A real-time monitoring and feedback control system for electrostatic powder coating

Abstract

The application discloses a kind of enamel electrostatic powder coating real-time monitoring and feedback control systems, it is related to industrial enamel coating technical field, the system includes: data acquisition and monitoring module, feedback control decision module, actuator control module, spraying path planning and adjustment module, parameter optimization and self-adaptive adjustment module and multi-objective collaborative regulation strategy module;System is by gathering spraying enamel electrostatic powder state parameter, deposition error signal and environmental interference data, constructs standardization state vector and identifies exception, realizes to the closed loop control and dynamic adjustment of spray gun electrode voltage, powder supply, spray gun path parameter;Introduce particle swarm optimization and other intelligent algorithms to optimize controller and process parameters, combined with multi-objective collaborative strategy automatically balance coating uniformity and deposition efficiency, realize spraying quality stability under complex working condition.The application has self-learning, self-adaptive, scalable advantage, significantly improve the intelligent level and production performance of enamel electrostatic powder spraying process.

Description

Technical Field

[0001] This invention relates to the field of industrial enamel coating technology, and in particular to a real-time monitoring and feedback control system for electrostatic powder coating of enamel. Background Technology

[0002] Enamel electrostatic spraying, as an important surface treatment process, is essentially a complex dynamic control system involving multiple variables and disturbances. This process is widely used in the automotive, home appliance, and construction industries, and places stringent requirements on coating uniformity, spraying efficiency, and energy consumption control. However, due to the diverse shapes of the objects being sprayed and the volatile operating conditions, the coordinated control of various execution units, real-time parameter adjustment, and suppression of environmental disturbances during the spraying process become key factors affecting product quality and production efficiency.

[0003] In existing spraying control systems, mainstream solutions mostly rely on fixed control programs or manual experience to set control parameters, lacking a real-time closed-loop feedback adjustment mechanism, making it difficult to achieve dynamic adaptive control of the spraying process. Especially when facing complex curved surfaces, spatial obstructions, and fluctuating environmental conditions, the system control precision is insufficient, easily leading to quality problems such as uneven coating, overspray, and underspray. In addition, although some studies have introduced sensor monitoring, image recognition, or coating thickness detection modules, these functional units lack efficient linkage and coordination with control decision-making and execution mechanisms, failing to form a complete automated closed-loop control system.

[0004] Current control strategies generally suffer from the following shortcomings: First, the coupling between data acquisition, analysis, and execution is low, and information transmission delays lead to lag in control response; second, control logic is mostly statically set, lacking adaptive optimization capabilities based on real-time data; and third, path planning and spraying parameter adjustment are not coordinated under a unified control framework, failing to fully reflect the advantages of process control systems in multivariate coordinated optimization.

[0005] To address the aforementioned issues, this invention provides a real-time monitoring and feedback control system for enamel electrostatic powder coating. Through multi-source sensing, dynamic modeling, real-time feedback, and closed-loop control of actuators, combined with collaborative self-optimization of path and process parameters, it achieves intelligent control and adaptive operation of the entire coating process, significantly improving the system's control accuracy and stability. Summary of the Invention

[0006] To address the above problems, this invention provides a real-time monitoring and feedback control system for enamel electrostatic powder coating, which solves the problems of lack of real-time feedback control, disconnect between path and deposition state, and insufficient parameter optimization capability in the existing technology during enamel electrostatic powder coating process.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a real-time monitoring and feedback control system for enamel electrostatic powder coating, comprising the following modules:

[0008] The data acquisition and monitoring module includes a spray gun parameter sensing unit, a particle deposition monitoring unit, and an environmental state sensing unit, which respectively collect spraying state parameters, deposition error signals, and environmental interference parameters; the signal processing and anomaly detection unit organizes and generates standardized state vectors and weighted error signal vectors.

[0009] The feedback control decision module consists of an error calculation and reference signal reconstruction unit, an amplitude and phase compensation unit, a control algorithm unit, and an output signal adjustment unit. It sequentially receives and processes the standardized state vector and the weighted error signal vector to obtain the final execution control signal.

[0010] The actuator control module splits and maps the final execution control signal to different control objects, respectively driving the high-voltage electrostatic output control unit to adjust the spray gun electrode voltage, the powder supply and air pressure control unit to adjust the powder flow rate and the delivery gas pressure, and the motion path execution unit to control the spray gun's motion posture and trajectory speed.

[0011] The spraying path planning and adjustment module generates the initial pose control path of the spray gun based on the three-dimensional shape of the workpiece and the desired deposition specifications, and adjusts the path parameters in real time based on the coating deposition feedback signal during the spraying process.

[0012] The parameter optimization and adaptive adjustment module is responsible for intelligent tuning, dynamic coordination and knowledge accumulation of control parameters and process settings, so as to maintain stability and optimal coating performance under changing working conditions.

[0013] The multi-objective coordinated adjustment strategy module is used to comprehensively balance various control objectives during operation and adjust control parameters and execution parameters in real time.

[0014] Compared with the prior art, the advantages of the present invention are:

[0015] This invention constructs a data acquisition and monitoring module that integrates spray gun parameters, particle deposition, and environmental conditions. For the first time, it introduces a multi-channel synchronous acquisition and signal fusion mechanism in enamel electrostatic powder coating. Through standardized state vector construction and abnormal interference identification, it not only improves the integrity and consistency of state data, but also effectively eliminates the influence of external disturbances on the coating process. It solves the problem that traditional single-point monitoring methods are difficult to capture systematic errors, and provides highly reliable basic data for subsequent feedback control.

[0016] This invention designs a multi-layer feedback control structure that includes state reconstruction, gain adjustment, control algorithm and output correction. It not only corrects the spraying parameters in real time based on the instantaneous error signal, but also has the advantages of strong disturbance suppression capability, fast adjustment response and high control accuracy. In particular, it can maintain good coating consistency and uniformity even on complex workpiece surfaces and under extreme environmental conditions.

[0017] This invention refines the feedback control signal into three dimensions of execution actions: high-voltage power supply regulation, powder supply / air pressure linkage control, and path motion control. Combined with high-precision sensors and fast-response actuators, it achieves millisecond-level delivery of control commands, constructing a complete "command-execution-feedback" closed-loop chain. The spray gun electrode voltage, powder flow rate, and movement trajectory are highly coordinated, greatly reducing over-spraying and under-spraying, improving material utilization, and reducing the recycling burden.

[0018] This invention effectively solves the problem of uneven coating caused by the "Faraday cage" effect in complex irregular workpieces or edge areas through the fusion mechanism of offline path generation and online trajectory fine-tuning. The online trajectory adjustment unit can dynamically correct the spray gun speed, attitude and dwell time according to the coating deviation, so as to realize "re-spraying" and "avoiding spraying" as needed.

[0019] This invention introduces an improved particle swarm optimization algorithm and a multi-objective collaborative optimization strategy to perform global optimization search and dynamic adjustment of controller parameters and process parameters. The algorithm supports self-learning and online updates, and can adapt to different workpieces, powder types or environmental conditions in real time, avoiding repeated manual trial and error tuning, greatly shortening the debugging time and improving system stability. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is the system architecture diagram of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings. The described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to describe selected embodiments of the invention.

[0023] Please refer to Figure 1 , Figure 1 This is an architecture diagram of a real-time monitoring and feedback control system for enamel electrostatic powder coating provided by an embodiment of the present invention, which includes the following modules:

[0024] The data acquisition and monitoring module includes a spray gun parameter sensing unit, a particle deposition monitoring unit, and an environmental state sensing unit, which respectively collect spraying state parameters, deposition error signals, and environmental interference parameters. After the collected data is processed by the signal processing and anomaly detection unit for noise reduction filtering, standardization conversion, and anomaly analysis, a standardized state vector and a weighted error signal vector are generated.

[0025] The spray gun parameter sensing unit collects spraying state parameters through a high-voltage voltage sensor, a corona discharge current sensor, and an atomization and delivery air pressure sensor. The specific spraying state parameters are: high-voltage electrode voltage U(t), current I(t), and air pressure P(t) during the spraying process. The unit performs deviation analysis between the spraying state parameters and the set target parameters to determine whether the spray gun is in a stable working state, and then determines whether the working point of the enamel electrostatic powder coating equipment is normal. The working point refers to the set point of the operating state of the electrostatic spray gun body and its high-voltage power supply and pneumatic interface, which is the basis for the judgment of the control logic.

[0026] It should be noted that the high-voltage sensor is connected to the output terminal of the high-voltage power module of the spray gun, and the corona discharge current sensor is embedded in the discharge electrode path of the spray gun; the atomization and delivery air pressure sensor is installed in the air supply pipe of the spray gun; U(t) is used to determine whether the spray gun has reached the expected electric field strength; I(t) is used to reflect the spraying corona intensity and powder charging efficiency; P(t) is used to confirm whether the pneumatic delivery is stable; in addition, the operating point of the spraying equipment is the real-time operating status point of the electrostatic spray gun and its high-voltage power supply and air supply interface, and the determination is based on... All parameters are within the set threshold range; if the working point deviates, it is determined as "abnormal operation of the spray gun". This state will trigger the feedback control module to adjust the spraying parameters and issue an alarm signal.

[0027] The particle deposition monitoring unit continuously acquires the coating thickness using a laser thickness gauge deployed near the enamel electrostatic powder spray gun. Simultaneously, it captures spray coverage images using an industrial camera and extracts uniformity features from the spray images. By introducing a deposition rate estimation model based on current fluctuations, it indirectly calculates the change in powder amount per unit time (i.e., deposition rate per unit time), comprehensively reflecting the coating quality and powder utilization efficiency.

[0028] The deposition rate per unit time is shown in the formula:

[0029]

[0030] in, Deposition rate per unit time Let be the current fluctuation value per unit time, and k be the charging efficiency constant. Unit of time.

[0031] The collected coating thickness, uniformity characteristics, and deposition rate per unit time are compared with the target value to generate a deposition error signal.

[0032] The environmental condition sensing unit is used to detect interference factors in the spraying environment that may affect the stability of the coating. It uses a digital humidity sensor to obtain the relative humidity RH(t) and a thermoelectric sensor to obtain the ambient temperature. And the electric field disturbance sensor detects background interference voltage not originating from the spray gun. These three factors together constitute environmental disturbance parameters;

[0033] When any of the environmental interference parameters exceeds the set tolerance threshold, this state is marked as "external interference activated" and the subsequent control module is triggered to dynamically adjust parameters such as spraying voltage or air pressure to maintain coating stability.

[0034] The signal processing and anomaly detection unit performs unified preprocessing, anomaly detection, and weighted error signal generation on the collected spraying state parameters, deposition error signals, and environmental interference parameters. First, the three types of raw data are organized into a spraying state parameter vector S(t), a deposition effect parameter vector D(t), and an environmental interference parameter vector E(t), respectively. These three vectors are then unified to form the original state dataset. Then, the input dataset is fused based on the time window, as follows:

[0035] First, a moving average filter is used to denoise the three types of state vectors respectively, eliminating transient biases caused by high-frequency acquisition fluctuations and equipment jitter. Then, a Z-score-based standardization method is used to normalize the state vectors to zero mean and unit variance. Eliminate the deviation between different dimensions;

[0036] For the standardized state vector, three anomaly detection strategies are further designed:

[0037] 1. Set dynamic control limits for each parameter in the standardized state vector, and construct an upper limit based on historical averages and process fluctuation thresholds. and lower limit ,when Any parameter term in the range When this occurs, it is considered an out-of-bounds anomaly and its deviation magnitude is recorded;

[0038] 2. Calculate the perturbation growth rate of each state parameter and compare it with its sliding historical mean. If the rate of change increases or decreases abnormally, it is marked as a trend drift anomaly.

[0039] III. Establishing the State Residual Function ,in, The residual value of the i-th state parameter at time t reflects the difference between the current value and the predicted value; The value of the i-th parameter is the actual observed value. The estimated value is derived from the autoregressive sliding window prediction; when the residual Exceeding the seatbelt threshold When this occurs, it is determined to be a soft fault risk.

[0040] All abnormal results identified by the above three anomaly detection strategies will be uniformly encoded into a multidimensional error signal vector. Each item in the vector not only represents the anomaly type of the corresponding indicator, but also includes the specific deviation value, trigger time and confidence score, which are used to express specific scenarios, including "instantaneous drop in spray gun voltage", "uneven powder deposition" and "excessive interference from workshop temperature".

[0041] To further enhance the response accuracy of the feedback loop, this unit constructs a parameter priority matrix. Where m is the dimension of the control variables and n is the dimension of the error signal, the multidimensional error signal vector is weighted, and each element in W... This matrix represents the influence strength of the j-th error signal on the i-th control variable. It can be set based on historical control data, expert experience, and statistical analysis of fault impact. After being applied to the multidimensional error signal vector, this matrix outputs a weighted error signal vector. While retaining multi-source anomaly information, it highlights the impact of key control variables, thereby driving the feedback adjustment mechanism of the subsequent control strategy generation module in a more targeted manner.

[0042] The feedback control decision module consists of an error calculation and reference signal reconstruction unit, an amplitude and phase compensation unit, a control algorithm unit, and an output signal adjustment unit. It sequentially receives and processes the standardized state vector and the weighted error signal vector to obtain the final execution control signal.

[0043] The error calculation and reference signal reconstruction unit is based on the standardized state vector output by the data acquisition and monitoring module. and weighted error signal vector The current state of the system is further processed to construct a smoother and more dynamically adaptive control reference signal. Based on this, the instantaneous error signal vector that the controller can directly use is calculated. .

[0044] First, introduce a sliding time window. Historical state sequence The current reference state trajectory is generated using a weighted quadratic spline interpolation function. The interpolation weight vector Based on the current weighted error signal Dynamically adjust and prioritize enhancing the fitting accuracy of high-weight outliers; the expression for generating the reference trajectory is:

[0045]

[0046] In the formula, Let be the quadratic spline basis function with the control node at time i. As a weighting factor, satisfying The above methods ensure that the reference signal not only reflects historical trends but also has real-time adjustment capabilities to meet control requirements under different disturbance levels.

[0047] Obtain the current reference state vector Then, the unit further integrates with the real-time state vector. By comparison, the instantaneous error signal vector of the system is calculated:

[0048]

[0049] in, The instantaneous error signal vector, This is the real-time state vector. The current reference state vector; vector Each dimension It indicates the instantaneous deviation of the corresponding state parameters, specifically including the spray gun output voltage, current, powder deposition thickness, and environmental disturbance intensity; the error signal of this unit is generated based on an adaptively constructed reference trajectory, which has stronger time consistency and anti-interference capability.

[0050] To enhance the controller's response efficiency to abnormal changes, additional measures were taken. We perform weighted mapping to construct an adjustment signal that can be used in the incentive control strategy model, as shown in the equation:

[0051]

[0052] in, To adjust the signal, For parameter priority matrix This is the instantaneous error signal vector, with dimension m representing the number of parameters in the state vector. This vector not only preserves the amplitude characteristics of the error information; ultimately, this unit outputs two core variables: the adaptive reference state trajectory. and weighted control error signal .

[0053] The amplitude and phase compensation unit is located in the signal preprocessing stage of the feedback control module. It is used to correct the system amplitude and phase frequency distortion caused by transmission delay, sensing lag, and dynamic mismatch in the spraying control chain before the control error signal enters the controller calculation. By designing a compensation filter in the frequency domain to reverse the gain attenuation and phase lag generated during the transmission process, the entire control channel is made equivalent to an ideal linear system, thereby enhancing the controller's response speed and adjustment accuracy to error signals.

[0054] First, through frequency domain experiments during the trial operation phase, the frequency response model of the spraying control path was obtained, denoted as the transfer function. This represents the response relationship between the input of the spray gun control parameters and the output of the deposition error signal. This function can be decomposed into an amplitude-frequency response. Phase frequency response The specific formula is as follows:

[0055]

[0056] Based on this model, a compensation filter is constructed. Its goal is to achieve amplitude-frequency inverse gain and phase inverse feedforward, such that:

[0057]

[0058] This makes the entire transmission chain tend to have unity gain and zero phase lag within the control frequency band, thereby achieving instantaneous amplification and phase alignment of the error signal under linear response conditions.

[0059] When the filter is applied to the error signal And generate the compensated control reference input. This ensures that the signal received by the controller is no longer distorted due to the dynamic lag of the spray gun-deposition-sensing channel, thus improving the dynamic controllability and stability margin of the closed-loop control system.

[0060] For multi-gun channel scenarios, this unit constructs a channel amplitude and phase compensation matrix. The filter coefficients of each channel are adjusted online to maintain consistent response bandwidth and minimize cross-interference in the multiple-input multiple-output system. The compensation formula is extended as follows:

[0061]

[0062] in, and These represent the Fourier transform and its inverse, respectively, used to perform frequency domain filtering operations.

[0063] It should be noted that the accuracy and response stability of the compensation filter are limited by the accuracy of the transfer function model and the sampling frequency. Therefore, in the initial deployment phase, the system employs a method of experimental identification, model fitting, and parameter self-tuning to... and Iterative optimization is performed to ensure that the compensation effect meets the performance requirements for rapid convergence of deposition errors within the main operating frequency band, and the finally outputted error signal vector is the compensated error signal vector. .

[0064] The control algorithm unit is responsible for receiving the error signal after it has been processed by the amplitude and phase compensation unit. And combined with the current reference state vector Generate main control output signal This system is used to guide the spraying equipment to achieve precise adjustment of coating thickness. Considering the significant nonlinear disturbances, operating point drift, and environmental interference during electrostatic powder spraying, a Linear Active Disturbance Rejection Control (LADRC) strategy is selected to construct the control mainline. Its core advantages lie in its weak model dependence and strong disturbance rejection robustness. The LADRC control structure consists of the following three parts:

[0065] The tracking differentiator TD is used to process the reference state vector. The mutations or discontinuous boundary values ​​present in the data are smoothed using a first- or second-order dynamic structure to prevent high-frequency excitation components from entering the controller. The smoothed reference trajectory is denoted as... It retains the expected behavioral trend while eliminating signal jitter.

[0066] The Extended State Observer (ESO) uses feedback and control signals to estimate the current system state (including deposition thickness and deposition rate) and the total disturbance term (including unmodeled dynamics, powder supply instability, and environmental wind disturbance). The ESO is typically designed in a second- or third-order form, and the state estimate is expressed as follows: Deposition thickness estimate, Thickness change rate The total disturbance estimate, and all the estimation results constitute the extended state vector. This provides an observational basis for the feedback law.

[0067] Linear State Error Feedback (LSEF) is used to calculate the desired trajectory. With estimated state The deviation between them is used to construct a feedback law to generate the control output:

[0068]

[0069] in, These are proportional gain and disturbance compensation gain, respectively. The feedback adjustment process has structural decoupling, fast response and active disturbance suppression capabilities. The overall control strategy is designed as a control gain adjustable structure, which automatically adjusts according to the system transient response and control accuracy requirements during operation, so that the system can maintain a stable deposition thickness for different workpiece structures.

[0070] In addition, to improve output controllability, this unit also introduces a limiting function. And a control rate constraint to limit sudden changes in control quantities or exceeding physical boundaries:

[0071]

[0072] Ultimately, this unit outputs the main control variable. This signal will be transmitted to the output signal adjustment unit for pre-processing and is the core command stream for adjusting the operating status of the spray gun.

[0073] The output signal conditioning unit is responsible for receiving the main control output signal generated by the control algorithm unit. Before the command is sent to the actuator module, a series of safety restrictions and response adaptation operations are completed, ultimately generating an execution command stream that can directly drive the spray gun actuator. .

[0074] Control signal safety limiting mechanism: generated by the controller Ideal control commands may exceed the operating range allowed by the spray gun hardware in engineering practice, posing risks such as overpressure, overcurrent, and overheating. Therefore, physical limiting constraints are first applied to ensure that all control variables are within the set safe range.

[0075]

[0076] in, These are the lower and upper limits of the physical constraints for each execution quantity (such as electrostatic voltage, powder supply valve opening, guide rail speed, etc.). This process ensures that all sent signals do not cause equipment overload, operational instability, or control failure.

[0077] Dynamic adaptation mechanism for actuators: Different types of actuators have their own specific dynamic response characteristics. To ensure that the actual response matches the control intention... Figure 1 To, further Implement the following compensation adjustments:

[0078] The slope limiter suppresses sudden signal changes and prevents equipment shock or coating scattering caused by a sharp increase in command. The expression is as follows:

[0079]

[0080] The hysteresis compensation term pre-compensates for factors such as valve response delay and voltage rise inertia, improving the tracking accuracy of the control response; the bias correction term compensates for static errors and constructs output zero-point alignment; the feedforward gain factor adds gain correction based on the trend direction of the control quantity, improving the response bandwidth.

[0081] The above mechanisms work together to The final execution control signal after dynamic matching is obtained:

[0082]

[0083] in, The combined effect of all compensation functions can be considered a "device-friendly control mapping function"; furthermore, to achieve closed-loop data management, this unit will, after each signal correction, The time-series values ​​are synchronously transmitted back to the data acquisition and monitoring module as a control signal execution log; the final output signal conditioning unit will then execute the control signal. Transmitted to the actuator module.

[0084] The actuator control module receives the final execution control signal output by the feedback control module, splits and maps it to different control objects, and drives the high-voltage electrostatic output control unit to adjust the spray gun electrode voltage, the powder supply and air pressure control unit to adjust the powder flow rate and the delivery gas pressure, and the motion path execution unit to control the spray gun's motion posture and trajectory speed. Together, they act on the electro-pneumatic-mechanical components of the spraying equipment to realize the spatial distribution control and thickness adjustment of the powder deposition process.

[0085] In the high-voltage output control unit, the control signal components for the parameter section in the feedback control module are analyzed. And map it in real time to the high voltage output setpoint of the spray gun electrode. This unit adjusts the charge state and migration driving force of powder particles to precisely control the deposition behavior of enamel electrostatic powder on the workpiece surface; it also constructs a high-voltage electrode control interface layer, which is connected to a high-voltage generation module to achieve closed-loop adjustment of the electric field strength between the spray gun electrode needle and the workpiece.

[0086] Specifically, control signals After signal decoding and safety verification, the upper and lower limits of electrostatic output are first set through a dynamic limiting module, while a voltage slope limiter is introduced. This ensures stable voltage changes, preventing sudden changes that could lead to powder buildup instability or spray gun puncture. Subsequently, the adjusted electrostatic command drives the high-voltage power supply through a closed-loop interface, causing its output voltage to... Closely track and control setpoints; to ensure closed-loop performance, a local feedback control loop is constructed using integrated voltage / current sensors to monitor the electrical status of the spray gun end in real time. When the output current is detected... If the voltage exceeds the set threshold, the system will automatically reduce the voltage output and activate the overcurrent protection mechanism.

[0087] During the dynamic operation of the spraying process, if the control system detects insufficient powder deposition in a certain area, the feedback decision module will improve... This enhances the charge level of powder particles, increases the electric field driving force for their migration to the workpiece, and achieves enhanced local deposition; conversely, if the deposition is too thick, the spray gun electrode voltage is reduced to limit the adhesion of excess powder.

[0088] The high-voltage control output of this unit The data will be synchronously transmitted to the data acquisition and monitoring module for recording the operation of spraying electrical parameters and tracing subsequent faults. At the same time, the response delay and adjustment deviation of the electrostatic control interface will be periodically checked to ensure that the control closed loop has high bandwidth, high precision and low latency performance.

[0089] The powder supply and gas pressure control unit is used to analyze the powder supply and gas drive components in the output signal of the feedback control decision module, i.e. and These commands are then converted into powder supply motor drive commands and atomizing / conveying gas pressure regulation commands, respectively, to construct a complete powder conveying and atomizing spray control path. This unit controls the spray concentration and velocity distribution of powder particles, directly determining the coating thickness uniformity and material utilization rate.

[0090] control signals After adjustment, it serves as the command for the powder feeder, used to control the drive frequency of the screw powder feeder or vibratory powder feeder. This frequency determines the amount of powder delivered per unit time, which in turn adjusts the actual powder flow rate at the spray gun nozzle. To prevent powder supply blockage or overshoot, a speed limiter and slope constraint module are introduced before signal transmission to ensure smooth powder supply adjustment. The system integrates a powder flow sensor and uses feedback to build a closed-loop regulation to maintain the powder supply consistent with the set value.

[0091] control signals Deduced as delivery air pressure With atomizing air pressure The adjustment command acts on the electronically controlled proportional pressure regulating valve to control the gas flow rate and pressure, thereby affecting the propagation range and atomization quality of the powder jet. In addition, the air pressure configuration is dynamically adjusted according to process requirements. When the powder is insufficient in complex contours or deep concave areas, the atomizing air pressure is increased to improve the particle jet speed and dispersion. In flat areas, the air pressure can be reduced to reduce powder rebound and waste.

[0092] By constructing a complete powder supply-gas pressure coordinated control mechanism, the kinetic energy matching between powder particles and gas is adjusted to control their distribution and deposition coverage during flight. In addition, the local feedback loop monitors the gas pressure, flow rate, and powder concentration parameters in real time to ensure the ejection quality and actuator response performance. Finally, the powder supply and gas pressure control unit outputs the adjusted powder supply drive signal and gas pressure setting signal.

[0093] The motion path execution unit is used to parse the spatial path control command portion of the output signal from the feedback control decision module. It is then converted into real-time trajectory scheduling instructions for the mechanical actuators in the spraying equipment, driving the spray gun to run on the workpiece surface according to the desired trajectory, thus completing the coating control in the spatial dimension.

[0094] The main objects controlled by this unit include industrial robot arms, linear guides, rotary platforms, and other motion control devices. The input signals exist in the form of desired spray gun position, attitude angle, instantaneous velocity path parameters, forming a three-dimensional motion command set. The unit is equipped with a trajectory interpolator and a kinematic solver to convert discrete commands into a continuous, smooth, and physically reachable sequence of joint commands.

[0095] To ensure the coordination between the spray gun path and powder supply / voltage control, this unit constructs a spraying synchronization mechanism to constrain the dynamic relationship between the spray gun position, powder concentration, and electric field strength. When the path passes through the edge of the workpiece or a deep recess, if the data acquisition module detects insufficient deposition, the motion path execution unit responds to the deceleration command issued by the control module to reduce the local passing speed, thereby increasing the actual deposition thickness in that area. If overspray is detected in a local area, the path execution unit immediately increases the spray gun speed to shorten the action time and suppress overspray.

[0096] It should be noted that this unit supports two path scheduling modes: offline path execution mode, which generates the spraying trajectory in advance based on the workpiece CAD model and process requirements, stores it as a standard path file, and executes it in a fixed sequence during the spraying process; and online path dynamic correction mode, which corrects the running path or local trajectory segments in real time based on the deposition error signal and position deviation fed back by the monitoring module, so as to achieve adaptive spraying.

[0097] Finally, the motion path execution unit outputs real-time control commands to the robot controller or motion platform driver; at the same time, it transmits the actual position and speed information of the spray gun back to the data acquisition and monitoring module in real time to assist in the subsequent analysis of the deviation relationship between deposition thickness and path execution.

[0098] The spraying path planning and adjustment module is responsible for generating the initial pose control path of the spray gun based on the three-dimensional shape of the workpiece and the desired deposition specifications. During the spraying process, it adjusts the path parameters in real time based on the coating deposition feedback signal to achieve dynamic optimization of deposition uniformity and material utilization.

[0099] The offline path generation unit constructs an initial spraying path point set based on the 3D CAD model of the workpiece in the spraying task, the expected thickness mapping matrix, and the spray width characteristic parameters, and outputs a path control vector sequence.

[0100] The path planning employs a geometric algorithm based on voxel division and surface scanning, and integrates process rule constraints to ensure that all critical surface regions are effectively covered, while avoiding repeated powder application or missed areas. The generated path vector sequence includes parameters such as pose points, passing speed, and preset attitude angles, and is uploaded to the main control module before spraying to provide initial scheduling basis for the motion path execution unit.

[0101] The online trajectory adjustment unit receives the spatial distribution error signal of deposition thickness output by the particle deposition monitoring unit during the spraying process. Combined with the offline path point set, it identifies the deviation area in real time, executes the fine-tuning strategy, and generates the correction path instruction.

[0102] Specific adjustment strategies include: for areas with significant insufficient thickness, dynamically reducing the spray gun's throughput speed, increasing the number of repetitions at that path point, or appropriately shortening the distance between the spray gun and the workpiece; for areas detected as excessively thick, increasing the spray gun's throughput speed, reducing repetitions at path points, or temporarily disabling the path segment in that area; applying a smoothing function to all trajectory fine-tuning operations to avoid actuator oscillations or position error propagation caused by path jumps; and triggering the adjustment frequency according to the spraying batch or cycle to avoid unnecessary waiting or false triggering responses caused by frequent path reconstruction.

[0103] The multi-axis coordination and obstacle avoidance unit is used to coordinate and schedule the multi-axis joints of the spraying actuator during path planning and adjustment, and to detect physical or constraint conflicts that may occur during path adjustment.

[0104] The inputs include the corrected path control vector, the painting environment model, and the robot motion constraints; the inverse kinematics solver generates joint space commands, and an obstacle avoidance detector runs at each path point. Once a path point is detected to be approaching or crossing the obstacle volume boundary, path replanning or local avoidance is performed.

[0105] In extreme cases, it supports setting fault-tolerant rules or marking areas for manual respraying to avoid the risk of posture singularities or collisions caused by forced path adjustments; finally, the path control commands output by the module and the control commands output by the feedback control decision module jointly drive the motion path execution unit to form a multi-channel fusion control chain of spatial path and process control, thereby improving the system's ability to control the consistency of coatings on complex workpiece surfaces.

[0106] The parameter optimization and adaptive adjustment module provides the system with continuous learning and performance self-improvement capabilities. It is responsible for intelligent tuning, dynamic coordination, and knowledge accumulation of control parameters and process setpoints, ensuring the system maintains stability and optimal coating performance under varying operating conditions. This module forms a loosely coupled closed loop with the control decision module and the actuator control module through parameter vectors and state feedback, and possesses independent calculation, periodic update, and event triggering mechanisms.

[0107] The controller parameter self-tuning unit takes the LADRC control structure in the feedback control decision module as the target, constructs a joint objective function of multiple performance indicators around the key parameters (i.e., observer gain and control gain), and uses an improved hybrid particle swarm optimization algorithm (M-PSO) for global search and local adaptive fine-tuning.

[0108] The input is performance feedback, namely the current control error signal, coating error convergence speed, system response waveform, and disturbance recovery time; the output is the optimal control parameter set, which is written into the LADRC execution structure through the controller parameter update interface; a hybrid evaluation mode of "simulation + short-time measurement" is adopted to verify particle performance and avoid pure model error; chaotic disturbance and adaptive inertial weight adjustment mechanism are introduced in the optimization iteration to improve the convergence rate and suppress premature convergence trap; the self-tuning process is supported by events (such as powder type switching), periods (such as before the first spraying of the day), or trend triggers (such as a significant increase in control error); the results are synchronously written into the parameter update and knowledge accumulation unit for long-term experience backtracking and rapid transfer.

[0109] The multi-parameter collaborative optimization unit is responsible for constructing a multi-objective optimal control strategy for the spray gun voltage setpoint, powder supply rate, spray gun movement speed, and air pressure control value, and outputting a vector of process setpoint parameters, as follows:

[0110] The inputs are the current particle deposition distribution error, powder consumption rate per unit area, and total spraying time. The objective function is constructed using the square of coating uniformity deviation, powder utilization rate, and spraying efficiency as indices. An improved multi-objective particle swarm optimization algorithm (MOPSO) is adopted, which can be switched to alternative algorithms such as CPO to improve global optimization capabilities. The output is a set of parameters in the optimal Pareto front solution set that are close to the expected weight combination. After smoothing, the parameters are pushed to the actuator control module to adjust the system behavior in real time. An optional "online fine-tuning" mode is available: when the monitored indicators deviate continuously, the optimal neighborhood is maintained by local updates within a limited calculation step. It exhibits better adaptive capabilities than static strategies in high humidity, low temperature, or complex irregular part scenarios, such as automatically increasing the voltage to offset the decrease in charging efficiency.

[0111] The parameter update and knowledge accumulation unit, acting as a central coordinator, is responsible for distributing optimization results to each target module in an orderly manner and recording and managing parameter evolution trajectories in the background. It smoothly injects control parameter sets and process parameters into the controller and execution modules through linear interpolation buffering to prevent abrupt disturbances. It archives and manages the results of each optimization iteration along with corresponding operating condition labels (such as powder type, temperature and humidity, and workpiece number) to form an operating condition-parameter mapping table. When the similarity between the current operating condition and historical optimization conditions exceeds a threshold, the corresponding historical optimal parameter scheme can be invoked without re-iteration. It records long-term control and process indicator trends, providing a basis for model correction for system evaluation. It allows engineers to adjust optimization target weights, set iteration step sizes, and enable specific parameter versions, supporting the integration of human experience and algorithmic decision-making.

[0112] It should be noted that this module uses "controller parameter vector" and "process setting parameter vector" as the core outputs, and transmits them synchronously to the feedback control decision module and the actuator control module as inputs, to build a dynamic optimization closed loop across modules, multiple objectives, and multiple time domains, providing adaptive enhancement capabilities and long-term evolution support for the entire spraying system.

[0113] The multi-objective coordinated adjustment strategy module is used to comprehensively balance various control objectives during system operation, adjust control and execution parameters in real time, and achieve dynamic coordination and optimality between quality indicators and production efficiency.

[0114] The performance monitoring and evaluation unit receives performance index data streams from the data acquisition and monitoring module, and, in conjunction with the current spraying process objectives, quantifies in real time the degree to which each control objective is met.

[0115] Input data includes deposition rate, coating thickness standard deviation, and material utilization rate; by filtering out the impact of short-term fluctuations through the data trend within the objective function window, the current target satisfaction vector is output, indicating whether each performance target is met; the indicator judgment logic is based on the set threshold range and judges whether the trend continues to deteriorate; the output results are used as the criteria for weight adjustment and parameter redistribution and are passed to the target weight decision unit.

[0116] The target weight decision unit dynamically adjusts the weight configuration of multi-objective control based on performance evaluation results and current process priorities, driving the system strategy to shift towards the optimal preference direction:

[0117] The input consists of a performance satisfaction vector and a historical set of optimal parameter candidates. A fuzzy rule engine or an experience-driven mapping function is used to output the current weight vector. When uniformity indicators are found to be degrading and approaching a quality threshold, their weights are appropriately increased while efficiency weights are decreased to prioritize coating consistency. Simultaneously, a set of parameter adjustment strategies is determined, clarifying the adjustment direction and magnitude for each parameter. The optimized adjustment direction can be matched with the Pareto parameter set output by the parameter optimization module to select the reference solution that best suits the current weight configuration. This unit forms a complete adjustment strategy package, which is submitted to the coordination and execution unit for implementation, entering the strategy execution closed loop.

[0118] The coordination and execution unit is responsible for implementing parameter update tasks across modules and continuing to monitor system performance after adjustments, forming a real-time control closed loop: receiving parameter adjustment instruction sets and weight adjustment schemes, and distributing them to the corresponding modules;

[0119] For the feedback control decision module, dynamically modify the LADRC control objective or feedback gain;

[0120] Write the set values ​​for spray gun voltage, powder supply rate, and moving speed to the actuator control module;

[0121] For the parameter optimization module, update the multi-objective optimization weight vector to guide its new round of output;

[0122] To avoid system disturbances, all instruction execution adopts a trapezoidal buffer injection method (with transition time and slow change rate set).

[0123] After implementation, continuously monitor the performance results. If the target indicators improve as expected, maintain the strategy; otherwise, further fine-tune or implement a strategy rollback mechanism.

[0124] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations will be apparent to those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A real-time monitoring and feedback control system for enamel electrostatic powder coating, characterized in that, Includes the following modules: The data acquisition and monitoring module includes a spray gun parameter sensing unit, a particle deposition monitoring unit, and an environmental state sensing unit, which respectively collect spraying state parameters, deposition error signals, and environmental interference parameters; the signal processing and anomaly detection unit organizes and generates standardized state vectors and weighted error signal vectors. The feedback control decision module consists of an error calculation and reference signal reconstruction unit, an amplitude and phase compensation unit, a control algorithm unit, and an output signal adjustment unit. It sequentially receives and processes the standardized state vector and the weighted error signal vector to obtain the final execution control signal. The actuator control module splits and maps the final execution control signal to different control objects, respectively driving the high-voltage electrostatic output control unit to adjust the spray gun electrode voltage, the powder supply and air pressure control unit to adjust the powder flow rate and the delivery gas pressure, and the motion path execution unit to control the spray gun's motion posture and trajectory speed. The spraying path planning and adjustment module generates the initial pose control path of the spray gun based on the three-dimensional shape of the workpiece and the desired deposition specifications, and adjusts the path parameters in real time based on the coating deposition feedback signal during the spraying process. The parameter optimization and adaptive adjustment module is responsible for intelligent tuning, dynamic coordination and knowledge accumulation of control parameters and process settings, so as to maintain stability and optimal coating performance under changing working conditions. The parameter optimization and adaptive adjustment module includes: The controller parameter self-tuning unit takes the LADRC control structure in the feedback control decision module as the target, constructs a joint objective function of multiple performance indicators around the key parameters, and uses an improved hybrid particle swarm optimization algorithm for global search and local adaptive fine-tuning. The input is performance feedback, namely the current control error signal, coating error convergence speed, system response waveform, and disturbance recovery time; the output is the optimal control parameter set, which is written into the LADRC execution structure through the controller parameter update interface; a hybrid evaluation mode of "simulation + short-time measurement" is adopted to verify particle performance and avoid pure model error; chaotic disturbance and adaptive inertia weight adjustment mechanism are introduced in the optimization iteration to improve the convergence rate and suppress premature convergence trap; the self-tuning process is supported to be triggered by events, periods, or trends; the results are synchronously written into the parameter update and knowledge accumulation unit for long-term experience backtracking and rapid transfer; The process multi-parameter collaborative optimization unit is responsible for constructing a multi-objective optimal control strategy for the spray gun voltage setpoint, powder supply rate, spray gun moving speed and air pressure control value, and outputting a vector of process setpoint parameters. The parameter update and knowledge accumulation unit is responsible for distributing the optimization results to each target module in an orderly manner, and recording and managing the parameter evolution trajectory in the background; it smoothly injects the control parameter set and process parameters into the controller and execution module through linear interpolation buffering; it archives and manages the results of each optimization iteration along with the corresponding operating condition label to form an operating condition-parameter mapping table. The multi-objective coordinated adjustment strategy module is used to comprehensively balance various control objectives during operation and adjust control parameters and execution parameters in real time.

2. The real-time monitoring and feedback control system for enamel electrostatic powder coating according to claim 1, characterized in that: The spray gun parameter sensing unit collects spraying status parameters through a high voltage sensor, a corona discharge current sensor, and an atomization and delivery air pressure sensor. It analyzes the deviation between the spraying status parameters and the set target parameters to determine whether the spray gun is in a stable working state, and then determines whether the working point of the spraying equipment is normal. The particle deposition monitoring unit continuously acquires the coating thickness using a laser thickness gauge, while simultaneously capturing sprayed coating images using an industrial camera and extracting uniformity features from the sprayed images. It then introduces a deposition rate estimation model based on current fluctuations to indirectly estimate the change in powder application per unit time, i.e., the deposition rate per unit time. The collected coating thickness, uniformity features, and deposition rate per unit time are compared with the target value to generate a deposition error signal. The environmental state sensing unit is used to detect interference factors in the spraying environment. It uses a digital humidity sensor to obtain relative humidity, a thermoelectric sensor to obtain ambient temperature, and an electric field disturbance sensor to obtain background interference voltage from sources other than the spray gun. The three together constitute the environmental interference parameters. The signal processing and anomaly detection unit organizes the spraying state parameters, deposition error signals, and environmental interference parameters into spraying state parameter vectors, deposition effect parameter vectors, and environmental interference parameter vectors, respectively, and unifies them into an original state dataset. The input dataset is then fused using a time window, denoised using a moving average filter, and normalized to a standardized state vector using a Z-score-based standardization method. For the standardized state vector, three anomaly detection strategies are further designed: all anomaly results detected by the three strategies are uniformly encoded into a multidimensional error signal vector; a parameter priority matrix is ​​constructed to weight the multidimensional error signal vector, and this matrix, applied to the multidimensional error signal vector, outputs a weighted error signal vector.

3. The real-time monitoring and feedback control system for enamel electrostatic powder coating according to claim 1, characterized in that: The error calculation and reference signal reconstruction unit constructs a smoother and more dynamically adaptive control reference signal based on the standardized state vector and the weighted error signal vector, and calculates the instantaneous error signal vector accordingly. It introduces a historical state sequence over a sliding time window and uses a weighted quadratic spline interpolation function to generate the current reference state trajectory, i.e., the current reference state vector. After obtaining the current reference state vector, it compares it with the real-time state vector, calculates the instantaneous error signal vector, performs weighted mapping processing, constructs an adjustment signal that can be used for the excitation control strategy model, and finally outputs the adaptive reference state trajectory and the weighted control error signal. The amplitude and phase compensation unit is used to correct the amplitude and phase frequency distortions of the system before the weighted control error signal enters the controller for calculation. It uses a compensation filter designed in the frequency domain to inversely correct the gain attenuation and phase lag generated during transmission, making the entire control channel equivalent to an ideal linear system. Through frequency domain experiments during the trial operation phase, the frequency response model of the spraying control path is obtained, denoted as the transfer function. This function can be decomposed into amplitude-frequency response. Phase frequency response Based on this, a compensation filter is constructed. Acting on the error signal; The control algorithm unit is responsible for receiving the error signal processed by the amplitude and phase compensation unit, and generating the main control output signal by combining it with the current reference state vector. The linear active disturbance rejection control strategy is selected to construct the control main line, and the amplitude limiting function and control rate constraint are introduced to limit the sudden change of control quantity or exceeding the physical boundary, and finally output the main control quantity. The output signal conditioning unit is responsible for receiving the output signal from the main controller, applying physical limiting constraints to the ideal control commands generated by the controller, and performing corresponding compensation and correction on the physical limiting constraints according to the specific dynamic response characteristics of different types of actuators to obtain the final execution control signal after dynamic matching.

4. The real-time monitoring and feedback control system for enamel electrostatic powder coating according to claim 1, characterized in that: The high-voltage electrostatic output control unit is used to analyze the control signal components related to electrostatic parameters in the feedback control module. And map it in real time to the high voltage output setpoint of the spray gun electrode. By adjusting the charge state and migration driving force of powder particles, the deposition behavior of powder on the workpiece surface can be controlled. The powder supply and air pressure control unit is used to analyze the powder supply and gas drive components in the output signal of the feedback control decision module, i.e. and control signals After adjustment, it serves as the powder feeder to execute commands and control the drive frequency. This frequency determines the amount of powder delivered per unit time, thereby adjusting the actual powder flow rate at the spray gun nozzle. Control signals Deduced as delivery air pressure With atomizing air pressure The adjustment command acts on the electronically controlled proportional pressure regulating valve to control the gas flow rate and pressure, thereby affecting the propagation range and atomization quality of the powder jet; by constructing a powder supply-pressure coordinated control mechanism, the kinetic energy matching between powder particles and gas is adjusted to control their distribution state and deposition coverage during flight. Finally, the powder supply and pressure control unit outputs the adjusted powder supply drive signal and pressure setting signal. The motion path execution unit is used to parse the spatial path control command portion of the output signal from the feedback control decision module. It is then converted into real-time trajectory scheduling instructions for the mechanical actuators in the spraying equipment, driving the spray gun to run on the workpiece surface according to the desired trajectory, thus completing the coating control in the spatial dimension.

5. The real-time monitoring and feedback control system for enamel electrostatic powder coating according to claim 1, characterized in that: The spraying path planning and adjustment module includes: The offline path generation unit constructs an initial spraying path point set based on the 3D CAD model of the workpiece in the spraying task, the expected thickness mapping matrix, and the spray width characteristic parameters, and outputs a path control vector sequence. The path planning adopts a geometric algorithm based on voxel partitioning and surface scanning, and integrates process rule constraints to generate a path vector sequence. The online trajectory adjustment unit receives the spatial distribution error signal of deposition thickness output by the particle deposition monitoring unit during the spraying process, and, in conjunction with the offline path point set, identifies the deviation area in real time and executes the fine-tuning strategy to generate a correction path instruction. The multi-axis coordination and obstacle avoidance unit is used to coordinate and schedule the multi-axis joints of the spraying actuator during path planning and adjustment, and to detect physical or constraint conflicts that may occur during path adjustment. It uses an inverse kinematics solver to generate joint space instructions and runs an obstacle avoidance detector at each path point. Once a path point is found to be close to or cross the obstacle volume boundary, path replanning or local avoidance is performed. In extreme cases, it supports setting fault tolerance rules or marking as manual respraying areas to avoid attitude singularities or collision risks caused by forced path adjustment.

6. The electrostatic powder coating real-time monitoring and feedback control system according to claim 1, characterized in that: The multi-objective coordinated regulation strategy module includes: The performance monitoring and evaluation unit receives performance index data streams from the data acquisition and monitoring module, combines them with the current spraying process objectives, and quantifies the degree of satisfaction of each control objective in real time; it filters out the impact of short-term fluctuations by analyzing the data trends within the objective function window, and outputs the current objective satisfaction vector; the index judgment logic is based on the set threshold range, and it judges whether the trend is continuously deteriorating; the output results serve as the criteria for weight adjustment and parameter redistribution. The target weight decision unit dynamically adjusts the weight configuration of multi-objective control based on the performance evaluation results and the current process priority, driving the system strategy to shift to the optimal preference direction; The coordination and execution unit is responsible for implementing parameter update tasks across modules and continuing to monitor system performance after adjustments, forming a real-time control closed loop: receiving parameter adjustment instruction sets and weight adjustment schemes, and distributing them to the corresponding modules.

Images