100results about How to "Rich data base" patented technology

The invention relates to the technical field of unmanned aerial vehicles, in particular to a mooring unmanned aerial vehicle safety management system which comprises a multi-source data acquisition module, a data fusion and analysis module, an intelligent risk assessment module, a control instruction generation module and an execution mechanism module. The multi-source data acquisition module acquires flight and mooring cable data of the unmanned aerial vehicle by using various sensors, and transmits the flight and mooring cable data to the data fusion and analysis module through a time division multiple access protocol with low packet loss rate and low delay; the intelligent risk assessment module judges a risk level by combining a risk assessment model based on a support vector machine and a Bayesian network algorithm according to the information, and generates prompt information if the risk level exceeds a preset range; the control instruction generation module generates instructions according to a risk prompt and a preset strategy, ranks the instructions and sends the instructions to the execution mechanism module; a flight control system in the execution mechanism module cooperates with a mooring cable electric take-up and pay-off device and the like, real-time adjustment and control over the flight attitude of the unmanned aerial vehicle and the cable state are achieved, and it is ensured that tasks are safely and stably executed in the complex environment.

The invention provides a lithium battery anomaly detection method and system based on deep learning, and relates to the technical field of data processing, and the method comprises the steps: obtaining an original data set of a lithium battery in a plurality of working states; performing multi-layer preprocessing on the original data set to obtain a basic data set; performing step-by-step feature extraction on the basic data set to obtain a feature data set; dividing the characteristic data set into a charging state data set, a discharging state data set and a static working state data set according to the working state of the battery; obtaining a decomposition data set according to the state data set; constructing a multi-dimensional state model of the lithium battery, and inputting the decomposition data set into the multi-dimensional state model to obtain a matching result set; performing comprehensive analysis according to the matching result set, and generating anomaly discrimination data; according to the anomaly discrimination data, analyzing detection information representing an anomaly position and severity, and generating anomaly identification information; the method has a good anomaly detection effect.

The invention provides a lottery store violation abnormity identification method and system, and relates to the technical field of image identification, and the method comprises the steps: collecting multi-source heterogeneous image data in real time through an edge computing device; preprocessing the collected multi-source heterogeneous image data to form a stereo image data set containing space, depth and thermodynamic information; constructing a multi-task deep learning model, wherein the multi-task deep learning model comprises a face recognition module, a behavior analysis module and an environment detection module; inputting the three-dimensional image data set into the trained multi-task deep learning model, and executing face recognition, behavior analysis and environment detection tasks at the same time; based on an execution result of the multi-task deep learning model, generating a violation score by applying a fuzzy comprehensive evaluation method, and weighting by considering historical violation records; and generating a lottery store site investigation plan based on the violation scoring result.

The invention discloses a multi-device cooperative management system based on cloud edge cooperation, and relates to the technical field of Internet of Things communication, and the system comprises a device management module which is used for connecting edge devices, carrying out the centralized management of the edge devices, and carrying out the optimal scheduling based on the functions and distribution positions of different edge devices; and the data acquisition module is connected with the equipment management module. According to the multi-device collaborative management system based on cloud edge collaboration, a comprehensive and accurate data basis is provided for multi-device collaborative management, and information exchange and collaborative work between devices are promoted; the optimization decision-making unit can formulate an equipment optimization management strategy according to the operation scheme generated by the big data analysis unit and the prediction result obtained by the prediction analysis unit, and continuously performs optimization adjustment according to the real-time operation condition of each piece of edge equipment; therefore, the system can adapt to continuously changing equipment operation environments and task requirements, and more efficient multi-equipment collaborative management is realized.

The invention relates to the technical field of artificial intelligence, in particular to a charging pile construction planning method and system based on artificial intelligence. The method comprises the following steps: acquiring existing power infrastructure data and multi-source heterogeneous data, wherein the multi-source heterogeneous data comprises traffic data, geographic data, vehicle data and charging behavior data; performing time sequence feature extraction on the multi-source heterogeneous data, and performing cross-time scale fusion on different data source features by using a preset hierarchical attention network model to obtain factor weight data; and constructing a long-term charging demand prediction model through a sequence-to-sequence model according to the factor weight data and the multi-source heterogeneous data. According to the site selection method based on the risk minimum set coverage problem, initial site selection scheme data is generated by combining multi-objective optimization of the coverage rate, the service radius and the accessibility, and more scientific charging pile layout is achieved.

The invention discloses a self-adaptive intestinal function adjusting system based on biological feedback, and relates to the technical field of intestinal function adjustment, the self-adaptive intestinal function adjusting system comprises a signal acquisition module, a data analysis module, an adjustment feedback module and a user interaction module; the signal acquisition module adopts a multi-sensor fusion algorithm and a multi-source sensor to acquire intestinal function state information; wherein the multi-source sensor comprises a pressure sensor, a peristaltic sensor and a nerve sensor. The intestinal function self-adaptive adjusting system based on biological feedback can comprehensively collect biological signals of intestinal tracts, including pressure, peristalsis, nervous system signals and chemical substance secretion signals; the multi-dimensional monitoring mode can reflect the real state of the intestinal function more accurately, the limitation of a single sensor is avoided, and a rich and accurate data basis is provided for subsequent analysis and adjustment.

The invention discloses a traumatic orthopedic nursing data intelligent analysis method, and relates to the field of medical data analysis. A traumatic orthopedic nursing data intelligent analysis system constructed by the method comprises a scheme management module, a data acquisition module, a data processing module, a nursing stage analysis module, a nursing data analysis module and a nursing scheme optimization module. According to the method, a complete trauma orthopedic nursing scheme library is constructed, comprehensive nursing schemes of different trauma types are covered, a rich data basis is provided for subsequent analysis, and the comprehensiveness and applicability of the nursing schemes are improved. Through the nursing stage analysis model, the wound type and the nursing stage of the patient can be corrected and confirmed, and the analysis accuracy is remarkably improved.

The invention discloses a threat intelligence aggregation and attack abnormal feature intelligent generation method based on a large language model. The method comprises the following steps: acquiring network security intelligence of an open source threat intelligence community platform in real time; cue words of the large language model are optimized according to predefined multi-dimensional feature attributes, and multi-dimensional feature attribute extraction of threat intelligence is carried out on network security intelligence; according to the extracted multi-dimensional feature attributes of the threat intelligence, constructing a net-shaped graph relation structure database of the threat intelligence; constructing a threat intelligence knowledge graph according to the net-shaped graph relation structure database; and querying the threat intelligence knowledge graph, marking attack behavior features, connecting attack behavior feature points, and generating attack behavior feature sub-graphs. According to the method, the security team can quickly respond to emerging threats, and the capability of coping with network security threats is improved.

The invention discloses a surrounding rock deformation prediction method for tunnel construction in a soft rock area, belongs to the technical field of surrounding rock deformation prediction, and can solve the problem of low accuracy of surrounding rock deformation prediction in the soft rock area in the prior art. The method comprises the following steps: S1, according to a surrounding rock monitoring network of a tunnel, respectively adding adaptive weight items to a plurality of loss functions in a target function of a gradient boosting decision tree algorithm to obtain an improved gradient boosting decision tree algorithm; the plurality of loss functions are in one-to-one correspondence with a plurality of monitoring points in the monitoring network; s2, using surrounding rock monitoring data collected by a surrounding rock monitoring network to train an improved gradient boosting decision tree algorithm to obtain a tunnel surrounding rock deformation prediction model; and S3, predicting the surrounding rock deformation condition of the tunnel by using the surrounding rock deformation prediction model. The method is used for predicting surrounding rock deformation.

The invention provides an industrial robot safety monitoring method and system, and relates to the technical field of safety monitoring, and the method comprises the steps: arranging a multi-mode sensor array to collect the state information of a robot; performing empirical mode decomposition on the vibration signal to obtain an intrinsic mode function reflecting the abrasion degree of the robot transmission system; performing wavelet transform on the temperature signal to obtain a temperature feature set reflecting the heat dissipation condition of the robot; through knowledge extraction, knowledge representation and knowledge fusion technologies, semantic representation and linking are carried out on various knowledge in robot components, performance parameters, degradation modes and maintenance measures, and a management knowledge graph corresponding to the robot is constructed; and performing reasoning mining on the management knowledge graph based on a rule mining algorithm of path sorting, extracting degradation characteristics corresponding to each component of the robot in the management knowledge graph by mining a reasoning path formed by frequently co-occurring entities and relationships, and predicting a health state trend of the robot in combination with a pre-constructed time sequence prediction model.

The invention provides a power equipment state temperature monitoring method and system. The method comprises the following steps: acquiring heat indexes and energy consumption data of to-be-monitored equipment at a plurality of specified moments in a preset time window; the heat index is a weighted value according to the equipment temperature, the equipment energy consumption and the equipment operation state; arranging the popularity index and the energy consumption data at the specified moment according to a time sequence, and constructing a popularity time sequence and an energy consumption time sequence; performing dynamic analysis on the energy consumption time sequence to obtain an energy consumption time sequence feature set; analyzing the popularity time sequence, and extracting a popularity time sequence feature set; according to the energy consumption time sequence feature set, carrying out weight adjustment and combination on the extracted heat time sequence features to form a combined heat time sequence feature set; and according to the merged heat time sequence characteristics, whether the working condition of the to-be-monitored equipment is abnormal or not is identified through the preset heat critical value and the preset change interval, so that the reliability of state temperature monitoring of the power equipment is effectively improved.

The invention relates to the technical field of battery monitoring and management, and discloses a power battery monitoring and management system and method based on the Internet, and the system comprises a data collection unit, a preprocessing unit, a data transmission unit, a feature extraction unit, a data analysis unit, a fault diagnosis unit, and a remote management unit. In battery state evaluation, the charge state is calculated by combining an ampere-hour integral method with an open-circuit voltage method, and errors are calibrated and reduced; the health state is calculated through the internal resistance change rate, maintenance or replacement requirements can be prompted in time, voltage, temperature, charge and health state parameter trends are analyzed through setting observation points, faults including potential faults are efficiently diagnosed in combination with normal threshold value comparison and change rate comparison, and finally a user can remotely check the visual battery state. When a fault occurs, an instruction can be sent remotely, control such as power limiting and balance operation starting is achieved, and convenience and high efficiency are achieved.

The invention relates to the technical field of water quality tracking and monitoring, and discloses a water environment water quality tracking and monitoring system and a method thereof.The water environment water quality tracking and monitoring system comprises a data acquisition unit, a preprocessing unit, a data analysis unit, a pollution analysis unit and an early warning display unit. The water environment condition can be reflected comprehensively and accurately; abnormal value judgment and replacement are performed on the water quality data of the acquisition points and the time points, so that the reliability and the stability of the data are improved; by constructing a regional water quality vector matrix and a fixed-point water quality vector matrix, indexes such as variable coefficients are calculated from space and time dimensions, and acquisition points with local pollution or abnormal fluctuation are accurately found out; according to the invention, the system determines the pollution level, displays the abnormal collection point and the pollution level to related personnel in time, facilitates the related personnel to take countermeasures rapidly, improves the efficiency and level of water environment management, and provides powerful support for the protection and treatment of the water environment.

The invention provides an equipment information management method and system based on a multi-modal large model in the technical field of equipment information management. The method comprises the following steps: S1, acquiring historical multi-modal data to construct a data set, and dividing the data set into a training set, a verification set and a test set; s2, creating a multi-modal large model based on a Vi T module, a Bert module, a numerical feature construction module, a feature fusion module and an output module, and setting hyper-parameters; s3, training and evaluating the multi-modal large model through the training set, the verification set and the test set; step S4, compressing the multi-modal large model which passes the evaluation; and S5, collecting real-time multi-modal data of installed equipment in the management area by each area server, and inputting the real-time multi-modal data into the multi-modal large model to obtain an equipment information analysis report. The method has the advantages that the efficiency, the quality, the traceability and the resource utilization rate of equipment information management are greatly improved, and then the reliability of equipment operation and maintenance is greatly improved.

The invention relates to the technical field of data analysis and evaluation, in particular to a color fastness evaluation method and system, and the method comprises the following steps: setting test conditions to simulate external factors encountered in the actual use or processing process of textiles; under the simulated test condition, treating the textile sample for a preset time; acquiring a multispectral image of the textile by adopting a multispectral imaging technology, and capturing color change of the textile; developing a self-adaptive image enhancement algorithm, and performing preprocessing; comparing spectral information of the multi-spectral image of the textile before and after processing based on chromatic aberration evaluation analysis so as to quantify the color fastness; analyzing the multispectral image after image enhancement by using a deep learning technology, extracting color attenuation features, and comprehensively evaluating a color fastness score; according to the color fastness evaluation method, basic color difference information is considered, advanced features such as textures, edges and local color modes of the textile are further included, and the accuracy and reliability of color fastness evaluation are remarkably improved.

The invention discloses an automatic conversion analysis method and system based on an implicit common cause fault model, and relates to the technical field of power system fault analysis, and the method comprises the steps: obtaining and marking power equipment, forming a plurality of equipment sets, setting a monitoring period for operation monitoring, and storing data into an operation record library; extracting feature data of the equipment, calculating a correlation index between the equipment according to the category of the equipment, and calculating the correlation between the same category of equipment and different categories of equipment by adopting different strategies; utilizing an implicit common cause fault model to calculate the probability that the reference equipment and other equipment are abnormal at the same time under each abnormal cause, including an independent failure probability, a common cause failure probability and a comprehensive probability; and according to the comparison of the comprehensive probability and a preset threshold value, the monitoring period of the equipment exceeding the threshold value is adjusted, and according to the abnormal reason, the monitoring period of all the associated equipment is also adjusted, so that the effective monitoring and fault prevention of the power system are realized.

The invention discloses an investigation and design enterprise asset multivariate heterogeneous data fusion and knowledge graph construction method, which comprises the following steps: acquiring enterprise asset data information including structured data and unstructured data; invalid and missing value removal processing is carried out on the structured data, and meanwhile, the cleaned numerical data is converted into a text format; performing text data standardization processing and optimization on the unstructured data and the structured data after the text format, and constructing word vectors while performing text disambiguation standardization; respectively extracting features of the structured data and the unstructured data by using a decoder BERT model; fusing all kinds of data characteristics by using a Lite Transform to construct a knowledge graph; the problems that existing investigation and design enterprises are huge in data scale and uneven in data quality, corresponding design histories and collected field related information cannot be rapidly obtained from historical data, and consequently, early-stage investigation data are repeated, and later-stage design is inconsistent are solved.

The invention discloses a traditional Chinese medicine analysis and identification system, method, equipment and medium based on a large model, and the system comprises a data collection module which is used for collecting multi-dimensional data of traditional Chinese medicines, a data preprocessing module which is used for carrying out the cleaning, standardization processing and feature extraction of the collected data, and a data analysis module. The data processing module is used for analyzing and processing received data by utilizing a pre-trained deep large model, the result output module is used for outputting an analysis result to a user in a visual manner, and the analysis result output module comprises a visual chart, a detailed report form and an intelligent feedback module according to the feedback of the user on the analysis result and the actual application condition. Through multi-dimensional data acquisition, advanced preprocessing, deep large model analysis and the like, the practicability is improved by modules of accurate identification, intelligent feedback, collaborative analysis and the like of the traditional Chinese medicine, data updating and the like ensure that the system advances with the times, powerful technical support is provided for industrial development of the traditional Chinese medicine, and the quality and safety of the traditional Chinese medicine are guaranteed.

The invention belongs to the technical field of space planning, and particularly relates to a land space planning evaluation method based on deep learning, and the method comprises the steps: constructing a scene graph structure, generating an initial scene graph, updating the scene graph through an iteration planner, recording and feeding back, and enabling the scene graph structure to visually present the relation between planning objects. The complex territorial space layout can be conveniently analyzed; the method comprises the steps of extracting physical characteristics of a current scene graph, deploying a planning evaluation model to generate label distribution, determining evaluation indexes to calculate spatial planning loss, optimizing the planning evaluation model, deeply mining current scene graph information, facilitating accurate grasping of relation characteristics between objects, providing accurate evaluation for spatial planning, comprehensively considering planning constraints, and improving the planning efficiency. Reliable decision support is provided; a dynamic updating mechanism is adopted to discover a space planning problem in time, and rationality and feasibility of space planning are improved.

The invention discloses a clinical test data analysis system based on a large model, which comprises a data acquisition module, a preprocessing module, a training module, an analysis module and an output module, data acquisition is efficient and supports multiple modes, preprocessing adopts an intelligent algorithm, model training is advanced and has efficient indexes, analysis covers multiple aspects and has evaluation indexes, and result output is diversified and supports customization. The invention further comprises a corresponding data analysis method, related equipment and a medium, and each link is provided with an innovative technology and an evaluation mode to guarantee the performance and the quality; the system realizes efficient data acquisition and integration, intelligent processing and quality improvement, constructs a powerful analysis model, provides various visual result display and interpretation, has real-time updating and personalized service capabilities, can accurately evaluate the curative effect and safety, provides a scientific basis for clinical decisions, assists in improving the efficiency and accuracy of medical data analysis, and has a wide application prospect. Different user requirements are met, and development of the medical industry is promoted.

The invention discloses an intensive culture precise feeding control method based on fuzzy-MPC control, and relates to the technical field of aquaculture intelligent control, and the method comprises the steps: employing a sensing system, collecting multi-source sensing data including fish activity images, healthy growth states and water quality environment data in real time, and forming a comprehensive state data set; and based on the acquired multi-source sensing data, constructing an MTL-LSTM-SAT multi-task learning model. According to the method, the multi-task learning model fusing the LSTM and the soft attention mechanism is constructed, the long-term dependency relationship and dynamic association among the water quality, the feeding and the fish state can be captured, the attention mechanism enables the model to be automatically focused on a key time step, the perception ability of the water quality fluctuation and the feeding influence stage is enhanced, and the accuracy of the water quality fluctuation and feeding influence stage is improved. Meanwhile, double-task collaborative prediction of the growth speed and the disease proportion is achieved through the shared feature layer, a prospective state sequence of the future 7 days is output, and a quantitative and reliable prediction basis is provided for feeding decision making.

The invention relates to the technical field of private domain traffic, in particular to a private domain traffic guiding method and system based on collaborative filtering, and the method comprises the following steps: S1, collecting user behavior data: collecting the behavior data of a user in a private domain platform in real time through client equipment; s2, constructing a user portrait: extracting the interest preference of the user, generating a user interest label, and forming the user portrait; s3, interest extension and cross-domain recommendation: constructing a relation graph between the user and the content, identifying association between the user interest node and the cross-domain content, calculating the edge weight between the user and the content node, and extending the interest boundary of the user in the private domain traffic; and S4, recommended content pushing: pushing cross-domain recommended content to the user, and meanwhile, optimizing the display sequence of the content in combination with the situation information of the user. According to the invention, the diversity and personalization of the recommended content can be enhanced, and the user experience and the diversity of the recommendation effect are further improved.

The present invention discloses a method for intelligently generating threat intelligence aggregation and attack anomaly features based on a large language model, including obtaining network security intelligence of an open source threat intelligence community platform in real time; optimizing the prompt words of the large language model according to predefined multidimensional feature attributes, and extracting multidimensional feature attributes of threat intelligence from network security intelligence; constructing a mesh graph relationship structure database of threat intelligence according to the extracted multidimensional feature attributes of threat intelligence; constructing a threat intelligence knowledge graph according to the mesh graph relationship structure database; querying the threat intelligence knowledge graph, marking attack behavior features, connecting attack behavior feature points, and generating an attack behavior feature subgraph. The method of the present invention enables the security team to quickly respond to emerging threats and improves the ability to respond to network security threats.

The utility model discloses a robot sensor platform with two degrees of freedom and a damping function. The robot sensor platform comprises a damping base, a spherical hinge, a spring, an upper stepping motor, a damping connecting plate, a sensor support, a sensor clamping jaw and a lower stepping motor. The damping device is characterized in that the damping base and the damping connecting plate are connected through a spherical hinge and a spring, and the spherical hinge is a telescopic spherical hinge with damping and is matched with the spring to achieve double damping; the upper stepping motor is fixedly installed on the damping connecting plate, and an output shaft of the upper stepping motor is connected with the sensor support. One side of the sensor support is fixed to a lower stepping motor through a groove, and an output shaft of the lower stepping motor is connected with a sensor clamping jaw. According to the platform, the operation range of the sensor is increased through the two stepping motors, and the precision is ensured while the degree of freedom of the sensor is increased through buffering and damping of the damping assembly.

The present invention relates to the field of data processing technology, and specifically provides a data mining method and deep learning system applied to smart industrial services. The present invention innovatively extracts material flow state characteristics and equipment operation quantitative characteristics by deeply mining the production factors in the automated production line image data, providing a rich data basis for the optimization of production line efficiency, and by retrieving the initial screening production line optimization annotation images that are compatible with these characteristics, two initial screening sets are formed to ensure the pertinence and feasibility of the optimization plan. At the same time, the concept of common weights for efficiency optimization is introduced to comprehensively evaluate the value of each initial screening annotation image for the current production line efficiency optimization, so as to accurately mine the target production line optimization annotation image for direct guidance of the efficiency optimization and upgrade of the industrial production line, which not only improves the accuracy and efficiency of the optimization process, but also brings significant economic benefits and competitiveness improvement to smart manufacturing enterprises.

The invention discloses a strip network railway base station data processing method and system. The method comprises the following steps: obtaining observation data of a plurality of IGS base stations; obtaining observation data of a plurality of standard reference stations in a preset range around the strip network railway reference station, wherein the standard reference stations comprise a national station, a provincial station and a Chinese tectonic environment monitoring network station; and based on the observation data of the IGS reference station, combining the observation data of the standard reference station to perform hierarchical calculation on the plurality of railway reference stations to obtain space coordinate data of each railway reference station. Observation data of an IGS base station and standard base stations such as surrounding national stations and provincial stations are obtained, and a rich data basis is provided for railway base station coordinate measurement and calculation. A hierarchical resolving mode is adopted, the coordinate precision is improved, error accumulation is avoided, and the result reliability is enhanced; the method can adapt to different precision requirements and complex environments, considers local differences, and meets the requirement of a strip network railway for high-precision coordinate data.

The invention discloses an electro-deposition preparation device and system for a graphene-twin crystal copper-based composite material. The preparation method of the graphene-twin-crystal copper composite material comprises the following steps: carrying out electro-deposition in a composite electroplating solution containing graphene and copper ions through a pulsed power supply by taking a copper-phosphorus plate as an anode, namely, directly obtaining the graphene-twin-crystal copper composite material on a cathode substrate. According to the preparation method, multiple parameters in the electrodeposition process are automatically adjusted through the upper computer, preparation of the graphene-twin crystal copper composite material with excellent dispersity and high homogenization degree is achieved, and further, multiple different parameter combinations can be set, so that the multi-layer multi-gradient composite material with different graphene volume fractions and different twin crystal lamellar thicknesses is achieved. The method has potential application value in the technical field of flexible electronics and the like.

This invention discloses a method for detecting the aging state of asphalt pavement by integrating hyperspectral imagery and deep learning, relating to the field of deep learning technology. The method includes: acquiring and preprocessing hyperspectral images of the target road section to obtain surface reflectance images; dynamically identifying and masking vehicles and their shadows in the images using a spectral angle matching method to extract effective pixel spectral datasets; performing hybrid pixel decomposition on the effective pixel spectral datasets using a non-negative matrix factorization algorithm to obtain asphalt endmember spectra and aggregate endmember abundance maps; inputting the asphalt endmember spectra into a pre-trained oxidation degree assessment model to output asphalt oxidation degree values; overlaying the aggregate endmember abundance map with near-infrared band images and inputting it into a pre-trained exposure rate segmentation network to output a binary mask of aggregate exposure areas and calculate the aggregate exposure rate. This method solves the technical problem that existing technologies cannot quickly and non-destructively detect the aging state of asphalt pavement.

The embodiment of the invention provides an insurance recommendation text generation method and device and related equipment, and belongs to the technical field of large language models. The method comprises the following steps: acquiring historical claim settlement data and external event data of a target object, and constructing multi-source fusion data according to the historical claim settlement data and the external event data; inputting the multi-source fusion data into a pre-trained text generation model, so that the text generation model generates an original recommendation text according to the multi-source fusion data; comparing the multi-source fusion data with the original recommended text to obtain deviation data; inputting the deviation data and the original recommended text into a text generation model, so that the text generation model corrects the original recommended text to obtain a target recommended text; and pushing the target recommendation text to the target object. According to the embodiment of the invention, the highly personalized insurance recommendation text can be generated.

The invention discloses an unmanned aerial vehicle water-air dual-medium adaptive communication optimization method and related equipment. The method comprises the following steps: acquiring unmanned aerial vehicle environment sensing information; determining the current communication medium state of the unmanned aerial vehicle according to the unmanned aerial vehicle environment sensing information; the communication medium state comprises an air state, an underwater state and a water-air interface crossing state; generating a target communication strategy according to the current communication medium state of the unmanned aerial vehicle, and controlling the unmanned aerial vehicle to execute the target communication strategy; the communication strategies comprise an air communication strategy, an underwater communication strategy and a water-air dual-medium communication strategy; and dynamically monitoring the communication link performance index of the unmanned aerial vehicle, and dynamically optimizing the target communication strategy in combination with the environment change information and the communication link performance index. The technical problem that the communication performance of the unmanned aerial vehicle on the water-air interface is sharply reduced can be effectively solved, smooth transition and high-reliability maintenance of the communication link in the double-medium environment are achieved, and the method can be widely applied to the technical field of unmanned aerial vehicle communication.