38results about How to "Easy to explore" patented technology

The invention relates to the technical field of surface acoustic wave devices, and discloses a lithium tantalate single crystal film-based Loff wave photonic crystal resonator, which comprises a functional substrate and a piezoelectric film layer on the upper surface of the functional substrate, an interdigital transducer is arranged in the middle area of the upper surface of the piezoelectric film layer, and photonic crystals are arranged on the two sides of the interdigital transducer; the piezoelectric thin film layer is a lithium tantalate single crystal thin film, and the tangential direction of the lithium tantalate single crystal thin film is any one of Y cuts, including 0-degree Y, 36-degree Y, 42-degree Y, 50-degree Y, 64-degree Y and 128-degree Y cuts. The photonic crystal is used for replacing a reflecting grating in a traditional surface acoustic wave resonator, higher sound velocity and lower longitudinal loss are achieved through a material system constructed by combining the functional substrate and the lithium tantalate single crystal film, and the photonic crystal resonator has the advantages of being simple in structure, small in size and easy to prepare and integrate. And compared with a Bragg reflection grating resonator with the same size, the resonator has a higher Q value.

The invention discloses a closed-loop training device and a closed-loop training method for an in-vitro neural chip to control obstacle avoidance of a trolley, and belongs to the field of neurobiology and in-vitro brain-computer interfaces, and the closed-loop training device shapes a biological neural network structure and regulates and controls a discharge mode in a closed-loop training mode, so that the device makes a correct decision in different obstacle situations to control obstacle avoidance of the trolley. According to the device, by means of sensory electrical stimulation coding, the neuron culture can'sense 'the obstacle condition faced by the trolley in the environment, and then motion control over the external trolley is achieved by collecting and decoding the discharge signals of the neuron. According to the method, training electrical stimulation is adopted to provide feedback for a neural network, a system applies reward or punishment electrical stimulation to the neural network according to whether a trolley successfully avoids an obstacle or not, a synaptic structure and a discharge mode of the neural network are adaptively shaped and optimized, and the safety of the trolley is improved. And a network structure capable of autonomously controlling the trolley to avoid obstacles in different obstacle environments is gradually formed.

The invention discloses a variable resource-considered uncorrelated parallel machine scheduling method for a rocket tank assembly process. According to the method, for multi-variety and small-batch rocket tank assembly tasks, the influence of skill constraints and skill proficiency of workers and cooperation efficiency of worker combinations on scheduling optimization is comprehensively considered, and an irrelevant parallel machine scheduling model with the minimum completion time as the target is constructed. By solving the scheduling optimization model, an optimal scheduling scheme is obtained, reasonable matching among the tasks, the work center and the workers is ensured, the utilization rate of the work center and the cooperation efficiency of the workers are improved, and therefore the completion time of the assembly tasks is effectively shortened. And on the premise that various constraint conditions are met, the optimal scheduling scheme is applied to the rocket tank assembly process, so that the assembly efficiency is improved.

The invention relates to a logic comprehensive sequence optimization space exploration method and system for reinforcement learning based on reward correction, and belongs to the technical field of electronic design automation. A logic comprehensive process is embedded into a reinforcement learning algorithm based on reward correction, and the logic comprehensive process is modeled as a reinforcement learning environment; and the logic comprehensive sequence optimization space is automatically explored. A reinforcement learning algorithm is improved for a logic comprehensive scene, feature extraction and action coding are carried out, the circuit state expression precision is improved, and the search space is expanded. According to the method, by adding award correction, the exploration ability in the early stage of training is improved, a self-adaptive award function is obtained through intelligent agent training, performance reduction caused by limitation of priori knowledge on the circuit optimization direction is reduced, and a better logic comprehensive area result is obtained under the condition that delay constraints are met.

The invention relates to the field of communication warning, in particular to a real-time monitoring system and method based on OLT equipment warning, and the method comprises the following steps: S1, collecting equipment response time, packet loss rate, state information of OLT equipment and disconnection data of a user, and carrying out the preprocessing of the collected data; s2, analyzing and processing the collected data, and identifying a complex fault mode and a multi-factor problem; s3, alarm information is generated for the Pi ng measurement alarm, the port abnormity alarm, the optical modem offline alarm and the comprehensive alarm, operation and maintenance personnel are informed of the generated alarm information through automatic processing, and processing suggestions are provided; and S4, real-time state display is performed through the data acquisition module, and comprehensive data display, historical data query and data interaction are performed through the data processing module. The fault response speed can be improved, the network stability is enhanced, the operation and maintenance efficiency is improved, and the user experience is improved.

The invention discloses a method for evaluating residual stress distribution of a material microstructure. The method comprises the following steps of: 1, acquiring a material back scattering picture; 2, converting the back scattering picture into an RGB (Red, Green and Blue) picture; 3, performing grid partitioning on the RGB image and establishing a finite element model; 4, acquiring mechanical property parameters of the material; and 5, calculating the microstructure residual stress of the material and drawing a residual stress distribution cloud picture. According to the method, the back scattering picture of the material is converted, the finite element model is established, the mechanical property parameters of the material are combined, the microstructure residual stress of the material is calculated, and the microstructure residual stress distribution cloud picture is obtained, so that the residual stress distribution of the microstructure of the material can be intuitively presented; later analysis of material fracture behaviors, exploration of stress concentration problems and regulation and control of material microstructures are facilitated; the microcosmic residual stress of the structure generated by the material in different test states can be efficiently and accurately obtained at low cost.

The invention belongs to the technical field of office automation, and discloses an intelligent office process automation service system and method based on reinforcement learning, and the system comprises a data obtaining module which is used for obtaining office process structure data; the data processing module is used for preprocessing the office process structure data to obtain an office interaction feature data set; the task time evaluation module is used for training according to the office interaction feature data set to obtain a task allocation matching degree prediction model, and predicting an office task allocation matching degree based on the task allocation matching degree prediction model; the work intelligent distribution module is used for performing task distribution on the employees by utilizing an intelligent distribution algorithm according to the predicted office task distribution matching degree; the feedback early warning adjustment module is used for establishing a feedback channel, acquiring the work feedback rate of the employees in real time, and judging whether adjustment early warning information needs to be generated or not according to the work feedback rate of the employees; the ability and task requirements of the employees are matched more reasonably, and the working efficiency is improved.

The invention relates to the technical field of roaming path optimization, in particular to a real-time roaming path optimization method, system and device based on a three-dimensional building scene. The method comprises the steps of positioning position information of a starting point and a target point of a roaming path, generating the roaming path in a three-dimensional scene according to the position information, continuing to carry out shortest path optimization on the generated roaming path, carrying out curve smooth optimization on the generated shortest roaming path, generating a visual roaming file along the optimized roaming path, and outputting the visual roaming file. Roaming scene display conforming to human visual aesthetics is realized; the system comprises a scene modeling and data input module, a path planning module, a path optimization module and a visual processing and display module. According to the invention, a stable and intelligent roaming path is automatically generated in a three-dimensional building scene according to a starting point and a target point; in combination with actual engineering application, manual workload is reduced, manual errors caused by proficiency are eliminated, path generation efficiency is high, and positioning calculation is accurate.

The invention discloses a system for analyzing reverse eye jump and cognitive brain network dysfunction of a Parkinson's disease patient, and relates to the technical field of medical diagnosis. The reverse eye jump data acquisition module is connected with the data collection module; the cognitive data acquisition and evaluation module is connected with the data collection module; the nuclear magnetic resonance imaging module is connected with the data collection module; the data preprocessing module is connected with the nuclear magnetic resonance imaging module; the independent component analysis module is connected with the data preprocessing module; the brain function connection analysis module is connected with the data preprocessing module and the independent component analysis module; and the statistical analysis module is respectively connected with the reverse eye jump data acquisition module, the cognitive data acquisition and evaluation module, the independent component analysis module and the brain function connection analysis module. According to the invention, the medical personnel can explore the relationship between the anti-glancing test and the cognition and brain network dysfunction, and the blank of the prior art is filled.

The invention discloses a system for analyzing double-task cost and brain micro-injury of a Parkinson type multi-system atrophy patient, and relates to the technical field of medical diagnosis, and the system comprises a gait collection module which is used for collecting single-task and double-task gait data of a participant, and processing the gait data to obtain gait feature data of the participant; the gait analysis module is used for analyzing the difference of the participants in the single-task gait aspect and the double-task gait aspect according to the gait feature data and calculating the double-task cost; the nuclear magnetic resonance imaging module is used for collecting nuclear magnetic resonance imaging data of participants; the morphological analysis module is used for performing voxel-based morphological analysis on the nuclear magnetic resonance imaging data and analyzing the correlation between the double-task cost of the participant and the grey matter under the brain curtain; and the diffusion tensor imaging analysis module is used for performing diffusion tensor imaging analysis on the nuclear magnetic resonance imaging data and analyzing the correlation between the double-task cost of the participant and the white matter under the brain screen.

This invention discloses a method for preparing HVC NPs for the diagnosis and treatment of Parkinson's disease and their applications. The probe HV in the HVC NPs responds to H2O2 and viscosity, utilizing near-infrared fluorescence imaging capabilities to achieve real-time monitoring of Parkinson's disease biomarkers, helping us better understand the occurrence and development of Parkinson's disease. With the assistance of RVG29, the HVC NPs effectively cross the blood-brain barrier. The curcumin within them can reduce cell apoptosis, lower oxidative stress levels, scavenge excess reactive oxygen species, combat inflammation, and reduce α-synuclein aggregation, demonstrating effective anti-Parkinson's disease effects at the cellular and experimental animal levels. This nanomaterial provides a new strategy for integrated research on the diagnosis and treatment of Parkinson's disease.

The utility model discloses a laboratory mouse multi-layer three-dimensional cage structure which comprises a supporting bottom plate for feeding laboratory mice, two ends of the supporting bottom plate extend to form mounting plates, two sides of the supporting bottom plate are provided with side plates, the top of each mounting plate is provided with a separation net, each mounting plate is provided with a ventilation plate, and the other side of each mounting plate is provided with a mounting groove. An inserting groove is formed below the mounting groove, a turning cover plate is mounted in the inserting groove, and a supporting layer is inserted in the inserting groove; a water feeder is installed on the side plate, a feeding opening is further formed in the side plate, and a feeding trough is installed outside the side plate; the multi-layer three-dimensional layout is achieved through the slot type supporting layer, the space utilization rate is increased, the water feeder and the feeding system are independent, the requirement of experimental mice for the activity space is met, and separation net holes are matched with the transparent observation plate to form the three-dimensional activity space; due to the inclined plane design of the feeding trough and the sealing structure of the flip head, the feed residue rate is reduced; the inserting groove type supporting layer is matched with a supporting block locking structure, and the excrement leakage risk is completely eradicated.

The present invention provides a camouflaged object semantic segmentation method based on decision-level feature fusion modeling, characterized in that the method comprises: S1 inputting an image to be segmented, using a camouflaged object image with pixel-level labels to supervise the model training, and obtaining the multi-scale feature extraction result of the image to be segmented through the supervised training; S2 based on the multi-scale feature extraction result, the camouflaged object in the image to be segmented is processed by global feature extraction and edge feature guidance to obtain high-resolution mask features; S3 based on the high-resolution mask feature result, the decision-level feature fusion method is used to focus on the high-resolution mask features, and the prediction result of the model is output. The present invention can comprehensively and accurately predict camouflaged objects.

The invention discloses a method and application for evaluating the mineralization potential of granite based on rare earth element characteristics, and relates to the technical field of mineral deposit exploration. The method for evaluating the mineralization potential of granite based on rare earth element characteristics can make full use of the whole rock geochemical characteristics of granite to evaluate the mineralization potential of granite. Starting from the mineralization parent rock of the ion adsorption type rare earth ore in South China, according to the enrichment multiples of rare earth oxides of the ore body and the mineralization granite, the application takes the lowest industrial grade as the evaluation standard, and establishes the whole rock rare earth element characteristic evaluation standard by analyzing the threshold value of the whole rock rare earth oxide content and the ratio of the light and heavy rare earth oxide content, so as to judge the mineralization potential of granite, and can screen out granite with mineralization potential, which is convenient for further exploration, and is conducive to narrowing the exploration scope, shortening the exploration cycle, and reducing material and labor costs; quickly and efficiently screening granite with potential; and greatly improving the accuracy of mineralization prediction.

The application provides a method for identifying the inflammatory activity grade of ulcerative colitis based on dynamic graph multi-instance learning, which comprises data collection, data preprocessing, dynamic graph multi-instance learning model construction for inflammation activity diagnosis and model evaluation. The data collection mainly comprises collecting pathological images of ulcerative colitis patients, digitizing through a digital scanner, and excluding a part of pathological images containing problems such as blur, discoloration and abnormal staining, and finally determining the grading label of each pathological image by senior gastrointestinal pathology experts. The data preprocessing mainly comprises processing the digital pathological images, converting the digital pathological images into image blocks, then using a visual basic model to extract features from each image block, and finally constructing a dynamic graph multi-instance learning model for training. The application promotes the exploration of the internal relationship of the image blocks input into the WSI, and improves the prediction effect of the model.

The application belongs to the technical field of material structure characterization, and relates to an in-situ reaction device based on electromagnetic induction heating and diffuse reflection infrared detection and application thereof. The reaction device comprises an infrared laser, an in-situ reaction cell and an infrared detector. The infrared laser is used for providing infrared laser for a reaction sample of the in-situ reaction cell. The infrared detector is used for detecting diffuse reflection light generated by diffuse reflection after the reaction sample absorbs the infrared laser. A diffuse reflection assembly is arranged between the infrared detector and the in-situ reaction cell. The diffuse reflection assembly is used for collecting the diffuse reflection light generated by diffuse reflection after the reaction sample absorbs the infrared laser into the infrared detector. An electromagnetic induction heating assembly is arranged in the in-situ reaction cell. The application is used for exploring the influence of an alternating magnetic field on intrinsic structure of a material in an electromagnetic induction heating process and research on a reaction path and a reaction mechanism of a catalyst surface reactant.

The present invention discloses a method for dynamic economic and emission optimization scheduling in combination with a plug-in vehicle power generation unit, which belongs to the field of microgrid scheduling and includes the following steps: using the power demand in each time period and the number of plug-in vehicles that can be connected to the power grid as input parameters of the model; constructing a power generation unit optimization scheduling model; using a peak shaving and valley filling algorithm to allocate the charging and discharging of plug-in vehicles, and reasonably allocating the charging and discharging behaviors of plug-in vehicles to different time periods according to the power demand during peak and valley periods; using a multi-objective genetic algorithm to optimize the power generation of the power generation unit; and obtaining the power generation of the power generation unit at each moment and the number of plug-in vehicles connected to the power grid through optimization and solution. The present invention can effectively balance the power load and significantly improve the safety and stability of the power grid operation by coordinating the charging and discharging behaviors of plug-in vehicles with power generation units, discharging plug-in vehicles during peak periods and charging during valley periods through the coordinated scheduling of the charging and discharging behaviors of plug-in vehicles and power generation units.

The invention discloses a composite material hot press molding production scheduling method based on a man-hour uncertainty dynamic transmission mechanism. In order to solve the problems that in the hot press molding process of the composite material, the working hours have uncertainty and are easily transmitted and amplified among multiple batches and multiple devices, a working hour uncertainty dynamic transmission mechanism is constructed, and the propagation rule of the workpiece machining time uncertainty between a batch layer and a machine layer is described; on this basis, constraints including an equipment capacity constraint and a workpiece and equipment matching constraint are comprehensively considered, a scheduling optimization model with minimization of the maximum completion time as an optimization target is established, and an optimization adjustment strategy is deduced; and finally, designing a mixed meta-heuristic algorithm to solve the scheduling optimization model in combination with an optimization adjustment strategy, thereby realizing efficient search of an optimal scheduling scheme. According to the method provided by the invention, the influence of uncertainty on the design of the production scheduling scheme can be effectively reduced, and the scheduling efficiency and stability of the composite material hot press molding process are improved.

The invention discloses a device for simulating the space weathering effect of lunar rock and a simulation method thereof.The device for simulating the space weathering effect of lunar rock comprises a vacuum chamber, a carrying table, a vacuum degree control module, a temperature control module, a pulse laser module, a high-energy ion beam module, a microscope module and a control module; the interior of the vacuum cabin is hollow to form a vacuum chamber; the carrying table is arranged in the vacuum cabin; the vacuum degree control module is arranged on the vacuum cabin; the temperature control module, the pulse laser module and the high-energy ion beam module are all connected with the vacuum chamber; the microscope module is arranged in the vacuum cabin and directly faces the carrying table; the control module is electrically connected with the vacuum degree control module, the temperature control module, the pulse laser module and the high-energy ion beam module. A plurality of modules are integrated and accurately controlled to simulate the space weathering action on the lunar rock, so that a scene under the combined action of a plurality of space weathering factors is realized, and the authenticity of a ground simulation test is improved.

The present invention discloses a method for predicting formation pressure based on seismic data, and relates to the field of petroleum geological exploration and seismic data inversion technology. The method comprises the following steps: S1, determining a prediction formula for formation pressure; S2, preprocessing of seismic velocity data bodies; S3, calculation and normalization of various parameter data bodies; S4, optimization and adjustment of core parameters; S5, verification and improvement of prediction results. The present invention provides a method for predicting formation pressure based on seismic data. The method can be used efficiently over a large area with only seismic velocity data, is suitable for estimating the pressure of dense layers, is simple and efficient, can obtain the trend of pressure changes, and is conducive to the exploration and development of oil and gas resources.

The invention belongs to the field of analytical chemistry, and particularly relates to a living cell in-situ detection method based on glass electrospray ionization mass spectrometry. The system provided by the invention is composed of a glass slide part for culturing cells, a glass electrospray part and a mass spectrum online detection part. The invention provides a novel living cell in-situ detection method based on glass spray ionization mass spectrometry aiming at the vacancy in the single cell in-situ rapid detection direction. The method not only can realize in-situ detection of cells, but also can monitor and analyze change of substances in the cells through atmospheric pressure mass spectrometry; the method fills the technical blank in the field; the method provided by the invention can realize real-time, continuous and online analysis of the sample without pretreatment of the sample, and plays an important role in the aspects of substance structure analysis and mechanism exploration. Therefore, when the probe is used for in-situ detection of living cells, substances in the cells can be better explored, and practical application of the probe is promoted.

The present invention discloses a continuum robot motion control method based on deep reinforcement learning, which applies the deep reinforcement learning DDPG algorithm to the continuum robot motion control, and trains the control strategy by making the robot continuously interact with the environment. The present invention realizes the autonomous motion control of the continuum robot by constructing a virtual training environment for the continuum robot, establishing a DDPG algorithm model, constructing a reward function calculation model, accumulating an experience replay pool, training a deep reinforcement learning network, and deploying the trained model to the actual robot system. The DDPG algorithm in the present invention adopts a deterministic strategy, has good network convergence, reduces excessive Q estimation values ​​by using a target network, and adds random action noise, which is conducive to exploring the optimal strategy; priority experience replay is used during training, which can speed up the training speed and improve the training effect; and through the domain randomization method, the deviation between the training environment and the real environment is reduced, and the success rate of model deployment is improved.

The invention discloses a visualization system based on a multispectral optical fiber. The visualization system comprises an optical fiber bundle system, an imaging system and a sensing system, the optical fiber bundle system comprises a visible light image transmission bundle, an infrared image transmission bundle and an ultraviolet image transmission bundle; the visible light image transmission bundle comprises a visible light beam and a white light guide beam; the imaging system comprises an input end and an imaging end; the front end of the optical fiber bundle system is connected with the input end of the imaging system, and the rear end of the optical fiber bundle system is connected with the imaging end of the imaging system. The imaging system is arranged, can replace the function of an electronic camera to achieve the watching effect, and is suitable for some environments which are sensitive to electromagnetic interference and cannot use electronic equipment; through multi-band setting, integrated observation from ultraviolet light to visible light to infrared light can be realized, and safer and more visual images can be brought to technicians. Meanwhile, the ultraviolet image transmitting bundle is arranged, so that electric power accidents are reduced; and the epoxy resin layer is arranged between the optical fiber bundle system and the flange connecting part, so that good air tightness is ensured.

This invention relates to a dieless spinning machine and spinning process, belonging to the field of dieless spinning machines. The dieless spinning machine and spinning process of this invention include a chuck mechanism, a spindle mechanism, a tailstock mechanism, at least two sets of external spinning mechanisms, and at least two sets of internal spinning mechanisms. Each set of external spinning mechanisms includes an external spinning wheel, and each set of internal spinning mechanisms includes an internal spinning wheel. The internal and external spinning wheels perform synchronous interpolation movements along the axial and radial directions. The radial gap between the internal and external spinning wheels is controlled and matched in real time according to the axial variation of the preset wall thickness of the wheel rim blank. Dynamic pressure support replaces the fixed mandrel, working together to achieve dieless spinning forming of the wheel rim blank. This invention has advantages such as significantly reducing production costs, improving production flexibility, improving forming quality, expanding the range of applicable materials, and facilitating automation integration.

The invention relates to the technical field of surface acoustic wave devices, and discloses a lithium niobate single crystal film-based Loff wave phononic crystal resonator, which comprises a functional substrate and a piezoelectric film layer located on the upper surface of the functional substrate, an interdigital transducer is arranged in the middle area of the upper surface of the piezoelectric film layer, and phononic crystals are arranged on the two sides of the interdigital transducer; the piezoelectric thin film layer is a lithium niobate single crystal thin film, and the tangential direction of the lithium niobate single crystal thin film is any one of X-cut, including X-cut and Y-direction of-15 degrees to 15 degrees. The photonic crystal is used for replacing a reflecting grating in a traditional surface acoustic wave resonator, higher sound velocity and lower longitudinal loss are achieved through a material system constructed by combining the functional substrate and the lithium niobate single crystal film, and the photonic crystal resonator has the advantages of being simple in structure, small in size and easy to prepare and integrate. And compared with a Bragg reflection grating resonator with the same size, the resonator has a higher Q value.

The invention provides a flood control scheduling method combining a convolutional neural network and Monte Carlo tree search, which comprises the following steps of: firstly, collecting and sorting original data of historical flood control scheduling and simulated flood control scheduling, and obtaining a flood characteristic matrix data set by using a designed data conversion method; the flood characteristic matrix data set is used for training two convolutional neural networks, namely a strategy network and a value network, the strategy network is used for predicting the delivery flow, and the value network is used for evaluating the comprehensive flood control effect; and finally, taking the optimal comprehensive flood control effect as a target, and combining Monte Carlo tree search with a strategy network and a value network to find an optimal flood control scheduling scheme. When the method is applied to an actual flood control dispatching process, the outbound flow of the dispatching process is stable, and flood with different frequencies can be coped with; meanwhile, the flood control dispatching effect is improved to a certain extent.

The present invention discloses a composite pharmaceutical preparation, a preparation method and an application thereof. The composite pharmaceutical preparation comprises the following components by weight: 50 to 70 parts of neomycin sulfate, 10 to 30 parts of ursolic acid and 10 to 30 parts of poplar flower extract. The composite pharmaceutical preparation has excellent delaying and blocking effects on the disease of Babylonia snails, and can significantly reduce and delay the mortality rate of Babylonia snails with the disease.

The invention discloses a dynamic economy and emission optimization scheduling method combined with a plug-in type automobile power generation unit, which belongs to the field of micro-grid scheduling and comprises the following steps of: taking electric quantity requirements of each time period and the number of plug-in type automobiles capable of being accessed to a power grid as input parameters of a model; constructing a power generation unit optimization scheduling model; a peak clipping and valley filling algorithm is adopted to distribute charging and discharging of the plug-in type automobile, and charging and discharging behaviors of the plug-in type automobile are reasonably distributed to different time periods according to the electric quantity requirements of the peak period and the valley period; using a multi-target genetic algorithm to optimize the power generation amount of the power generation unit; and through optimization solution, the generating capacity of the generating unit at each moment and the number of plug-in automobiles connected to the power grid are obtained. According to the invention, through the charging and discharging behaviors of the plug-in type automobile and the cooperative scheduling of the power generation unit, the plug-in type automobile is discharged in the peak period, and is charged in the valley period, the power load can be effectively balanced, and the safety and stability of the operation of the power grid are obviously improved.

A reinforcement learning system, method, and computer program code for controlling an agent to perform multiple tasks while interacting with an environment. The system learns options, where an option comprises a sequence of primitive actions performed by the agent under the control of an option policy neural network. In implementations, while the agent interacts with the environment, the system discovers options useful for multiple different tasks by a meta-learning reward for training the option policy neural network.

The present invention provides a ship navigation track prediction method and system based on AIS, which relates to the technical field of ship navigation track prediction, including: taking the parameters to be optimized by the vSVR model as genes and randomly generating an initial population, obtaining the ship AIS data and inputting it into each vSVR model individual of the initial population for training; dividing the initial population into multiple evolutionary groups according to the similarities and differences of the individual output prediction values, and determining the individual fitness value according to the deviation between the prediction value and the actual value; using the fitness value as the basis for selecting operations in the genetic algorithm, performing independent evolution and overall evolution alternately on the evolutionary group, and iterating multiple times to determine the navigation track prediction model; inputting the AIS data of the ship to be tested into the navigation track prediction model to perform navigation track prediction. The present invention finds the optimal model parameters by dividing the evolutionary groups, performing independent evolution alternately on the evolutionary groups, and transmitting the settings of excellent individuals and overall evolution, thereby improving the prediction accuracy of the ship's navigation track.