30results about How to "Simulate the real" patented technology

The invention provides a system frequency response estimation method and system based on a fault position, and belongs to the technical field of new energy power grids, and the method comprises the steps: constructing a new energy multi-operation-state model; solving the node voltage of the state model after the fault based on the superposition principle and iterative calculation, establishing a mapping model of the fault position and the low-cross cluster, obtaining the initial value of the active power vacancy, and comprehensively considering the difference between the initial value of the active power vacancy and the active recovery rate between stations at different fault positions. Calculating the system-level total active power vacancy considering the station low-voltage crossing characteristic difference; and constructing a system frequency response model based on the frequency modulation capability difference caused by the fault position, and estimating the system frequency response by adopting the response model. According to the invention, quantitative description of active power vacancy and system equivalent frequency modulation capability loss after disturbance at different fault positions is realized, so that the accuracy of frequency response analysis is improved.

This application discloses a simulation device and method for multiple unidirectional impact load tests under gradient confining pressure, comprising a track, a specimen, and several impactors. The track includes a horizontal section, one end of which is connected to a ramp section. The specimen is located on the horizontal section, and a gradient confining pressure device is provided outside the specimen to pressurize it. Several impactors, acting as power input sources, are released from the ramp section at different heights above the ground and move along the track, sequentially impacting the specimen to apply multiple unidirectional impact loads. The impactors can be nested sequentially so that subsequent impactors can pass over earlier impactors, avoiding collisions between them. This application enables the specimen to be subjected to gradient confining pressure and multiple unidirectional impact loads to simulate the blasting damage effect of multiple blasting loads on surrounding rock with continuously changing stress states in underground engineering.

The invention discloses a photo-induced phase change protein and application thereof. The protein with the phase change function is provided for the first time, and the amino acid sequence of the protein is shown as SEQ ID NO.7; the invention further provides the IDR polypeptide with the phase change function, and the amino acid sequence of the IDR polypeptide is shown as SEQ ID NO.5. The invention also provides an application of the PE1 protein in photo-induced phase transition. The tool has the advantages of being high in space-time precision, good in biocompatibility, capable of integrating physiological signals and the like, can be used for controllable phase separation research, local enrichment, material construction and the like, and has wide prospects in the fields of life science research, drug development and synthetic biology.

The invention discloses a foam curtain generation and performance characterization device and method and application. The device comprises a foam curtain injection system used for generating a foam curtain, a combustion system used for heating the foam curtain and a motion capture system used for capturing the motion trail of the foam curtain. The foam curtain injection system comprises a foam mixed liquid tank body, a pump body and a foam curtain injector, the pump body is arranged between the foam mixed liquid tank body and the foam curtain injector, and the pump body is connected with the foam curtain injector through a pipeline; the combustion system comprises a combustion gas cylinder, a combustion improver cylinder, a combustion mixer and a combustor, the combustion gas cylinder and the combustion improver cylinder are respectively connected with the input end of the combustion mixer through pipelines, and the combustor is connected with the output end of the combustion mixer. The device can monitor the motion trail of the marking section foam curtain in real time and control the angle of combustion flame, so that the marking section foam curtain receives heat radiation all the time, the actual working condition is simulated more truly, and the actual heat insulation effect is reflected.

This invention relates to the field of experimental equipment technology, and particularly to an NMR combustion-explosion dynamic load multi-field coupling experimental device and method. The technical solution includes a control module, an impact module, a detection module, and a vibration reduction module; wherein, the control module includes an operating platform, a transmission line, an engine, a fuel tank, an oxygen tank, connecting pipes, a high-temperature resistant fuel nozzle, and a high-temperature and pressure resistant gas sealing pipe; the operating platform is located behind the engine. This invention achieves graded control of dynamic load through combustion-explosion-electromagnetic composite loading, integrates simultaneous detection of multiple physical fields such as nuclear magnetic resonance, acoustic waves, and resistivity, sets up multi-level vibration reduction and dissipation structures to absorb residual kinetic energy, adopts a high-temperature and high-pressure resistant sealing design to ensure safety, simulates ground temperature, ground stress, and pore fluid fields to reproduce the deep in-situ multi-field coupling environment, and is centrally controlled and automatically reset by the operating platform, thereby realizing a laboratory simulation of deep dynamic disasters with flexible and adjustable parameters.

A cloud-native resource scheduling optimization method based on predictive perception and deep reinforcement learning, the specific process includes: collecting cluster operation data and constructing scheduling samples; generating future load prediction results based on historical load sequences; constructing a prediction-enhanced scheduling state representation; determining the target node using continuous preference output and action mask; performing task allocation and life cycle management; updating the scheduling policy based on a multi-objective reward function; achieving policy convergence through experience replay and target network updates; online deployment and incremental reasoning. The present application introduces the future load prediction results of the nodes into the scheduling state representation, so that the scheduling decision is not only based on the current node resource state and task demand, but also combines the resource pressure change trend in the future period of time, thereby improving the early identification ability of the scheduler for potential high-load nodes.

The application belongs to the field of power system management, and specifically discloses a charging load probability space-time distribution prediction method considering random fluctuations of commuting demand. A multi-dimensional travel strategy set is constructed to simulate the comprehensive decision of users in energy, time and space. Uncertainty is introduced from the demand side, and Monte Carlo sampling is used to generate demand fluctuation samples to describe the random fluctuations of the load from the root. An utility evaluation model is constructed based on cumulative prospect theory to correct the limited rationality of users. A multi-layer nested discrete choice model is used to divide the decision into three levels of charging, departure and path, and a heuristic algorithm is used to iterate to traffic network equilibrium. In the equilibrium state, the strategy selection proportion is combined with the demand sample to input the dynamic traffic network model, and the space-time probability distribution of the charging load is output. The probability space-time distribution of the evening peak electric vehicle charging load is more accurately predicted, and a scientific basis is provided for charging facility planning and power grid dispatching.

The application discloses a protein conformation change simulation method based on vector elastic network dynamics, and belongs to the cross field of biophysical simulation and computational mechanics. Protein atom structure is discretized into a coarse-grained particle elastic network model; a particle dynamics control equation is established; a path unit concept is introduced to perform time domain incremental linearization processing on a simulation process; particle displacement is solved based on a central difference method; virtual reverse motion is used to decouple rigid body displacement and pure deformation of the particle; internal force vectors in a local coordinate system are solved based on a virtual work principle; the internal force is returned to a global coordinate system through a coordinate conversion matrix and the particle state is updated; external loads are applied, and continuous dynamics trajectories of the protein are iteratively calculated and output. The application solves the problems that traditional elastic network models cannot process large-scale nonlinear conformation changes of proteins and molecular dynamics simulation is low in efficiency, and can accurately capture non-harmonic motion and transient response of the protein when performing a biological function.

The invention provides a test method and device based on a dynamic network environment, and relates to the technical field of vehicle-mounted application test.The method is applied to electronic equipment, the electronic equipment is provided with an automatic test tool, and the framework of the automatic test tool comprises a network simulation layer, a test case execution layer and a test verification layer; the method comprises the steps that a test case execution layer obtains a target test case used for testing a target vehicle-mounted application; the test case execution layer analyzes the target test case, and generates M page operation instructions in one-to-one correspondence with the M page operations and N network switching instructions in one-to-one correspondence with the N network requirements; the test case execution layer sequentially executes the M page operation instructions and the N network switching instructions according to the operation time sequence so as to test the target vehicle-mounted application in the target abnormal network environment indicated by at least one network requirement in the N network requirements.

The invention relates to the technical field of processor soft error fault analysis, and discloses a processor soft error fault mode analysis system which comprises a fault generation module, a fault injection module, a data receiving module, a data processing module, a result display module and a controlled operation module. The fault generation module, the fault injection module, the data receiving module, the data processing module, the result display module and the controlled operation module work cooperatively to jointly establish a fault mode-system response-task influence association model; the system surpasses traditional single-point fault injection, provides highly configurable and reproducible fault excitation, simulates soft errors of a processor caused by space radiation, supports in-situ test capability of multi-physics field coupling, realizes application-oriented closed-loop correlation modeling, supports systematic test and deep analysis, and has a wide application prospect. And a full-digital efficient analysis assembly line is constructed, so that the evaluation efficiency and the iteration speed are greatly improved.

The application belongs to the field of aircrafts, and particularly relates to a multi-blade tilt-rotor aircraft ground vehicle-mounted wind blowing test mechanism and a test method. The test platform expands the existing aircraft ground test method, can more truly simulate various states of the aircraft in flight, and the response states are closer to the aircraft in flight. Meanwhile, compared with the wind tunnel test, the aircraft is in a moving state, and the air is in a natural state, which is closer to the real flight environment of the aircraft, and provides a more reliable test scheme for the aircraft ground test stage.

The invention relates to the technical field of generator excitation simulation, and discloses an RT-LAB-based generator excitation small signal semi-physical simulation method and detection system, and the simulation method comprises the following steps: building a single machine infinite bus simulation system and a generator excitation system model in RT-LAB software; a semi-physical simulation platform loaded with RT-LAB software is electrically connected with the AVR excitation regulator through a conversion card; based on the generator excitation system model established in RT-LAB software, determining the optimal working parameters of the PSS of the AVR excitation regulator; and based on the optimal working parameters of the PSS of the AVR excitation regulator, adjusting the control parameters of the single-machine infinite-bus simulation system and the generator excitation system model so as to evaluate the response performance of the AVR to various generator faults. According to the method, a small signal model is built through RT-LAB, and the test accuracy is ensured by utilizing the real-time capability of the step length of 15 microseconds.

The utility model discloses a wax deposition determination flash filter relates to the petroleum chemical industry detection equipment field, including filter main part, two groups of filter inlet joint, bar hollow support, filter screen and netted support, through the combination design of bar hollow support and filter screen, netted support makes filter screen installation stable, can effectively resist high pressure fluid impact, prevents the collapse, through the pull bar hollow support, takes out the whole filter screen and netted support whole, has realized the quick sampling of deposit and the convenient replacement of filter screen, and filter screen design can replace, makes the device can nimblely adapt to the multiple crude oil system of different wax crystal size, promotes the experimental flexibility and repeatability, provides high reliability data support for wax deposition mechanism research and prevention and control technology development, widens the application range.

The utility model relates to a kind of sunlight simulation device and lifting sunlight lamp, comprising: light assembly, including first light source and multiple second light sources, multiple second light sources are arranged along circumference interval, and first light source is located at the central position of multiple second light sources;Light guide plate is attached with light assembly, and light guide plate is fixedly provided with light collecting cylinder, and light collecting cylinder is opened with two ends through light guide hole along axial direction, and first light source is located at one end of light collecting cylinder, and the light of first light source is emitted outward along light guide hole.The utility model is to solve the problem that the simulation effect of traditional simulation device to sunlight is poor and the structure is complex, and the maintenance cost is higher.

This utility model relates to a microscopic observation device for soil particles in shaking table tests, belonging to the technical field of civil engineering experimental equipment. It includes a vibrating base that supports the entire device and provides controlled vibration power; a fully transparent model box is mounted on the vibrating base to hold the soil particle sample to be observed, and the bottom layer of the fully transparent model box has symmetrically arranged wave-absorbing sponge layers on its left and right sides; a miniature camera assembly is mounted on the vibrating base, vibrating synchronously with the fully transparent model box; the miniature camera assembly includes a camera with high-speed continuous shooting capability, and the frame rate of the miniature camera assembly is not less than 300fps; a horizontal electric guide rail is mounted on the vibrating base; and a vertical electric lifting column is mounted on the horizontal electric guide rail. This utility model effectively eliminates the relative displacement problem caused by the asynchronous movement of the camera equipment and the model box in shaking table tests.

The invention relates to a transient detection method for the thickness of deposited ash. The method comprises the following steps: firstly, acquiring actually measured temperatures of a first thermocouple arranged on the inner side of a pipe wall and a second thermocouple arranged in the pipe wall and a flue gas temperature; on the basis of the temperatures, thermophysical parameters under the current working condition are calculated, an assumed ash deposition thickness is initialized, and then the convective heat transfer coefficient between the flue gas and the ash layer is solved. Then, a transient temperature field control equation from the pipe wall to the outer surface of the ash layer is established in combination with the assumed thickness and the heat exchange coefficient; and iteratively solving the equation through an explicit central difference method to obtain the simulated temperature of the second thermocouple position. And finally, comparing the simulated temperature with an actual measurement value of the second thermocouple: if an error is smaller than a set threshold value, outputting the current assumed thickness as a detection result; otherwise, according to the condition that the simulation value is higher than or lower than the measured value, the assumed ash deposition thickness is correspondingly increased or decreased, and the calculation process is repeated until convergence is achieved. According to the invention, starting from unstable heat conduction, the heat conduction process under variable working conditions can be simulated more truly, and transient and accurate online detection of the soot deposition thickness is realized.

The application provides a system frequency response estimation method and system based on fault location, and belongs to the technical field of new energy power grids, and comprises the following steps: constructing a new energy multi-operation state model; solving the node voltage after the fault based on the superposition principle and iterative calculation on the state model, establishing a mapping model of the fault location and the low penetration cluster, obtaining the initial value of the active power shortage, and comprehensively considering the initial value of the active power shortage between stations and the difference in active power recovery rate under different fault locations, and calculating the total active power shortage of the system level considering the difference in the low penetration characteristics of the stations; constructing a system frequency response model based on the difference in the frequency modulation capability caused by the fault location, and estimating the system frequency response by using the response model. The application realizes quantitative description of the active power shortage after disturbance of different fault locations and the loss of equivalent frequency modulation capability of the system, so that the accuracy of frequency response analysis is improved.

The invention discloses a test scene intelligent simulation method, system and device for a real-time control system and a medium, and the method comprises the steps: receiving a test demand, carrying out the analysis of the test demand based on a large language model and a test domain knowledge graph, and generating a structured test demand; calling a corresponding generative artificial intelligence model, and generating a cross-modal test data set associated in time sequence and semantics; based on a structured test requirement, a cross-modal test data set and a test domain knowledge graph rule, automatically generating a test script and arranging a time sequence test scene process so as to drive the real-time control system and collect response of the real-time control system; and analyzing a test execution result, optimizing decision logic of data generation or protocol simulation through reinforcement learning based on the analysis result, and updating the decision logic to the test domain knowledge graph. Compared with the prior art, full-link automation of real-time control system testing from scene construction and execution to autonomous evolution is achieved.

The invention provides a housing policy intelligent decision auxiliary method and system based on a large language model, and belongs to the technical field of housing policy decision support, and the method comprises the steps: constructing a market theory hypothesis based on a housing filtering mechanism; inputting proxy attributes and market variable data; based on an intelligent agent modeling technology, a simulation model for simulating housing decision behaviors of different income groups is constructed and operated, and market dynamics are deduced under various policy scenarios; key indexes such as transaction activeness, housing quality distribution and group structure in the simulation process are extracted and visualized; and finally, inputting the structured indexes into a large language model, and generating an intelligent decision-making auxiliary text containing scene analysis, result induction and policy suggestion. According to the method, the dynamic deduction ability of computational simulation and the semantic understanding ability of artificial intelligence are fused, refined and intelligent evaluation of the housing policy effect can be realized, and the scientificity and the foresight of housing policy design are effectively improved.

The utility model relates to the technical field of automatic testing, and discloses vending cabinet full-automatic testing equipment, which comprises a frame, a limiting piece, a reset piece, a lifting frame, a driving device and a control assembly, the lower side part of the frame is provided with a placing cavity, the limiting piece is connected with the frame, and a lock catch is suitable for being placed on the limiting piece in a sliding manner; the reset piece is arranged on the limiting piece and is suitable for abutting against the lock catch, and the lifting frame is movably arranged in the frame; the driving device is fixed on the frame; the driving device is connected with the lifting frame, and the driving device is configured to drive the lifting frame to move downwards to push the lock catch to overcome the bias pressure of the reset piece to enter the cabinet lock or drive the lifting frame to move upwards to be separated from the lock catch; the control assembly is electrically connected with the driving device and the cabinet lock so as to control the driving device and the cabinet lock to generate corresponding actions. The vending cabinet full-automatic test equipment has remarkable advantages in the aspects of dynamic characteristic test capability, parameter control precision, data acquisition and analysis and the like.

The invention provides a heavy ion irradiation test terminal for simulating a wood star environment and a test method thereof, and belongs to the technical field of space environment ground simulation. The problems that in the prior art, simulation factors are single, the simulation factors are seriously disjointed with the actual environment, and step-by-step tests cannot reveal the synergistic effect and the acceleration failure mechanism are solved. The device comprises a three-dimensional Helmholtz coil, a vacuum cabin, a vacuum high-low temperature sample table and a vacuum four-dimensional displacement table, the three-dimensional Helmholtz coil is arranged outside the vacuum cabin and used for generating a three-dimensional controllable magnetic field in a sample area in the vacuum cabin, and the vacuum four-dimensional displacement table is located in the vacuum cabin and used for generating a three-dimensional controllable magnetic field in a sample area in the vacuum cabin. And a vacuum high-low temperature sample table is mounted on the vacuum four-dimensional displacement table. Through the unified comprehensive monitoring and control system, the environment parameters and the irradiation parameters are subjected to time sequence programming and synchronous control, a complex task environment profile can be reproduced, and methodology upgrading from single-factor superposition to multi-physics coupling test is achieved.

This application discloses a bench test system and method, belonging to the field of testing technology. The bench test system includes a heat exchange mechanism, a first connecting pipe, a second connecting pipe, a first valve, and a flow detector. The second heat exchange pipe of the heat exchange mechanism contains a heat exchange fluid. The inlet end of the first connecting pipe is adapted to communicate with the coolant outlet of the engine under test, and the outlet end of the first connecting pipe is connected to the inlet end of the first heat exchange pipe of the heat exchange mechanism. The inlet end of the second connecting pipe is connected to the outlet end of the first heat exchange pipe, and the outlet end of the second connecting pipe is adapted to communicate with the coolant inlet of the engine under test, wherein the coolant outlet is connected to the coolant inlet. The flow detector is installed on either the first or second connecting pipe. The first valve is installed on either the first or second connecting pipe, wherein the opening degree of the first valve is adjustable. Using the technical solution of this application can improve the accuracy of engine bench test results.

The invention relates to the technical field of rock damage simulation, in particular to a rock crack propagation simulation method under the action of pulse fracturing, which comprises the following steps: introducing a phase field variable, and constructing a phase field method model; establishing a friction damage physical mechanism based on a rate state dependent friction law and a sliding law; introducing a friction damage physical mechanism into the phase field method model, establishing a coupling relationship between a friction state variable and the phase field method model, and constructing a phase field-friction damage coupling control equation set; constructing a rock loading model under the action of pulse fracturing; solving a phase field-friction damage coupling control equation set by adopting an adaptive alternating iteration algorithm introducing a dynamic local grid; a friction damage physical mechanism is introduced into a phase field method model, a rock crack network under the action of pulse fracturing can be simulated more truly and accurately, a dynamic local grid adaptive method is adopted, and waste of computing resources is reduced while rock crack propagation simulation precision is guaranteed.

The invention particularly relates to a nuclear power plant simulator communication system simulation method based on an NS-3 simulator, and belongs to the field of industrial control system communication simulation, and the method comprises the following steps: S1, network structure modeling: establishing an acquisition node, a transfer node and an operator station node in the NS-3 simulator, and establishing a communication link between the nodes; a network bridging function is configured on the transfer node; s2, protocol and communication simulation: adopting a CSMA protocol to simulate local area network communication, and configuring link parameters; simulating a ZMQ publishing and subscribing mechanism in a UDP (User Datagram Protocol) broadcasting mode; setting data processing logic at the transfer node and introducing processing delay; s3, performance test and analysis: simulating data flow under different data generation periods and different network working conditions, and monitoring and recording key performance indexes of a communication link; in the simulation process, the stability and robustness are tested by improving the load. According to the invention, end-to-end simulation and performance testing of the full-range analog machine communication system of the nuclear power plant are realized.

The invention relates to a construction method and application of a plateau pulmonary hypertension rat model, and belongs to the technical field of animal disease model construction. The method comprises the following steps: firstly, selecting healthy male SD (Sprague Dawley) rats which are 5-6 weeks old and have the weight of 220 + / -10g, and adaptively feeding the rats in a standard SPF (Specific Pathogen Free) animal room for 7 days; dividing the rats into a healthy control group and a plateau hypoxia group; feeding the rats of the healthy control group in a normal-oxygen normal-temperature standard environment for 28 days; placing the rats of the plateau hypoxia group in the multifunctional plateau environment experiment system for intermittent exposure for 28 days; the intermittent exposure method specifically comprises the following steps: placing the rats in a simulated plateau low-oxygen-low-temperature-low-pressure composite stress environment in 12 hours of a daily illumination period; if the dark period is 12 hours, the rat is moved out to a normal-oxygen normal-temperature standard environment to be recovered. The rat model shows typical pulmonary arterial hypertension characteristics, all indexes are stable and repeatable, and a reliable basis is provided for research.

The utility model discloses a co-extrusion plastic wood grain shaping device which comprises a compression molding structure, capsule-shaped groove bodies 112 are symmetrically formed in the positions, on the input side and the output side of the upper side of a conveying belt, of a front wall plate and a rear wall plate of a conveyor in the compression molding structure, and a first bearing seat is arranged in the groove body 112 on the input side in a sliding mode and rotationally connected with a veneering rolling pipe; the veneering rolling pipe is communicated with the heat preservation liquid channel through a heat preservation connecting pipe; a second bearing seat is arranged in the output side groove body 112 in a sliding mode, fixes a spring piston and is rotationally connected with an embossing roller, a lifting hole channel where the spring piston is located is communicated with the heat preservation liquid channel, and a rolling and sticking brush is arranged on the upper side of the embossing roller and movably arranged above the output side of the conveyor through a third bearing seat. The device can dynamically adjust the pressure of the embossing roller in real time according to the temperature change of a co-extrusion material through the cooperation of the veneering rolling pipe, the heat preservation liquid channel and the spring piston, solves the texture problem caused by the temperature change in the traditional process, and improves the texture quality and stability of a co-extrusion plastic wood product.

The invention relates to the technical field of cell culture, and discloses a tracheal epithelial cell and application thereof, the cell is a guinea pig tracheal epithelial immortalized cell, and the preservation number is CGMCC No. 46763. In the invention, the cell has good susceptibility to respiratory pathogens, and has excellent transfection efficiency. The cells can be used for separating or culturing respiratory pathogens, preparing respiratory pathogen vaccines, screening and / or preparing medicines for preventing and / or treating diseases caused by the respiratory pathogens, serving as host cells for researching the respiratory pathogens, detecting the respiratory pathogens for a non-diagnostic purpose and constructing organoids. The method has high scientific research value and practical application potential.

The present application relates to the field of rock mechanics and explosion impact technology, in particular to a numerical simulation method and system for the breaking process of ore rock under explosion load based on FDEM, comprising the following steps: using finite element grid software Gmsh to construct an ore rock specimen geometric model and grid division to obtain a triangular grid; setting material parameters and explosion load boundary conditions according to the ore rock specimen geometric model and the triangular grid; simulating the explosion load application process based on the parameters and the boundary conditions to obtain explosion load simulation information, and deriving and establishing a normal contact force and a tangent contact force calculation function according to the simulation information; based on the above calculation function and the updated node coordinates and velocity of the triangular element, obtaining the updated coordinate and velocity information of the triangular element nodes, and then generating a cloud chart, counting the crack condition and analyzing the energy change to obtain the crack propagation, stress field distribution and energy evolution results of the ore rock under the explosion load. The present application provides a quantitative basis for the selection of on-site blasting parameters, and improves the blasting efficiency and accuracy.

The invention relates to the technical field of shoe measurement, in particular to a shoe sole anti-skid performance testing device which comprises a shell, a testing platform, a movable frame and a shoe tree. The testing platform is slidably installed on the shell in the horizontal direction. The movable frame is hinged to the shell, and a pressure sensor used for sensing the pressure of the movable frame in the first direction is arranged between the movable frame and the shell. The shoe tree is sleeved with the experiment shoe and comprises an instep portion and a sole portion, the instep portion is installed on the movable frame, the movable frame enables the angle of the shoe tree to be adjustable, and then the contact mode of the experiment shoe and the testing platform is adjustable. The sole part of the shoe tree is divided into a plurality of mutually independent blocks, so that the proper blocks can be selected to step on the experimental shoe according to the stress states of different foot types. The rotating frame rotates to adjust the contact area of the experiment shoe and the test platform so as to switch the front sole, the heel or the whole sole to be in contact with the test platform, the anti-skid performance in different states during walking can be simulated, and the actual use scene is better met.

The application relates to a collision risk early warning method and system based on an unmanned ship front-view camera, wherein the method comprises the following steps: based on the front-view camera collected image data in the real scene and the virtual scene, a real-virtual fusion collision risk data set is constructed; under the parameter combination of multiple main ship lengths, widths, speeds, braking performances, drifting performances and personnel reaction times, risk range parameters are calculated; the risk level of each collision risk image in the real-virtual fusion collision risk data set is judged and labeled through the relationship between the obstacle position and the risk range parameters; based on the labeled real-virtual fusion collision risk data set, a collision risk early warning model is trained through a convolutional neural network architecture; the trained model is deployed into an unmanned ship computing terminal, image data returned by the unmanned ship camera is processed in combination with the motion state of the unmanned ship, and collision risk early warning is realized. Compared with the prior art, the application has the advantages of high early warning accuracy.