75results about How to "Avoid features" patented technology

This application discloses a method and apparatus for constructing a multi-scale adaptive gating MambaPlus network, belonging to the field of artificial intelligence and machine learning technology. The method includes: initializing the network configuration and constructing the basic structure; preprocessing the input data to generate a standard dataset; mapping the input data to the hidden space via an input mapping layer, and extracting backbone features from the Mamba backbone; constructing at least two parallel scale branches in the hidden space to obtain a multi-scale feature set; inputting the backbone features and multi-scale features into an adaptive gating module, dynamically allocating weights and adaptively fusing them through a hierarchical gating mechanism to generate fused features; further enhancing the features through cross-scale attention and feedforward enhancement, and then superimposing the residuals to generate the final discriminative features; finally, completing category prediction and model training evaluation. This application, while retaining the advantages of Mamba's long-range dependency modeling, addresses the problems of insufficient utilization of multi-scale information, poor adaptive feature fusion, and low robustness in complex scenarios.

The invention relates to the technical field of data processing and state monitoring, and discloses an intelligent ring main unit fault real-time diagnosis method and system, and the method comprises the steps: constructing a fault feature over-complete dictionary; collecting and preprocessing high-frequency operation data streams of the ring main unit in real time; carrying out online sparse decomposition and manifold mapping based on the dictionary, and converting a high-dimensional signal into a low-dimensional sparse coefficient and manifold features; and a lightweight classification model is used for rapid reasoning, and fault early warning and data uploading are triggered when abnormity is detected. The system comprises a data acquisition module, a preprocessing unit, an edge computing engine, an intelligent diagnosis module and a communication gateway. According to the method, depth dimension reduction and feature enhancement are realized through a manifold sparse coding technology, the bandwidth pressure and the calculation overhead are reduced on the premise of ensuring the fault feature integrity, and the real-time performance and the accuracy of diagnosis are improved.

The application discloses a sludge biological drying multi-target proteinase regulation method based on bacterial chemotaxis, relates to the technical field of sludge resource disposal, and comprises the following steps: S1, constructing a microfluidic detection system based on the chemotaxis of thermophilic bacteria, and monitoring and quantifying the metabolic preference degree of the thermophilic bacteria to different amino acids in a sludge biological drying process; S2, selecting and proportioning proteinases with a targeted degradation function according to the metabolic preference degree of the thermophilic bacteria; and S3, decomposing refractory proteins into specific amino acids preferred by the thermophilic bacteria through multi-target regulation of the proteinases. The application optimizes the metabolic environment of the thermophilic bacteria to improve the utilization rate of protein organic matter and accelerate the sludge biological drying process.

The utility model relates to a dust catcher functional joint, belong to dust catcher technical field, including pipe lower joint, pipe lower joint on be equipped with with pipe lower joint intercalation limit pipe upper joint, pipe upper joint with pipe lower joint interval be equipped with lock catch spare, adopt lock catch spare and realize pipe upper joint and pipe lower joint between the separation and the assembly, pipe upper joint on be equipped with a plurality of extension pipe upper joint lower end's electric rod, through electric rod realizes pipe lower joint lower end part's circuit intercommunication. Pipe lower joint lower part be equipped with with pipe lower joint intercalation formula connection's transition electric plug, transition electric plug with electric rod circuit intercommunication formula intercalation. Have the characteristics such as high integration, convenient operation, connection reliability is high, safe durable and assembly process good. Solveed the connection stability, the operation convenience and the electric power transmission reliability problem that the dust catcher encountered in use.

The application belongs to the technical field of positioning and measurement, and particularly relates to a target positioning method and device combining CSI coarse positioning and Y-type CCD fine correction, which comprises the following steps: S1, CSI acquisition equipment is arranged in a target area, and CSI data between a sending end and a receiving end is acquired; S2, a Y-type CCD sensor is geometrically calibrated and a coordinate system is corrected, and coordinate reconstruction is completed; S3, the CSI data is preprocessed, noise is removed, and input is normalized; S4, time domain features of the CSI are extracted by using a time domain branch; S5, frequency domain features of the CSI are extracted by using a frequency domain branch; S6, a gating fusion module is used to fuse the time domain and frequency domain features, and the approximate position coordinates of a target are output; S7, the Y-type CCD sensor is controlled to optically acquire a target area, and point light source coordinate information is acquired; and S8, error correction is performed on the data acquired by the Y-type CCD, and the accurate position coordinates of the target are output.

This invention belongs to the field of photocrosslinking probe technology, specifically relating to a GUSB photocrosslinking probe, its preparation method, and its application. The GUSB photocrosslinking probe has the general chemical formula C0. 13 H 18 N2O6-R, wherein R is any one of the oxalic group elements or peptide bonds. The R group covalently connects the glucuronic acid recognition core and the photosensitive unit in the GUSB photocrosslinking probe, and the R group contains no more than 4 atoms, resulting in a small size and low steric hindrance. The GUSB photocrosslinking probe provided by this invention uses glucuronic acid as the recognition core, combined with a small-volume diazacyclopropane photosensitive unit and a terminal alkyne group, to achieve µM-level in-situ covalent capture of active GUSB in complex systems such as cells or tissues. The probe of this invention significantly improves positioning accuracy, sensitivity, data availability, quantification, and enrichment, while also possessing advantages such as modular synthesis, mild conditions, and strong system compatibility, effectively addressing the technical shortcomings of existing technologies that struggle to accurately obtain GUSB activity in situ under low-dose conditions.

The utility model belongs to the field of quick assembly, provide a portable installation structure and comb grass machine, including machine body, lithium battery, apron and turnover framework. Compared with prior art, the utility model's advantage lies in setting positioning portion and anti -drop portion at every lithium battery bottom, cooperation positioning slot and mounting groove in machine body, so that multiple lithium batteries can be automatically guided and abutted to positioning when assembling, realize quick, accurate positioning installation, significantly improve the assembly efficiency, apron and turnover framework are stacked along the vertical direction in turn, make full use of space, the overall structure is compact, form multistage mechanical pressure locking to lithium battery, effectively prevent lithium battery from loosening or falling off because of vibration or external force during use or movement, improve the safety and stability of equipment operation, the structure avoids the use of a large number of screws, improves the assembly and disassembly efficiency, provides the convenience for subsequent maintenance and replacement.

The application provides a speech enhancement method and device based on a hidden space Schrodinger bridge, equipment and a medium. The method comprises the following steps: obtaining degraded speech; encoding the degraded speech into degraded latent variables in a hidden space through a target encoder, the target encoder being used to converge different types of degraded speech to a distribution close to corresponding clean latent variables in the hidden space, and the target encoder having the function of maintaining the energy characteristics of the power spectrum of an audio signal; processing the degraded latent variables through a target generation network based on a Schrodinger bridge to obtain clean latent variables corresponding to the degraded latent variables; and decoding the clean latent variables through a target decoder to obtain clean speech corresponding to the degraded speech. The application can solve the problem that the related art cannot realize the structured reconstruction of missing spectral information in a high-noise and low signal-to-noise ratio environment, and effectively improve the speech enhancement effect.

The invention discloses an NFC power supply detection device and method and an NFC system, the NFC power supply detection device comprises an NFC chip, an NFC antenna module and a control module, and the NFC antenna module supplies power to the control module when receiving an electromagnetic signal; and the control module obtains the power supply state of the NFC antenna module according to the power supply voltage output by the NFC antenna module, and sends the power supply state to the NFC chip, so that the NFC chip gives out a prompt. According to the technical scheme, through the synergistic effect of the NFC chip, the NFC antenna module and the control module, the power supply state of the NFC antenna module can be monitored and fed back in real time, communication interruption or functional faults caused by insufficient power supply are avoided, the stability of the system is improved, the equipment can timely prompt a user to adjust the position of the equipment under the condition of insufficient power supply, and the user experience is improved. Continuous operation of the equipment is ensured, and the reliability and user experience of the equipment are improved.

This invention relates to the field of medical image processing technology, specifically to an automatic segmentation and recognition method and system for cancer images based on the fusion of multimodal images. The method includes: acquiring a multimodal medical image set containing computed tomography (CT) images, magnetic resonance imaging (MRI) images, and positron emission tomography (PET) images; generating spatially aligned data blocks through three-dimensional spatial registration and voxel alignment; extracting fusion features using an improved encoder-decoder network optimized based on the characteristics and spatial dependencies of different modal channels; obtaining enhanced semantic feature maps through multi-level context-aware processing; identifying suspected cancer regions through a region candidate generation network; and simultaneously outputting pixel-level segmentation masks and cancer type probabilities through a segmentation network, ultimately generating a three-dimensional segmentation and recognition result. This method can accurately capture multimodal-specific features, clearly distinguish the boundaries between lesions and normal tissue, simplify the processing flow, reduce errors, and achieve accurate and efficient segmentation and recognition of cancer lesions.

This invention relates to a method and system for monitoring and enhancing the power of GPS P(Y) signals based on civilian signal tracking. The method employs a carrier tracking loop to stably track civilian satellite navigation signals, generating a local carrier with the same frequency and phase as the GPS satellite navigation digital signal. The GPS satellite navigation digital signal is then orthogonally mixed with the local carrier to output the in-phase branch signal u. I (n) and the orthogonal branch signal u Q (n); the in-phase branch signal u I (n), Orthogonal branch signal u Q (n) Perform low-pass filtering and incoherent integration respectively to obtain the average power P of the baseband signal in the in-phase branch. IB Average power P of baseband signal in orthogonal branch QB The average power P of the baseband signal in the same phase branch IB Average power P of baseband signal in orthogonal branch QB The power enhancement state of the current P(Y) signal is determined by comparing it with the weak-state power threshold of the in-phase branch, the strong-state power threshold of the in-phase branch, the strong-state power threshold of the quadrature branch, and the weak-state power threshold of the quadrature branch.

The invention belongs to the technical field of 3D projection, and particularly relates to an immersive deduction system and method based on 3D projection fusion, and the system comprises a sensing module which is used for collecting the rapid motion data and height parameters of an actor, a motion pre-judgment module which is used for judging the motion parameters of the actor, and a control module which is used for controlling the motion pre-judgment module and the motion pre-judgment module. The motion pre-judgment module is in communication connection with the sensing module and comprises a motion association sub-model and a motion time sequence prediction sub-module, and the motion association sub-model establishes a mapping relation among the body height of the actor, the motion amplitude and the track span. And data is provided for follow-up action pre-judgment and special effect adaptation, and the problem that the projection special effect is disjointed with the action of the actor due to action acquisition omission or insufficient data precision is avoided.

This invention discloses a frequency-domain guided multi-scale deformable alignment method and system for UAV target detection. The method first acquires and preprocesses a dataset of UAV aerial images. Next, it improves the YOLOv8 model, performing target detection based on the preprocessed UAV aerial image dataset and outputting the target detection results. Finally, the improved YOLOv8 model is integrated into a pre-configured training environment, using training and validation images from the UAV aerial image dataset for training and validation. This invention significantly enhances the model's ability to perceive small targets, thereby significantly improving the localization and classification accuracy of small targets, reducing detection errors, and improving overall accuracy.

The application provides a vehicle hibernation method, medium, vehicle and product, relates to the technical field of vehicle control, and the method comprises the following steps: generating a hibernation preparation instruction and sending the hibernation preparation instruction to an electronic control unit on the vehicle under the condition that the vehicle meets hibernation conditions; receiving confirmation information fed back by the electronic control unit based on the hibernation preparation instruction, performing state detection on the vehicle based on the confirmation information, and obtaining a detection result; wherein the confirmation information is used to indicate that the electronic control unit has completed a hibernation preparation task indicated by the hibernation preparation instruction; and if the detection result indicates that the vehicle is in a standby state, performing a whole-vehicle hibernation operation. The technical scheme of the application sends a hibernation preparation instruction to the electronic control unit first, so that the electronic control unit has sufficient time to complete the hibernation preparation task, effectively avoids the task execution risk caused by direct forced hibernation, improves the reliability and robustness of the vehicle hibernation process, and guarantees the integrity and safety of the vehicle function.

The utility model relates to a communication equipment fault detection device, which comprises a detector main body and a U-shaped protective cover, the surface of the detector main body is provided with a display and a controller, the U-shaped protective cover is installed on the outer surface of the detector main body in a damping sliding manner, and the surface of the U-shaped protective cover is provided with a visual window A; a turning cover is rotationally installed on the side, close to the controller, of the upper surface of the U-shaped protective cover, a visual window B is formed in the surface of the turning cover, and when the turning cover is turned to the upper surface of the U-shaped protective cover, the visual window B corresponds to the visual window A. Through the arrangement of the flip cover, when the flip cover is turned over to the surface of the U-shaped protective cover, the display can be observed through the visual window B and the visual window A, normal use is not affected, when the flip cover is turned over to the position above the controller, the flip cover and the U-shaped protective cover protect the controller and the display respectively, and therefore the situation that the controller is touched by mistake can be avoided; and unnecessary damage or functional abnormality to the detector main body is avoided.

The application provides a Chinese-oriented named entity recognition method and system, and belongs to the technical field of natural language processing; comprising: an encoding stage step, context encoding of input Chinese text; a feature extraction stage, a double-branch structure of global semantic branch and local boundary branch in parallel is adopted, wherein the local boundary branch carries out boundary feature enhancement through multi-granularity convolution combined with a learnable Gaussian Laplacian operator; a feature fusion and prediction stage step, global and local features are fused through a gating mechanism, and multi-label classification prediction is carried out based on a preset six-tuple grid label system; a decoding stage step, named entities are obtained through bidirectional search strategy decoding according to predicted relationship probability; through boundary perception feature extraction and enhanced label system, the recognition accuracy and recall rate of nested entities and discontinuous entities in Chinese text are significantly improved.

The application discloses an end plate bracket, a vehicle equipment cabin and a railway vehicle, relates to the technical field of rail transit, and the end plate bracket comprises a main body frame, at least two avoiding structures and at least two connecting beams. The main body frame comprises a base plate and a supporting plate integrally arranged on the base plate, and the supporting plate is used for connecting a vehicle body; the at least two avoiding structures are integrally arranged in the main body frame, and the avoiding structures are used for avoiding a sand box and a connecting device of a sanding device; the at least two connecting beams are connected with the at least two avoiding structures in correspondence, and the connecting beams are used for connecting the vehicle body. The end plate bracket solves the problems of low integration level and low modularization degree of a traditional end plate bracket and poor connection stability and reliability of the end plate bracket and the vehicle body.

The invention belongs to the technical field of power distribution network leakage current detection, and particularly relates to a CNN and FiLM-based leakage current type intelligent identification method. Comprising the following steps: S1, collecting real-time operation data of a typical power supply area, obtaining leakage current waveforms and related environment characteristic parameters under different working conditions, and constructing an original leakage current sample data set; s2, a CNN-based leakage current classification model is constructed, a FiLM condition modulation module is introduced, a multi-task learning framework is used at the tail of the network, a main task is leakage current type identification, and an auxiliary task is grounding system discrimination; s3, using a weighted cross entropy loss function to alleviate a class sample imbalance problem; an OneCycleLR dynamic learning rate scheduling strategy is introduced; s4, evaluating the performance of the model on the test set, and using the accuracy and the confusion matrix as evaluation indexes; according to the method, high-precision identification of multiple types of faults such as single-phase grounding, arc type electric leakage and direct current system electric leakage is realized, different grounding systems can be adapted, and the accuracy and robustness of system diagnosis are improved.

This application discloses a composite fixing structure for fixing a PCB, comprising: a carrier plate with multiple adsorption holes; a first abutment edge and a second abutment edge fixedly disposed on the periphery of the carrier plate; multiple pressure blocks disposed on the first abutment edge; a venting plate fixedly disposed below the carrier plate, the venting plate including a normally venting area and a first venting area; a first through hole in the first venting area selectively connected to or disconnected from a vacuum pump, and the first through hole in the first venting area opens or closes progressively from the side closer to the normally venting area to the side farther away from the normally venting area; and a pressure plate movably disposed above the carrier plate; the projection of the pressure plate on the carrier plate in a first direction is longer than the length of the carrier plate in the first direction, and the projection at least covers a portion of the first venting area. This composite fixing structure organically integrates vacuum adsorption and mechanical pressing, and can adaptively adjust the adsorption area and pressing position according to the PCB size, ensuring fixing reliability.

The application provides a display panel and a display device. The display panel has a display area and a binding area located on one side of the display area. The display panel comprises an array substrate and a first inorganic layer. The array substrate has a plurality of array metal layers arranged in a stack. The plurality of array metal layers are stacked to form a plurality of binding pads in the binding area. The first inorganic layer is located on the side of the array substrate provided with the array metal layers. The first inorganic layer is provided with a plurality of first openings. Each first opening corresponds to an intermediate area of each binding pad. The first inorganic layer covers an edge area of each binding pad. The application can effectively avoid the edge area of the binding pad from being corroded, thereby avoiding the problem of poor conduction of the binding pad.

A deep learning-based method for monitoring anesthesia depth based on high time-frequency resolution bispectral density includes the following steps: acquiring single-channel EEG data, preprocessing it, and dividing it into time segments; extracting features from the EEG data, and calculating a high time-frequency resolution wavelet bispectral matrix by combining variational mode decomposition and synchronous squeeze wavelet transform; constructing and training a deep neural network classification model based on partially convolutional neural networks and gated recurrent units; for test data, inputting the extracted features into the model, and selecting the state with the highest output probability (awake, anesthesia maintenance, or awakening) as the final prediction result for the test data; this invention uses single-channel EEG signals as input, greatly reducing the difficulty of acquiring clinical signals, and the proposed high time-frequency resolution bispectral features combined with a dedicated deep learning network have a strong ability to capture complex nonlinear EEG dynamic features and cross-subject generalization adaptability; therefore, this invention can perform high-precision, multi-classification, and real-time automatic monitoring of anesthesia depth.

This invention relates to the field of power chip testing technology and discloses a power chip testing data processing system and method based on big data. The method performs spatial and temporal correction on raw time-series observation data synchronously acquired by multiple testing devices to form a three-dimensional observation data volume. A virtual sensing grid is constructed within this volume, and a continuously spatially distributed gridded parameter field is generated through interpolation and iterative resampling. The local geometric features of this field are identified and adaptively divided into feature sub-regions, mapped to a graph structure, and a hierarchical feature map is generated through topological agglomeration. This map drives a parameter generation network to synthesize a high-resolution parameter field. The synthesized field is then fused with the original gridded parameter field across scales to obtain an enhanced observation data volume. Physical decoupling and signal reconstruction are performed based on a physical constraint model, and the test results are output. This invention achieves intelligent identification and enhanced reconstruction of the internal features of the test data, improving the resolution and physical reliability of the results.

This disclosure provides a method, system, electronic device, and storage medium for offline deployment of AI capabilities, relating to the field of computer technology. The method includes: obtaining target resources corresponding to a target AI capability based on AI capability resource mapping information; differentially encapsulating the target resources according to offline environment information to obtain an offline resource package adapted to the offline environment information, the offline resource package including multiple resource sub-packages, the resource sub-package types including at least one of dependent software, image files, and AI model files; and importing the offline resource package into an integrated repository module to achieve offline deployment of the target AI capability. This disclosure, through differential encapsulation for different offline environments, enables dynamic resource adaptation, avoiding storage waste caused by full resource import and deployment failures and functional abnormalities caused by resource incompatibility with the environment during offline deployment, effectively improving the success rate of offline AI capability deployment and the compatibility of AI functions after deployment.

The invention relates to the technical field of epilepsy detection, and discloses an epilepsy prediction model algorithm based on an artificial intelligence technology, which comprises the following steps: an epilepsy electroencephalogram data set is collected, the data set comprises five subsets, each subset comprises 100 single-channel electroencephalogram segments, the length of each segment is 23.6 seconds, two of the subsets are from healthy volunteers, and each subset is from the healthy volunteers; the other three subsets are from epilepsy patients. According to the epilepsy prediction model algorithm based on the artificial intelligence technology, frequency spectrum components of signals are extracted through fast Fourier transform, and significant features are mined through feature selection based on LASSO. Manhattan distance is used to assess importance and weight LASSO penalty terms of features. Due to the form of a linear function, the Manhattan distance is insensitive to noise, so that the Manhattan distance can keep stable performance in a noisy environment.

The invention discloses a self-powered piezoelectric ceramic atomization device, and relates to the technical field of gas-liquid reaction equipment. The self-powered piezoelectric ceramic atomization device comprises a reaction kettle and a self-powered piezoelectric ceramic atomizer, the self-powered piezoelectric ceramic atomizer comprises a shell, a piezoelectric atomization module and a fluctuation power generation module. The piezoelectric atomization module comprises a liquid storage tank, a liquid inlet pipe and a piezoelectric ceramic atomization sheet; the fluctuation power generation module comprises a fluctuation pressure sensor, a cantilever pipe arranged on the fluctuation pressure sensor and a small metal ball arranged in the cantilever pipe. When liquid fluctuation is generated in a reaction kettle to impact the self-powered piezoelectric ceramic atomizer, kinetic energy of liquid drives a metal ball in a cantilever pipe to fluctuate in a reciprocating mode, pressure is transmitted to a fluctuation pressure sensor to generate mechanical energy, electric energy is formed through triggering and used for driving a piezoelectric atomization module, and energy self-supply, high efficiency and energy conservation are achieved. After the piezoelectric atomization module is powered on, liquid is atomized into micron-sized tiny liquid drops, and the contact area of the gas phase and the liquid phase is increased.

The application discloses an ECG signal acquisition method and device based on an rPPG signal, equipment and a medium, relates to the technical field of computer software, and comprises the following steps: obtaining a PPG signal based on an rPPG signal, and generating rPPG conditional coding through a PPG encoder. The low-frequency, high-frequency and spatial domain ECG tokens of the mask state are used as the current mask token, the rPPG conditional coding is used as the constraint, and the mask reconstruction is completed through a bidirectional Transformer. The mask is gradually removed according to the confidence, and the reconstruction is circularly performed until all the masks are removed to obtain target tokens. The low-frequency and high-frequency feature tokens in the target tokens are subjected to corresponding decoding and inverse short-time Fourier transform to generate a frequency domain ECG waveform; and the spatial domain feature tokens are subjected to an ECG decoder to obtain a spatial domain ECG waveform. The target ECG signal is obtained by fusing the two waveforms, and a high-precision and high-fidelity ECG signal can be obtained based on the rPPG signal.

The utility model discloses a kind of gas leakage detection devices of being convenient for carrying, including detector body, connecting pipe, install the top of detector body, detection probe, install one end of connecting pipe;For achieving the protection mechanism that storage protection is carried out to detection probe, and protection mechanism is arranged at the side of detector body, the utility model is set through the setting of protection mechanism, the active protection of detection probe is realized, when device is in carrying or idle state, protection mechanism can store detection probe in protection structure, effectively avoid detection probe and external object from colliding by physical isolation, to protect its internal precision component from mechanical damage, the setting of protection mechanism is in keeping the portability of device, also significantly improve the durability of detection probe, ensure that detection probe can long-term maintain high detection accuracy and sensitivity, avoid performance attenuation or function failure caused by collision, so as to can prolong its service life.

This invention proposes a vehicle-to-everything (V2X) interaction system, an electronic device, and a vehicle. The V2X interaction system includes a SOC module, an MCU module, and a touch display module. The SOC module is developed based on QM safety level requirements, and the MCU module is developed based on ASIL B safety level requirements. According to the V2X interaction system proposed by this invention, when controlling the vehicle through the V2X interaction system, the MCU module developed based on ASIL B safety level requirements performs security verification on the software data generated and transmitted between them. This can avoid the problem of erroneous or unresponsive functions caused by abnormal virtual buttons on the screen, thereby improving the accuracy of vehicle control and vehicle safety.

The application discloses a kind of based on adaptive time-frequency modeling light-weight sound signal enhancement method, medium and equipment, it is related to sound signal processing technology, based on adaptive time-frequency modeling light-weight sound signal enhancement method mainly includes: constructing sound signal noise reduction model;Utilize sound signal data training set to train sound signal noise reduction model and obtain trained sound signal noise reduction model;Utilize trained sound signal noise reduction model to carry out noise reduction to target sound signal, obtain the sound signal after noise reduction.It can realize the effective enhancement of sound signal in complex noise environment by implementing the based on adaptive time-frequency modeling light-weight sound signal enhancement method, medium and equipment provided by the application, improve the expression ability of model, environmental adaptability and signal restoration precision, reduce the cost of calculation.

The utility model relates to air sampling technical field, concretely relates to a kind of air sampling multistage dehumidification device for rail transit, including sampling collection pipe, filter module, U-shaped tube, gas analysis module in turn along gas flow direction;It further includes primary water droplet flow guide structure, secondary extrusion structure, tertiary discharge structure, respectively for promoting the condensation of water vapor in sampling collection pipe into water droplet and flow guide discharge, for extruding the adsorbed water of filter sponge adsorption in filter module, for collecting the water drop of filter sponge seepage or extrusion.The utility model adds primary water droplet flow guide structure, secondary extrusion structure, tertiary discharge structure in correspondence in sampling collection pipe, filter module, U-shaped tube, carries out flow guide collection to water vapor water droplet etc., multiple protection is realized, prevent water droplet flow into gas analysis module, thus can avoid affecting gas monitoring work and avoid causing economic loss.