73results about How to "Easy landing" patented technology

The invention discloses a power coupling-considered capacity ratio estimation method for a network-constructing / network-following multi-machine hybrid system, and the method comprises the steps: building a GFL-GFM hybrid system model, combining VSG power coupling small signal modeling and a dynamic relative gain matrix (DRGA), and quantifying the coupling strength of active / reactive power to output voltage; the frequency domain impedance characteristic of the GFM converter is simplified, and an ideal voltage source series equivalent reactance equivalent model is constructed; deriving an analytical relationship between a generalized short-circuit ratio (gSCR) and a capacity ratio based on the model, and generating a ratio formula in combination with engineering boundary sub-scenes to form a ratio range suitable for multiple scenes; for coupling time-varying characteristics, a closed-loop mechanism of'mixed mode-coupling sensing-correction calculation-execution feedback 'is constructed, and the ratio is dynamically adjusted. According to the method, coupling quantization and multi-machine coordination control are deeply fused, the problem that the external characteristic model precision and the multi-working-condition adaptability are difficult to consider under complex environment coupling is effectively solved, and the engineering applicability of the hybrid fan system is enhanced.

The invention provides an unmanned aerial vehicle for intelligent identification of a polar ice rink. The unmanned aerial vehicle comprises an aerial vehicle main body, a solar battery pack, rudder wings, a spiral propeller and an observation cabin, the aircraft is powered by solar energy, so that long endurance time can be achieved; the rudder wings are symmetrically installed on the two sides of the aircraft body to achieve lifting and steering of the aircraft body. The spiral propeller provides propulsive force and cooperates with the rudder wings to complete lifting steering motion. A plurality of functional modules are arranged in the observation cabin, and the functions of ice rink recognition, aircraft and transport ship positioning, data transmission, state monitoring and the like can be achieved. Illumination warning lamps are arranged on the front edges of the two wings of the aircraft, so that ice rink identification and aircraft landing under the extremely night condition are facilitated. The invention also provides a polar ice rink intelligent identification method of the unmanned aerial vehicle. The sea ice morphology around the navigation channel can be observed in detail and accurately, accurate sea ice data can be obtained, the polar transport ship is helped to sail, a captain is assisted to make a decision, and the risk of distress is reduced.

A cage culture unmanned ship-borne unmanned aerial vehicle integrated inspection device disclosed by the present invention comprises an unmanned ship body, an inspection unmanned aerial vehicle, a docking platform, a battery replacement mechanism, a holding mechanism and a hatch cover component, a docking cabin is arranged in the unmanned ship body, the docking platform and the battery replacement mechanism are both arranged in the docking cabin, and the hatch cover component is arranged above the docking cabin. The battery replacing mechanism is arranged on one side of the docking platform, and the holding mechanisms are arranged on the two sides of the unmanned ship body. The parking platform is used for clamping and releasing the inspection unmanned aerial vehicle, and the battery replacement mechanism is used for implementing battery replacement operation on the inspection unmanned aerial vehicle; the enclasping mechanism is used for enclasping a handrail of the net cage to achieve parking. According to the invention, the water surface moving capability of the unmanned ship and the air flying capability of the inspection unmanned aerial vehicle are combined, all-directional inspection of a net cage culture area can be realized, and detection of environment data inside and outside the net cage is realized.

The invention relates to a method for supporting secure transmission and verification of large data files in a cloud environment, and belongs to the technical field of file transmission. The method comprises the steps of a preparation stage, file fragmentation encryption, confusion processing, file uploading and downloading, file decryption and combination and the like. The method is characterized in that a file is divided into a plurality of file fragments, each fragment is symmetrically encrypted to generate file fragment metadata, then confusion fragments and encrypted file fragments are uploaded to a cloud server at the same time, and due to the fact that the encrypted file fragments lose sequence information, the sequence information is encrypted and then stored, and the file fragments are stored in the cloud server. The user needs to access and read; and hiding the sequence information of the file fragments in the graph, obtaining the sequence information by using an algorithm, and finally carrying out file merging, so that the secure transmission of the file is completed. According to the method, under the condition of distributed deployment, the transmission efficiency can be improved, the file transmission safety can be improved, and safe transmission of file data is achieved.

The small multi-unmanned aerial vehicle nest comprises a box body and a box cover, the box body and the box cover are rotationally connected, a lifting mechanism is arranged in the box body and comprises a lifting plate, and top plates are fixedly connected to the four corners of the top of the lifting plate through supporting columns respectively. The top plate is provided with a plurality of lifting parking aprons, the lifting parking aprons are fixedly connected with the telescopic ends of lifting air cylinders, and the lifting air cylinders are fixed to a lifting plate. According to the small multi-unmanned aerial vehicle nest provided by the utility model, the plurality of lifting parking aprons are arranged on the top plate, so that a plurality of unmanned aerial vehicles can be accommodated; and the top plate and the lifting parking apron can move up and down under the action of the lifting mechanism and the lifting air cylinder, so that the unmanned aerial vehicle can land conveniently.

The invention belongs to the technical field of process form design, and particularly relates to a lightweight process form intelligent prediction method and system. The method comprises the following steps: S1, obtaining process entity information and component entity information with a uniform structure from each business system, and forming a data set; s2, generating a model dictionary, and performing enhancement training on the model dictionary through a machine learning method according to the data set to obtain a data dictionary; the data dictionary has a data matching function; s3, according to the requirements of different enterprises, importing a data dictionary into the system for providing lightweight data model service; the data model service comprises a prediction component list and an output process form. The method has the advantages that the user can conveniently build the form, intelligent prediction is provided for the building process, the training process does not depend on a large amount of business data, and the method can quickly adapt to continuously changing business requirements of enterprises.

The invention relates to a training data generation method and device, a readable medium and electronic equipment, and aims to automatically generate training data of a model, improve the model training efficiency and reduce the cost. The method comprises the following steps: acquiring sample data; determining a data generation template according to the sample data; generating question data according to first data in the sample data and the data generation template, and determining answer data according to second data except the first data in the sample data; and combining the question data and the answer data into training data.

The invention belongs to the technical field of computer networks, and discloses a user interest estimation method based on multi-structure feature crossover, which comprises the following specific steps: step 1, sample construction: the quality of a training sample often decides an upper limit of a model generalization ability, and in order to ensure the completeness and reliability of data, the quality of the training sample is determined by the upper limit of the model generalization ability; behavior logs of the seasonal living users in the master station are extracted, and sampling is carried out to construct a data set; 2, feature engineering: performing fine-grained division on the general features; the user interest estimation method based on multi-structure feature intersection is adopted, the whole method is visual in logic and easy to land, the deep feature relation can be excavated, the coverage rate of user preferences is higher after combination, the accuracy rate is ensured through model offline posteriori, certain self-optimization ability is achieved, and the user interest estimation method is suitable for being popularized and applied. And a better promotion effect is achieved on formulating strategies such as recommendation, search and push in a support platform, delineating user groups and carrying out precise marketing and refined operation.

The application discloses a long spiral drilling machine concrete automatic pouring control method and system, the method comprises the following steps: in response to the pipeline filling rate and / or the concrete pressure reaching the preset value, the drill pipe automatic lifting mode is started; according to the feedforward speed and the feedback correction amount generated based on the concrete pressure deviation, the lifting speed is calculated, the winch equipment is controlled to lift the drill pipe according to the lifting speed, and the real-time correction of the lifting speed is carried out in the lifting process; the total lifting length is divided into multiple preset lengths, when the drill pipe is lifted by the preset length each time, the pile filling rate in the preset length range is calculated, if the pile filling rate is lower than the preset filling rate, the underfilling backfilling is carried out, after the backfilling is completed, the drill pipe returns to the original position and continues to lift until the concrete pouring is completed. The pouring process of the concrete is changed from the experience type manual adjustment to the closed loop control based on the pile filling rate and / or the concrete pressure, the underfilling and overfilling phenomena are avoided, and the piling quality is higher and more stable.

The application relates to a carbon / ceramic brake disc with wear resistance improved by silicon-chromium alloy synergistic modification and a preparation method thereof. The carbon / ceramic brake disc is prepared through key steps of preparing a carbon fiber preform, a carbon / carbon composite material, ceramic powder and a carbon / ceramic composite material, and the process is through gradient melting impregnation and segmented heat treatment process, thereby avoiding the composition segregation problem caused by the traditional silicon powder embedding melting impregnation method; the structured laying of the ceramic powder positively superimposes the capillary force in the brake disc and the furnace pressure; the ceramic production efficiency is increased by 30%, and the waste rate is reduced to below 1%. In the structure aspect, the porosity of the brake disc prepared by the application is reduced to 2%-3%; in the material composition aspect, the ceramic phase in the matrix is formed by Si-Cr-C three elements to form crystals, so that the wear resistance of the composite material is increased by three times; the brake disc prepared by the application has excellent wear resistance and friction performance, and is suitable for the aerospace field.

The present invention provides an AC relay control circuit and a control method, which are applied to a high-voltage DC line. The control circuit includes a pre-charging module and a low-voltage control module. The pre-charging module is respectively connected to an external power supply module, a low-voltage control module and a power conversion module. The pre-charging module includes at least a first AC relay, a second AC relay and a capacitor charging unit. The first AC relay and the second AC relay are connected in series in the high-voltage DC line. The first AC relay and the second AC relay are used to pre-charge the capacitor charging unit, and the capacitor charging unit is used to store electric energy. The low-voltage control module is used to set the action sequence of the first AC relay and the second AC relay, and control the shut-off of the first AC relay and the second AC relay. The present invention effectively reduces the volume and cost of the power conversion system by using a small AC relay, and improves the power density, safety and flexibility of the system.

The present invention provides a 5G terminal discovery method based on spectrum energy detection, which belongs to the field of radio signal positioning, including: using a wireless receiver to receive and collect data; frequency shifting and downsampling the data; calculating the frame synchronization position of the data; determining the time slot position of the uplink signal in the current frame; converting the time domain data into frequency domain data; performing median filtering after obtaining the frequency domain energy of each OFDM symbol; converting the RE level energy into the RB level energy; respectively accumulating the spectrum energy of all the first 7 and last 7 OFDM symbols in a time slot; and performing energy judgment. The receiver of the present invention can directly calculate the frame header position in the baseband synchronization without relying on GPS for synchronization calibration. Only a short amount of data is needed to complete the synchronization, and there is no need to merge and desynchronize multiple frames of data. After synchronization, the energy of the entire signal bandwidth can be calculated without complex decoding process or time-frequency domain correlation.

The present invention relates to the field of construction technology, and in particular to a dust reduction device for highway construction and a method for using the same. The dust reduction device comprises a spray assembly and an oscillation assembly; the spray assembly comprises a positive charge generating device, a pressurizing device, an insulating water tank, a pipeline and a spray head; a conductive sheet is arranged in the insulating water tank, which is electrically connected to the positive charge generating device and allows water droplets sprayed from the spray head to have positive charges; the oscillation assembly comprises an alternating electric field generating device and an electrode plate group, and the electrode plate group comprises a first electrode plate and a second electrode plate electrically connected to the alternating electric field generating device and generates an alternating electric field; the present invention forms an alternating electric field by exchanging the direction of the electric field between the first electrode plate and the second electrode plate, and forms oscillation by repeatedly changing the force direction of the droplets with positive charges in the alternating electric field, thereby reducing the surface tension of the droplets and facilitating the adsorption of dust; the large area is negatively charged, and the positive charge can facilitate the landing of the droplets.

The application discloses a color video restoration method based on a low-rank quaternion tensor and a space-time SVTV constraint, and comprises the following steps: acquiring a color video sequence to be restored, judging a video format and completing format analysis, converting YUV format into RGB format, constructing a corresponding pure quaternion tensor representation, and generating damaged video data; building a neural network parameterized low-rank quaternion tensor continuous function model, constructing a neural representation driven space-time SVTV regular term, and restraining a saturation gradient and a brightness gradient in different dimensions; fusing data consistency constraints and regular constraints, constructing a finite-dimensional solvable joint optimization model, and completing solvability verification; taking the damaged video as input, iteratively solving an optimal prediction tensor through a gradient descent algorithm or an Adam optimization algorithm, fusing a predicted pixel and an original known pixel, and outputting a complete restored video after data restoration and format adaptation. The application can realize high-precision, high-robustness and high-stability restoration of a missing area of a video, and is suitable for various video damage scenes.

The invention discloses an intelligent traffic signal lamp intelligent control system, which comprises a signal lamp controller, a sensor module, a data processing module, an intelligent decision module and a communication module, and is characterized in that the signal lamp controller is used for receiving traffic data and adjusting signal lamp timing according to a preset algorithm; the sensor module comprises a vehicle detection sensor and a flow monitoring sensor and is used for collecting traffic flow, vehicle speed and traffic flow information in real time. The data processing module is used for processing data acquired by the sensor module and analyzing traffic conditions; the intelligent decision-making module intelligently adjusts a timing scheme of the signal lamp according to an analysis result of the data processing module; according to the wireless traffic signal lamp, the signal lamp body is controlled by adopting a wireless control method, so that the problem that the signal lamp body is bound by a cable is effectively solved.

The utility model relates to the technical field of fire fighting truck supporting facilities, in particular to a fire fighting truck-based water supply pump device used in cooperation with a forest fire fighting system, which is used in cooperation with a fire fighting truck with a carriage and comprises a water supply tank, a lifting platform and a water spraying unmanned aerial vehicle with a water storage tank, the liquid supply end comprises a funnel-shaped connecting nozzle, a liquid supply pipe connected with the bottom end of the connecting nozzle and fixed to the lifting table, an electromagnet arranged at the bottom of the connecting nozzle and arranged outside the liquid supply pipe in a sleeving mode, and a liquid supply pump connected between the water supply tank and the liquid supply pipe. The liquid receiving end comprises a liquid inlet hose, an annular electromagnet and an iron ring. Water can be flexibly supplemented for the water spraying unmanned aerial vehicle parked on the lifting platform, the water spraying unmanned aerial vehicle does not need to carry a water pipe to be connected with the ground, operation moving flexibility is good, and the problem that in the prior art, an unmanned aerial vehicle needs to carry a water pipe to be connected with the ground, and consequently moving of the unmanned aerial vehicle is limited is solved.

The invention discloses a method and a system for protecting portrait feature data of multilayer space mapping, which are used for scenes such as user portraits / article portraits and the like. According to the method, a three-layer mapping architecture of a scalar input layer (S)-a middle random layer (R)-a target output layer (C) is adopted, and multi-stage irreversible transformation processing is carried out on original scalar data. The method comprises the following steps: firstly, mapping a scalar into an intermediate vector by utilizing a one-way function of cryptography security and combining a secret key and an offset; secondly, distance features of the intermediate vector are extracted, direction features are discarded, and the distance features are converted into two-dimensional space coordinates through a deterministic mapping algorithm; according to the method, the irreversibility of reversely deducing the scalar quantity from the coordinates is ensured by erasing the angle information of the middle layer; the space isolation in different business scenes is realized by introducing the offset; according to the method, a deterministic rejection sampling algorithm based on HMAC is preferably adopted, the problem of center aggregation existing in a traditional mapping method is solved, uniform distribution of coordinates in a set area is achieved, sensitive feature value scalar quantities of all dimensions in portrait feature vectors are mapped into two-dimensional coordinate points, and the two-dimensional coordinate points are aggregated to form an object portrait feature matrix; the method is used for cross-system analysis, clustering, similar retrieval and visual display, and meanwhile plaintext feature value leakage can be avoided.

The present application relates to the technical field of weather forecast, especially a strong convective rainfall short-term forecast method and system based on multi-band radar observation, the method comprising the following steps: obtaining multi-band radar observation data, and performing feature extraction on the multi-band radar observation data to construct a multi-band radar differentiated feature set; performing convective system development stage identification based on the multi-band radar differentiated feature set to obtain a convective evolution stage identification result; dynamically and differentially fusing the multi-band radar differentiated feature set according to the convective evolution stage identification result to generate a multi-band radar fusion feature set; constructing an adaptive short-term forecast model, and inputting the multi-band radar fusion feature set into the adaptive short-term forecast model to output a strong convective rainfall short-term forecast result. The present application supports regional emergency scheduling, and matches the fine and time-sensitive business needs of the short-term warning of strong convective rainfall in coastal urban clusters.

The invention discloses a method for detecting the performance degradation degree of a photovoltaic panel, and the method comprises the steps: carrying out the testing of the open-circuit voltage and short-circuit current of the photovoltaic panel under a standard testing condition, obtaining the values of the open-circuit voltage and short-circuit current of the photovoltaic panel, carrying out the feature zooming, and obtaining the mapping values of the open-circuit voltage and short-circuit current; performing a color change test on the surface material of the photovoltaic panel to obtain a color change quantitative value of the surface material of the photovoltaic panel, and performing feature scaling to obtain a color change quantitative mapping value; performing electroluminescent imaging detection to obtain the ratio of the normal luminous area of the surface of the photovoltaic panel to the surface area of the photovoltaic panel, and performing feature scaling to obtain an electroluminescent imaging mapping value; and finally, calculating the performance degradation degree of the photovoltaic panel through each mapping value. According to the method, the performance degradation degree of the photovoltaic panel is visually represented in a numerical form, the detection result is accurate, the detection process does not depend on a complex algorithm, huge calculation process and calculation amount do not exist, and implementation and landing are facilitated.

The utility model provides a device facilitating landing and transferring of an unmanned aerial vehicle, and belongs to the technical field of unmanned aerial vehicle landing and transferring. Comprising a bottom plate and a pull rod connected with the bottom plate in a drawing mode, self-locking type universal wheels are arranged at the bottom end of the bottom plate and the bottom end of the pull rod, the damping platform comprises a damping shock absorber arranged on the top face of the bottom plate and a landing platform arranged on the damping shock absorber; the unmanned aerial vehicle installation component comprises sliding rails arranged on the two side edges of the landing platform, a fixing piece is fixed to the two sliding rails in an adjustable mode, and the fixing piece is used for installing and fixing various unmanned aerial vehicle landing racks of different specifications and shapes. Damage to the unmanned aerial vehicle during landing of the unmanned aerial vehicle landing gear is conveniently relieved, and the unmanned aerial vehicle is indirectly protected; and meanwhile, the unmanned aerial vehicle landing gear with various specifications and shapes can be adaptively fixed conveniently.

The invention discloses a snowfall inversion method based on a millimeter wave cloud radar and a microwave radiometer, and relates to the technical field of radar snowfall inversion, and the method comprises the following steps: S1, data input: millimeter wave cloud radar data and microwave radiometer data are input; s2, multi-source equipment data space-time dimension matching is carried out; s3, phase state classification of extremely cold particles, and elimination of non-snowfall phase state interference; s4, inverting the snowfall rates of different height layers; s5, the accumulated snowfall of the whole-layer snowfall rate is obtained through time integration; and S6, outputting a result. According to the invention, the complementary advantages of high temporal-spatial resolution of the millimeter wave cloud radar and abundant atmospheric environment parameters of the microwave radiometer are brought into full play, the inherent defect of single equipment inversion is solved, and the precision and reliability of snowfall inversion in the extremely cold environment are greatly improved; and a data support is provided for practical applications such as meteorological monitoring and early warning in extremely cold areas, and prevention and reduction of snowstorm disasters.

The sealing material guiding device comprises a material collecting part and a material guiding part which are connected in sequence, the end, corresponding to the upstream, of the material collecting part is provided with a feeding port, and the end, corresponding to the downstream, of the material guiding part is provided with a discharging port; the discharge port is configured to correspond to the position of the conveying line; the material guiding part comprises a feeding channel and a material crushing assembly movably arranged in the feeding channel. The material crushing assembly can impact large coal in the feeding channel through self movement to crush the large coal into small coal, and the small coal leaves the feeding channel from the window body under the pushing of the material crushing assembly, so that the problem of space blockage caused by the large coal is solved; the effects of cleaning large coal materials and smoothing the conveying line can be achieved, the requirements for equipment parameters such as power of related components of the material crushing assembly can be lowered, therefore, the equipment cost is effectively controlled, the cost lowering effect is remarkable, low cost of overall transformation of the conveying line is achieved, and the technology can be applied and popularized easily.

The invention discloses an automatic parking apron for an unmanned aerial vehicle, relates to the technical field of unmanned aerial vehicle parking aprons, and aims to solve the problems that most of existing unmanned aerial vehicle parking aprons are designed in a plane shape, external wind blows over the unmanned aerial vehicle in the landing process of the unmanned aerial vehicle, the unmanned aerial vehicle is easily influenced by the external wind during landing, the landing difficulty of the unmanned aerial vehicle is increased, and meanwhile, the unmanned aerial vehicle is difficult to land. The mode of arranging a windproof baffle is adopted, the influence of external wind on landing of the unmanned aerial vehicle is reduced, however, the unmanned aerial vehicle is likely to make contact with the baffle in the landing process, consequently, the unmanned aerial vehicle is unbalanced, and the unmanned aerial vehicle is seriously damaged, an adjustable wind shielding plate is designed, when the unmanned aerial vehicle lands, external wind is shielded for the unmanned aerial vehicle, and the unmanned aerial vehicle is prevented from being damaged. And meanwhile, when the unmanned aerial vehicle makes contact with the stopping disc, under the gravity effect of the unmanned aerial vehicle, the stopping disc drives the wind shielding plate to descend together, collision between the wind shielding plate and the unmanned aerial vehicle is effectively prevented, and the safety of the unmanned aerial vehicle is protected.

The invention relates to the technical field of computer vision, and discloses an image forgery detection model interpretability analysis method and system, and the method comprises the steps: extracting the middle layer features of an image forgery detection model to be interpreted, and generating a semantic feature map based on the non-negative matrix decomposition of sparsity constraint; performing feature importance pre-screening on the semantic feature map to obtain an important feature map; for each important feature map, positioning a high activation area and extracting an image block, and analyzing a dominant frequency component and a bandwidth of the image block corresponding to the high activation area; constructing a band elimination filter in the frequency domain of the original image, generating a disturbance image, and endowing each feature with an initial weight based on a decision distance; and performing grouping fusion on the features according to the symbols of the initial weights, and respectively generating final visual saliency evidence graphs which support counterfeiting and reality. According to the invention, through combination of the non-negative matrix factorization and the frequency domain disturbance shielding method, the interpretability of the image forgery detection model can be effectively improved.

The utility model discloses an apis cerana breeding device which comprises a beehive base with an upward opening, and the side portion of the beehive base is provided with an inclined inlet and outlet allowing apis cerana to enter the beehive base conveniently. The first beehive is arranged on the upper side of the beehive base, the downward end of the first beehive is provided with a hexagonal containing sinking groove facilitating nesting of Chinese bees, top plates on the two sides of the long diagonal line of the inner side of the first beehive incline downwards, and a center comb foundation groove is formed in the long diagonal line of the top plate of the inner side of the first beehive. A plurality of comb foundation grooves are uniformly distributed in the first beehive from two sides of the central comb foundation groove to the outside, and comb foundation strips are respectively embedded in the central comb foundation groove and the comb foundation grooves of the first beehive; due to the arrangement of the hexagonal beehive and the herringbone top plate, after Chinese bees enter the first beehive, the Chinese bees preferentially start to make honeycombs at the long diagonal line, namely the center of the herringbone top plate, and after the central honeycombs are made, the Chinese bees sequentially make honeycombs towards the two sides, so that the Chinese bees can more quickly make the honeycombs by concentrating force downwards, and the quality of making the honeycombs is improved.

The present invention provides a micron-level dry mist dust suppression hierarchical management system, which relates to the field of environmental protection technology, and aims to solve the technical problem that the existing dust control in the prior art mostly adopts 10μm dry mist to suppress dust. This single management mode causes dust above 10μm to be unable to settle at all, and the expected dust suppression effect cannot be achieved. The device includes a first-level management system, a second-level management system and a third-level management system centered on a dust generating point, surrounding the dust generating point and sequentially arranged outside the dust generating point, wherein: the first-level management system includes a plurality of first micro-mist nozzles, and the diameter of water mist particles generated by the first micro-mist nozzles is less than 50μm; the second-level management system includes a plurality of second micro-mist nozzles, and the diameter of water mist particles generated by the second micro-mist nozzles is less than 30μm; the third-level management system includes a plurality of third micro-mist nozzles, and the diameter of water mist particles generated by the third micro-mist nozzles is less than 10μm. The present invention is used for hierarchical management of dust at an unorganized emission site.

The invention relates to the field of waste gas treatment, in particular to an industrial waste gas purification device which comprises a spraying box, a partition plate, a filter screen, a sprayer and an air pump, the partition plate, the filter screen, the sprayer and the air pump are fixedly connected into the spraying box, the partition plate and the filter screen are horizontally arranged, the partition plate is located below the filter screen, silk threads of the filter screen are made of elastic materials, and the sprayer is located above the filter screen; the rolling ball, the rotating shaft, the push rod and the rotating motor are located in the spraying box, the rotating motor is fixedly connected with the spraying box, the rotating motor and the rolling ball are located below the sprayer and above the filter screen, the lower surface of the rolling ball is in pressing contact fit with the upper surface of the filter screen, one end of the push rod is connected with the rolling ball, and the other end of the push rod is connected with the rotating shaft. The other end of the push rod is fixedly connected with the rotating shaft; the rolling ball rolls the upper surface of the filter screen to form a lower pit; the input end of the air pump is connected with the top of the space below the partition plate, and the output port of the air pump communicates with the bottom of the space above the partition plate. The contact efficiency of waste gas and water mist can be improved, waste of water resources is reduced, and the purification cost is reduced.