A high-pressure spraying system and method for unmanned aerial vehicles based on a reciprocating oscillating drive mechanism

Abstract

This invention discloses a high-pressure spraying system and method for unmanned aerial vehicles (UAVs) based on a reciprocating swing drive mechanism. The system includes a fuselage linkage, a fixed frame, a drive assembly, a transmission assembly, a high-pressure direct-injection nozzle assembly, a housing, and a water pipe. The fixed frame is connected to the UAV fuselage via the fuselage linkage. The drive assembly and transmission assembly are both mounted on the fixed frame, and the housing covers the outside of the drive assembly and transmission assembly. One end of the water pipe is connected to a water source, and the other end is connected to the high-pressure direct-injection nozzle assembly. The drive assembly includes a motor and a motor rocker arm, with the motor's output end fixedly connected to one end of the motor rocker arm. The transmission assembly includes a rotating shaft and a nozzle rocker arm. The rotating shaft is fixedly mounted on the other end of the motor rocker arm, and the nozzle rocker arm has a groove adapted to the rotating shaft, allowing the rotating shaft to slide within the groove. The high-pressure direct-injection nozzle assembly is fixedly connected to the end of the nozzle rocker arm furthest from the groove. This invention not only meets the requirements of high-pressure rinsing but also significantly improves operational efficiency and scene adaptability.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) technology, and specifically to a high-pressure spraying system and method for UAVs based on a reciprocating oscillating drive mechanism. Background Technology

[0002] In recent years, the application of drones in agricultural and forestry plant protection, municipal operation and maintenance, and industrial inspection has rapidly become widespread, with spraying drones being one of the most widely used categories. Currently, most mainstream spraying drones are designed for plant protection scenarios, using low-pressure atomizing nozzles with water pressure typically below 1MPa. They are mainly used for atomizing and spraying pesticides and foliar fertilizers, and the water impact force is extremely low, making them completely unsuitable for cleaning operations that require high-impact water jets, such as washing dust off photovoltaic panels and removing stains from exterior walls.

[0003] To meet the needs of cleaning operations, some drone solutions equipped with high-pressure nozzles have emerged in the industry, but they generally suffer from the following technical shortcomings: Limited spraying range: Most existing high-pressure cleaning drones use fixed high-pressure nozzles, which can only cover a very small area with a single spray, resulting in very low operating efficiency and the problem of missed spraying.

[0004] Design flaws of the swing mechanism: In order to expand the spraying range, some existing solutions have added a swing drive mechanism to the nozzle. However, they generally adopt complex structures such as multi-motor linkage and hydraulic drive. Not only is the weight too large, which greatly reduces the payload and endurance space of the drone, but it is also prone to jamming and high failure rate, making it difficult to adapt to complex outdoor operating environments.

[0005] The operating parameters cannot be adaptive: the water pressure and swing parameters of the existing solution are mostly preset manually, and cannot be automatically adjusted according to the material of the working surface and the type of dirt. This can easily lead to problems such as excessive water pressure damaging the working surface (such as flexible photovoltaic panels and exterior wall coatings) and insufficient water pressure failing to remove stubborn stains, resulting in poor cleaning effect consistency.

[0006] Unstable output pressure: When working at heights, differences in atmospheric pressure at different working heights and fluctuations in pump efficiency caused by changes in the remaining water volume of the water supply source can cause the actual water pressure at the nozzle to deviate from the preset value, further affecting the stability of the cleaning effect. Most existing solutions lack a targeted dynamic calibration mechanism. Summary of the Invention

[0007] This invention aims to solve at least one of the technical problems existing in the prior art. To this end, this invention proposes a high-pressure spraying system and method for drones based on a reciprocating oscillating drive mechanism. The system achieves stable reciprocating oscillation of the nozzle through a crank-slider mechanical structure, and combines multi-scenario parameter adaptive adjustment and dynamic pressure calibration mechanisms. While meeting the high-pressure flushing requirements, this significantly improves operational efficiency and scene adaptability, and reduces equipment weight and failure rate.

[0008] To solve the above problems, the technical solution adopted by the present invention is as follows: A high-pressure spraying system for unmanned aerial vehicles (UAVs) based on a reciprocating swing drive mechanism includes a fuselage link, a fixed frame, a drive assembly, a transmission assembly, a high-pressure direct injection nozzle assembly, a housing, and water pipes. The mounting frame is connected to the drone body via a fuselage linkage. The drive and transmission components are both mounted on the mounting frame, and the outer shell covers the outside of the drive and transmission components. One end of the water pipe is connected to the water supply source, and the other end is connected to the high-pressure direct injection nozzle assembly. The drive assembly includes a motor and a motor rocker arm, with the output end of the motor fixedly connected to one end of the motor rocker arm; the transmission assembly includes a rotating shaft and a nozzle rocker arm, with the rotating shaft fixedly mounted on the other end of the motor rocker arm, and the nozzle rocker arm having a groove adapted to the rotating shaft, with the rotating shaft slidably embedded in the groove; the high-pressure direct injection nozzle assembly is fixedly connected to the end of the nozzle rocker arm away from the groove. The motor rotation drives the motor rocker arm to rotate around the motor output end. The rotating shaft slides in the slide groove and pushes the nozzle rocker arm to swing back and forth. The high-pressure direct injection nozzle assembly swings back and forth synchronously with the nozzle rocker arm. The high-pressure direct injection nozzle assembly and motor dynamically adjust the output water pressure and speed according to the material type and dirt type of the area to be cleaned.

[0009] Preferably, the high-pressure direct injection nozzle assembly includes a direct injection nozzle body and a high-pressure connector. One end of the high-pressure connector is connected to the direct injection nozzle body, and the other end is sealed to a water pipe. The water outlet end of the direct injection nozzle body has a constricted structure.

[0010] Preferably, the motor is a stepper motor with adjustable speed, used to adjust the reciprocating swing frequency of the nozzle rocker arm.

[0011] A high-pressure spraying method for drones based on a reciprocating oscillating drive mechanism, employing the aforementioned high-pressure spraying system for drones, includes the following steps: The high-pressure spraying system of the drone is detachably and fixedly connected to the drone body via a fuselage linkage, and the water pipes are sealed and connected to the water supply source and the high-pressure direct spray nozzle assembly respectively. The remote-controlled drone flies to the area to be cleaned and starts the water supply source to supply water to the high-pressure direct spray nozzle assembly at a preset water pressure value, so that it outputs a high-pressure concentrated water jet. Start the motor and drive the motor rocker arm to rotate around its output end at a preset speed. Drive the rotating shaft at the end of the motor rocker arm to slide in the slide groove of the nozzle rocker arm, converting the rotational motion of the motor into the reciprocating oscillating motion of the high-pressure direct injection nozzle assembly, so that the high-pressure concentrated water jet reciprocates and sweeps the area to be cleaned. The system identifies the material type and dirt type of different areas to be cleaned, and adjusts the water pressure based on the current actual water supply pressure to adapt to different material types. It matches the corresponding target oscillation frequency by using a dirt-frequency mapping table pre-stored in the UAV control system, and adjusts the motor from the preset speed to the corresponding speed to match the target oscillation frequency.

[0012] Preferably, when supplying water to the high-pressure direct injection nozzle assembly at a preset water supply pressure value, the process includes: The system acquires the drone's current operating altitude and the remaining water volume of the water source in real time. Based on the atmospheric pressure difference corresponding to the operating altitude and the deviation of the water pump output efficiency corresponding to the remaining water volume, the preset water supply pressure value is dynamically compensated and calibrated to serve as the current actual water supply pressure value.

[0013] Preferably, dynamic compensation calibration includes: Multiple sets of altitude pressure compensation benchmark values ​​corresponding to the operating altitude range and multiple sets of water pump efficiency compensation coefficients corresponding to the remaining water volume range of the water supply source are stored in the UAV control system in advance. Based on the current operating altitude of the drone, the corresponding altitude pressure compensation benchmark value is matched, and the preset water supply pressure value is corrected for the first time to obtain the initial calibration pressure value. Based on the current remaining water volume of the water supply source, the pump efficiency compensation coefficient for the corresponding range is matched, and the initial calibration pressure value is corrected a second time to obtain the calibrated pressure value.

[0014] Preferably, when identifying material type and dirt type, the following are included: Pre-store a library of material-dirt related features in the drone control system; The visible light images and spectral reflectance data of the area to be cleaned are collected using image acquisition devices and spectral sensors mounted on the drone. Texture features and spectral features are extracted from visible light images and spectral reflectance data, respectively, and then matched with surface texture features and spectral reflectance features in the material-dirt association feature library to obtain the material type and preliminary dirt type. Call the identification confidence threshold for dirt under the corresponding material. If the initial judgment of the dirt type matching degree is higher than the identification confidence threshold, the final dirt identification result is directly output. If the matching degree is lower than the identification confidence threshold, control the drone to approach the area to be cleaned until it reaches the preset close identification distance, re-collect spectral reflectance data for secondary identification, until the matching degree meets the identification confidence threshold requirement.

[0015] Preferably, when adjusting water pressure to suit different material types, the following are included: A material-water pressure mapping table is pre-stored in the drone control system, which stores the maximum water pressure tolerance threshold and the optimal cleaning water pressure range for different material types. Based on the identified material type, the corresponding optimal cleaning water pressure range is selected, and the median value of the range is used as the initial adjustment water pressure value. Based on the identified type of fouling, a corresponding offset correction amount is generated, and the initial regulating water pressure value is offset to obtain the target regulating water pressure value. The pressure deviation between the target regulating water pressure value and the preset water supply pressure value is obtained, and the pressure deviation value is added to the current actual water supply pressure value to complete the water pressure regulation. Among them, the superimposed water pressure value does not exceed the highest water pressure threshold of the corresponding material.

[0016] Preferably, when generating the corresponding offset correction amount, the following is included: A dirt-water pressure offset mapping table is pre-stored in the UAV control system, which stores the basic water pressure offset corresponding to different dirt types; Based on the identified current type of fouling, match the corresponding baseline water pressure offset; The crosswind speed and flight attitude tilt angle parameters of the UAV are collected in real time. Based on the water column offset corresponding to the crosswind speed and the nozzle water outlet angle deviation corresponding to the flight attitude tilt angle, the basic water pressure offset is calibrated a second time to obtain the final offset correction amount.

[0017] Preferably, when performing a secondary calibration of the basic water pressure offset, the following steps are included: Real-time acquisition of crosswind speed values ​​in the current operating environment of the drone, and acquisition of the current roll angle and pitch angle of the drone as flight attitude tilt parameters; The mapping relationship between crosswind speed and water pressure compensation coefficient, as well as the mapping relationship between water outlet angle deviation and water pressure correction coefficient under different flight attitude tilt angles, are stored in the UAV control system in advance. The crosswind speed value collected in real time is matched to obtain the corresponding first calibration coefficient, and the flight attitude tilt angle parameter collected is matched to obtain the corresponding second calibration coefficient. The basic water pressure offset is calibrated using the first calibration coefficient and the second calibration coefficient to obtain the final offset correction.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention employs a crank-slider transmission structure driven by a single stepper motor. Through the mechanical cooperation of the motor rocker arm, rotating shaft, and nozzle rocker arm with a sliding groove, the rotational motion of the motor is converted into a stable reciprocating oscillation of the nozzle. This eliminates the need for complex structures such as multi-motor linkage and hydraulic drives, reducing the overall weight by more than 40% compared to existing oscillation solutions. This significantly reduces the impact on the drone's payload and endurance. The purely mechanical transmission structure, without complex electrical or hydraulic components, offers enhanced resistance to sandstorms and water mist interference during outdoor operations, reducing the failure rate by more than 60% and making it suitable for complex outdoor work scenarios. Furthermore, the groove length can be flexibly customized, allowing adjustment of the nozzle oscillation amplitude without modifying the overall structure, adapting to different spray width requirements and enhancing its versatility.

[0019] The high-pressure direct-injection nozzle, featuring a constricted outlet structure, achieves increased water pressure through a hydrodynamic structure, delivering a concentrated water jet with high impact force. The water pressure can reach 1~15MPa, which is more than 10 times that of traditional atomizing nozzles for agricultural drones. It can meet the high-pressure cleaning needs of operations such as washing dust from photovoltaic panels, removing stains from exterior walls, and cleaning municipal facilities. Combined with the nozzle's reciprocating oscillating design, the coverage area of ​​a single spray is 3~5 times that of a fixed high-pressure nozzle. It retains the strong flushing power of the high-pressure water jet while avoiding the problems of narrow coverage, missed sprays, and low work efficiency of fixed nozzles, greatly improving work efficiency.

[0020] With a built-in image and spectral multi-sensor fusion recognition mechanism, it can automatically identify the material type and dirt type of the working surface, and automatically match the corresponding optimal water pressure and nozzle oscillation frequency: for fragile materials (such as flexible photovoltaic panels and exterior wall paint), it automatically limits the maximum water pressure to avoid damage; for stubborn stains, it automatically reduces the oscillation frequency to extend the water column residence time and improve the rinsing force; for light dust, it automatically increases the oscillation frequency to speed up the operation. It can adapt to the needs of all scenarios without manual adjustment of parameters, which not only avoids damage to the working surface caused by human operation errors, but also ensures the consistency of cleaning effect in different scenarios.

[0021] A dual-pressure dynamic calibration mechanism was constructed: on the one hand, the preset water pressure can be calibrated in two dimensions according to the real-time operating height and the remaining water volume of the water supply source, so as to offset the efficiency fluctuation caused by the difference in atmospheric pressure at high altitude and the remaining water volume of the water pump, and control the water pressure fluctuation within ±5% throughout the entire operation cycle; on the other hand, the water pressure can be compensated a second time according to the real-time crosswind speed and the drone's flight attitude tilt angle, so as to offset the water column energy loss caused by crosswind and attitude changes, further ensuring that the actual output pressure of the nozzle accurately matches the working condition requirements, and avoiding the problem of unstable cleaning effect caused by pressure fluctuation.

[0022] The entire spraying system is a modular design that can be attached to drones, allowing for quick assembly and disassembly. When not performing cleaning operations, the modules can be removed so that the drones can perform other tasks such as plant protection and inspection, effectively improving the drone reuse rate.

[0023] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0024] Figure 1 This is an overall structural diagram of the high-pressure spraying system for unmanned aerial vehicles (UAVs) according to an embodiment of the present invention; Figure 2 This is a cross-sectional view of the high-pressure spraying system for unmanned aerial vehicles according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating the steps of a high-pressure spraying method using a drone according to an embodiment of the present invention; Figure 4 This is a flowchart of the dynamic compensation calibration process for the preset water supply pressure value according to an embodiment of the present invention; Figure 5 This is a flowchart illustrating the identification process for material type and dirt type in an embodiment of the present invention.

[0025] The following are the reference numerals: 1. Body connecting rod; 2. Fixing frame; 3. Outer shell; 4. Water pipe; 5. Motor; 6. Motor rocker arm; 7. Rotating shaft; 8. Nozzle rocker arm; 9. Slide groove; 10. High-pressure direct injection nozzle assembly; 11. Direct injection nozzle body; 12. High-pressure connector. Detailed Implementation

[0026] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0027] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0028] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0029] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0030] Example 1, see Figure 1 , 2 The present invention provides an overall structural diagram and cross-sectional view of a high-pressure spraying system for unmanned aerial vehicles (UAVs). Figure 1 , 2The high-pressure spraying system for a drone based on a reciprocating swing drive mechanism shown includes a fuselage link 1, a fixed frame 2, a drive assembly, a transmission assembly, a high-pressure direct spray nozzle assembly 10, a housing 3, and a water pipe 4. The mounting frame 2 is connected to the drone body via the fuselage linkage 1. The drive assembly and transmission assembly are both mounted on the mounting frame 2. The outer shell 3 covers the outside of the drive assembly and transmission assembly. One end of the water pipe 4 is connected to the vehicle-mounted water supply source of the drone, and the other end is connected to the high-pressure direct injection nozzle assembly 10. The drive assembly includes a motor 5 and a motor rocker arm 6, with the output end of the motor 5 fixedly connected to one end of the motor rocker arm 6; the transmission assembly includes a rotating shaft 7 and a nozzle rocker arm 8, with the rotating shaft 7 fixedly mounted on the other end of the motor rocker arm 6, and the nozzle rocker arm 8 having a groove 9 adapted to the rotating shaft 7, with the rotating shaft 7 slidably embedded in the groove 9; the high-pressure direct injection nozzle assembly 10 is fixedly connected to the end of the nozzle rocker arm 8 away from the groove 9; The rotation of motor 5 drives motor rocker arm 6 to rotate around the output end of motor 5. The rotating shaft 7 slides in the slide groove 9 and pushes the nozzle rocker arm 8 to swing back and forth. The high-pressure direct injection nozzle assembly 10 swings back and forth synchronously with the nozzle rocker arm 8. The water output of the high-pressure direct injection nozzle assembly 10 is a concentrated water column. Among them, the high-pressure direct injection nozzle assembly 10 and the motor 5 dynamically adjust the output water pressure and speed according to the material type and dirt type of the area to be cleaned.

[0031] Specifically, the entire spraying system serves as an external functional module for the drone. A mounting frame 2, which carries the functional components, is connected to the main body of the drone via a fuselage linkage 1. An outer shell 3 protects the drive and transmission components. The water supply path is as follows: a high-pressure water source mounted on the drone is directly connected to the high-pressure direct-spray nozzle assembly 10 via a water pipe 4, creating a concentrated, high-impact water jet. This solves the problem of insufficient water pressure in traditional agricultural drone atomization spraying, which cannot meet the needs of cleaning operations.

[0032] The system uses a crank-slider type transmission structure to convert the rotational motion of motor 5 into the reciprocating oscillation of the nozzle: Power input: The stepper motor of the drive component outputs rotational power, which drives the motor rocker arm 6, which is fixed to its output end, to perform circular rotational motion; Motion conversion: The rotating shaft 7 fixed at the end of the motor rocker arm 6 is embedded in the long strip groove of the nozzle rocker arm 8. When the motor rocker arm 6 rotates, the rotating shaft 7 will slide back and forth in the groove 9, and at the same time drive the nozzle rocker arm 8 to make periodic left and right reciprocating swings with the support point that is hinged to the fixed frame 2 as the fulcrum. Action output: The high-pressure direct injection nozzle assembly 10 is fixed to the swing end of the nozzle rocker arm 8 and performs reciprocating sweeping motion synchronously with the rocker arm.

[0033] During operation, the drone flies to the area to be cleaned (such as photovoltaic panels, building exterior walls, etc.), the water supply system supplies high-pressure water to the nozzles, and the nozzles output concentrated high-pressure water jets to wash away the stains. Motor 5 drives the nozzle to swing back and forth, allowing the high-pressure water jet to form a sweeping path on the working surface. This retains the cleaning impact of the high-pressure water jet while greatly expanding the coverage area of ​​a single spray, solving the problems of narrow spray range and low work efficiency of fixed high-pressure nozzles. The cleaning effect can be optimized for different scenarios by adjusting the stepper motor speed and the nozzle oscillation frequency, and by dynamically matching the water pressure parameters according to the surface material and dirt type.

[0034] In one possible embodiment, the groove 9 is a long strip-shaped through groove, the extension direction of the long strip-shaped groove is matched with the length direction of the nozzle rocker arm 8, and the sliding stroke of the rotating shaft 7 in the long strip-shaped groove is adapted to the reciprocating swing angle of the nozzle rocker arm 8.

[0035] Specifically, the slide 9 is the core adapter structure that realizes the motion conversion of "motor 5 rotational motion → nozzle reciprocating oscillation", and is used to eliminate the path difference between circular motion and oscillation motion: The motor rocker arm 6 drives the rotating shaft 7 in a complete circular motion. The motion of the rotating shaft 7 includes both tangential displacement (perpendicular to the motor rocker arm 6) and radial displacement (along the length of the motor rocker arm 6). If the slide groove 9 were a fixed circular hole, the radial displacement would be blocked by the structure, and power could not be transmitted. However, the elongated through groove that matches the length of the nozzle rocker arm 8 provides sliding space for the radial displacement of the rotating shaft 7, transmitting only the tangential thrust to the nozzle rocker arm 8, driving it to swing smoothly around the hinge point, and avoiding problems such as structural jamming and stress concentration that could damage components.

[0036] The length of the slide groove 9 (i.e., the sliding stroke of the rotating shaft 7) directly determines the maximum oscillation amplitude of the nozzle: When the motor rocker arm 6 rotates, the rotating shaft 7 will slide to both ends of the slide groove 9 respectively. At this time, the nozzle rocker arm 8 just reaches the designed maximum left and right swing angle. Depending on the required spray coverage width, the length of the corresponding chute 9 is matched: if a larger spray range is required, the length of the chute 9 is increased to expand the sliding stroke of the rotating shaft 7 and increase the nozzle swing angle; if a small area of ​​precise spraying is required, the length of the chute 9 is shortened to limit the swing amplitude.

[0037] This invention can precisely limit the swing range of the nozzle using only the mechanical structure of the slide groove 9, without the need for additional electronic limiters, angle sensors, or other components. It has a simple and reliable structure, low cost, and low failure rate. Furthermore, the parameters of the slide groove 9 can be flexibly customized according to different operating scenarios to adapt to different spraying needs.

[0038] In one possible embodiment, the high-pressure direct injection nozzle assembly 10 includes a direct injection nozzle body 11 and a high-pressure connector 12. One end of the high-pressure connector 12 is connected to the direct injection nozzle body 11, and the other end is sealed to the water pipe 4. The water outlet end of the direct injection nozzle body 11 has a constricted structure to increase the water outlet pressure.

[0039] Specifically, the high-pressure direct injection nozzle assembly 10 is the core execution component for realizing the output of high-pressure concentrated water jets. It solves the pain points of traditional agricultural drone nozzles, such as severe atomization and insufficient water impact force, which cannot meet the needs of cleaning operations. Ultimately, it outputs a concentrated water jet with high impact force and no dispersion, which is suitable for the stain washing needs of photovoltaic panels, building exterior walls and other scenarios.

[0040] The high-pressure connector 12 serves as a transitional connection structure between the water supply system and the nozzle body. On the one hand, it achieves high-pressure sealing: preventing high-pressure water flow from leaking or depressurizing at the connection point and ensuring that the pressure at the water supply end can be fully transmitted to the nozzle flow channel. On the other hand, it achieves structural rigidity: firmly connecting the flexible high-pressure wear-resistant water pipe to the rigid nozzle body, preventing the water pipe 4 from being pulled loose or falling off when the nozzle swings back and forth. At the same time, the inner diameter of the flow channel is perfectly matched with the water pipe 4 and the nozzle body, avoiding additional water flow resistance and pressure loss caused by sudden changes in pipe diameter.

[0041] The constriction structure utilizes Bernoulli's principle of fluid mechanics to achieve pressure boosting: the main flow channel inside the nozzle is designed with a constant diameter, maintaining a stable high pressure state within the channel. When the water flows to the constriction position at the outlet, the flow cross-sectional area suddenly shrinks. Under the same water supply flow conditions, the flow velocity increases dramatically, and a large amount of the static pressure is converted into dynamic pressure. The final sprayed water column possesses extremely high impact kinetic energy. Furthermore, the constriction's guiding effect prevents the water from scattering and atomizing, forming a concentrated columnar jet, significantly improving flushing capacity.

[0042] The direct injection nozzle body 11 achieves increased water pressure through mechanical structure design alone, without the need to increase the output power of the water supply pump, thus reducing the load pressure on the power supply and water supply system of the drone; at the same time, it can be flexibly adapted to different scenarios by adjusting the orifice diameter parameters of the constriction.

[0043] In one possible embodiment, the fixed frame 2 is provided with a swing support seat, and the end of the nozzle rocker arm 8 away from the slide groove 9 is hinged to the swing support seat. The nozzle rocker arm 8 swings back and forth with the hinge point as the fulcrum.

[0044] Specifically, the swing support is the mechanical basis for the stable reciprocating swing of the nozzle rocker arm 8. It is used to limit the movement trajectory of the rocker arm and ensure the stability of power transmission and the accuracy of spraying action.

[0045] The swing support base, as a rigid support component fixed on the system mounting frame 2, bears the weight of the nozzle rocker arm 8 and the nozzle assembly. At the same time, it provides a fixed positioning reference for the swing of the rocker arm, preventing the rocker arm from shifting or swaying in the front-back or up-down directions during movement. This ensures that the spraying direction and swing path of the nozzle always meet the design expectations and prevents problems such as spraying deviation and loss of control of the working range.

[0046] The end of the nozzle rocker arm 8 away from the slide groove 9 is hinged to the swing support (i.e., connected by a rotatable pin). On the one hand, it can firmly fix the end position of the rocker arm, and on the other hand, it completely releases the rotational freedom of the rocker arm around the hinge point, which will not hinder the swing of the rocker arm, ensure the smooth power transmission, and avoid structural jamming.

[0047] The hinge fulcrum is set at the end of the nozzle rocker arm 8 away from the slide 9 (i.e. away from the power input end). The thrust input by the rotating shaft 7 at the end of the slide 9 will form a rotational torque around the hinge point, which drives the entire rocker arm to swing using the lever principle. When the motor 5 drives the rotating shaft 7 to push the end of the slide 9 to move, the rocker arm will swing in an arc with the hinge point as the center. The power transmission efficiency is high, and the swing trajectory is completely fixed, and the control precision of the swing angle is higher.

[0048] The above structure is simple and reliable. It achieves both stable fixation and smooth swing of the rocker arm through a combination of rigid support and hinge. It does not require additional complex guide structures, has a low failure rate, and is easy to maintain and disassemble later.

[0049] In one possible embodiment, the motor 5 is a stepper motor with adjustable speed, used to adjust the reciprocating swing frequency of the nozzle rocker arm 8.

[0050] Specifically, stepper motors are special motors with extremely high controllability and precision. The output speed can be precisely controlled by adjusting the frequency of the input pulse signal, and the speed is not affected by load changes (such as changes in nozzle water pressure, airflow disturbances, etc.). They do not have the speed drift problem of ordinary DC motors, and can ensure that the nozzle oscillation frequency is stable and controllable for a long time, avoiding problems such as uneven spray coverage and missed sprays.

[0051] The stepper motor's rotational speed is linearly positively correlated with the nozzle's oscillation frequency. Each time the stepper motor completes a full 360° rotation, it drives the motor rocker arm 6 to rotate synchronously one revolution, which corresponds to driving the nozzle rocker arm 8 to complete one complete reciprocating swing (left swing to the limit → right swing to the limit → reset complete cycle). Therefore, the higher the speed of the motor 5, the more rotations it completes per unit time, the higher the oscillation frequency of the nozzle, and the more sweeping times per unit time.

[0052] By adjusting the motor speed, the oscillation frequency can be changed to flexibly match different operational needs. When there are stubborn stains on the working surface, the motor speed can be reduced by 5 and a low oscillation frequency can be used to allow the high-pressure water jet to stay at a single point for a longer time, thereby increasing the rinsing power and ensuring the cleaning effect. When the working surface is covered with light pollution such as floating dust, the motor speed can be increased by 5, a high oscillation frequency can be used to increase the sweeping speed and expand the working coverage area per unit time. It can also be linked with the drone flight control system to automatically match the swing frequency according to the drone's flight speed: the faster the flight speed, the higher the swing frequency should be to avoid problems such as spray gaps and missed sprays.

[0053] This invention can adjust the swing frequency simply by adjusting the electrical control parameters, without modifying the mechanical structure such as the slide 9 and the rocker arm. It has low adaptation cost and high flexibility, and can also realize digital control of operating parameters to adapt to the operating needs of different scenarios.

[0054] In one possible embodiment, the water pipe 4 is a high-pressure wear-resistant hose, and the pressure resistance of the high-pressure wear-resistant hose matches the water supply pressure range of the high-pressure direct injection nozzle assembly 10.

[0055] Specifically, the pressure resistance of the hose is matched to the water supply pressure range of the nozzle, which is based on the reliability design of high-pressure water delivery: After the high-pressure water flow from the water supply pump is output, it needs to be transmitted to the high-pressure nozzle through a hose. If the hose pressure resistance is lower than the system working pressure, pipe bursting and pressure leakage will occur, resulting in insufficient water pressure at the nozzle and failure to form an effective impact water jet. A safety margin of 20% to 50% is reserved in the design. For example, when the maximum working water pressure of the system is 10MPa, the hose pressure resistance will be selected to be 12 to 15MPa. This can fully cover extreme scenarios such as water hammer impact during the start and stop of the water supply pump and pressure fluctuations under high-altitude conditions, avoid pressure leakage, and ensure that the pressure at the water supply end can be fully transmitted to the nozzle end.

[0056] The use of a flexible hose instead of a rigid hose is to accommodate the reciprocating oscillation characteristics of the nozzle: one end of the hose connects to a water supply fixed to the drone body, and the other end connects to the nozzle that oscillates with the high frequency of the rocker arm. The hose can bend freely in sync with the nozzle's oscillation without hindering the oscillation action, and it avoids the problems of bending and breaking, loosening and leaking of rigid hoses. The wear-resistant properties are to accommodate the wear and tear of long-term use: when the nozzle oscillates at high frequency, the hose will bend repeatedly and rub against the outer shell 3 / fixed frame 2. At the same time, in outdoor operation scenarios, it will come into contact with corrosive media such as sand, dust, pesticides / cleaning agents, etc. Wear-resistant materials can significantly improve the hose's wear resistance and corrosion resistance, extend its service life, and reduce maintenance costs.

[0057] In one possible embodiment, the housing 3 has a swing opening through which the nozzle rocker arm 8 passes, and the size of the swing opening is adapted to the reciprocating swing range of the nozzle rocker arm 8.

[0058] Specifically, the core function of the outer shell 3 is to seal and cover the precision and vulnerable parts such as the motor 5 and the transmission shaft 7, preventing sand, dust, pesticide / cleaning liquid mist and rain from entering the interior during outdoor operations, and at the same time preventing collision damage to the internal structure during flight; while the nozzle rocker arm 8 needs to pass through the inside of the outer shell 3 to connect to the nozzle, so a special swing port is required.

[0059] The opening size and swing range are precisely matched: if the opening is too small, the rocker arm will hit the edge of the opening when it swings to the left and right extreme positions, causing jamming, structural wear, or even overload and burnout of the motor 5; if the opening is too large, it will greatly reduce the sealing and protection capability of the outer shell 3, and sand and water mist can easily enter and corrode internal components and jam the transmission structure.

[0060] The swing opening is typically designed as an arc-shaped groove that matches the swing trajectory of the rocker arm. The length of the groove just covers the entire swing stroke of the rocker arm from its left limit angle to its right limit angle. The width of the groove is only 1-2mm larger than the thickness of the rocker arm, allowing for assembly tolerances and thermal expansion and contraction. This design does not hinder the swing of the rocker arm and minimizes the gap in the opening, maximizing the protective effect. In some scenarios, dustproof brushes and elastic sealing strips are also added to the inside of the swing opening to further fill the gap between the rocker arm and the opening, improving the level of protection.

[0061] This embodiment achieves both the protection requirements of the outer shell 3 and the movement requirements of the rocker arm by simply using a slotted structure with precisely matched dimensions. It does not require the addition of complex dynamic sealing and motion guiding structures, resulting in low cost, high reliability, and suitability for the lightweight and highly durable operational requirements of drones.

[0062] Example 2, see Figure 3 The present invention also provides a step-by-step diagram of a high-pressure spraying method for drones, and provides, for example, Figure 3 The high-pressure spraying method for drones based on a reciprocating oscillating drive mechanism, as shown, includes the following steps: Step S1: Connect the high-pressure spraying system of the drone to the drone body in a detachable and fixed manner through the fuselage linkage 1, and seal and connect the water pipe 4 to the vehicle-mounted water supply source of the drone and the high-pressure direct spray nozzle assembly 10 respectively to complete the system assembly. Step S2: The remote-controlled drone flies to the cleaning area and starts the vehicle-mounted water supply of the drone to supply water to the high-pressure direct injection nozzle assembly 10 at a preset water pressure value, and outputs a high-pressure concentrated water jet through the high-pressure direct injection nozzle assembly 10. Step S3: Start motor 5. Motor 5 drives motor rocker arm 6 to rotate around its output end at a preset speed. The rotating shaft 7 at the end of motor rocker arm 6 slides in the slide groove 9 of nozzle rocker arm 8, converting the rotational motion of motor 5 into the reciprocating swing motion of nozzle rocker arm 8. Step S4: The high-pressure direct injection nozzle assembly 10 reciprocates synchronously with the nozzle rocker arm 8, causing the high-pressure concentrated water column to reciprocate and sweep the area to be cleaned, thus performing large-area high-pressure cleaning spray. Step S5: Identify the material type and dirt type of different areas to be cleaned using the image acquisition device and spectral sensor mounted on the drone; Step S6: Based on the identified material type, adjust the water pressure according to the current actual water supply pressure value to adapt to different material types; Step S7: Based on the identified dirt type, match the corresponding target oscillation frequency using the dirt-frequency mapping table pre-stored in the UAV control system, and adjust motor 5 from the preset speed to the corresponding speed to match the target oscillation frequency.

[0063] Specifically, the design of detachable fixed connection makes the spraying system an optional external functional module for drones. It can be quickly disassembled when not in operation, and the drone can also perform other tasks such as plant protection and inspection, which greatly improves the reusability of drones. The requirement of sealed connection is to avoid leakage and pressure loss when high-pressure water supply occurs, and to ensure that the pressure at the water supply end can be effectively transmitted to the nozzle end, avoiding pressure loss.

[0064] The preset benchmark water supply pressure can quickly output a stable high-pressure concentrated water column. Compared with traditional atomized spraying, the impact kinetic energy of the concentrated water column is more than 10 times that of atomized water droplets, which can directly wash away stubborn stains and meet the core needs of cleaning operations.

[0065] Driven by a single motor 5, the circular rotation of motor 5 is directly converted into the reciprocating swing of nozzle rocker arm 8 through a crank-slider transmission structure. It does not require complex multi-motor linkage, hydraulic drive and other structures. It is lightweight, has a low failure rate and is perfectly suited to the payload limit and high reliability requirements of drones.

[0066] The high-pressure water jet sweeps back and forth synchronously with the nozzle, expanding the original single-point high-pressure flushing coverage to a surface coverage. The width of a single operation is 3 to 5 times that of a fixed high-pressure nozzle. It retains the strong flushing ability of the high-pressure water jet while greatly improving the operation efficiency, solving the problems of narrow operating range and extremely low efficiency of traditional fixed high-pressure nozzles.

[0067] The differentiating advantage of this method lies in its ability to automatically match operation parameters according to the operation scenario, eliminating the need for manual adjustment. Scene perception logic: The drone automatically identifies the material (such as tempered glass photovoltaic panels, flexible thin film photovoltaic panels, exterior wall paint / tiles, etc.) and type of dirt (such as dust, bird droppings, cement residue, pesticide crystals, etc.) of the work surface through the visual acquisition device on the drone, so as to realize the automatic perception of the work scene.

[0068] Water pressure adaptive adjustment: Different materials have different pressure resistance capabilities. The system will automatically adjust the water supply pressure according to the recognition results: for example, reduce the water pressure to avoid damage to easily damaged flexible thin film photovoltaic panels, and increase the water pressure to enhance the cleaning power for high pressure resistant exterior wall tiles, maximizing the cleaning effect without damaging the working surface.

[0069] Oscillation frequency adaptive adjustment: The system has a pre-stored optimal oscillation frequency mapping table corresponding to different types of dirt. For stubborn dirt, it automatically reduces the motor speed and oscillation frequency, allowing the high-pressure water jet to stay on the dirt surface for a longer time and ensuring the rinsing effect. For easy-to-clean floating dirt, it automatically increases the speed and oscillation frequency to speed up the sweeping and improve the work efficiency.

[0070] This method lowers the barrier to entry for operators through standardized operating procedures and achieves full-scenario operation adaptability through intelligent adaptive adjustment. Compared with traditional manually operated high-pressure cleaning drones, it greatly improves operating efficiency and cleaning effect, while avoiding damage to the working surface caused by misoperation.

[0071] In one possible embodiment, when supplying water to the high-pressure direct injection nozzle assembly 10 at a preset water supply pressure value in step S2 above, the following is included: The system acquires the drone's current operating altitude and the remaining water volume of the water source in real time. Based on the atmospheric pressure difference corresponding to the operating altitude and the output efficiency deviation of the water pump corresponding to the remaining water volume, the preset water supply pressure value is dynamically compensated and calibrated, and the calibrated pressure value is used as the current actual water supply pressure value.

[0072] Specifically, this embodiment is an automatic calibration mechanism that ensures the long-term stability of the output pressure of the high-pressure nozzle. It solves the problem that the actual water pressure at the nozzle deviates from the preset value during high-altitude operation and water consumption of drones, ensuring the consistency of the rinsing effect throughout the entire operation cycle and avoiding problems such as excessive pressure damaging the working surface and insufficient pressure failing to clean properly.

[0073] The preset water supply pressure is the calibrated value under standard ground conditions. In actual operation, it will be affected by two types of variable factors, resulting in deviations: Pressure deviation caused by operating altitude (which can be obtained through terrain radar): The external atmospheric pressure is different at different altitudes, and the pressure difference inside and outside the nozzle will change with altitude, and the impact force of the water jet will also fluctuate.

[0074] Pressure deviation caused by residual water volume: The output efficiency of the water supply pump is directly related to the positive pressure value at the inlet: When the residual water volume of the water supply source is sufficient, the water pressure at the inlet is high, and the output efficiency of the water pump can reach more than 90% of the rated value; when the residual water volume is insufficient, negative pressure appears at the inlet, and the actual output efficiency of the water pump will decrease by 10% to 30%, and the output pressure under the same power will be lower than the preset value, resulting in insufficient water impact force.

[0075] This embodiment uses two pre-calibrated mapping tables, namely "operating altitude - pressure compensation value" and "remaining water volume - efficiency deviation compensation value", pre-stored in the UAV control system. During operation, the current operating altitude is obtained in real time from the UAV control system, and the remaining water volume is obtained from the water level sensor of the water supply system. The compensation amount for the corresponding two dimensions is obtained by looking up the tables respectively. The two compensation values ​​are superimposed on the preset reference water supply pressure value to obtain the calibrated actual output pressure, ensuring that the actual water outlet pressure at the nozzle end remains stable near the designed target pressure value.

[0076] The above mechanism can achieve dynamic pressure calibration by simply reusing the existing sensor data of the drone. It is low-cost and highly reliable, and can control the fluctuation of the outlet water pressure within ±5% throughout the entire operation cycle. This greatly improves the consistency of cleaning effect under different working conditions and avoids the operation cost of frequent manual parameter adjustments.

[0077] In one possible embodiment, see [reference] Figure 4 The flowchart for dynamic compensation calibration of the preset water supply pressure value includes the following steps: Multiple sets of altitude pressure compensation benchmark values ​​corresponding to the operating altitude range and multiple sets of water pump efficiency compensation coefficients corresponding to the remaining water volume range of the water supply source are stored in the UAV control system in advance. Based on the current operating altitude of the drone, the corresponding altitude pressure compensation benchmark value is matched, and the preset water supply pressure value is corrected for the first time to obtain the initial calibration pressure value. Based on the current remaining water volume of the water supply source, the pump efficiency compensation coefficient for the corresponding range is matched, and the initial calibration pressure value is corrected a second time to obtain the calibrated pressure value.

[0078] Specifically, this embodiment is a low-cost embedded solution for the aforementioned dynamic pressure compensation requirement. Its core is to achieve high precision and high reliability of pressure calibration without increasing hardware costs or consuming too much flight control computing power. It perfectly adapts to the characteristics of UAV embedded control systems with limited computing power and high requirements for response speed.

[0079] The reason for not using a real-time formula to calculate the compensation value, but instead pre-calibrating the interval parameters and storing them in the system, is due to two considerations: Computing power optimization: Pressure compensation involves complex calculations such as fluid mechanics and pump characteristics. If real-time calculation is performed during operation, it will consume a lot of flight control computing power and may affect flight safety. However, if the compensation values ​​of each interval are pre-calibrated, only a table lookup is needed during operation. The computing power consumption is less than 1 / 10 of that of real-time calculation, and the response speed can reach the millisecond level, which will not burden the flight control system.

[0080] Reduce calibration costs: Divide the continuous height and remaining water volume values ​​into several intervals (usually every 2 to 5 meters is a height interval, and every 10% of the remaining water volume is a water volume interval). Only the compensation value at the endpoint of each interval needs to be calibrated to cover the entire working condition. The workload of factory calibration and subsequent maintenance calibration is only 1 / 5 of that of full continuous sampling, which greatly reduces production and maintenance costs.

[0081] The two calibration logics are completely decoupled, allowing for independent adjustment without affecting each other, resulting in higher calibration accuracy. First calibration: The gravity pressure loss and atmospheric pressure deviation caused by height are fixed difference errors and are unrelated to the current output pressure value. Therefore, a fixed compensation benchmark value is directly used for correction: for example, the compensation benchmark value corresponding to the 10-15 meter working height range is 0.6MPa. 0.6MPa is directly added to the preset water supply pressure value to offset the pressure loss caused by height and obtain the initial calibration pressure value.

[0082] Second calibration: The pump efficiency deviation is a proportional error, which is directly proportional to the current output pressure value. Therefore, a coefficient multiplication correction is used: for example, the pump efficiency corresponding to the remaining water volume of 20%~30% is 87%, and the compensation coefficient is 1 / 0.87≈1.15. By directly multiplying the initial calibration pressure value by this coefficient, the pressure loss caused by the decrease in pump efficiency can be offset, and the final calibrated pressure value can be obtained.

[0083] Taking a real-world scenario as an example: the preset target water supply pressure is 8MPa, the compensation benchmark value matched by the current operating height is 0.6MPa, and after the first correction, the initial calibration value of 8.6MPa is obtained; the compensation coefficient matched by the remaining water volume is 1.15, and after the second correction, the final output value of 8.6×1.15≈9.89MPa is obtained. Finally, the actual water pressure at the nozzle end can be stably maintained within the error range of 8MPa±3%.

[0084] The above solution balances calibration accuracy, computing power cost, and calibration cost, making it a superior solution for pressure compensation in embedded scenarios: stable calibration can be achieved by pre-storing parameters, and the range parameters can be flexibly adjusted according to different operating scenarios, making it highly adaptable.

[0085] In one possible embodiment, see [reference] Figure 5 The flowchart for identifying material type and dirt type, in step S5 above, includes identifying the material type and dirt type of different areas to be cleaned, including: The material-dirt association feature library is stored in advance in the drone control system. The material-dirt association feature library stores the surface texture features of common material types in different operating scenarios, the spectral reflectance features of common dirt types, and the recognition confidence thresholds corresponding to different material and dirt combinations. The drone acquires visible light images of the area to be cleaned using its image acquisition device, and simultaneously acquires spectral reflectance data of the area to be cleaned using its onboard spectral sensor. Texture features are extracted from visible light images and matched with surface texture features in the material-dirt association feature library to obtain the material type; Spectral features are extracted from the spectral reflectance data and matched with spectral reflectance features in the material-dirt association feature library to obtain an initial judgment of the dirt type. Combined with the material identification results, the identification confidence threshold of the dirt under the corresponding material is called. If the matching degree of the initial judgment of the dirt type is higher than the identification confidence threshold, the final dirt identification result is directly output. If the matching degree is lower than the identification confidence threshold, the drone is controlled to approach the area to be cleaned until it reaches the preset close identification distance, and spectral reflectance data is collected again for secondary identification until the matching degree meets the identification confidence threshold requirement.

[0086] Specifically, this embodiment uses a multi-sensor fusion and confidence-adaptive verification scheme to achieve high-accuracy scene recognition without human intervention, providing a precise decision-making basis for subsequent adaptive adjustment of water pressure and oscillation frequency, thereby avoiding problems such as damage to the work surface and inadequate cleaning caused by recognition deviation.

[0087] Storing material and dirt characteristics together rather than separating them independently is to reduce the probability of misidentification. Different materials have fixed types of common dirt (for example, photovoltaic panels are often covered with bird droppings, dust, and silicone sealant residue, while exterior walls are often covered with cement stains, dust, and peeling paint). The associated features can directly exclude recognition results that do not fit the scenario. The three types of pre-stored information in the database each have their own specific function: Material surface texture characteristics: The texture differences of different materials (tempered glass, flexible film, ceramic tile, concrete exterior wall, etc.) are stable and are not affected by excessive surface dirt coverage. They can be accurately matched through ordinary visible light images. Dirt spectral reflectance characteristics: Dirt of different components has a unique "spectral fingerprint". Even stains with similar appearances (such as bird droppings and paint splatter, dust and mold) have obvious differences in spectral reflectance curves, which can achieve precise differentiation at the molecular level. Material-specific confidence thresholds: Different materials have different levels of background interference. For example, smooth glass has less interference, so the recognition threshold is set higher to avoid false positives; rough concrete exterior walls have more texture interference, so the threshold is appropriately lowered to avoid missed positives, thus adapting to the recognition characteristics of different materials.

[0088] A combination of a visible light camera and a spectral sensor is used to balance cost and recognition accuracy. Visible light cameras are low-cost and provide stable imaging. They are specifically designed to collect surface texture information to identify materials and can output results quickly without additional computing power. The spectral sensor is specifically designed to collect spectral reflectance data of material surfaces, penetrate surface dirt to capture its compositional characteristics, and solve the problem that the naked eye / visible light cannot distinguish dirt with similar appearances.

[0089] This process balances recognition efficiency and accuracy, avoiding unnecessary job interruptions: First round of identification: First, the material type is determined by visible light texture matching, and then the initial judgment result and matching degree score of dirt are obtained by spectral feature matching; Confidence verification: Call the exclusive confidence threshold corresponding to the material. If the initial judgment is that the matching degree meets the standard, the recognition result is directly output without affecting the normal operation process. Most common scenarios can be recognized in this step. Secondary close-range recognition: If the matching degree is lower than the threshold (usually because the dirt coverage is thin or the spectral data is affected by atmospheric interference due to the distance), the system automatically controls the drone to approach the work surface to the preset close-range recognition distance (usually 1 to 2 meters, based on distance sensor detection), eliminates atmospheric interference, and re-collects spectral data until the matching degree reaches the standard, ensuring the recognition accuracy in special scenarios.

[0090] In one possible embodiment, when adjusting the water pressure to suit different material types in step S6 above, the following is included: A material-water pressure mapping table is pre-stored in the drone control system. The material-water pressure mapping table stores the highest water pressure tolerance threshold and the optimal cleaning water pressure range corresponding to different material types. Based on the material type of the current area to be cleaned, the corresponding optimal cleaning water pressure range is selected, and the median value of the range is used as the initial adjustment water pressure value. Based on the currently identified type of fouling, a corresponding offset correction amount is generated, and the initial regulating water pressure value is offset to obtain the target regulating water pressure value. The pressure deviation between the target regulating water pressure value and the preset water supply pressure value is obtained, and the pressure deviation value is added to the current actual water supply pressure value to complete the water pressure regulation. Among them, the superimposed water pressure value does not exceed the highest water pressure threshold of the corresponding material.

[0091] Specifically, the median value of the optimal cleaning water pressure range is taken as the initial adjustment water pressure value, and the initial adjustment water pressure value is dynamically offset according to the type of dirt to achieve fine-grained adaptation to the scene: If the stains are identified as stubborn dirt such as bird droppings, cement residue, or grime, the offset correction is positive (usually +0.5~2MPa), increasing the water pressure to enhance the flushing force. If the stain is identified as light dirt such as dust or pollen, the offset correction is negative (usually -0.5~1MPa), which reduces water pressure and energy and water consumption.

[0092] The revised target water pressure will still be limited by the highest tolerance threshold, and there will be no risk of overpressure.

[0093] Water pressure adjustment is achieved by adding a deviation value to the actual water supply pressure that has been dynamically calibrated, rather than directly adjusting the preset reference value. For example, if the preset water supply pressure is 8MPa, the target adjustment water pressure is 9MPa, and the pressure deviation is 1MPa, then 1MPa is added to the current actual water supply pressure to complete the water pressure adjustment. This process combines the calibration results of the previous height and water volume compensation, so that the actual water pressure at the nozzle end is always stable near the target adjustment water pressure value.

[0094] In one possible embodiment, generating the corresponding offset correction includes: A dirt-water pressure offset mapping table is pre-stored in the UAV control system. The dirt-water pressure offset mapping table stores the basic water pressure offset corresponding to different dirt types. Based on the identified type of dirt, a corresponding baseline water pressure offset is matched (if it is a stubborn type of dirt, the baseline water pressure offset is positive to appropriately increase the water pressure and thus improve the cleaning effect; if it is a non-stubborn type of dirt, the baseline water pressure offset is negative to appropriately reduce the water pressure and thus reduce resource waste). The crosswind speed and flight attitude tilt angle parameters of the UAV are collected in real time. Based on the water column offset corresponding to the crosswind speed and the nozzle water outlet angle deviation corresponding to the flight attitude tilt angle, the basic water pressure offset is calibrated a second time to obtain the final offset correction amount.

[0095] Specifically, pressure calibration for different types of dirt is pre-defined in the drone control system to create a dirt-water pressure offset mapping table: for stubborn dirt such as bird droppings, cement residue, and grime, a positive base offset is set to appropriately increase water pressure to ensure flushing power; for easy-to-clean non-stubborn dirt such as dust and pollen, a negative base offset is set to reduce water pressure and reduce water and electricity consumption.

[0096] The static baseline offset is the calibration value under ideal conditions of no wind on the ground and the drone hovering horizontally. In actual outdoor operations, two variables will cause the kinetic energy of the water column impacting the working surface to deviate from the expectation: Crosswinds can deflect and disperse high-pressure water jets, significantly reducing the actual impact force reaching the work surface. The roll and pitch attitude of the drone during flight will change the distance between the nozzle and the working surface and the angle of water jet incidence, which will also lead to changes in the actual impact energy.

[0097] Therefore, it is necessary to introduce these two real-time parameters to dynamically calibrate the basic offset and offset the pressure deviation caused by the environment and flight conditions.

[0098] In one possible embodiment, the secondary calibration of the baseline water pressure offset includes: Real-time acquisition of crosswind speed values ​​in the current operating environment of the drone, and acquisition of the current roll angle and pitch angle of the drone as flight attitude tilt parameters; The mapping relationship between crosswind speed and water pressure compensation coefficient, as well as the mapping relationship between water outlet angle deviation and water pressure correction coefficient under different flight attitude tilt angles, are stored in the UAV control system in advance. The crosswind speed value collected in real time is matched to obtain the corresponding first calibration coefficient, and the flight attitude tilt angle parameter collected is matched to obtain the corresponding second calibration coefficient. The basic water pressure offset is calibrated using the first calibration coefficient and the second calibration coefficient to obtain the final offset correction.

[0099] Specifically, crosswinds cause the concentrated high-pressure water jet to shift laterally and disperse the water flow. The greater the wind speed, the greater the energy loss of the water jet reaching the work surface, and the actual flushing force will decrease significantly under the same water pressure. The roll and pitch attitude of the drone during operation will change the relative distance between the nozzle and the working surface and the incident angle of the water column. The larger the tilt angle, the longer the effective impact distance of the water column, the more deviated the incident angle, and the higher the energy loss.

[0100] Both types of parameters can be directly obtained through the drone's built-in wind speed sensor and flight control attitude sensor, without requiring additional hardware costs, thus meeting the drone's lightweight and low-power design requirements.

[0101] Using pre-defined mapping relationships instead of real-time computation is to adapt to the characteristics of limited computing power and high response speed requirements of UAV embedded control systems: During the research and development phase, the energy loss rate of water column under different crosswind speeds will be simulated in laboratory / flight test scenarios. The corresponding water pressure compensation ratio that needs to be supplemented will be calibrated to form a crosswind speed-water pressure compensation coefficient mapping table. The higher the wind speed, the larger the corresponding compensation coefficient. Simultaneously, the water pressure loss caused by the nozzle water outlet angle deviation under different roll and pitch angles is simulated, and the corresponding correction coefficients are calibrated to form a flight attitude tilt angle-water pressure correction coefficient mapping table. The larger the tilt angle, the larger the corresponding correction coefficient.

[0102] During operation, the corresponding coefficients can be quickly obtained by simply looking up a table. The computing power consumption is less than 1 / 10 of that of real-time fluid dynamics calculation. The response speed is fast and it will not consume too much flight control computing power, thus affecting flight safety.

[0103] The first calibration coefficient (crosswind compensation coefficient) is a proportional coefficient, which is used to offset the energy loss of water column caused by crosswind. For example, the coefficient is 1.1 for a crosswind of 3 m / s, which means that 10% water pressure compensation needs to be added on the basis of the basic offset. The second calibration coefficient (attitude correction coefficient) is also a proportional coefficient. Its function is to offset the water pressure loss caused by changes in flight attitude. For example, a 10° pitch angle corresponds to a coefficient of 1.08, which means that an additional 8% water pressure compensation is required. The final offset correction is obtained by superimposing the calibration results of the two types of coefficients on the basic water pressure offset. This not only matches the cleaning pressure requirements of the current type of dirt, but also offsets the dynamic interference of the current environment and flight status, ensuring that the actual water pressure output by the nozzle can accurately adapt to the current working conditions, and the water pressure error can be controlled within ±5%.

[0104] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A high-pressure spraying method for unmanned aerial vehicles (UAVs) based on a reciprocating oscillating drive mechanism, characterized in that, The use of a drone high-pressure spraying system includes the following steps: The high-pressure spraying system of the drone is detachably and fixedly connected to the drone body via a fuselage linkage, and the water pipes are sealed and connected to the water supply source and the high-pressure direct spray nozzle assembly respectively. The remote-controlled drone flies to the area to be cleaned and starts the water supply source to supply water to the high-pressure direct spray nozzle assembly at a preset water pressure value, so that it outputs a high-pressure concentrated water jet. Start the motor and drive the motor rocker arm to rotate around its output end at a preset speed. Drive the rotating shaft at the end of the motor rocker arm to slide in the slide groove of the nozzle rocker arm, converting the rotational motion of the motor into the reciprocating oscillating motion of the high-pressure direct injection nozzle assembly, so that the high-pressure concentrated water jet reciprocates and sweeps the area to be cleaned. The system identifies the material type and dirt type of different areas to be cleaned, and adjusts the water pressure based on the current actual water supply pressure to adapt to different material types. It matches the target oscillation frequency by matching the dirt-frequency mapping table pre-stored in the UAV control system, and adjusts the motor from the preset speed to the corresponding speed to match the target oscillation frequency. When supplying water to the high-pressure direct injection nozzle assembly at a preset water supply pressure value, the following is included: The system obtains the drone's current operating altitude and the remaining water volume of the water source in real time. Based on the atmospheric pressure difference corresponding to the operating altitude and the output efficiency deviation of the water pump corresponding to the remaining water volume, the preset water supply pressure value is dynamically compensated and calibrated to serve as the current actual water supply pressure value. When adjusting water pressure to suit different material types, including: A material-water pressure mapping table is pre-stored in the drone control system, which stores the maximum water pressure tolerance threshold and the optimal cleaning water pressure range for different material types. Based on the identified material type, the corresponding optimal cleaning water pressure range is selected, and the median value of the range is used as the initial adjustment water pressure value. Based on the identified type of fouling, a corresponding offset correction amount is generated, and the initial regulating water pressure value is offset to obtain the target regulating water pressure value. The pressure deviation between the target regulating water pressure value and the preset water supply pressure value is obtained, and the pressure deviation value is added to the current actual water supply pressure value to complete the water pressure regulation. Among them, the superimposed water pressure value does not exceed the maximum water pressure threshold of the corresponding material; The high-pressure spraying system for drones includes a fuselage linkage, a mounting frame, a drive assembly, a transmission assembly, a high-pressure direct injection nozzle assembly, a housing, and water pipes. The mounting frame is connected to the drone body via a fuselage linkage. The drive and transmission components are both mounted on the mounting frame, and the outer shell covers the outside of the drive and transmission components. One end of the water pipe is connected to the water supply source, and the other end is connected to the high-pressure direct injection nozzle assembly. The drive assembly includes a motor and a motor rocker arm, with the output end of the motor fixedly connected to one end of the motor rocker arm; the transmission assembly includes a rotating shaft and a nozzle rocker arm, with the rotating shaft fixedly mounted on the other end of the motor rocker arm, and the nozzle rocker arm having a groove adapted to the rotating shaft, with the rotating shaft slidably embedded in the groove; the high-pressure direct injection nozzle assembly is fixedly connected to the end of the nozzle rocker arm away from the groove. The motor rotation drives the motor rocker arm to rotate around the motor output end. The rotating shaft slides in the slide groove and pushes the nozzle rocker arm to swing back and forth. The high-pressure direct injection nozzle assembly swings back and forth synchronously with the nozzle rocker arm. The high-pressure direct injection nozzle assembly and motor dynamically adjust the output water pressure and speed according to the material type and dirt type of the area to be cleaned.

2. The high-pressure spraying method for unmanned aerial vehicles according to claim 1, characterized in that, When performing dynamic compensation calibration, the following are included: Multiple sets of altitude pressure compensation benchmark values ​​corresponding to the operating altitude range and multiple sets of water pump efficiency compensation coefficients corresponding to the remaining water volume range of the water supply source are stored in the UAV control system in advance. Based on the current operating altitude of the drone, the corresponding altitude pressure compensation benchmark value is matched, and the preset water supply pressure value is corrected for the first time to obtain the initial calibration pressure value. Based on the current remaining water volume of the water supply source, the pump efficiency compensation coefficient for the corresponding range is matched, and the initial calibration pressure value is corrected a second time to obtain the calibrated pressure value.

3. The high-pressure spraying method for unmanned aerial vehicles according to claim 1, characterized in that, When identifying material type and dirt type, the following are included: Pre-store a library of material-dirt related features in the drone control system; The visible light images and spectral reflectance data of the area to be cleaned are collected using image acquisition devices and spectral sensors mounted on the drone. Texture features and spectral features are extracted from visible light images and spectral reflectance data, respectively, and then matched with surface texture features and spectral reflectance features in the material-dirt association feature library to obtain the material type and preliminary dirt type. Call the identification confidence threshold for dirt under the corresponding material. If the initial judgment of the dirt type matching degree is higher than the identification confidence threshold, the final dirt identification result is directly output. If the matching degree is lower than the identification confidence threshold, control the drone to approach the area to be cleaned until it reaches the preset close identification distance, re-collect spectral reflectance data for secondary identification, until the matching degree meets the identification confidence threshold requirement.

4. The high-pressure spraying method for unmanned aerial vehicles according to claim 1, characterized in that, When generating the corresponding offset correction, the following are included: A dirt-water pressure offset mapping table is pre-stored in the UAV control system, which stores the basic water pressure offset corresponding to different dirt types; Based on the identified current type of fouling, match the corresponding baseline water pressure offset; The crosswind speed and flight attitude tilt angle parameters of the UAV are collected in real time. Based on the water column offset corresponding to the crosswind speed and the nozzle water outlet angle deviation corresponding to the flight attitude tilt angle, the basic water pressure offset is calibrated a second time to obtain the final offset correction amount.

5. The high-pressure spraying method for unmanned aerial vehicles according to claim 4, characterized in that, When performing a secondary calibration of the basic water pressure offset, the following is included: Real-time acquisition of crosswind speed values ​​in the current operating environment of the drone, and acquisition of the current roll angle and pitch angle of the drone as flight attitude tilt parameters; The mapping relationship between crosswind speed and water pressure compensation coefficient, as well as the mapping relationship between water outlet angle deviation and water pressure correction coefficient under different flight attitude tilt angles, are stored in the UAV control system in advance. The real-time collected crosswind speed values ​​are matched to obtain the corresponding first calibration coefficient, and the collected flight attitude tilt angle parameters are matched to obtain the corresponding second calibration coefficient. The basic water pressure offset is calibrated using the first calibration coefficient and the second calibration coefficient to obtain the final offset correction.

Images