Paint supply robot

Abstract

The invention relates to the technical field of coating automation, discloses a coating supply robot, and aims to solve the problems of existing coating supply in the aspects of rheological property change, multi-component mixing and activation period management and realize accurate supply. The system is characterized in that coating parameters are obtained in real time through a rheological property sensing module, and an intelligent control unit applies adaptive model predictive control, activation period predictive and mixing proportion dynamic calibration and an automatic cleaning and maintenance module. By the adoption of the technical scheme, accurate supply of coating flow, pressure and mixing proportion can be achieved, the activation period is effectively managed, blockage is prevented, the coating quality and production efficiency are improved, and the operation cost is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coating automation, and particularly relates to a paint supply robot. BACKGROUND

[0002] Industrial robots, as the core component of advanced manufacturing industry, are multi-joint manipulators or multi-degree-of-freedom mechanical devices widely used in various industrial production processes. With its excellent automation capability, precise motion control and strong load bearing capacity, industrial robots have become an indispensable tool to realize efficient and high-quality industrial processing and manufacturing functions. In many industrial fields, such as electronic product manufacturing, modern logistics management, fine chemical production and automobile spraying, industrial robots play a crucial role and significantly improve production efficiency, product consistency and safety. Especially in the field of paint application, traditional manual coating operation not only has low efficiency, but also is easily affected by human factors, leading to fluctuations in coating quality. At the same time, the working environment often accompanies harmful substances such as volatile organic compounds, which pose a potential threat to the health of operators. Therefore, the introduction of industrial robots for paint operation can effectively avoid the above-mentioned drawbacks, realize the automation, standardization and intelligentization of the coating process, and thus ensure the uniformity of the coating, improve the material utilization rate and improve the working environment.

[0003] Currently, the paint application robot system in the prior art mainly focuses on realizing high-precision trajectory control and spraying path planning. Such a system usually consists of a multi-axis robot arm, a spraying end effector (such as a spray gun), and a paint supply unit based on preset parameters. Specifically, the robot arm can reproduce complex coating paths through precise programming, ensuring that the paint can be accurately applied to the designated location. The paint supply unit usually includes a paint tank, a metering pump, a pressure regulator and the corresponding pipeline system, and its main task is to provide a continuous and constant paint flow or pressure for the spray gun. In many conventional coating applications, such as furniture painting, household appliance shell coating, etc., the prior art solutions have shown significant advantages, successfully solving the problems of poor consistency, low efficiency and part of dangerous environment operation of manual operation, and promoting the automation development of coating process. By separating the coating task from manual operation, the prior art solutions effectively reduce production costs and improve overall production efficiency.

[0004] However, with the continuous development of related technologies and the increasingly stringent requirements of industrial application scenarios on coating quality, material performance, and production flexibility, the above-mentioned existing technical solutions based on fixed parameter supply gradually show limitations in coping with new challenges due to some inherent characteristics at the principle level, thereby leading to deep-seated technical contradictions. The reason lies in that although traditional coating supply systems can provide relatively stable flow or pressure, they often cannot dynamically sense and compensate for the slight fluctuations of the physicochemical properties of the coating itself in the actual production process. Specifically, the viscosity, thixotropy, and other rheological parameters of many industrial coatings, especially high-performance or functional coatings, are easily affected by factors such as environmental temperature, shear history, batch differences, and storage time, and thus change. When the coating viscosity deviates, even if the supply system maintains a constant pressure, the actual coating flow through the nozzle will also fluctuate; conversely, if a constant flow is maintained, the pressure before the nozzle will abnormally increase or decrease, thereby affecting the atomization characteristics (for spraying applications), coating uniformity, and final coating thickness. This non-linear response in the coating supply link ultimately leads to instability in coating quality, such as local over-thickness leading to sagging, local under-thickness leading to insufficient hiding power or performance degradation.

[0005] Furthermore, for two-component (2K) or multi-component (MC) reactive coatings, accurate component mixing ratio is a core prerequisite for ensuring coating curing performance, mechanical strength, and durability. In the prior art, volume or weight ratio mixing is usually used, but in actual feeding process, if the viscosity or flow rate of each component is affected by the environment and deviates slightly, the traditional metering pump system cannot dynamically compensate in real time, thereby leading to imbalance in the mixing ratio. This deviation in the mixing ratio not only directly affects the curing kinetics of the coating, which may lead to incomplete curing, sticky surface, or performance indicators that cannot meet design requirements, but also may cause waste of expensive materials. In addition, multi-component coatings usually have a limited "activation period", and once mixed, their viscosity will increase sharply over time until solidification. When dealing with such materials, traditional feeding systems lack fine management of the activation period and effective cleaning mechanism for residual materials in the feeding path, which easily leads to solidification and blockage of materials in the pipeline or spray gun, requiring time-consuming cleaning and even component replacement when necessary, seriously affecting production efficiency and maintenance cost. The existing system mainly focuses on the path accuracy of the robot arm, but relatively ignores the profound influence of the complex rheological properties of the coating itself on the feeding link, limiting the robustness and adaptability of the entire coating system. This deep-seated technical contradiction lies in the disconnection between the physical and chemical properties of the coating itself and the dynamic adaptability of the existing feeding system, rather than just the accuracy of the robot motion control.

[0006] Therefore, how to develop a coating supply robot capable of intelligently sensing the change of coating rheological properties, dynamically adjusting the supply parameters in real time, and effectively managing the mixing and supply process of multi-component coatings, so as to ensure accurate, stable and efficient supply of coatings under complex working conditions, has become a key challenge and technical problem to be solved for those skilled in the art. SUMMARY

[0007] The purpose of the present application is to overcome the deep-seated technical contradictions in the existing coating supply system in dealing with the dynamic change of coating rheological properties, especially the management of multi-component coating mixing and activation period, so as to ensure accurate, stable and efficient supply of coatings in complex industrial application scenarios, and to provide a coating supply robot.

[0008] The purpose of the present application is achieved by the following technical solution: a coating supply robot, comprising the following constituent modules: a coating storage and supply unit, a rheological property sensing module, an intelligent control and data processing unit, a dynamic mixing and conveying unit, and an automatic cleaning and maintenance module. The modules work together to form a closed-loop controlled coating supply system.

[0009] The coating storage and supply unit is provided with at least one main coating storage tank and at least one curing agent storage tank or diluent storage tank for storing various coating components to be supplied. The outlet of each storage tank is connected to its corresponding precision metering pump through an independent pipeline system. The precision metering pump adopts a volumetric pump driven by a servo motor, for example, it can be a planetary gear pump or a progressive cavity pump with high precision metering capability. The drive servo motor of the planetary gear pump is configured with a high-resolution rotary encoder to realize accurate control of pump speed and flow feedback. The metering accuracy of the pump can reach within ±0.5%, and the repeatability error is less than 0.2%. The precision metering pump can accurately output the set flow of coating components according to the instructions issued by the intelligent control and data processing unit. On the suction inlet pipeline or outlet pipeline of the precision metering pump, temperature sensors and pressure sensors are configured to monitor the initial state of the coating components and the pressure fluctuation during the pumping process. The temperature sensor uses a platinum resistance thermometer with a measurement accuracy of ±0.1℃. The pressure sensor is a piezoresistive sensor with a measurement range of 0 to 50 MPa and an accuracy of 0.25% FS.

[0010] The rheological property sensing module is a key component for solving the core technical contradiction of the present application, which is installed on the outlet pipeline of the precision metering pump, adjacent to the inlet end of the dynamic mixing and conveying unit, and is used for real-time and online detection of the rheological parameters of the coating components or mixed coating. The module includes one or more non-contact or micro-contact rheological sensors. As a preferred embodiment, the rheological sensor adopts a viscosity sensor based on the torsional resonance principle of a micro-electro-mechanical system (MEMS), the resonant element of which is in contact with the fluid, and the dynamic viscosity of the fluid is indirectly or directly calculated by measuring the response of the resonant frequency, damping coefficient or Q value to the change of the fluid viscosity and density. The MEMS viscosity sensor has the characteristics of small volume, fast response, high integration and insensitivity to environmental disturbances, and its viscosity measurement range is 0.1 mPa·s to 10000 mPa·s, the measurement accuracy is within ±2% FS, and the response time is less than 100 milliseconds. In addition, the rheological property sensing module is also integrated with a high-precision temperature sensor for compensating the influence of temperature on the viscosity of the fluid during the viscosity measurement process, and providing accurate temperature data for subsequent activation period prediction. The high-precision temperature sensor adopts a thermistor array, and its spatial resolution and time response speed can meet the accurate monitoring requirements of the internal temperature field of the fluid. The module transmits real-time rheological parameters (such as dynamic viscosity, density) and temperature data to the intelligent control and data processing unit through a standard industrial communication interface, such as EtherCAT or PROFINET.

[0011] The intelligent control and data processing unit is the core intelligent decision center of the present application, and its hardware composition includes an industrial-grade multi-core processor unit (for example, an embedded processor based on ARM Cortex-A series architecture or an Intel Atom series processor), equipped with a high-speed data acquisition card, a large-capacity random access memory (RAM) and a non-volatile memory (for example, an industrial-grade solid state disk). The processor unit runs a real-time operating system (RTOS), such as QNX or VxWorks, to ensure the deterministic response time of sensor data acquisition and actuator control. The intelligent control and data processing unit is embedded with complex software algorithm modules, including:

[0012] Data preprocessing and fusion module: this module is responsible for receiving raw data from the rheological property sensing module, pressure sensor, temperature sensor and precision metering pump encoder. By applying digital signal filtering algorithms, such as Kalman filtering or extended Kalman filtering, the sensor data is denoised and multi-sensor data fusion is performed to eliminate measurement noise and improve data accuracy and reliability. The module is also responsible for time stamp synchronization to ensure the time consistency of all collected data, providing accurate basic data for subsequent real-time control.

[0013] Material state estimation and pot life prediction module: Based on the real-time collected coating temperature, viscosity, density data and pre-stored coating physicochemical parameter database (which contains the viscosity-temperature curve, density-temperature curve and curing kinetics parameters such as Arrhenius equation parameters or more complex polymerization reaction kinetics model parameters of different batches and different formula coatings at different temperatures), the real-time viscosity, density and potential shear thinning or thixotropy of the current coating are accurately estimated by using recursive least squares method or nonlinear state observer. For two-component or multi-component reactive coatings, according to the pot life characteristics of the mixed coating, real-time temperature and time accumulation effect, the residual pot life of the coating in the pipeline and the gun is accurately predicted by integrating the chemical reaction kinetics model and the energy accumulation model, and the potential curing risk is evaluated. The prediction results will be the key basis for triggering the automatic cleaning and maintenance module or adjusting the feeding speed.

[0014] Adaptive model predictive control (MPC) algorithm module: This module is the core control strategy to achieve dynamic and accurate coating supply. The MPC algorithm module uses the established mathematical model of coating rheological properties and dynamic response of the supply system (which considers the nonlinear effects of pipeline resistance, pump efficiency, coating viscosity and temperature on flow and pressure), according to the real-time coating state parameters provided by the material state estimation and pot life prediction module (such as the estimated viscosity value), combined with the preset target coating flow, pressure and mixing ratio, to predict the dynamic behavior of the coating supply system in the future (for example, 200 milliseconds to 500 milliseconds prediction time domain). Based on this prediction, the MPC algorithm determines the optimal servo motor speed of the precision metering pump, valve opening and other control variables to minimize the control error, overshoot and wear on the actuator in the prediction time domain, while meeting system operation constraints (such as maximum pressure, minimum flow). The algorithm can compensate for fluctuations in coating viscosity, temperature and other physical and chemical properties in real time, ensuring that the coating flow and pressure output to the spray gun always remain at the set target value under various operating conditions, or for multi-component coatings, ensuring accurate mixing ratio. For example, when the coating viscosity is detected to rise, the MPC algorithm will predict a decrease in flow or an increase in pressure, and actively increase the speed of the corresponding precision metering pump or adjust the valve opening to compensate for the change, thereby maintaining constant flow output or target pressure.

[0015] Dynamic calibration module for mixed proportion: designed for two-component or multi-component coatings. Based on the real-time viscosity and density estimates of each component provided by the material state estimation and pot life prediction module, and the preset mixing proportion (e.g., volume ratio or weight ratio), this module dynamically adjusts the output flow of each precision metering pump. For example, if the viscosity of component A increases, causing its flow to decrease slightly at the same pump speed, while the viscosity of component B remains unchanged, this module will calculate and adjust the servo motor speed of the precision metering pump for component A in real time to ensure that the actual output flow ratio of components A and B strictly meets the preset mixing proportion, thereby overcoming the mixing deviation caused by the difference or fluctuation of the rheological properties of each component. This dynamic calibration mechanism ensures the consistency of the curing performance of the coating and the quality of the final coating.

[0016] Fault diagnosis and early warning module: this module continuously monitors all sensor data and actuator states. Using machine learning models (e.g., support vector machines or neural networks) trained based on historical data or rule-based expert systems, it identifies abnormal conditions such as sensor drift, pump wear, pipeline blockage, valve failure, etc. When potential faults or system parameters deviate from the safety range are detected, this module triggers audible and visual alarms and displays detailed diagnostic information on the human-machine interface. It can even automatically adjust the operating mode or shut down according to preset strategies to avoid equipment damage or product quality problems.

[0017] Dynamic mixing and delivery unit for precise mixing and delivery of coating components from different precision metering pumps to the spray gun. For single-component coating systems, this unit can only include delivery pipelines and a spray gun. For two-component or multi-component coating systems, this unit further includes a dynamic mixer and a mixed coating delivery pipeline. The dynamic mixer is a servo motor driven rotary mixer with multiple spiral blades or stirring elements inside and can accurately control the stirring speed through the servo motor. The mixer inlet is connected to the pipeline of each component, and the inlet valve is controlled by the intelligent control and data processing unit. The mixer can achieve different degrees of shear mixing by adjusting the stirring speed according to the type and mixing requirements of the coating, ensuring the full and uniform mixing of each component, and avoiding the negative impact of excessive shear on the performance of the coating. At the outlet end of the dynamic mixer, a mixed coating temperature sensor and an optional mixed coating viscosity sensor are configured to verify the mixing effect and monitor the progress of the pot life. The mixed coating delivery pipeline is made of materials with smooth inner walls, corrosion resistance, and low adsorption, such as polytetrafluoroethylene (PTFE) or ultra-high molecular weight polyethylene (UHMW-PE) lined pipe. The spray gun is a professional coating spray gun with high atomization efficiency, and its parameters (such as spray width, flow rate, atomization pressure) can be digitally adjusted by the intelligent control and data processing unit or controlled by the robot.

[0018] An automated cleaning and maintenance module is designed to solve the problem of short activation period and easy solidification of multi-component coatings, and to realize rapid cleaning of switching between different coatings. The module includes one or more cleaning liquid storage tanks (e.g., storing solvents, diluents or special cleaning agents respectively), a high-pressure cleaning pump, a series of electromagnetic switching valves, and special cleaning pipelines and waste liquid collection devices. The cleaning pipeline is designed with cross connection points with the coating supply pipeline and the internal design of the dynamic mixer. When the material state estimation and activation period prediction module of the intelligent control and data processing unit predicts that the coating activation period is about to end, or the system is in a long idle state, or the coating type needs to be switched, the automated cleaning and maintenance module will automatically start according to the preset cleaning program. The cleaning program usually includes the following steps: first, the residual coating in the pipeline to be cleaned is emptied to the waste liquid collection device through the electromagnetic switching valve; second, the high-pressure cleaning pump is started to inject the cleaning liquid into the pipeline and the mixer at high flow rate and high pressure, and the residual coating in the pipeline and the mixer is removed by physical flushing and chemical dissolution; then, multiple rounds of flushing with different cleaning liquids can be performed, and compressed air can be introduced for purging to ensure that the pipeline is completely dry and free of residues, preventing contamination between different coatings. The electromagnetic switching valve is made of corrosion-resistant material and has the characteristics of fast response and reliable sealing. The high-pressure cleaning pump can provide adjustable cleaning pressure of 0.5 MPa to 5 MPa to ensure cleaning effect.

[0019] The working principle of the present application is as follows:

[0020] At the initial stage of coating supply, the intelligent control and data processing unit loads the coating component information, target flow, target pressure and mixing ratio required for the current coating task from the coating formulation database. The precision metering pump starts to deliver each coating component to the dynamic mixing and conveying unit according to the set parameters.

[0021] During the supply process, the rheological property sensing module continuously monitors the viscosity, density and temperature of each coating component or mixed coating flowing through the pipeline in real time, and transmits these data to the intelligent control and data processing unit at high speed.

[0022] The data preprocessing and fusion module of the intelligent control and data processing unit processes and fuses the received multi-source heterogeneous sensor data in real time, generating reliable and low-noise real-time system state data.

[0023] Subsequently, the material state estimation and activation period prediction module accurately estimates the real rheological state of the current coating based on the fused data and the built-in coating physical and chemical model, and real-time predicts the remaining activation period of the mixed coating for multi-component coatings, providing key information for subsequent control and cleaning decisions.

[0024] An adaptive model predictive control (MPC) algorithm module receives target parameters from a coating task setting and real-time coating state estimation values from a material state estimation and pot life prediction module. The MPC algorithm calculates and optimizes the servo motor speed of each precision metering pump and the opening of the related valve online according to the system dynamic model inside it. For example, when the coating viscosity is detected to decrease due to the increase of ambient temperature, the MPC algorithm can predict that the flow will increase at the original pump speed, so it will actively reduce the pump speed to accurately maintain the set target flow; conversely, when the viscosity increases, the pump speed will be increased. For multi-component coatings, the dynamic calibration module of the mixing ratio will independently adjust the speed of each precision metering pump according to the real-time viscosity and density difference of each component to ensure the accuracy of the mixed coating component ratio, thereby ensuring the consistency of the final curing performance and coating quality.

[0025] A dynamic mixing and conveying unit mixes the components of the coating according to the instructions of the intelligent control and data processing unit at a specific speed after receiving the components of the coating, ensuring uniform mixing of the components. The mixed coating is conveyed to the spray gun, which applies it in the best atomized state.

[0026] When the coating task is completed, the system is idle for a long time, or the material state estimation and pot life prediction module predicts that the pot life of the mixed coating is about to be exhausted, the intelligent control and data processing unit will trigger the automatic cleaning and maintenance module. The module automatically performs cleaning, emptying and purging operations of the pipeline and mixer according to the preset cleaning program to prevent coating solidification and clogging and prepare for the next coating task or switching to a different coating.

[0027] The man-machine interaction and system integration module displays the system running state, coating parameters, fault information, etc. to the operator in real time through a visual interface, and supports the operator to set parameters and manage tasks. At the same time, through standard industrial communication protocols, it exchanges data and coordinates instructions with external robot controllers, production management systems or supervisory control and data acquisition systems to realize seamless integration and advanced automation of the entire coating application process.

[0028] The present application has the following advantages: 1. Real-time adaptability of coating supply: by introducing an online rheological property perception module and an adaptive control algorithm based on model prediction, the present application can dynamically perceive and compensate for changes in coating viscosity, density and other rheological parameters in real time, ensuring that the actual supply flow, pressure and atomization state of the coating remain constant or within the set target range under the influence of environmental temperature fluctuations, coating batch differences or storage time, thereby greatly improving the stability and consistency of coating quality and effectively avoiding coating defects caused by fluctuations in coating properties in traditional systems.

[0029] 2. Extreme precision and robustness of multi-component paint mixing: The present invention realizes extreme precise control of the mixing ratio of two or more component paints by monitoring the rheological properties of each component paint independently and adjusting the output of each precision metering pump in real time by a dynamic calibration module. This is not affected by fluctuations in the viscosity or density of each component. This is crucial for ensuring the curing performance, mechanical strength and durability of reactive paints, and significantly reduces material waste caused by mixing errors.

[0030] 3. Intelligent management of activation period and anti-clogging ability: The material state estimation and activation period prediction module of the present invention can accurately predict the remaining activation period of the mixed paint based on real-time temperature and paint kinetics model. Combined with the automatic cleaning and maintenance module, the system can automatically start the cleaning program before the activation period is exhausted or during the idle period, effectively preventing the solidification and clogging of paint inside the pipeline and mixer, significantly reducing maintenance time and cost, and improving the availability and production efficiency of the equipment.

[0031] 4. High integration and intelligence of the system: The present invention integrates the functions of paint storage, rheological sensing, precision metering, dynamic mixing, intelligent control, automatic cleaning and human-computer interaction into a unified and intelligent paint supply robot system. Through high-speed industrial network, the modules exchange data and coordinate instructions, realize closed-loop control, fault diagnosis and early warning of the whole process of paint supply, and have high automation, intelligence and flexible production capacity, which can adapt to more complex and demanding industrial painting application scenarios.

[0032] 5. Improve production efficiency and reduce operating costs: Due to the improvement of the stability and precision of paint supply, the product failure rate is reduced, and the automatic cleaning mechanism reduces manual cleaning and equipment downtime, thereby significantly improving production efficiency. Through precise material control and waste reduction, as well as reducing maintenance frequency, the present invention also directly reduces operating costs. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 A system structure schematic diagram of a paint supply robot of the present invention.

[0034] Figure 2 A functional module block diagram of the intelligent control and data processing unit of the present invention.

[0035] Figure 3 A working flowchart of a paint supply robot of the present invention.

[0036] In the figure, 100, paint supply robot; 110, paint storage and supply unit; 111, main paint storage tank; 112, curing agent storage tank; 113, diluent storage tank; 114, precision metering pump; 115, servo motor; 116, high-resolution rotary encoder; 117, metering pump temperature sensor; 118, metering pump pressure sensor; 119, connecting pipeline; 120, rheological property sensing module; 121, rheological sensor; 122, high-precision temperature sensor; 130, intelligent control and data processing unit; 131, industrial-grade multi-core processor unit; 132, data preprocessing and fusion module; 133, material state estimation and pot life prediction module; 134, adaptive model predictive control algorithm module; 135, mixed proportion dynamic calibration module; 136, fault diagnosis and early warning module; 140, dynamic mixing and conveying unit; 141, dynamic mixer; 142, mixed paint conveying pipeline; 143, spray gun; 144, mixed paint temperature sensor; 145, mixed paint viscosity sensor; 150, automatic cleaning and maintenance module; 151, cleaning liquid storage tank; 152, high-pressure cleaning pump; 153, electromagnetic switching valve; 154, cleaning pipeline; 155, waste liquid collection device; 160, human-computer interaction and system integration module. DETAILED DESCRIPTION

[0037] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0038] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor are within the scope of protection of the present application.

[0039] It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0040] It should be noted that: similar reference numerals and letters represent similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0041] In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship commonly understood by those skilled in the art, or the orientation or positional relationship commonly understood by those skilled in the art, only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, the terms "first", "second" and the like are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0042] In the description of the present application, it should be noted that unless otherwise specified and limited, the terms "set", "mount", "connected", "connected" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be connected inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0043] The present application provides a coating supply robot, which aims to revolutionize the traditional coating supply mode, by introducing an advanced real-time rheological property perception, model prediction based adaptive control strategy and integrated dynamic mixing and intelligent cleaning mechanism, to ensure the accurate, stable and efficient supply of coatings in harsh industrial application scenarios, especially significantly improving the intelligent level of multi-component coating mixing and activation period management. The coating supply robot system integrates multiple functional modules, which work together through a high-speed and reliable industrial communication network, and together form a closed-loop controlled and highly automated coating supply system, as shown in the overall structure example Figure 1

[0044] Specifically, the core components of the present coating supply robot 100 include a coating storage and supply unit 110, a rheological property perception module 120, an intelligent control and data processing unit 130, a dynamic mixing and conveying unit 140, and an automatic cleaning and maintenance module 150. In addition, in order to facilitate operation and external system integration, the robot is also equipped with a human-computer interaction and system integration module 160.

[0045] ​The paint storage and supply unit 110 is the base component of the paint supply robot, responsible for the storage, pretreatment and preliminary metering of various paint components. This unit is provided with at least one main paint storage tank 111 and at least one curing agent storage tank 112 or thinner storage tank 113 to store the base paint, curing agent and thinner for viscosity adjustment or cleaning as needed. These storage tanks are usually made of high-quality stainless steel (such as SUS304 or SUS316L), with mirror-polished inner walls to ensure the cleanliness and low adsorption of the paint. Each storage tank is equipped with a liquid level sensor and a stirring device. The liquid level sensor monitors the remaining amount of paint in the tank in real time and triggers a low liquid level alarm or an automatic replenishment mechanism. The stirring device (e.g., a paddle stirrer or a magnetic stirrer) can maintain the uniformity of the paint components according to the properties of the paint, preventing sedimentation or stratification, which is particularly important for paint that is prone to sedimentation or has high viscosity. The capacity of the storage tank can be customized according to specific application requirements, with a common capacity range of 20 liters to 200 liters, and is usually integrated with a temperature control jacket or heating band to maintain the paint temperature at the optimal storage and pumping temperature, such as 20°C ± 2°C, which is crucial for the stability of the subsequent rheological properties.

[0046] The outlet of each storage tank is connected to its corresponding precision metering pump 114 through an independent connecting pipeline 119. The precision metering pump 114 is a key actuator for precise flow control, which uses a volumetric pump driven by a servo motor 115, such as a planetary gear pump or a progressive cavity pump with high precision metering capability. The planetary gear pump is known for its non-pulsating, high shear stability and excellent handling capability for high viscosity fluids. The internal gears are made of wear-resistant and corrosion-resistant materials (such as tungsten carbide, zirconia ceramic or high molecular engineering plastics such as PEEK) for precision machining, ensuring long-term precision and reliability. The progressive cavity pump (screw pump) is suitable for specific paint systems due to its low shear force, smooth delivery and adaptability to solid-containing paints. The drive servo motor 115 of the precision metering pump is equipped with a high-resolution rotary encoder 116, which is usually an incremental or absolute encoder with a resolution of up to 262144 pulses per revolution (2^18 PPR) or even higher, achieving high precision control of pump speed and providing real-time speed and angle feedback, which is directly related to the output flow of the paint. The metering accuracy of the pump can be stably maintained within ±0.5%, with a repeatability error of less than 0.2%, which is crucial for the mixing accuracy of multi-component paints. The precision metering pump can accurately output the set flow of paint components according to the digital instructions from the intelligent control and data processing unit 130.

[0047] A temperature sensor 117 and a pressure sensor 118 are strategically installed on the inlet or outlet line of the precision metering pump 114 to monitor the initial state of the coating components and the pressure fluctuation during pumping. The temperature sensor 117 is preferably a platinum resistance thermometer (RTD), such as Pt100, with a measurement accuracy of ±0.1°C and a fast response time of 0.5 seconds. Its probe is usually directly immersed or closely attached to the pipe wall to ensure that the measured value truly reflects the fluid temperature. The pressure sensor 118 is a piezoresistive sensor with a measurement range of 0 to 50 MPa, an accuracy of 0.25% FS, and good linearity and overload resistance. Real-time data from these sensors not only monitor the working state of the pump, alerting to pipeline blockage or pump wear, but also provide important process parameters for subsequent rheological property compensation and control strategy optimization.

[0048] The rheological property sensing module 120 is the core component of the present application to solve the deep technical contradiction of dynamic changes in coating rheological properties. This module is usually installed on the outlet line of the precision metering pump 114, close to the inlet of the dynamic mixing and delivery unit 140, to ensure that it can detect the rheological parameters of the coating components or mixed coatings before mixing or during pumping in real time. This module contains one or more non-contact or micro-contact rheological sensors 121. As a preferred embodiment, the rheological sensor 121 is designed based on the torsional resonance principle using micro-electro-mechanical systems (MEMS) technology. Its core resonant element, such as a cantilever beam with a micro-channel or a torsional vibrator, is precisely machined on a silicon substrate and contacts the fluid sample through microfluidic technology. When the fluid flows through or contacts the resonant element, the viscosity, density, and other parameters of the fluid will affect its resonant frequency, damping coefficient, or quality factor (Q value). By accurately measuring the changes in these resonant characteristic parameters and combining the built-in physical model, the system can indirectly or directly calculate the dynamic viscosity of the fluid. For example, when the fluid viscosity increases, the damping of the resonant element increases, causing the resonance peak to broaden and the Q value to decrease. This MEMS viscosity sensor has the significant features of small size, fast response speed (usually less than 100 milliseconds), high integration, low power consumption, and insensitivity to external vibration or temperature disturbance, making it particularly suitable for online and continuous monitoring. Its viscosity measurement range can cover a wide range of 0.1 mPa·s to 10,000 mPa·s, sufficient to meet the measurement needs of most industrial coatings, and the measurement accuracy can be within ±2% FS, with a repeatability better than 1%. This real-time and high-precision viscosity data is crucial for the intelligent control and data processing unit to perform adaptive control.

[0049] Further, the rheological property sensing module 120 is also highly integrated with a high-precision temperature sensor 122 for precisely compensating the significant influence of temperature on fluid viscosity during viscosity measurement and providing accurate temperature data for subsequent activation period prediction. The high-precision temperature sensor 122 preferably adopts a thermistor array (such as NTC type) with millisecond-level fast response speed and sub-millimeter-level spatial resolution, which can accurately capture the subtle changes in the internal temperature field of the fluid, thereby ensuring that the viscosity measurement value is a true reflection at the precise temperature. The array is usually strategically placed near the fluid channel of the MEMS sensor to achieve local temperature measurement. The rheological property sensing module 120 transmits real-time rheological parameters (such as dynamic viscosity, density) and temperature data to the intelligent control and data processing unit 130 in a high-speed and deterministic manner through standard industrial communication interfaces such as EtherCAT or PROFINET. These protocols ensure real-time, reliability, and low latency of data transmission, providing a solid data link for closed-loop control.

[0050] The intelligent control and data processing unit 130 is the core intelligent decision-making hub of the coating supply robot, responsible for all sensor data acquisition, processing, analysis, model prediction, and precise control of actuators. Its hardware composition includes an industrial-grade multi-core processor unit 131, which usually adopts an embedded processor based on ARM Cortex-A series architecture (such as NXP i.MX 8M Plus series or NVIDIA Jetson series) or Intel Atom series processor, usually configured with 4 to 8 processing cores, with a main frequency of up to 1.5 GHz to 2.0 GHz, to provide powerful parallel computing power. The processor unit is equipped with a high-speed data acquisition card that supports multi-channel analog and digital signal input, such as 16 channels, 16-bit resolution, and a sampling rate of up to 1 MSPS, ensuring real-time and high-precision acquisition of sensor data. At the same time, it is integrated with a large-capacity random access memory (RAM), such as 8GB to 16GB of DDR4 RAM, for high-speed data caching and algorithm running, and a large-capacity non-volatile memory, such as 128GB to 512GB of industrial-grade solid-state drive (SSD), for storing operating systems, applications, coating parameter databases, and historical running data. The processor unit runs a real-time operating system (RTOS) such as QNX or VxWorks, which is known for its deterministic scheduling, low interrupt latency, and task priority management capabilities, ensuring deterministic response time for sensor data acquisition and actuator control, meeting the strict real-time requirements of industrial control.

[0051] The intelligent control and data processing unit 130 is embedded with complex software algorithm modules that work together to achieve intelligent control of coating supply:

[0052] Data pre-processing and fusion module 132: This module is the entrance of data stream, responsible for receiving raw data stream from flow behavior sensing module 120, metering pump pressure sensor 118, metering pump temperature sensor 117, and high-resolution rotary encoder of precision metering pump servo motor 116. In order to eliminate measurement noise and improve data accuracy and reliability, advanced digital signal filtering algorithms are applied inside this module. For example, for sensor data with Gaussian noise characteristics, Extended Kalman Filter (EKF) or Unscented Kalman Filter (UKF) can be used, which can estimate the true value of system state (such as viscosity, pressure, temperature) and update its uncertainty online. For scenarios with non-Gaussian noise or outliers, nonlinear filtering methods such as median filtering or wavelet denoising can be used. This module is also responsible for multi-sensor data fusion, through weighted averaging, complementary filtering or Bayesian inference-based methods, effectively integrating measurement data from different types of sensors but reflecting the same physical quantity, such as correlating the viscosity measured by the rheological sensor with pump pressure and flow data to obtain a more robust and accurate estimate of the paint state. This module also undertakes the important function of time stamp synchronization, using Precision Time Protocol (PTP) or Network Time Protocol (NTP) to ensure that all collected data has strict temporal consistency, which provides accurate basic data for subsequent real-time control, event sequence analysis and fault diagnosis.

[0053] Material state estimation and pot life prediction module 133: Based on the real-time paint temperature, viscosity, and density data provided by the data pre-processing and fusion module, as well as the pre-stored paint physical and chemical parameter database, this module accurately estimates the true rheological state of the current paint and predicts the remaining pot life of reactive multi-component paint. The paint physical and chemical parameter database is a dynamically updated, structured data set that contains viscosity-temperature curves (usually fitted with Arrhenius equation or Vogel-Fulcher-Tammann (VFT) equation), density-temperature curves, and curing kinetics parameters (such as Arrhenius equation parameters, or more complex polymerization reaction kinetics models such as Kamal-Sourour model, or Auto-Catalytic model-based parameters) of different batches and formulations of paint at different temperatures. This module uses Recursive Least Squares (RLS) or nonlinear state observers (such as sliding mode observer or Luenberger observer) to combine these real-time data and pre-set models to accurately estimate the true viscosity, density, shear thinning characteristic parameters (such as power-law fluid index), thixotropy (time-dependent viscosity change), and curing degree of the paint online.

[0054] For two-component or multi-component reactive coatings, the module accurately predicts the remaining pot life of the coating inside the piping and gun based on the pot life characteristics (usually defined as the time required for the viscosity to double from the initial value or the time required for a certain degree of cure) of the mixed coating, the real-time temperature, and the time accumulation effect by integrating a chemical reaction kinetics model and an energy accumulation model (e.g., using a temperature-time-temperature (TTT) diagram or an equivalent time model). For example, when the coating stays at a higher temperature, its curing speed will accelerate, and the pot life will correspondingly shorten. The module continuously monitors and predicts the progress of the pot life, and when the predicted remaining pot life is below a preset safety threshold (e.g., 10 minutes), or when the system is idle for more than a set time (e.g., 5 minutes), the prediction result will be a key basis for triggering the automated cleaning and maintenance module 150 or adjusting the feed speed to avoid the coating from solidifying and clogging in the piping.

[0055] Adaptive model predictive control (MPC) algorithm module 134: This module is the core control strategy to achieve dynamic and accurate coating supply. The MPC algorithm uses a pre-established mathematical model of the coating rheological properties and the dynamic response of the supply system. The model is a complex nonlinear dynamic system model that takes into account the piping resistance (such as Poiseuille's law, Darcy-Weisbach equation), the actual efficiency of the pump (which varies with speed, pressure, and viscosity), the nonlinear effects of coating viscosity and temperature on flow and pressure (e.g., by introducing a Shear-thinning or Newtonian fluid model), and the compressibility of the coating in the piping. Based on the real-time coating state parameters provided by the material state estimation and pot life prediction module 133 (such as the estimated current viscosity, density), combined with the target coating flow, target pressure, and target mixing ratio for multi-component coatings set by the coating task setting, the MPC algorithm will predict the dynamic behavior of the coating supply system for a certain period of time in the future (e.g., a prediction horizon of 200 milliseconds to 500 milliseconds, containing 5-10 control steps).

[0056] Based on this prediction, the MPC algorithm determines the optimal sequence of control variables, such as the servo motor speed of the precision metering pump, the opening of key valves (e.g., backflow valve, stop valve), by solving an online optimization problem. The optimization objective is usually set to minimize the control error (e.g., the deviation of actual flow rate from target flow rate, the deviation of actual pressure from target pressure), overshoot, and wear on the actuators (e.g., pumps and valves) within the prediction horizon. Meanwhile, the optimization process strictly follows the system operation constraints (e.g., maximum / minimum speed of the precision metering pump, maximum / minimum pipeline pressure, maximum / minimum flow rate). This control strategy can compensate for fluctuations in the physical and chemical properties of the paint, such as viscosity and temperature, in real time, ensuring that the flow rate and pressure of the paint output to the spray gun remain at the set target values under various operating conditions. For example, when the rheological property perception module 120 and the material state estimation and pot life prediction module 133 detect an increase in paint viscosity, the MPC algorithm will use its internal model to predict that the flow rate will decrease or the pressure will increase at the existing pump speed, and will actively and proactively increase the servo motor speed of the corresponding precision metering pump 114 to accurately compensate for the change, thereby maintaining a constant target flow rate or target pressure. This adaptive control mechanism, which combines feedforward and feedback, significantly outperforms traditional PID control and can handle more complex system nonlinearities and external disturbances.

[0057] Dynamic calibration of mixing ratio module 135: This module is designed for two-component or multi-component reactive coatings and is crucial to ensuring the consistency of the final coating quality. It dynamically adjusts the output flow rate of each precision metering pump 114 based on the real-time viscosity and density estimates of each component (e.g., base A, curing agent B) provided by the material state estimation and pot life prediction module 133, as well as the pre-set mixing ratio (e.g., volume ratio A:B = 2:1 or weight ratio A:B = 100:30). Different components of the paint may have different viscosities and densities even at the same temperature, and these parameters may fluctuate with batch, storage time, or temperature. For example, if the viscosity of component A decreases slightly due to an increase in ambient temperature, resulting in a slight increase in flow rate at the same pump speed, while the viscosity of component B remains unchanged, this module will calculate and adjust the servo motor speed of the precision metering pump for component A or reduce its set flow rate in real time to ensure that the actual output flow rate ratio of components A and B strictly meets the pre-set mixing ratio. The calibration process usually involves the following steps: first, calculate the pumping efficiency curve of each component at the current viscosity and density; second, calculate the target volume flow rate or mass flow rate of each component based on the target mixing ratio and the required total flow rate; finally, convert this target flow rate into real-time speed instructions for each precision metering pump. This dynamic calibration mechanism overcomes the mixing deviation caused by differences or fluctuations in the rheological properties of each component, ensuring the consistency of the curing performance, mechanical strength, and final coating quality of reactive coatings, thereby significantly reducing material waste and product failure rates caused by inaccurate mixing.

[0058] Fault diagnosis and warning module 136: This module continuously monitors all sensor data (such as temperature, pressure, flow rate, viscosity) and actuator status (such as pump speed, motor current, valve opening, agitator status). It internally fuses multiple diagnostic techniques, for example, utilizing machine learning models trained on historical operation data (such as Support Vector Machine (SVM) combined with Radial Basis Function (RBF) kernel, or Long Short-Term Memory Network (LSTM) for time series anomaly detection) to identify abnormal conditions such as sensor drift, pump wear, internal pipe fouling or blockage, valve sticking or leakage, and servo motor overload, etc. For example, by analyzing the correspondence between pump motor current and set speed, output pressure, it can be determined whether the pump has abnormal wear or efficiency decline. At the same time, this module also contains a rule-based expert system that defines upper and lower threshold values, rate of change thresholds, and logic condition combinations for key parameters, which are used to identify more direct failure modes. When potential failures or system parameters deviate from the safe operating range, this module will immediately trigger multi-level audible and visual alarms, such as sounding an alarm through an industrial buzzer, and displaying detailed diagnostic information (such as fault code, fault location, recommended treatment measures) through warning lights or pop-up windows on the human-machine interface 160. In more serious cases, this module can automatically adjust the operating mode (such as switching to backup mode, reducing feed speed) or safety shutdown according to the pre-set fault handling strategy, to avoid further damage to the equipment, paint waste, or product quality problems. All fault events, diagnostic results, and handling actions are recorded in detail in the log file for subsequent analysis and maintenance.

[0059] The dynamic mixing and delivery unit 140 is used to precisely mix and deliver the paint components from different precision metering pumps 114 to the spray gun 143. For single-component paint systems, this unit can only contain the delivery pipeline and the spray gun 143. For two-component or multi-component paint systems, this unit further contains a dynamic mixer 141 and a mixed paint delivery pipeline 142. The dynamic mixer 141 uses a servo motor driven rotary mixer with multiple replaceable helical blades or stirring elements inside, such as inclined paddle, turbine paddle or dispersion disc, which can be replaced according to the viscosity and mixing requirements of the paint. The stirring speed of the mixer can be precisely controlled by the intelligent control and data processing unit 130, usually ranging from 50 RPM to 3000 RPM, to achieve different degrees of shear mixing. The pipeline of each component paint is connected to the inlet of the mixer respectively, and the flow is controlled by independent electromagnetic or pneumatic valves. The opening and closing of these inlet valves and the opening degree are precisely controlled by the intelligent control and data processing unit 130. The mixer can achieve sufficient and uniform mixing of each component paint by adjusting the stirring speed according to the type of paint and mixing requirements, while avoiding the negative impact of excessive shear on the performance of the paint (such as thixotropy and leveling). The mixer body is usually made of stainless steel or corrosion-resistant alloy, and can be equipped with a temperature control jacket to precisely control the temperature during mixing, because temperature has a significant impact on the curing rate of reactive paint.

[0060] At the outlet end of the dynamic mixer 141, a mixed paint temperature sensor 144 and an optional mixed paint viscosity sensor 145 are configured to verify the mixing effect and monitor the progress of the activation period in real time. The mixed paint temperature sensor 144 is similar to the previous temperature sensor type to ensure measurement accuracy. The mixed paint viscosity sensor 145 can be an online viscosity meter based on vibration principle, providing real-time viscosity data of the mixed paint as an additional input for verification of mixing uniformity and prediction of activation period. The mixed paint delivery pipeline 142 is made of materials with smooth inner walls, corrosion resistance and low adsorption, such as polytetrafluoroethylene (PTFE), ultra-high molecular weight polyethylene (UHMW-PE) lined pipe or PFA (perfluoroalkoxy) material, which can effectively reduce the residence, adsorption and shear degradation of the paint in the pipeline, ensuring the stability of the delivery and the original performance of the paint. The inner diameter and length of the pipeline are optimized according to the flow and installation space to minimize the pressure drop and dead zone. The spray gun 143 is a professional paint spray gun with high atomization efficiency, which can be air atomizing, airless atomizing or electrostatic spray gun. Its key parameters (such as spray width, flow, atomization pressure, fan angle) can be precisely adjusted by digital control valves or pressure regulators controlled by the intelligent control and data processing unit 130, or can be cooperatively controlled with external mechanical arms to achieve synchronous optimization of spraying path and spray gun parameters. The spray gun is usually equipped with quick disassembly and cleaning interfaces for easy maintenance.

[0061] The automated cleaning and maintenance module 150 aims to address the short activation period and easy solidification of multi-component coatings, and to achieve rapid and thorough cleaning of different coatings, thereby greatly improving the availability and production efficiency of the equipment. The module includes one or more cleaning liquid storage tanks 151, such as solvent (e.g. acetone, toluene), diluent or special water-based cleaning agent, to meet the cleaning needs of different coatings. Each cleaning liquid storage tank is equipped with a liquid level sensor and a liquid supplement interface. A high-pressure cleaning pump 152 is responsible for delivering cleaning liquid to the pipeline system, which can provide adjustable cleaning pressure of 0.5 MPa to 5 MPa and flow rate of 5 liters to 20 liters per minute to ensure sufficient flushing force and dissolution capacity. A series of electromagnetic switching valves 153 constitute a complex pipeline network, which are usually made of corrosion-resistant materials (such as PTFE, SS316L) with millisecond-level fast response speed and reliable zero-leakage sealing characteristics. The cleaning pipeline 154 is designed with strategic cross-connection points inside the coating supply pipeline and dynamic mixer, such as through three-way valves or multi-way valves, so that the cleaning liquid can be directed to inject into all paths of the coating flow. The waste liquid collection device 155 is used to collect residual coatings and cleaning liquids discharged during the cleaning process, which is usually equipped with a liquid level sensor and overflow protection.

[0062] When the material state estimation and activation period prediction module 133 of the intelligent control and data processing unit 130 predicts that the coating activation period is about to end (e.g. less than 5 minutes remaining), or the system is in a long idle state (e.g. more than 10 minutes of stop feeding), or the coating type needs to be switched, the automated cleaning and maintenance module 150 will automatically start according to the preset cleaning program. The cleaning program is usually a multi-stage, precise automated sequence, the steps of which are as follows:

[0063] First, by controlling the electromagnetic switching valve 153, the residual coating in the pipeline to be cleaned (especially the mixed coating in the dynamic mixer 141 and the mixed coating delivery pipeline 142) is emptied to the waste liquid collection device 155 by gravity or low-pressure air blowing to maximize the recovery or isolation of un-solidified coatings.

[0064] Second, start the high-pressure cleaning pump 152 to inject the selected first cleaning liquid (e.g. solvent) into the pipeline and the inside of the dynamic mixer 141 at high flow rate and high pressure. The cleaning liquid physically flushes and chemically dissolves the inner wall of the pipeline and the inside of the mixer through the preset cleaning pipeline 154 and the nozzles distributed in the key areas (such as the inside of the mixer, the inside of the spray gun), removing the residual coating. The flushing usually lasts for several minutes and can be repeated multiple times.

[0065] Subsequently, to ensure thorough cleaning and prevent contamination between different cleaning fluids or coatings, the system can perform multiple rounds of flushing with different cleaning fluids, for example, first with a strong solvent, then with a diluent, and finally with clean water or deionized water. After each round of cleaning, both the pipeline and the mixer can be introduced with clean compressed air for purging to expel residual cleaning fluids and accelerate the drying process, ensuring complete drying of the pipeline interior without residue.

[0066] This automated cleaning process greatly reduces the need for manual cleaning, reduces maintenance intensity and potential hazards, and ensures cleanliness during coating switching, thereby significantly improving equipment availability and production efficiency.

[0067] The working principle of the coating supply robot 100 of the present application is as follows:

[0068] At the beginning of coating supply, the intelligent control and data processing unit 130 loads the coating component information required for the current coating task, the target flow, the target pressure, and the accurate mixing ratio, etc. key parameters from the coating formula database through its human-computer interaction and system integration module 160 or external production management system. The precision metering pump 114 starts to deliver each coating component (such as the main coating, the curing agent) to the dynamic mixing and conveying unit 140 at the set initial flow rate under the accurate driving of the servo motor 115 according to these set parameters.

[0069] During the entire supply process, the rheological property sensing module 120 continuously monitors the dynamic viscosity, density, and temperature of each coating component or mixed coating flowing through the pipeline in real time and online. These high-frequency collected data are transmitted to the intelligent control and data processing unit 130 with millisecond-level delay through high-speed industrial communication interface.

[0070] The data preprocessing and fusion module 132 inside the intelligent control and data processing unit 130 performs real-time high-precision processing on the received multi-source heterogeneous sensor data (including rheological parameters, pumping pressure, temperature, pump speed feedback, etc.). By applying advanced algorithms such as Kalman filtering, this module can effectively filter out measurement noise and synchronize and fuse the data from different sensors with time stamps, generating reliable, low-noise real-time system state data, providing a solid foundation for subsequent intelligent decision-making.

[0071] Subsequently, the material state estimation and activation period prediction module 133 accurately estimates the real rheological state of the current coating, such as its actual viscosity value and shear characteristics at the current temperature, based on the fused data and the built-in coating physical and chemical model database. For two-component or multi-component reactive coatings, this module will combine real-time temperature and accumulated reaction time to accurately predict the remaining activation period of the mixed coating inside the pipeline and the spray gun, and continuously assess the potential curing risk, providing key early warning information for subsequent control and cleaning decisions.

[0072] An adaptive model predictive control (MPC) algorithm module 134 receives target parameters (such as target flow rate, pressure, mixing ratio) from the coating task setting and real-time coating state estimation values (especially the estimated viscosity value) from the material state estimation and pot life prediction module 133. The MPC algorithm calculates and optimizes the servo motor speed of each precision metering pump 114 and the opening of related valves (such as backflow valve, cut-off valve) according to the internal dynamic model of the coating supply system (which includes physical parameters such as pump characteristics, pipeline resistance, coating rheological behavior, etc.) online. For example, when it is detected that the coating viscosity decreases due to the increase in ambient temperature, the MPC algorithm can accurately predict that the flow rate will increase at the original pump speed and may cause overshoot, so it will actively and proactively reduce the pump speed to accurately maintain the set target flow rate, ensuring that the output coating amount always meets the process requirements. Conversely, when the viscosity increases, the pump speed will be strategically increased for compensation. For multi-component coating systems, the mixing ratio dynamic calibration module 135 will independently and synchronously adjust the speed of each precision metering pump according to the difference between the real-time viscosity and density estimation values of each component, to ensure the accuracy of the mixed coating component ratio, thereby completely eliminating the mixing deviation caused by the fluctuation of the physical properties of each component and ensuring the consistency of the curing performance and coating quality of the final coating.

[0073] After receiving each component coating, the dynamic mixer 141 in the dynamic mixing and conveying unit 140 stirs and mixes at a specific speed according to the instructions of the intelligent control and data processing unit 130, ensuring that each component is fully and uniformly mixed to form a mixed coating with stable performance. The mixed coating is conveyed to the spray gun 143, which applies precise coating in the best atomization state to complete the coating task.

[0074] When the coating task is completed, the system is in a long idle state, or the material state estimation and pot life prediction module 133 predicts that the pot life of the mixed coating is about to be exhausted (reaches the preset safety threshold), the intelligent control and data processing unit 130 will immediately trigger the automatic cleaning and maintenance module 150. This module automatically performs cleaning, emptying, and purging operations of the pipeline and mixer according to the preset and strictly verified cleaning program. For example, it will first empty the residual mixed coating, then inject cleaning liquid (such as solvent, diluent) through the electromagnetic switching valve 153 into the cleaning pipeline 154 and the inside of the mixer through the high-pressure cleaning pump 152 for circulating flushing, and then perform compressed air purging to dry the pipeline to prevent coating solidification and blockage, and to fully prepare for the next coating task or switching to a different coating, greatly reducing maintenance costs and downtime.

[0075] The human-machine interaction and system integration module 160 displays the system running status, each coating parameter (such as viscosity, temperature, flow rate, pressure), current mixing ratio, remaining pot life, and fault information, etc. to the operator in real time through an intuitive visual interface (usually a high-resolution industrial touch screen). The operator can set parameters, manage tasks, select cleaning programs, and manually operate through the interface. At the same time, the module exchanges data and instructions seamlessly with external robot controllers, manufacturing management systems (MES) or supervisory control and data acquisition systems (SCADA) through standard industrial communication protocols (such as Modbus TCP / IP, EtherNet / IP, OPC UA), achieving highly automated and digitalized management of the entire coating application process, and thus deeply integrating the coating supply robot into the intelligent manufacturing production line.

[0076] Example One: Precision feeding and pot life management of high viscosity two-component polyurethane coating

[0077] In a specific embodiment, a coating supply robot of the present application is applied to high-quality surface spraying of automobile parts, using a high-viscosity, fast-curing two-component polyurethane coating, the base material (component A) of which has an initial viscosity of 2000 mPa·s (at 25°C), the curing agent (component B) of which has an initial viscosity of 500 mPa·s (at 25°C), and the designed mixing ratio of which is A:B = 3:1 (volume ratio), and the mixed pot life of which is about 45 minutes at 25°C. The target total coating flow rate is set to 500 ml / min. The ambient temperature may fluctuate in the range of 20°C to 30°C during the coating process.

[0078] The main coating tank 111 and the curing agent tank 112 of the coating storage and supply unit 110 are each equipped with a heating jacket and a temperature control system, and the temperatures of components A and B are preset and stabilized at 25.0°C±0.1°C. The precision metering pump 114 selects a planetary gear pump with a volume of 1.0 ml / revolution, and the servo motor 115 thereof is equipped with a high-resolution rotary encoder 116 with 262144 pulses per revolution. The metering accuracy of the pump is calibrated to ensure that the flow rate repeatability can reach ±0.2% under different viscosities. The pump outlet pipeline is respectively installed with a Pt100 temperature sensor 117 and a piezoresistive pressure sensor 118, with measurement accuracies of ±0.05°C and ±0.1% FS, respectively.

[0079] The rheological property sensing module 120 is installed on the pipeline before the A and B components enter the dynamic mixer 141. It uses a MEMS torsional resonant viscosity sensor 121 with a viscosity measurement range of 100 mPa-s to 5000 mPa-s, a measurement accuracy of ±1.5% FS, and a response time of less than 50 milliseconds. A high-precision temperature sensor 122 (NTC thermistor array) is tightly integrated inside the viscosity sensor, with a measurement accuracy of ±0.05°C. All data are transmitted in real time to the intelligent control and data processing unit 130 through EtherCAT.

[0080] The intelligent control and data processing unit 130 uses an ARM Cortex-A72 quad-core processor running VxWorks RTOS with 16 GB RAM and 256 GB industrial-grade SSD. Its internal software algorithm module is implemented as follows:

[0081] The data preprocessing and fusion module 132 uses an extended Kalman filter to fuse data from viscosity, temperature, pressure sensors and pump encoders, reducing measurement noise by more than 50% and ensuring that all data have a time synchronization accuracy of within 10 microseconds.

[0082] The material state estimation and activation period prediction module 133 pre-stores the viscosity-temperature curves (fitted by the Vogel-Fulcher-Tammann equation) and Kamal-Sourour curing kinetics model parameters of the polyurethane coating at different temperatures. The module estimates the actual viscosity of each component in real time and predicts the remaining activation period of the mixed coating based on the real-time temperature and time cumulative effect of the mixed coating. When the predicted remaining activation period is less than 10 minutes, the system will issue a warning signal, and if it is less than 5 minutes, it will automatically trigger the cleaning program.

[0083] The adaptive model predictive control (MPC) algorithm module 134 runs with a control period of 50 milliseconds and a prediction horizon of 250 milliseconds. Its internal mathematical model includes the power-law rheological model of each component coating, the actual flow-rate-speed-viscosity curve of the pump, and the pipeline resistance model. The MPC algorithm optimizes the servo motor speed of the A and B component precision metering pumps online based on the real-time estimated viscosity value and the set target flow rate. For example, when the viscosity of component A decreases to 1800 mPa-s due to an increase in ambient temperature to 28°C, the MPC algorithm will predict an increase in its flow rate and immediately calculate and execute instructions to reduce the speed of the A component pump to maintain its target flow rate at 375 ml / min.

[0084] The dynamic proportioning calibration module 135 works with the MPC to dynamically adjust the rotation speed of each precision metering pump according to the real-time estimated viscosity values of each component (e.g. 1800 mPa·s for component A and 500 mPa·s for component B) to ensure that the actual volumetric flow ratio of components A and B is always maintained at 3:1, even if the viscosity of the components fluctuates inconsistently.

[0085] The dynamic mixing and delivery unit 140 contains a servo motor driven rotary dynamic mixer 141 with a helical blade design inside, with stirring rotation speed adjustable between 500-1500 RPM to accommodate different flow rates and viscosities. A mixed paint temperature sensor 144 is installed at the outlet end of the mixer to monitor the progress of the pot life after mixing. The mixed paint delivery pipeline 142 uses a PTFE lined tube with an inner diameter of 6 mm to minimize paint residue. The spray gun 143 is an air-assisted electrostatic spray gun, with both the atomization pressure and the fan width digitally adjustable by the intelligent control and data processing unit 130.

[0086] The automated cleaning and maintenance module 150 is equipped with two cleaning fluid storage tanks 151, storing acetone and a special diluent respectively. A high-pressure cleaning pump 152 can provide a cleaning pressure of 3 MPa. When the material state estimation and pot life prediction module 133 predicts that the remaining pot life is less than 5 minutes or the system is idle for more than 5 minutes, the system automatically starts the cleaning program: first, the remaining paint in the mixer is emptied, then it is flushed with acetone at a pressure of 3 MPa for 2 minutes, followed by flushing with diluent for 1 minute, and finally, it is dried by blowing clean compressed air for 30 seconds, the entire cleaning cycle can be completed within 5 minutes.

[0087] The present application is not limited to the above embodiments, and those skilled in the art can make various modifications or changes to the technical solutions of the present application according to the technical solutions and concepts of the present application, all of which should fall within the scope of protection of the present application. For example, the rheological sensor in the rheological property sensing module can use other non-contact or micro-contact sensors, such as sensors based on ultrasonic waves, microfluidic chips or capacitive principles. The control algorithm in the intelligent control and data processing unit can also use other advanced control theories, such as reinforcement learning control, fuzzy control, etc., to further optimize the control performance. The cleaning program and cleaning fluid types of the automated cleaning and maintenance module can also be adjusted according to the characteristics of the paint and environmental regulations. The various technical features disclosed in the present application can be combined with each other to form new embodiments without conflict. The materials, sizes, models, etc. of all components of the present application can be adjusted according to actual application requirements as long as their functions and effects meet the technical solutions of the present application.

[0088] Although the present application has been described in detail with reference to the foregoing embodiments, the technical solutions recorded in the foregoing embodiments can be modified, or some of the technical features can be replaced by equivalent features, by those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A paint feeding robot (100), comprising: The paint storage and supply unit (110) is provided with at least one main paint storage tank (111) and at least one curing agent storage tank (112) or diluent storage tank (113) for storing various paint components to be supplied. The outlet of each storage tank of the paint storage and supply unit (110) is connected to its corresponding precision metering pump (114) through an independent pipeline system for accurately outputting paint components at a set flow rate. The rheological property sensing module (120) is installed on the outlet pipeline of the precision metering pump (114) and is used to detect the rheological parameters of the coating components or mixed coatings in real time and online, and transmit real-time rheological parameters and temperature data to the intelligent control and data processing unit (130). The intelligent control and data processing unit (130), as the core intelligent decision-making center of the paint supply robot (100), is connected to the paint storage and supply unit (110) and the rheological property sensing module (120). It receives sensor data from the rheological property sensing module (120) and the precision metering pump (114), processes and analyzes the data, and controls the flow rate of the precision metering pump (114) based on the processing results to achieve precise paint supply. The dynamic mixing and conveying unit (140), connected to the paint storage and supply unit (110) and the intelligent control and data processing unit (130), is used to precisely mix paint components from different precision metering pumps (114) and convey them to the spray gun (143). The automated cleaning and maintenance module (150) is connected to the dynamic mixing and conveying unit (140) and the intelligent control and data processing unit (130), and is used to automatically clean and maintain the paint pipeline and the dynamic mixer (141) under the command of the intelligent control and data processing unit (130).

2. The paint feeding robot according to claim 1, characterized in that: The intelligent control and data processing unit (130) has a complex software algorithm module embedded within it, which includes: The data preprocessing and fusion module (132) is used to receive raw data from the rheological property sensing module (120), the pressure sensor and temperature sensor of the precision metering pump (114) and the high-resolution rotary encoder (116), and to perform noise reduction processing on the sensor data and multi-sensor data fusion by applying digital signal filtering algorithm to eliminate measurement noise and improve data accuracy and reliability. The material state estimation and activation period prediction module (133) accurately estimates the actual viscosity and density of the current coating based on the real-time data provided by the data preprocessing and fusion module (132) and the pre-stored coating physicochemical parameter database. For multi-component reactive coatings, it accurately predicts the remaining activation period of the coating based on the activation period characteristics of the mixed coating, real-time temperature and time cumulative effect. The adaptive model predictive control (MPC) algorithm module (134) uses the established mathematical model of the rheological properties of the coating and the dynamic response of the supply system. Based on the real-time coating state parameters provided by the material state estimation and activation period prediction module (133), combined with the preset target coating flow rate, pressure and mixing ratio, it predicts the dynamic behavior of the coating supply system in the future period and determines the optimal servo motor (115) speed of the precision metering pump (114) and related valve opening and other control variables. The mixing ratio dynamic calibration module (135), designed specifically for two-component or multi-component coatings, dynamically adjusts the output flow rate of each precision metering pump (114) based on the real-time viscosity and density estimates of each component provided by the material state estimation and activation period prediction module (133) and the preset mixing ratio; and The fault diagnosis and early warning module (136) continuously monitors all sensor data and actuator status. It uses a machine learning model trained based on historical data or a rule-based expert system to identify abnormal conditions such as sensor drift, pump wear, pipeline blockage, and valve failure. When a potential fault is detected or system parameters deviate from the safe range, it triggers an alarm or automatically adjusts the operating mode.

3. The paint feeding robot according to claim 1, characterized in that: The rheological property sensing module (120) includes one or more non-contact or micro-contact rheological sensors (121) and a high-precision temperature sensor (122). The rheological sensor (121) is a microelectromechanical system (MEMS) viscosity sensor based on the torsional resonance principle. Its resonant element is in contact with the fluid. The dynamic viscosity of the fluid is calculated by measuring the response of the resonant frequency, damping coefficient or Q value to the changes in fluid viscosity and density. The MEMS viscosity sensor has the characteristics of small size, fast response and high integration. Its viscosity measurement range is 0.1 mPa·s to 10000 mPa·s, the measurement accuracy is within ±2%FS, and the response time is less than 100 milliseconds. The high-precision temperature sensor (122) uses a thermistor array to compensate for the influence of temperature on fluid viscosity during viscosity measurement.

4. The paint feeding robot according to claim 1, characterized in that: The precision metering pump (114) of the paint storage and supply unit (110) is a positive displacement pump driven by a servo motor (115), such as a planetary gear pump or a progressive cavity pump; the servo motor (115) is equipped with a high-resolution rotary encoder (116) to achieve precise control of pump speed and flow feedback; the metering accuracy of the precision metering pump (114) can reach within ±0.5%, and the repeatability error is less than 0.2%; a metering pump temperature sensor (117) and a metering pump pressure sensor (118) are configured on the inlet or outlet pipe of the precision metering pump (114) to monitor the initial state of the paint components and pressure fluctuations during the pumping process in real time.

5. A paint feeding robot according to claim 2, characterized in that: The material state estimation and activation period prediction module (133) is based on real-time collected coating temperature, viscosity, and density data, as well as a pre-stored coating physicochemical parameter database. The database contains viscosity-temperature curves, density-temperature curves, and curing kinetic parameters of different batches and formulations of coatings at different temperatures. The module uses recursive least squares or a nonlinear state observer to accurately estimate the true viscosity, density, and potential shear thinning or thixotropy of the current coating. For two-component or multi-component reactive coatings, the module accurately predicts the remaining activation period of the coating in the pipeline and inside the spray gun (143) based on the activation period characteristics of the mixed coating, real-time temperature, and time accumulation effect, by integrating chemical reaction kinetics and energy accumulation models, and assesses potential curing risks. The prediction results will serve as the key basis for triggering the automated cleaning and maintenance module (150) or adjusting the feeding speed.

6. A paint feeding robot according to claim 2, characterized in that: The adaptive model predictive control (MPC) algorithm module (134) utilizes an established mathematical model of the rheological properties of the coating and the dynamic response of the supply system. The model considers pipeline resistance, the efficiency of the precision metering pump (114), and the nonlinear effects of coating viscosity and temperature on flow rate and pressure. The MPC algorithm module (134) predicts the dynamic behavior of the coating supply system over a future period based on the real-time coating state parameters provided by the material state estimation and activation period prediction module (133), combined with preset target coating flow rate, pressure, and mixing ratio. Based on the prediction, the MPC algorithm solves an online optimization problem to determine the optimal control variables such as the servo motor (115) speed and valve opening of the precision metering pump (114), so as to minimize the control error and overshoot in the prediction time domain and satisfy the system operation constraints. The algorithm can compensate for the fluctuations in the physical and chemical properties of the coating such as viscosity and temperature in real time.

7. A paint feeding robot according to claim 1, characterized in that: The automated cleaning and maintenance module (150) includes one or more cleaning fluid storage tanks (151), a high-pressure cleaning pump (152), a series of electromagnetic switching valves (153), and dedicated cleaning pipelines (154) and waste liquid collection devices (155); the cleaning pipelines (154) are internally designed with cross-connection points with the paint supply pipelines and the dynamic mixer (141); when the intelligent control and data processing unit (130) predicts that the paint activation period is about to end, or the system is in a long-term idle state, When it is necessary to switch the type of coating, the automated cleaning and maintenance module (150) will automatically start according to the preset cleaning program; the cleaning program includes: draining the residual coating in the pipeline to be cleaned to the waste liquid collection device (155) through the electromagnetic switching valve (153); starting the high-pressure cleaning pump (152) to inject the cleaning liquid into the pipeline and the inside of the dynamic mixer (141) at high flow rate and high pressure to remove the residual coating; and being able to perform multiple rounds of rinsing with different cleaning liquids, and being able to introduce compressed air for purging.

8. A paint feeding robot according to claim 1, characterized in that: The dynamic mixing and conveying unit (140) includes a dynamic mixer (141) and a mixed coating conveying pipeline (142). The dynamic mixer (141) is a rotary mixer driven by a servo motor, which is equipped with multiple spiral blades or stirring elements inside. The stirring speed can be precisely controlled by the servo motor to achieve full and uniform mixing of each component coating. At the outlet end of the dynamic mixer (141), a mixed coating temperature sensor (144) and / or a mixed coating viscosity sensor (145) are configured to verify the mixing effect and monitor the progress of the activation period. The mixed coating conveying pipeline (142) is made of a material with a smooth inner wall, corrosion resistance, and low adsorption, such as polytetrafluoroethylene (PTFE) or ultra-high molecular weight polyethylene (UHMW-PE) lining pipe. The spray gun (143) is a professional coating spray gun with high atomization efficiency, and its parameters can be digitally adjusted by the intelligent control and data processing unit (130).

Images