38results about How to "Increase awareness" patented technology

The utility model relates to the technical field of educational toys for children, and discloses a cognitive educational toy for children, which comprises a shell, a groove is arranged in the shell, a drawer is slidably connected in the shell, a partition plate is fixedly connected in the drawer, jigsaw blocks are arranged in the drawer, splicing components are arranged on the side walls of the jigsaw blocks, and the splicing components are arranged on the side walls of the jigsaw blocks. A fixing assembly is arranged in the drawer; each splicing assembly comprises a first clamping block, the first clamping blocks are fixedly connected to the side walls of the jigsaw blocks, clamping grooves are formed in the jigsaw blocks, and first magnetic attraction blocks are fixedly connected to the interiors of the jigsaw blocks. According to the jigsaw puzzle, the grooves of different shapes are formed in the shell, the first clamping blocks slide into the clamping grooves of other jigsaw puzzle blocks, the first magnetic attraction blocks and the second magnetic attraction blocks are matched for fixing, and then patterns of different shapes corresponding to the shapes of the grooves are formed, so that the visual cognition of children on different shapes is improved, and the interestingness of children is improved. Through the structure, the spatial imagination and spatial reasoning ability of children can be cultivated.

The application belongs to the technical field of complex system organization modeling and simulation analysis, and particularly relates to a kind of airplane system fault simulation modeling method based on OPM, comprising the following steps: step one: selecting an airplane system to be simulated, determining the system function and system composition of the system;Step two: determining the OPM modeling object element according to the finished product, structural part and cable contained in the selected airplane system;Step three: defining and creating the OPD object process diagram of the system according to the system function of the selected airplane system and the finished product object;Step four: specifying the state of each object element in the airplane system to represent the actual working state of each finished product, structural part and cable in the system;Step five: taking the specified state of each object element as the state value of each object element in the OPD;Step six: matching the state change process for the object element with state;Step seven: connecting the object state with the matched process using the influence link in the OPD;Step ten: simulating and deducing the fault model.

The utility model relates to the technical field of sockets, in particular to a socket with an infrared plugging detection function, which comprises a bottom shell, an upper cover and a socket module, the bottom shell and the upper cover form a cavity for mounting the socket module, the socket module comprises an insulating seat body, the insulating seat body is fixedly mounted on the bottom shell, the socket module is provided with an infrared plugging detection device, and the infrared plugging detection device is connected with the upper cover. The infrared plugging detection device comprises an infrared pcb, an infrared transmitting tube and an infrared receiving tube, the infrared pcb is installed on the socket module, the infrared transmitting tube is installed on one side of the infrared pcb, the infrared receiving tube is installed on the other side of the infrared pcb, and the infrared transmitting tube, the infrared receiving tube and jacks of the socket module are located on the same straight line. By arranging the infrared plugging detection device, the plugging state of the plug can be accurately detected, timely and accurate information is provided for a user, and the problems of equipment failure and the like caused by the fact that the plug is not plugged well or unplugged accidentally are avoided.

The paper diaper with the soft leakage-proof partition edges comprises a body, the body comprises a liquid-permeable surface layer, an absorption part and a liquid-tight bottom layer, the leakage-proof partition edges extending in the longitudinal direction are symmetrically arranged on the skin side of the liquid-permeable surface layer, each leakage-proof partition edge comprises a leakage-proof partition edge non-woven fabric and a soft non-woven fabric, and the soft non-woven fabric is compounded on the leakage-proof partition edge non-woven fabric, so that the leakage-proof partition edges are made of non-woven fabric. The soft non-woven fabric covers the rubber root part of the non-woven fabric of the leakage-proof partition edge, so that the softness of the whole leakage-proof partition edge is improved, meanwhile, multi-layer material protection is formed, the liquid blocking capacity is improved, and the urine leakage risk is reduced.

This invention discloses a method for monitoring low-altitude targets in substations based on 5G-A integrated sensing technology, belonging to the field of integrated sensing security technology. The method includes: acquiring an electronic map and key equipment location information of the substation to construct a three-dimensional semantic map of the substation area; combining the kinematic features of a unique low-altitude target's continuous trajectory with the functional semantic annotations of key equipment in the three-dimensional semantic map of the substation area to perform behavioral pattern recognition and intent understanding on the unique low-altitude target, outputting target type labels, intent labels, and corresponding threat levels; comparing the target type labels, intent labels, corresponding threat levels, and the unique low-altitude target's continuous trajectory with an electronic fence to generate graded alarm events and countermeasure suggestions, and issuing linkage commands to defense equipment and a linkage platform to trigger corresponding actions. This invention achieves accurate mapping from target behavior to threat intent by extracting kinematic features and functional semantic annotations, improving the ability to recognize and predict hidden risks.

This utility model discloses a voice tremor nursing teaching model in the field of nursing teaching aids. It includes a three-dimensional humanoid model with a left lung model and a right lung model positioned at the chest area. The left lung model is covered with a silicone membrane, forming a pleural cavity model. A catheter is located at the top of the left lung model, with one end extending into the pleural cavity model. Air or liquid is injected into the pleural cavity through the catheter to simulate pneumothorax or pleural effusion. The right lung model includes a detachable and replaceable lung cavity model, lung tumor model, and bronchial obstruction model. A bronchial model is located between the left and right lung models, with a microphone connected to the top of the bronchial model. The two forked ends of the bronchial model at the bottom are connected to the left and right lung models, respectively. This device can replace different models to simulate pathological states, allowing students to intuitively experience the differences in voice tremor under different pathological conditions, improving teaching effectiveness and students' clinical skills training level.

This invention relates to a method for constructing a polycystic kidney disease (PCD) model and its applications. The method involves overexpressing the MYCN gene in a target animal to obtain a PCD model that leads to PCD-related phenotypes. The method includes the following steps: obtaining a first strain of mice with Rosa26 knock-in overexpressing the CAG-LSL-HA tag-MYCN-IRES-BFP-Wpre-polyA gene; obtaining a second strain of mice by inserting Cre-WPRE-polyA into the start codon of the Pax8 gene; and crossing the first and second strains of mice to obtain a MYCN-overexpressing PCD model. This invention employs various experimental methods for validation, including histopathological analysis, immunohistochemical staining, and Western blotting. The model provided by this invention overcomes the limitations of existing in vitro cell models, organoids, and existing animal models in terms of limited phenotypes. The established MYCN-overexpressing PCD animal model exhibits stable disease progression and a short disease cycle, which not only helps to elucidate the disease mechanism but also serves as an ideal platform for drug screening and efficacy evaluation.

This invention provides a method, apparatus, and electronic device for evaluating driving behavior and vehicle dynamic control functions. The method includes acquiring vehicle status information, road surface status information, vehicle dynamic control function status information, and driver input information; determining whether the vehicle dynamic control function switch is in an on or off state; if in an on state, determining whether the vehicle dynamic control function is activated; if in a off state, determining whether the triggering conditions for the vehicle dynamic control function are met based on the vehicle status information, road surface status information, and driver input information; if met, setting a virtual function activation flag; when the vehicle dynamic control function is activated or the virtual function activation flag is set, invoking multiple evaluation items and preset penalty rules; statistically analyzing the multiple evaluation items based on the vehicle status information and road surface status information; calculating the evaluation result based on the statistical results and penalty rules; and outputting the evaluation result.

The application discloses a task platform display method and device, equipment and computer program product, and relates to the field of computers. The method comprises the following steps: receiving a task platform starting operation; in response to the task platform starting operation, and the first task permission corresponding to the master account, displaying a first task platform, and the first task platform comprising a first quantity of first task options; receiving a platform switching operation, the platform switching operation being used for indicating switching from the first task platform to a second task platform; in response to the platform switching operation, and the second task permission corresponding to the master account, displaying the second task platform, and the second task platform comprising a second quantity of second task options, the first quantity being less than the second quantity. The method can improve the perception of target tasks of new users, thereby improving the efficiency of task interaction.

The application provides a mine underground roadway fire hazard simulation method, steps comprising: importing an underground roadway three-dimensional model into Unity software, building a three-dimensional scene in the Unity software; preparing a prefabricated body for dynamically generating objects, instantiating a personnel prefabricated body or a flame prefabricated body; simulating a smoke diffusion process; baking a navigation mesh, and enabling the personnel prefabricated body to walk on the navigation mesh; dynamically adjusting an escape route, and enabling the personnel to navigate and move according to the escape route; according to wind path navigation calculation, drawing a line segment indicating a wind path, performing ray detection on the line segment, comparing a ray direction with an attribute of a smoke object, and according to a comparison result, destroying or accelerating smoke diffusion. The method can specify personnel distribution and a fire starting position, plan an optimal escape path between each personnel position and each escape exit by comprehensively considering various factors, simultaneously, simulate a smoke diffusion process in real time, and recompute and adjust an escape path of each personnel of each mining level in real time.

The invention discloses a distribution duration adjustment method and electronic equipment, and relates to the field of distribution, and the method comprises the steps: determining a stop start road section when a user is recognized to start to stop in a traffic light region during distribution; identifying that the user starts to wait for the red light when the stop start section belongs to a planned route sequence, a continuous road section exists in the planned route sequence and a driving direction indicator light from the stop start section to the continuous road section is the red light; recording and identifying lamp waiting characteristics when the user starts to wait for the red light; when it is identified that the user is in delivery and stops staying, a staying stopping road section of the user is determined; when the stop ending road section is consistent with the stop starting road section in the last light waiting characteristic, identifying that the user finishes waiting for the red light; according to the stop ending time of the user and the stop starting time in the last light waiting characteristic, determining the actual red light waiting time; and the distribution duration is adjusted for the user according to the actual red light waiting duration, so that the distribution experience of the rider is improved, and the distribution is safe.

This invention relates to a method for detecting the time-frequency fingerprint characteristics of a frequency source, belonging to the field of frequency detection technology. The method acquires the clock signal of a frequency standard source, extracts a time-frequency fingerprint feature region from the curve of the clock signal's frequency stability changing with sampling time, and obtains the time-frequency fingerprint characteristics of the frequency source by detecting the time-frequency characteristics of this region. This fingerprint is then compared with the original, inherent time-frequency fingerprint characteristics to identify the type of the frequency standard source and the individual frequency source. This invention creatively proposes using the detection of a specific curve characteristic within the frequency source's stability changing with sampling time as a unique time-frequency fingerprint feature region to identify the type and individual characteristics of the frequency source. This achieves the application objective of uniquely identifying and determining the health status of a specific electronic device by detecting the time-frequency fingerprint characteristics of the frequency source. The method is unique and has good market application prospects.

The application discloses a Bluetooth earphone dynamic teaming method and system based on Auracast protocol, and belongs to the technical field of wireless communication and audio processing. The method comprises the following steps: embedding a scene semantic label in a broadcast assistant signal by a broadcasting device; collecting self context information by a listening device, calculating a matching weight with each broadcast source through a recommendation decision engine, and realizing intelligent sorting, filtering or automatic joining of a broadcast group. The broadcasting device supports creating a dynamic hierarchical broadcast group with role permissions, time-limited access codes and multiple audio sub-streams. In the overlapping coverage area, the listening device triggers intelligent switching based on signal quality and context matching degree, and realizes seamless connection by using pre-synchronization technology. In addition, the system synchronizes group state information through an auxiliary channel to realize state visualization and interaction. The application upgrades Auracast broadcast from "manual, static and equal" to "automatic, dynamic and intelligent", greatly improving the convenience, management accuracy and immersion experience in the multi-user audio sharing scene.

According to the air conditioner, the maximum operation frequency is preset, test running is carried out when the air conditioner runs for the first time after installation is completed, piping parameters of the air conditioner are determined in the test running process, and the actual maximum operation frequency of a compressor reaching the air conditioner matching number is determined according to the piping parameters and the preset maximum operation frequency; the actual maximum operation frequency is written into a storage module, the actual maximum operation frequency serves as the upper limit operation frequency of the compressor, and the actual maximum frequency of the compressor is the frequency capable of reaching the nominal air conditioner piconumber when the actual maximum frequency is matched with the piping parameters of the air conditioner after actual installation; the problems that the capacity of the air conditioner is attenuated due to pressure loss caused by the factors such as the length of the piping and the height difference, and the nominal power cannot be reached within the preset maximum operation frequency range are solved.

This invention discloses a method for measuring Rydberg atoms using micro-Doppler agility under a polarization-multiplexed microwave field, belonging to the field of quantum precision measurement technology. The method includes generating a dynamic coded sequence and an interleaved coded microwave field by aperiodically pairing and jumping the microwave polarization state with the carrier frequency; acquiring the target echo signal and applying it to the Rydberg atom ensemble with a reference microwave field to generate a quantum coherent state; applying a synchronous laser pulse to acquire a time-series optical response signal; obtaining the signal's fundamental frequency, envelope spectrum, and initial phase component through time-frequency transformation and feature separation; and performing parameter mapping and synthesis to generate three-dimensional micro-motion parameters. This invention uses an aperiodically agile polarization-frequency-multiplexed microwave field for active target detection and utilizes the Rydberg atom ensemble for quantum coherent detection and optical readout of the echo signal, enabling high-sensitivity, high-precision, and high-anti-interference measurement of the target's three-dimensional micro-motion parameters.

The invention discloses a strain of beer obligate anaerobic pectobacterium frisinensis, and the preservation number of the beer obligate anaerobic pectobacterium frisinensis is CGMCC (China General Microbiological Culture Collection Center) No. 7.552. The invention further discloses application of the pectobacterium in construction of a pectobacterium and lactic acid bacteria symbiotic system, the pectobacterium and lactic acid bacteria symbiotic system and application of the pectobacterium and lactic acid bacteria symbiotic system. A strain of obligate anaerobic pectobacterium CGMCC No.7.552 is separated and identified, a symbiotic test system is constructed, RT-PCR and GC-MS technologies are adopted, symbiotic behaviors, metabolites and pH changes of pectobacterium and lactic acid bacteria in different environments are qualitatively and quantitatively studied, and the pectobacterium and lactic acid bacteria in different environments can be identified. The method determines the influence of multiple factors on the growth of pectobacterium in the symbiotic environment of pectobacterium and lactic acid bacteria, including the activity of pectobacterium, the amount of initial contaminating bacteria and the pH value of the symbiotic environment, is beneficial to prevention and control of harmful microorganisms in the beer fermentation process, optimization of the beer production process and improvement of the beer quality, and has a wide application prospect.

The utility model provides a kind of active safety alarm system and sound-light alarm device.Active safety alarm system includes: balcony photovoltaic energy storage equipment, for detecting abnormal state and generating alarm information;Cloud platform, with balcony photovoltaic energy storage equipment wireless communication, for receiving and forwarding alarm information;Sound-light alarm device, independent of balcony photovoltaic energy storage equipment setting, with cloud platform wireless communication, for receiving alarm information and executing sound-light alarm;Wherein, balcony photovoltaic energy storage equipment periodically sends heartbeat signal to cloud platform and sound-light alarm device, and sound-light alarm device executes sound-light alarm when not receiving heartbeat signal within preset time.Such, realize the double safety alarm to balcony photovoltaic energy storage equipment, and sound-light alarm device is not limited by the protection level of balcony photovoltaic energy storage equipment, can realize effective remote abnormal alarm, improve the security and reliability of balcony photovoltaic energy storage equipment work.

A material processing apparatus including a laser processing station (425) configured to process a substrate (430) of crystalline material. The laser processing station (425) includes a laser configured to form subsurface laser damage (434) regions within an interior of the substrate (430), at least one translation stage configured to effect relative movement between the laser and the substrate (430), and a diffuse light source (438) configured to illuminate a top surface (433) of the substrate (430). The laser processing station (425) also includes an imaging device (442) configured to generate an image of the top surface (433) of the substrate (430). The apparatus also includes a computing device configured to analyze the image to identify a condition indicative of presence of uncracked regions in the interior of the substrate (430).

The invention relates to the technical field of agricultural big data and intelligent traditional Chinese medicine planting evaluation, and discloses a traditional Chinese medicine planting site selection evaluation method based on big data. According to the scheme, sunshine and soil salt content collection devices are deployed in candidate places, sampling is carried out daily to form two types of sequences, and a mean value is solved; normalizing a global minimum value and a global maximum value into standardized sunlight and low salt indexes, and carrying out arithmetic averaging to obtain baseline suitability; a perturbation step length is set according to the range and the candidate number, positive and negative disturbance and amplitude limiting are carried out on sunlight and salt, and four types of disturbance baselines are reconstructed after normalization; the sunlight sensitivity and the salt-containing sensitivity are added according to the ratio of the actual standardized step length to the suitability change, a penalty term is generated by taking the maximum value of the whole, the robustness suitability is obtained through deduction and amplitude limiting, and then quantile grading is used; and finally summarizing the longitude and latitude to form a structured result table.

The application discloses a reconstruction effect evaluation method based on post-fracturing coring near-mid-far field fracture identification, which comprises the following steps: performing post-fracturing coring work on high-deviation wells in a fracturing well adjacent area to obtain cores, and obtaining core fracture characterization parameters; distinguishing horizontal direction fractures from vertical direction fractures, and calculating fracture fractal dimension and fracture complexity; in combination with fracture stress sensitivity simulation experiment analysis, calculating and obtaining downhole hydraulic fracture conductivity; analyzing three-dimensional space positions of the fracturing well and the coring well, establishing a correlation chart based on fracturing technology, distance, azimuth analysis and fracture complexity and conductivity, establishing a fracturing reconstruction volume space conceptual model, and evaluating the reconstruction effect according to the model. The application has the beneficial effects that a correlation model of fracturing technology, distance, azimuth analysis and fracture complexity index and conductivity in three-dimensional space is established, the method is intuitive and reliable, and the limitations of indirect monitoring of fracture networks by means of microseismic and other methods are avoided.

The invention relates to a live broadcast interaction method and device and a storage medium. The method comprises the steps that first interaction prompt information is displayed in a target live broadcast room, the first interaction prompt information comprises a target interaction resource which is displayed based on a first display element and aims at a target article, and the target interaction resource is used for participating in an interaction activity corresponding to the target article; the target article is an article associated with the target live broadcast room; under the condition that the remaining available duration of the target interaction resource is smaller than a first preset duration, the first interaction prompt information in the target live broadcast room is updated to be second interaction prompt information, and the second interaction prompt information comprises the target interaction resource displayed based on a second display element. The second display element is different from the first display element. The interaction efficiency and interaction popularity of the target article can be improved, and circulation interaction of the target article is promoted.

The invention provides a panorama implantation mark method and related equipment, and when a user wants to look up panorama image data of a certain target geographic position, the target image position corresponding to the target geographic position corresponding to a user look-up instruction can be determined and implanted with a mark, so that the interaction experience of the user is enhanced, and the user experience is improved. And auxiliary information association and transmission are improved, so that the practicability and usability of the data are improved. Further, from the perspective of data production and maintenance, the marking is standardized marking of the panoramic basic points. The method can help a data manager to quickly identify and manage different panoramic basic points, can locate a specific position through marking, improves the data processing efficiency, can highlight the particularity of a target position for a complex panoramic scene through marking, helps a user to understand the importance of the position, strengthens the cognition of the user on the target geographic position, and improves the user experience. The lookup process is more efficient, the user experience is improved, and the application scene of the panoramic image data is expanded.

The application discloses a shale reservoir pressure solution strength characterization method and device, the method comprises the following steps: obtaining a digital mineral image of a resin block; performing mineral component analysis on the digital mineral image to determine mineral component information; determining the average particle size of quartz minerals according to the quartz mineral information; determining the average particle size of other minerals according to the information of other minerals except the quartz mineral information; determining the pressure solution strength of the shale reservoir to which the measured shale core belongs according to the quartz mineral information, the average particle size of quartz minerals and the average particle size of other minerals; and determining the pressure solution strength grade of the shale reservoir according to a preset shale reservoir pressure solution strength grade, so that the pressure solution strength of the shale reservoir can be quantitatively characterized, the influence mechanism of different pressure solution strengths on pore development is clear, the understanding of the pore development mechanism of deep shale reservoirs is improved, and the exploration and development of shale gas are better guided.

The invention belongs to the technical field of multi-modal combined imaging, and discloses a multi-modal combined imaging method and platform for shale nano-pore three-dimensional reconstruction and connectivity quantitative evaluation, and aims to realize shale organic pore and inorganic pore structure three-dimensional and quantitative characterization based on FIB-SEM and X-CT. High-resolution imaging and chemical imaging technologies are used for revealing the occurrence state of kerogen in mineral particles or at the boundaries, the mineral surface catalyzes organic matter to generate hydrocarbon, and the pore evolution mechanism is regulated and controlled.

The present application belongs to the technical field of hydrogeological parameter measurement, and discloses a kind of spatio-temporal quantitative expression method of hydrogeological parameter based on non-contact observation method, spatio-temporal quantitative expression method of hydrogeological parameter based on non-contact observation method includes: using the NMR geophysical prospecting method of different field source form, GPR geophysical prospecting method, TEM geophysical prospecting method collects different types, different structure geophysical data;And based on the different types, different structure geophysical data inversion obtains hydrogeological parameter;Based on D-S evidence theory, the hydrogeological parameter obtained by inversion is merged with the collected hydrogeological data to obtain the best estimate of the hydrogeological parameter.The present application merges the measured different types, different structure geophysical data and the collected hydrogeological data to obtain the best estimate of the hydrogeological parameter, and quantifies the uncertainty factor.

This invention discloses a CH@SHP099-NPs composite hydrogel, its preparation method, and its applications, belonging to the field of biomedical technology. Addressing the technical challenge of achieving long-term, safe, and precise metabolic intervention with SHP2, this invention constructs a hydrogel system composed of chondroitin sulfate and hyaluronic acid. This system can retain SHP099-loaded nanoparticles within the joint cavity, forming a long-acting intrasynovial drug library, thereby continuously inhibiting the SHP2-AMPK signaling axis. The CH@SHP099-NPs composite hydrogel prepared in this invention can effectively reprogram the metabolic state of macrophages, reduce glycolysis flux, restore synovial immune metabolic homeostasis, and significantly improve synovial inflammation, chondrocyte activity, and cartilage structural integrity in vivo. This invention proposes a mechanism-guided local delivery metabolic intervention strategy, which is expected to provide a new interventional approach for the treatment of osteoarthritis.

The application discloses a multi-agent-based data enhancement method and device, which comprises the following steps: a planning agent acquires task description and data description of sample data, wherein the data description comprises data distribution statistics and data feature details; the planning agent generates an initial zero-sample thought chain according to the data distribution statistics and the data feature details of the sample data, wherein the zero-sample thought chain comprises an intelligent agent type and an execution strategy of each execution level; the planning agent determines an execution level of a calling enhancement strategy intelligent agent based on the execution strategy of the zero-sample thought chain, and generates a new sample; the planning agent evaluates an enhancement effect of the new sample on a downstream deep learning model, and if the enhancement effect does not reach a set condition, adjusts an execution level order of the zero-sample thought chain and a data enhancement step in the execution level, and returns to the step of determining the execution level of the calling enhancement strategy intelligent agent based on the zero-sample thought chain, so as to improve the efficiency of the deep learning model.

The invention relates to an application of an SlCLE3 gene and polypeptide thereof in improvement of salt tolerance of tomatoes. Specifically, the nucleotide sequence of the SlCLE3 gene is as shown in SEQ ID NO: 1. According to the invention, an overexpression vector of the SlCLE3 gene is constructed, and an agrobacterium tumefaciens-mediated genetic transformation method is utilized to obtain a transgenic tomato homozygous strain overexpressed by the SlCLE3. Under the condition of salt stress, the transgenic line shows significantly enhanced salt tolerance, specifically, active oxygen accumulation is reduced, cell damage is reduced, photosynthetic indexes are improved, and relative conductivity is reduced. In addition, the damage of salt stress to tomato plants can be effectively relieved by exogenously applying the pCLE3 polypeptide. The invention provides a new gene resource and a breeding method for tomato salt-tolerant breeding.

The invention provides a simulation material for a lithosphere scale model and application of the simulation material. The lithosphere scale model comprises an upper crust, a lower crust, a lithosphere ground curtain and a soft flow circle ground curtain. The simulation material of the supercrust comprises the following raw materials: silica powder, quartz sand and carborundum; the simulation materials of the lower earth crust and the lithosphere mantle comprise the following raw materials: silica gel, quartz sand and carborundum; the simulation material of the soft flow ring mantle comprises any one of syrup, honey or a sodium iodide solution. The simulation material provided by the invention can quantitatively complete a lithosphere scale structure physical simulation experiment, and scans a lithosphere scale model through an industrial CT technology to obtain a structure evolution process image, so that the structure evolution characteristics are analyzed, the rheological structure and layered stretching deformation characteristics of the lithosphere can be better simulated, and the simulation material can be applied to the field of lithosphere engineering. And an effective technical means is provided for researching development and evolution of lithosphere structural deformation.

The invention discloses a lightweight underwater semantic segmentation method based on fine-grained attention and adaptive feature fusion, and belongs to the field of computer vision and underwater image processing. Aiming at the problems of underwater image illumination attenuation, complex background and difficulty in small target segmentation in aquaculture, the invention provides a lightweight underwater semantic segmentation network and a matched training method, which mainly comprise three core improvements: 1, designing a fine-grained multistage feature attention module; superficial layer feature extraction is enhanced through cascade cavity convolution and space and channel attention mechanisms; secondly, a self-adaptive feature fusion module is introduced, and space details are reserved and redundant features are suppressed in combination with wavelet down-sampling; thirdly, a Gaussian Dice loss function is provided, and the problem of unstable training caused by small target boundary offset is solved through Gaussian weighting. According to the method, the calculation overhead is remarkably reduced, and meanwhile, the segmentation precision in the complex underwater environment is effectively improved.