 303 is used to monitor the instantaneous paint flow rate of each spraying head 301, with a flow measurement accuracy of ±1.5%FS.

[0046] The pressure sensor installed at the inlet of control valve 302 is used to monitor the spraying pressure of the feed pipeline in real time. The pressure measurement range is 0 to 2.5 MPa, and the accuracy is ±0.5%FS.

[0047] The temperature and humidity sensor installed on the inner wall of the protective housing 102 is used to monitor the temperature and relative humidity of the spraying environment, with a temperature accuracy of ±0.5℃ and a humidity accuracy of ±3%RH.

[0048] The VOCs gas concentration sensor installed at the inlet of exhaust head 401 is used to detect the concentration of volatile organic compounds in the spraying exhaust gas in real time. The detection range is 0 to 1000 ppm and the response time is ≤5s.

[0049] The differential pressure switch and operating status sensor installed inside the fan 4 and the purification component 404 are used to monitor the fan speed, the degree of clogging of the purification filter, and the operating status of the equipment.

[0050] The aforementioned sensor sampling frequency is 10Hz to 100Hz. After the data is time-stamped, Kalman filtered for noise reduction, outlier removal, and dimensional standardization by control component 5, a standardized spraying state vector is generated. And send it to the subsequent modules.

[0051] The spraying object identification and process parameter library module automatically matches the corresponding spraying process parameter template based on the rotor model identified by the vision sensor, providing a benchmark for subsequent parameter adjustment and realizing full lifecycle management of process parameters. Specific implementation details are as follows:

[0052] B101: Intelligent Rotor Model Recognition

[0053] The module is based on a deep learning convolutional neural network (CNN) to build a rotor model recognition model. The model's pre-training dataset covers no less than 50 commonly used cast aluminum rotor models. The input is rotor images collected by a visual sensor, and the output is rotor model, specifications, and positioning features. The recognition accuracy is ≥99.5%, and the recognition response time is ≤200ms. For new rotor models, it supports transfer learning adaptation. Only 10 to 20 sample images from different angles are needed to complete the rapid iteration of the model without retraining the entire model, which greatly shortens the debugging cycle for new models.

[0054] B102: Structured Design of Process Parameter Library

[0055] The process parameter library uses an embedded SQLite database for structured storage. Each process parameter template contains four data segments, enabling the classification, management, and retrieval of parameters.

[0056] Basic Information Area: Stores rotor model, specifications, compatible motor power, entry time, version number, and operator information;

[0057] Process control area: Division of the storage end ring spraying area and spraying sequence, reference flow rate of each spray head. Optimal spraying distance With respect to allowable deviation range and reference spraying pressure The rotational speed matching relationship between the turntable and the placement pin, the single-piece spraying cycle time, the drying waiting time, the spraying trajectory parameters, and the pulse spraying timing parameters;

[0058] Quality threshold zone: Pass thresholds for coating coverage, uniformity, and edge sharpness; criteria for judging defect types; weighting coefficients for quality scores.

[0059] Safety Constraints: Storage of upper and lower safety limits for spraying pressure; Safety parameters such as concentration threshold, minimum fan speed, and motor overload protection threshold.

[0060] B103: Process Parameter Library Management Function

[0061] Version management: Each parameter modification automatically generates a new version number, records the modified content, the person who made the modification, and the modification time, and supports querying, rolling back, and comparing historical versions to avoid process anomalies caused by accidental parameter modifications;

[0062] Access Control: A three-tier access control system is set up. Operators can only access approved parameter templates and cannot modify them; process engineers can add and modify parameter templates, but must submit them for approval; administrators have the highest level of authority and can perform parameter approval, permission allocation, and version management to ensure the compliance of process parameters.

[0063] Intelligent recommendation: When the visual system recognizes a new model rotor that has not been entered into the parameter library, the module can automatically generate recommended process parameters by matching the three most similar mature process templates from the parameter library based on the rotor's outer diameter, end ring width, core length, number of slots, and other geometric features. Operators only need to make minor adjustments to complete the entry of parameters for the new model.

[0064] The adaptive adjustment module for spraying process parameters dynamically adjusts key spraying parameters based on real-time data provided by the state perception module and in conjunction with benchmark templates in the process parameter library. This adapts to different rotor specifications, coating characteristics, and environmental changes, ensuring consistent spraying quality. Specific implementation includes:

[0065] C101: Dynamic adjustment of spray flow rate

[0066] Based on the rotor model and spraying area identified by the vision sensor, control component 5 retrieves the corresponding reference flow rate from the process database. And based on the actual spraying distance fed back by the laser displacement sensor With optimal spraying distance To correct the deviation, a fuzzy PID control algorithm is used to correct the flow rate of each spray head 301 online.

[0067]

[0068] in The corrected spraying target flow rate is the final output of the fuzzy PID control algorithm. It serves as the set target for adjusting the opening of the spraying control valve and is the core control value for dynamic flow regulation. Indicates the actual spraying distance;

[0069] This represents the proportional gain, the proportional term adjustment parameter of a fuzzy PID controller, used for rapid response to distance deviation. The larger the deviation, the stronger the adjustment of the proportional term. This represents the integral gain, the integral term adjustment parameter of the fuzzy PID controller, used for accumulating historical deviations. The derivative gain represents the differential term adjustment parameter of the fuzzy PID controller, used to predict the changing trend of distance deviation; the corrected target flow rate. The flow control accuracy is achieved by adjusting the opening of the control valve 302, with real-time closed-loop feedback from each flow sensor, ensuring that the flow control accuracy is within ±2%.

[0070] C102: Online Identification and Pressure-Coordinated Control of Coating Viscosity

[0071] The module is based on the fluid mechanics orifice outflow formula to establish the spraying pressure P, paint flow rate Q, and paint viscosity. Mathematical model between them:

[0072]

[0073] in For flow coefficient, The flow area of ​​the valve orifice. The system identifies the real-time viscosity of the coating by using recursive least squares method based on real-time collected pressure and flow data. When the viscosity deviates from the preset range by ±10%, the system automatically adjusts the screw pump speed of the feeding device, corrects the spraying pressure, and maintains the pressure and flow characteristics within the preset range. Simultaneously, it adjusts the pulse spraying sequence of the spray head to ensure the paint atomization fineness is within the preset range. The optimal range is found to avoid poor atomization caused by batch differences in paint or temperature changes.

[0074] C103: Spraying trajectory and multi-axis linkage control

[0075] Trajectory interpolation planning: Based on visually recognized rotor position, angle, and contour data, the module uses NURBS non-uniform rational B-spline curves for smooth interpolation of the spraying trajectory. For the end ring face area, an equidistant annular spraying trajectory is generated, with the trajectory step distance automatically calculated according to the spraying width to ensure that the overlap rate of adjacent spraying paths is controlled within 30% to 50%. For the outer circle area of ​​the end ring, the system synchronously controls the intermittent indexing rotation of turntable 2, the uniform rotation of placement pin 201, and the axial uniform feed of spraying head 301. These three are synchronously controlled by electronic gears to generate a spiral upward spraying trajectory. The pitch can be adaptively adjusted according to the rotor outer diameter and spraying width, and the synchronous error is controlled. ;

[0076] Precision pulse spraying: For the iron core slot area, the system controls the high-frequency solenoid valve of the spray head 301 through a high-speed digital output module to achieve pulse spraying, with pulse width... Adjustable, the spraying timing is precisely synchronized with the rotation angle of the placement pin 201. Spraying is triggered only when the slot is aligned with the spray head, while the spraying pressure is slightly increased to ensure that the paint only covers the end face of the slot, preventing it from flowing into the slot and causing blockage. The synchronization accuracy of the slot spraying angle is high. ;

[0077] Real-time distance closed-loop control: During the spraying process, the laser displacement sensor provides real-time feedback on the spraying distance. The module adjusts the axial position of the spray head 301 in real-time via a servo driver, achieving real-time closed-loop control of the spraying distance and ensuring high control accuracy. When the distance deviation exceeds If this happens, immediately stop spraying to avoid gun collision or spraying failure.

[0078] The exhaust gas treatment and environmental protection linkage control module dynamically adjusts the operating parameters of the fan 4 and the purification component 404 according to the changes in the state of the spraying process, reducing energy consumption while ensuring effective collection and purification of exhaust gas. Specific implementation includes:

[0079] D101: Variable Frequency Drive (VFD) Hierarchical Closed-Loop Control for Fans

[0080] A segmented PID control algorithm is used to dynamically adjust the fan speed, which is divided into three control intervals:

[0081] Standby range: when VOCs concentration At that time, the wind turbine maintains the minimum base speed. Operation, maintaining only the cabin The slight negative pressure prevents exhaust gas from escaping and reduces standby power consumption;

[0082] Adjustment range: when At that time, linear PID control is used, according to the formula The target rotation speed is dynamically calculated to achieve closed-loop control of VOCs concentration, ensuring that the VOCs concentration in the chamber remains stable within a safe range.

[0083] Emergency Zone: When At that time, the fan immediately increased to its maximum safe speed. The system operates, simultaneously triggering audible and visual alarms, adjusting the spraying cycle, and extending the exhaust gas extraction time to ensure rapid discharge of exhaust gas from the chamber. The default setting is , The default setting is The parameters can be adjusted according to the characteristics of the coating; the fan frequency converter adopts vector control mode, and the speed adjustment range is [not specified]. Speed ​​regulation accuracy ≤ ±0.5Hz, response time ≤ 2s, achieving smooth and stepless speed adjustment.

[0084] D102: Full Lifecycle Management of Purification Components

[0085] Real-time acquisition of differential pressure signals at the inlet and outlet of the 404 purification component Based on pressure difference changes and cumulative operating time, a life decay model is used to predict the remaining service life of activated carbon adsorption filter cartridges and filter cotton. The remaining service life is displayed in real time on the touch screen as a percentage.

[0086] When the remaining lifespan In time, issue a replacement warning in advance; when If a blockage alarm is triggered, the system will automatically limit the spraying capacity until the filter element is replaced.

[0087] D103: Emergency Response Level for Waste Gas Treatment

[0088] Establish a three-level emergency response mechanism:

[0089] Level 1 alert: VOCs concentration exceeds But lower If the filter element pressure difference approaches the threshold, the system will issue an audible and visual warning, and the touch screen will display the warning information, without affecting the normal operation of the equipment.

[0090] Level 2 alarm: VOCs concentration exceeds If the spraying continues for 3 seconds, or if the filter element pressure difference exceeds the limit, or if the fan speed is abnormal, the system will issue an audible and visual alarm, suspend the spraying of new workpieces, stop the material supply after the current workpiece is finished spraying, maintain the fan at high speed, and send alarm information to the operator at the same time.

[0091] Level 3 Emergency Stop: If the VOCs concentration exceeds 200ppm for 5 seconds or the fan stops, the system will immediately trigger the emergency stop logic, cut off the material supply of control valve 302, stop the drive of the motion mechanism, keep the fan running, send an emergency alarm message, and retain all fault data.

[0092] The online coating quality assessment and closed-loop feedback module dynamically adjusts the operating parameters of the fan 4 and the purification component 404 based on the changes in the coating process, reducing energy consumption while ensuring effective collection and purification of exhaust gas. Specific implementation includes:

[0093] E101: Visual Inspection of Coating Quality

[0094] An improved YOLOv8 deep learning model was used to construct a coating defect detection model. The pre-trained dataset covered common coating defects such as pinholes, runs, orange peel, missed areas, over-spraying, and uneven edges. The defect recognition accuracy was [percentage missing]. Minimum identifiable defect size For coating thickness detection, the module uses a pre-established coating grayscale and thickness calibration curve to achieve indirect visual measurement of coating thickness, minimizing measurement error. Online assessment of coating thickness uniformity can be achieved without additional thickness measuring equipment. After spraying, the system controls turntable 2 to rotate the rotor to the inspection station, and the placement pin 201 drives the rotor to rotate at a uniform speed. The vision sensor synchronously acquires full-circle high-definition images, with the acquisition frame rate synchronized with the rotor speed, ensuring full-surface image acquisition without blind spots or overlap, and a single full inspection time. .

[0095] E102: Quality Feature Extraction and Evaluation Model

[0096] The system extracts coating coverage from the image. Coating thickness uniformity Edge sharpness Defect area Four core quality characteristics are used to construct a comprehensive evaluation model for spray coating quality:

[0097]

[0098] in , , , These are the weighting coefficients, set to [default value]. , , , It can be adjusted according to the process requirements of different rotors;

[0099] Supports automatic optimization of weight coefficients using the Analytic Hierarchy Process (AHP) based on historical qualified sample data, with a comprehensive scoring range of [missing information]. point.

[0100] E103: Quality Grading and Differentiation Compensation

[0101] Based on the comprehensive evaluation model for spray coating quality, the spray coating quality is divided into three levels, and a differentiated closed-loop compensation strategy is implemented:

[0102] excellent( ≥90 points): is judged as qualified, proceeds to the next workstation, and the system records all process and quality data and stores it in the traceability database;

[0103] qualified( (Points): If the work is qualified but has minor defects, record the defect type and location, and automatically fine-tune the spraying flow rate, pressure, and trajectory parameters of the corresponding area to avoid the same defects in subsequent workpieces;

[0104] To be resprayed ( <75 points): Automatically identify workpieces that need respraying, record their workstation information, defect type, defect location and range, and automatically plan the respraying path and parameters in the next spraying cycle. Only perform fixed-point respraying on the defect area, and control the coating thickness deviation after respraying within ±10%.

[0105] E104: Process parameter self-optimization

[0106] When three consecutive workpieces have the same type of defect, or five consecutive workpieces have a quality score below 90, the process parameter self-optimization program is automatically triggered. The program takes the overall spraying quality score as the objective function and spraying flow rate, pressure, distance, rotation speed, and trajectory overlap rate as optimization variables. It uses an improved particle swarm optimization (PSO) algorithm to optimize the parameters, with no more than 50 iterations, and generates optimized process parameters.

[0107] E105: Quality Lifecycle Traceability

[0108] Information such as coating quality data, process parameters, operators, production time, and rotor model for each workpiece is associated and stored to generate a unique workpiece traceability QR code. The QR code can be printed via touch screen or distributed to a marking machine for workpiece marking. It supports querying the entire production process data of the workpiece by scanning the code, realizing full life cycle traceability of coating quality.

[0109] The multi-station collaboration and safety interlock control module is responsible for coordinating the timing of various actuators in the spraying device to ensure that the equipment operates efficiently while possessing reliable safety protection capabilities. Specific implementation includes:

[0110] F101: Multi-mode operation and interlock control

[0111] Three device operating modes are configured, with hardware and software dual interlocking logic between each mode, allowing switching only after authorization verification.

[0112] Automatic mode: Normal production mode. It automatically completes the entire process of rotor positioning detection, path planning, spraying execution, quality inspection, and station switching according to the preset process flow and cycle time, without the need for manual intervention.

[0113] Manual mode: Debugging and maintenance mode. Operators can individually control the jog actions of each actuator through the touch screen, such as turntable rotation, spray head advance and retreat, valve opening and closing, fan start and stop, etc. All actions are equipped with speed limits and soft limit protection to avoid equipment collisions caused by misoperation.

[0114] Debugging Mode: Process debugging mode supports single-step execution of the spraying process. Individual processes such as spraying and testing can be executed separately, which is convenient for process debugging and parameter optimization of new rotor models. In debugging mode, some non-critical safety interlocks are shielded, while core safety protections such as emergency stop and protective door interlocks are retained.

[0115] F102: Multi-station timing collaborative precision control

[0116] A time-slice-based multi-task scheduling mechanism is adopted to divide the single-station spraying cycle into 6 core steps, with connection logic set between each step:

[0117] Step 1: Rotor positioning detection. The vision sensor confirms that the rotor is in place on each placement pin 201 and identifies the rotor model and positional deviation.

[0118] Step 2: Spraying path planning, generating spraying trajectory and process parameters based on the recognition results;

[0119] Step 3: Turntable positioning, drive turntable 2 to rotate the rotor to be sprayed to the spraying station;

[0120] Step 4: Spraying execution. Start the fan 4 to the target speed, open the control valve 302 and the spray head 301, and synchronously drive the placement pin 201 to rotate, and complete the spraying according to the planned trajectory.

[0121] Step 5: Quality inspection. After the spraying is completed, rotate the coating to the inspection station, collect images, and complete the quality scoring.

[0122] Step 6: Station switching. Turntable 2 rotates to move the rotor that has finished spraying out of the spraying area, while moving the next rotor to be sprayed in.

[0123] Once the painting process at the previous station is completed, the rotary table rotates simultaneously with the rotor positioning detection and path planning for the next station. The spray head moves to the starting position before painting begins, and the fan is raised to the target speed in advance, achieving zero waiting time for station switching. The single-station painting cycle time can be adjusted according to the rotor model. Dynamic adjustment within the range, switching time between each step .

[0124] F103: Attitude Coordination and Interference Protection

[0125] During the spraying process, the control component 5 dynamically adjusts the lifting height of the turntable 2 or the tilt angle of the placement pin 201 based on the distance data fed back by the laser displacement sensor in real time, so as to ensure that the spraying head 301 is always perpendicular to the end ring surface or the iron core slot, and maintains the best spraying angle and distance.

[0126] Real-time verification of the movement paths of the turntable, spray head, and placement pins to prevent collisions between moving parts. When a risk of motion interference or a spray distance deviation is detected, [the system will take action]. When this happens, the system automatically pauses spraying and issues an adjustment prompt.

[0127] F104: Graded Emergency Stop and Fault Reset

[0128] Two-level emergency stop logic is configured to adapt to different fault scenarios:

[0129] Global emergency stop: Triggered by the emergency stop button on the equipment body or the emergency stop button on the touch screen, the power supply to all actuators is immediately cut off after triggering, and only the power supply to the fan operation and control system is retained to perform the full-process emergency protection operation;

[0130] Emergency stop of process: triggered by safety conditions, only the spraying material supply and the drive of the motion mechanism are cut off, while the power supply to the fan operation and control system is maintained. After the fault is cleared, the production process can continue from the interrupted position without the need for a full reset.

[0131] The fault reset mechanism adopts a hierarchical reset mechanism: Level 1 early warning faults can be automatically reset; Level 2 alarm faults require manual confirmation by the operator before reset; Level 3 emergency stop faults require the operator to troubleshoot the fault and verification by the administrator before reset.

[0132] F105: Equipment Health Management and Preventive Maintenance

[0133] Real-time data collection of various moving parts, such as running time, motor load, number of start-stop cycles, and fault records, is used to establish an equipment health status assessment model. For key components such as turntable bearings, servo motors, spray valves, and fans, preventive maintenance plans are automatically generated based on cumulative running time and load status, and maintenance reminders are issued in advance, such as bearing lubrication, valve seal replacement, and motor maintenance, to reduce the probability of sudden equipment failures.

[0134] System Complete Operation Example: This embodiment takes the batch spraying production of a certain model of 4-pole motor cast aluminum rotor (rotor outer diameter 120mm, end ring width 15mm, core length 100mm, number of slots 36) as an example to explain in detail the full-process control implementation of control component 5:

[0135] Production preparation stage: Operators select the corresponding rotor model's process parameter template via touchscreen. The system retrieves the baseline process parameters for that model from the process parameter library: optimal spraying distance 150mm, baseline spraying pressure 1.2MPa, single-channel spray head baseline flow rate 150mL / min, turntable indexing speed 5r / min, placement pin rotation speed 30r / min, single-station spraying cycle time 30s, VOCs concentration low threshold 20ppm, high threshold 100ppm; simultaneously, the system completes self-testing of all sensors, resetting of actuators, and testing of safety functions. After the self-test shows no abnormalities, it enters standby mode.

[0136] Material loading and positioning inspection stage: The rotor to be painted is placed on the placement pin 201 of the turntable 2. The turntable rotates and sends the rotor to the vision inspection station. The multi-source data acquisition and state perception module controls the industrial vision sensor to acquire rotor images. The CNN model is used to identify the rotor model, placement position deviation, and end face angle. If the identification result is consistent with the preset model and the position deviation is ≤0.3mm, the system determines that the rotor is in place. At the same time, the module synchronously acquires real-time data from all sensors, and after preprocessing, generates a standardized state vector, which is sent to each functional module.

[0137] Adaptive adjustment stage of process parameters: The adaptive adjustment module of spraying process parameters dynamically adjusts the process parameters based on real-time data in the state vector: Based on the deviation between the actual spraying distance of 152mm fed back by the laser displacement sensor and the optimal distance of 150mm, the target flow rate of 154mL / min is calculated by the fuzzy PID algorithm, and the control signal is output to the control valve 302 to adjust the valve opening; Based on the real-time collected ambient temperature of 28℃ and relative humidity of 60%, the spraying pressure is compensated and adjusted to 1.25MPa; At the same time, based on the rotor position and angle recognized by vision, the annular spraying trajectory of the end ring end face, the spiral spraying trajectory of the end ring outer circle, and the pulse spraying trajectory of the iron core slot are generated to complete the parameter preparation before spraying.

[0138] During the spraying execution phase: The multi-station collaborative and safety interlock control module controls the turntable to rotate the rotor to be sprayed to the spraying station according to the preset rhythm; the exhaust gas treatment and environmental protection linkage control module controls the fan to start and increase to the basic speed of 800r / min based on the real-time VOCs concentration of 15ppm, maintaining a slight negative pressure in the chamber; subsequently, the spraying process parameter adaptive adjustment module synchronously controls the rotation of the placement pin, the start and stop of the spraying head and its feed, and the opening of the control valve to complete the spraying according to the planned trajectory: the end ring end face spraying adopts a circular equidistant trajectory with a path overlap rate of 40%; the outer circle of the end ring adopts a spiral upward trajectory with a pitch of 8mm; the iron core slot adopts pulse spraying with a pulse width of 20ms, synchronized with the rotor rotation angle; during the spraying process, the flow rate, pressure, and distance are controlled in real time in a closed loop. At the same time, when the VOCs concentration rises to 45ppm, the system automatically adjusts the fan speed to 1200r / min to ensure effective collection of exhaust gas in the chamber.

[0139] Online quality inspection and closed-loop feedback stage: After the coating is completed, the turntable rotates the rotor to the quality inspection station. The placement pin drives the rotor to rotate at a constant speed. The vision sensor collects high-definition images of the entire rotor surface. The online coating quality assessment and closed-loop feedback module extracts coating quality characteristics, detects surface defects, and calculates a comprehensive quality score of 94 points, which is judged as "excellent". The workpiece is qualified and enters the next station. The system records all process data and quality data of the workpiece and stores them in the traceability database.

[0140] Workstation switching and continuous production stage: After quality inspection is completed, the multi-workstation collaboration and safety interlock control module controls the turntable to rotate in an indexing manner, moving the rotor that has been sprayed out of the spraying area, while moving the next rotor to be sprayed into the spraying station. The above process is repeated to achieve multi-workstation continuous batch production.

[0141] When three consecutive workpieces have uneven coating thickness and the score is below 75, the system automatically triggers the process parameter self-optimization program to optimize the spraying flow rate and trajectory overlap rate parameters, and updates the process parameter library to achieve adaptive optimization of the process.

[0142] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor, characterized in that, include: A support frame (1) is provided, and a cone-bottom workbench (101) is installed on the top of the support frame (1). The surface of the cone-bottom workbench (101) is provided with uniformly distributed through holes, and a drain valve is provided at the bottom of the cone-bottom workbench (101). Turntable (2), the turntable (2) is rotatably connected to the surface of the cone-bottom worktable (101), and the surface of the turntable (2) is rotatably connected with evenly distributed placement pins (201) for placing the rotor to be sprayed; U-shaped spray frame (3), the U-shaped spray frame (3) is installed on the rear side of the surface of the cone bottom workbench (101), and multiple spray heads (301) are installed between the U-shaped spray frames (3). A protective housing (102) is installed above the support frame (1). A control valve (302) is installed on the rear side of the protective housing (102). The outlet of the control valve (302) is connected to the spray head (301) through multiple spray connection pipes (303). An exhaust head (401) is connected through the top of the protective housing (102). Fan (4), the fan (4) is installed at the bottom of the support frame (1), the fan (4) is connected to the exhaust head (401) through the exhaust pipe (402), and the air outlet of the fan (4) is connected to the purification component (404) through the air outlet pipe (403). Control component (5), which is mounted on the front side of the protective housing (102).

2. The intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor according to claim 1, characterized in that, The control component (5) includes a multi-source data acquisition and status perception module, a spraying object identification and process parameter library module, a spraying process parameter adaptive adjustment module, a waste gas treatment and environmental protection linkage control module, a spraying quality online assessment and closed-loop feedback module, and a multi-station collaboration and safety interlock control module. The multi-source data acquisition and state perception module is used to acquire multi-dimensional sensor data throughout the entire spraying process, perform timestamp synchronization, filtering and noise reduction, outlier removal and dimensional standardization, generate a standardized spraying state vector and send it synchronously to the corresponding module. The spraying object identification and process parameter library module is used to intelligently identify the rotor model based on the acquired rotor image, match the corresponding spraying process parameter template, and perform structured storage, full life cycle management and intelligent recommendation of process parameters. The adaptive adjustment module for spraying process parameters is used to dynamically adjust the spraying flow rate, spraying pressure, spraying trajectory, and multi-axis linkage parameters based on the spraying state vector and the reference process parameter template, so as to achieve adaptive optimization and precise control of the spraying process parameters. The exhaust gas treatment and environmental protection linkage control module is used to dynamically adjust the fan operating parameters based on the VOCs concentration and equipment operating status during the spraying process, and to perform full life cycle management of the purification components and emergency graded response for exhaust gas treatment. The online coating quality assessment and closed-loop feedback module is used for visual inspection and comprehensive evaluation of the coating quality after spraying, and performs differentiated closed-loop compensation and process parameter self-optimization based on the quality evaluation results. The multi-station collaboration and safety interlock control module is used to coordinate the timing of various actuators and the multi-station collaborative scheduling, while also building an equipment health status assessment model and a preventive maintenance mechanism.

3. The intelligent spraying device for the end ring and the core of the cast aluminum rotor of an electric machine according to claim 2, characterized in that, The multi-source data acquisition and state awareness module includes: Images of the rotor to be coated are acquired using industrial vision sensors to obtain rotor model, contour boundary, end face angle and placement deviation. The visual sampling frequency is no less than 30 frames / second and the recognition accuracy is ≤0.5mm. The linear distance between the spray head and the rotor end face is measured in real time using a laser displacement sensor, with a measurement range of 50mm to 300mm and an accuracy of ±0.2mm; the instantaneous paint flow rate of each spray head is monitored using a miniature flow sensor, with a flow measurement accuracy of ±1.5%FS. The spraying pressure of the material supply pipeline is monitored in real time by a pressure sensor. The pressure measurement range is 0 to 2.5 MPa, with an accuracy of ±0.5%FS. The spraying environment temperature and relative humidity are monitored by temperature and humidity sensors, with a temperature accuracy of ±0.5℃ and a humidity accuracy of ±3%RH. The concentration of volatile organic compounds in the spraying exhaust gas is detected in real time using a VOCs gas concentration sensor with a detection range of 0 to 1000 ppm and a response time of ≤5s. The fan speed, filter clogging level, and equipment operating status are monitored by differential pressure switches and operating status sensors. Configure synchronization timestamps for all acquisition channels to ensure time consistency of data from different sensors, and set the sampling frequency of each sensor to 10Hz to 100Hz; The collected raw data is processed by Kalman filtering for noise reduction, outlier removal and dimension normalization. Core features such as rotor model, profile parameters, spraying distance, flow rate of each channel, spraying pressure, temperature and humidity, VOCs concentration and fan operating status are extracted to generate a standardized spraying status vector.

4. The intelligent spraying device for the end ring and the core of the cast aluminum rotor of an electric machine according to claim 2, characterized in that, The spraying object identification and process parameter library module includes: A rotor model recognition model is built based on deep learning convolutional neural networks. The model's pre-training dataset covers no less than 50 commonly used cast aluminum rotor models for motors, and the recognition response time is ≤200ms. For new rotor models, transfer learning adaptation is supported, and 10 to 20 sample images from different angles are collected for rapid model iteration. The process parameter library is structured using an embedded SQLite database. Each process parameter template contains four data segments: basic information area, process control area, quality threshold area, and safety constraint area, which respectively store rotor basic specifications, spraying process control parameters, quality qualification judgment standards, and equipment safety constraint parameters.

5. The intelligent spraying device for the end ring and the core of the cast aluminum rotor of an electric machine according to claim 2, characterized in that, The adaptive adjustment module for spraying process parameters includes: The reference flow rate is retrieved based on the identified rotor model and the area to be sprayed. Based on the deviation between the actual spraying distance and the optimal spraying distance, the flow rate of each spraying head is corrected online using a fuzzy PID control algorithm. Real-time closed-loop feedback is provided through the flow sensor to ensure that the flow control accuracy is within ±2%. A mathematical model for spraying pressure, paint flow rate, and paint viscosity is established. By collecting pressure and flow rate data in real time, the real-time viscosity of the paint is identified online using the recursive least squares method. When the viscosity deviates from the preset range of ±10%, the speed of the feeding device is automatically adjusted to correct the spraying pressure. B-spline curves are used for smooth interpolation of the spraying trajectory. Equidistant ring spraying trajectories are generated for the end face of the end ring, and the overlap rate of adjacent paths is controlled between 30% and 50%. For the outer circle area of ​​the end ring, the turntable is rotated intermittently by electronic gear synchronous control, the placement pin rotates at a constant speed, and the spraying head is fed axially at a constant speed to generate a spiral upward spraying trajectory. For the slotted area of ​​the iron core, a high-frequency solenoid valve is controlled to achieve pulse spraying synchronized with the rotor's rotation angle, with the pulse width adjustable from 10 to 50 ms; during the spraying process, the spraying distance is fed back in real time by a laser displacement sensor.

6. The intelligent spraying device for the end ring and the core of the cast aluminum rotor of an electric machine according to claim 2, characterized in that, The waste gas treatment and environmental protection linkage control module includes: A segmented PID control algorithm is used to achieve dynamic hierarchical closed-loop control of the fan speed, divided into three control zones: standby zone, adjustment zone, and emergency zone. In the standby zone, when the VOCs concentration is ≤ low threshold, the fan maintains the minimum base speed. In the adjustment zone, when the VOCs concentration is between the low and high thresholds, linear PID control is used to dynamically calculate the target speed. In the emergency zone, when the VOCs concentration is ≥ high threshold, the fan immediately increases to the maximum safe speed, and an audible and visual alarm is triggered. The fan inverter uses vector control mode, and the speed adjustment range is [not specified]. Speed ​​regulation accuracy ≤ ±0.5Hz, response time ≤ 2s; The pressure difference signal at the inlet and outlet of the purification component is collected in real time. Based on the pressure difference change and cumulative operating time, the remaining service life of the filter element is estimated using a life decay model. When the remaining service life is ≤10%, a replacement warning is issued. When the pressure difference exceeds the maximum threshold, a blockage alarm is triggered and the spraying capacity is limited. A three-level emergency response mechanism is set up. The first level warning is when the VOCs concentration is below the threshold or the filter element pressure difference is close to the threshold, and the system issues an audible and visual warning. The level 2 alarm is triggered when the VOC concentration exceeds the threshold for 3 seconds, the filter element pressure difference exceeds the limit, or the fan speed is abnormal. The system will suspend the spraying of new workpieces and maintain the fan at high speed. A Level 3 emergency stop occurs when the VOC concentration exceeds 200 ppm for 5 seconds or the fan stops. The system immediately cuts off the material supply, stops the drive of the moving mechanism, keeps the fan running, and sends an emergency alarm.

7. The intelligent spraying device for the end ring and core of a cast aluminum rotor of an electric motor according to claim 2, characterized in that, The online coating quality assessment and closed-loop feedback module includes: An improved YOLOv8 deep learning model was used to build a coating defect detection model. The pre-trained dataset covered common coating defects such as pinholes, runs, orange peel, missed spraying, overspraying, and uneven edges. Four quality characteristics—coating coverage, coating thickness uniformity, edge clarity, and defect area—are extracted from the collected images to construct a comprehensive evaluation model for spraying quality. A comprehensive spraying quality score of 0 to 100 is generated through weighted calculation. Based on the comprehensive score, the spraying quality is divided into three levels: excellent, qualified, and needing to be resprayed. A differentiated closed-loop compensation strategy is implemented: a score of ≥90 is judged as excellent, and the data is stored in the traceability database; a score of 75≤score<90 is judged as qualified, and the spraying parameters of the corresponding area are automatically fine-tuned; a score<75 is judged as needing to be resprayed, and the respraying path and parameters are automatically planned to complete the fixed-point respraying of the defective area. When three consecutive workpieces have the same type of defect, or five consecutive workpieces have a quality score below 90, the process parameter self-optimization program is automatically triggered. The program uses the overall spraying quality score as the objective function and spraying flow rate, pressure, distance, rotation speed, and trajectory overlap rate as optimization variables. An improved particle swarm optimization algorithm is used to optimize the parameters and generate optimized process parameters.

8. The intelligent spraying device for the end ring and the core of the cast aluminum rotor of an electric machine according to claim 2, characterized in that, The multi-station collaboration and safety interlock control module includes: The equipment is configured with three operating modes: automatic, manual, and debugging. Hardware and software dual interlocking logic is set between each mode, and switching can only be performed after authorization verification. Automatic mode completes the fully automated production process, manual mode is used for equipment debugging and maintenance, and debugging mode is used for process parameter optimization. A time-slice-based multi-task scheduling mechanism is adopted to divide the single-station spraying cycle into six core steps: rotor positioning detection, spraying path planning, turntable positioning, spraying execution, quality inspection, and station switching. Connection logic is set between each step. During the spraying process, the height of the turntable or the tilt angle of the placement pin are dynamically adjusted based on the distance data fed back by the laser displacement sensor to ensure that the spray head always maintains the optimal spraying angle and distance; the movement path of each moving part is checked in real time, and when the risk of motion interference or the spraying distance deviation exceeds ±5mm is detected, the spraying is automatically paused and an adjustment prompt is issued. The system is configured with two levels of emergency stop logic: global emergency stop and process emergency stop. When the global emergency stop is triggered, all power supplies to the actuators are immediately cut off, while only the power supply to the fan and control system is kept on. The process emergency stop only cuts off the spraying material supply and the drive of the motion mechanism. After the fault is cleared, execution can continue from the interrupted position.

Images