912 results about "Time sequence" patented technology

The invention discloses a solar radiation space-time prediction method and system based on physical information constraint and a neural network, and the method comprises the steps: collecting multi-dimensional time sequence meteorological data, extracting high-dimensional time sequence dynamic features, converting geographic space data into a fuzzy set, and carrying out the defuzzification of the fuzzy set through an inference rule, thereby obtaining geographic space features; and a gating mechanism is adopted to realize deep fusion of the space-time features to generate high-dimensional space-time fusion features. In a model training stage, an energy conservation equation is introduced into an optimization process, a physical residual error is constructed by calculating a time derivative and a space derivative of a predicted value, a physical constraint total loss function is formed in combination with a data loss item, and model parameters are updated by using a gradient descent method. According to the method, the accuracy and reliability of a prediction result are remarkably improved while the calculation efficiency is ensured, and the method is particularly suitable for solar radiation prediction under complex meteorological conditions; according to the method, abnormal prediction caused by data noise can be effectively corrected, and a solution with physical rationality and data adaptability is provided for the fields of solar resource evaluation, photovoltaic power generation power prediction and the like.

The invention provides a mine abnormal event real-time identification method and system based on time sequence characteristics, and relates to the technical field of mode identification, and the method comprises the steps: carrying out the time-space alignment and semantic annotation of multi-modal monitoring data, and constructing a time sequence knowledge graph; calculating a dynamic association weight between entities, and analyzing a risk propagation path; predicting a risk situation based on a historical evolution rule; and dynamically generating a differential early warning strategy and establishing a closed-loop tracking system. According to the invention, early identification, accurate prediction and efficient disposal of mine safety risks can be realized, and the mine safety management level is improved.

The invention provides a feeder terminal fault detection method, system and device, a medium and a program product, and the method comprises the steps: obtaining multi-source heterogeneous data comprising feeder terminal operation data, a topological graph structure of a power distribution network and external sensing data, the multi-source heterogeneous data comprises at least one kind of structured or unstructured time sequence data, and the external sensing data comprises at least one kind of structured or unstructured time sequence data; the external sensing data comprises meteorological data, geographic space information and historical fault records; and preprocessing the multi-source heterogeneous data, inputting the preprocessed time-aligned multi-source heterogeneous data into the trained time-space diagram neural network model to extract features and perform joint modeling, and outputting a fault type classification result and a fault probability distribution result corresponding to each feeder terminal node in the power distribution network. According to the method, a multi-source heterogeneous data fusion and time-space diagram neural network modeling mechanism is introduced, so that the model has a dynamic response capability to complex environment changes, and the identification precision of a potential fault mode is effectively enhanced.

The invention provides a wind power prediction method and system based on time sequence decomposition and multi-model fusion, and the method comprises the steps: collecting the historical power generation power and meteorological data of a target wind power plant; decomposing the historical power generation power and the meteorological data to obtain a trend component, a seasonal component and a residual component; fusing with meteorological data to construct a trend feature matrix, a periodic feature matrix and a residual feature matrix; different modeling schemes are adopted to construct corresponding single models; dividing into a training set, a verification set and a test set according to a time sequence; performing training optimization on the single model by using the training set, the verification set and the test set, and constructing a wind power short-term power prediction model; and inputting the real-time meteorological data and the generated power to the wind power short-term power prediction model, and outputting the generated power prediction value of the target wind power plant, thereby effectively improving the comprehensiveness, accuracy and stability of model prediction.

The invention discloses a photovoltaic power generation power prediction method and system based on a large language model. The method comprises the following steps: converting historical power data and numerical weather forecast data into time sequence embedded representation; through cross-modal semantic alignment, semantic embedding representation is generated; constructing a natural language prompt containing task context information, encoding the natural language prompt into prompt embedding, combining prompt embedding with semantic embedding representation to form a fusion input sequence, inputting the fusion input sequence into a pre-trained large language model, and outputting implicit features; synchronously generating an initial power prediction result and a weather prediction result obtained by correcting the numerical weather prediction data through a parallel collaborative prediction mechanism; and taking the meteorological prediction result as a correction signal, performing joint optimization on the preliminary power prediction result, and outputting a power generation power prediction value. According to the method, the problem of deep fusion of heterogeneous data is effectively solved, and the prediction accuracy is improved.

The invention discloses an energy storage system state evolution trend prediction method based on multi-source data fusion. The method comprises the steps of terminal voltage, current and temperature time sequence data acquisition, time sequence segmentation normalization, multi-physics field coupling feature construction, trend prediction model construction and training and energy storage system state evolution trend prediction. According to the method, the distinguishing capacity of the model for charging and discharging physical characteristics is improved, meanwhile, the voltage change rate, the multi-dimensional feature vector of the differential internal resistance and the thermal-electric coupling effect and the explicit encoding electric-thermal-resistance coupling relation are constructed, the transient response and the temperature hysteresis effect can be effectively captured, and then the model can be used for analyzing the charging and discharging physical characteristics. A degradation-aware cross-cycle feature extraction and gating mechanism is adopted, short-term fluctuation and long-term trend are adaptively balanced in multi-scale prediction, the prediction conflict problem is relieved, finally, physical constraints based on the electrochemical law and the internal resistance temperature characteristic are embedded in a loss function, it is ensured that the prediction result is accurate in numerical value and conforms to the physical law, and the prediction accuracy is improved. And generation of physically impossible solutions is avoided.

The invention discloses a geological disaster early warning method and system based on multi-source data fusion and electronic equipment, and the method comprises the steps: carrying out the alignment of remote sensing data, sensor data and meteorological data in a space dimension and a time dimension, and obtaining multi-source data after the time-space alignment; performing noise elimination and missing value filling on the multi-source data after space-time alignment to obtain processed multi-source data; extracting multi-source features based on the processed multi-source data, and performing feature fusion on the multi-source features to obtain a multi-source spatio-temporal data cube; constructing a geological disaster prediction large model comprising a spatial feature extraction layer, a time sequence feature aggregation layer and a disaster classification and regression branch; inputting the multi-source spatio-temporal data cube into a trained geological disaster prediction large model for prediction, and obtaining a risk level classification result and a displacement change value; and performing geological disaster early warning according to the risk level classification result and the displacement change value. The geological disaster early warning accuracy can be improved.

The invention discloses a flow field measurement method based on an event camera, and the method comprises the steps: generating a PIV data set, each time sequence sample sequence comprising a plurality of frames of continuous particle images, a corresponding velocity vector field, and particle event data at all moments; establishing a flow field data acquisition device based on an event camera and a high-speed camera, acquiring real event data and real image data which are synchronous in time so as to adjust parameters of an event simulator, and verifying and updating particle event data in the PIV data set according to the adjusted event simulator so as to obtain a flow field data acquisition result; obtaining the updated PIV data set as a training data set; building an event camera optical flow method model, and training by adopting the training data set; and on the basis of the trained event camera optical flow method model, event sequences in two adjacent time periods are used as inputs to calculate a velocity vector field corresponding to a middle moment. According to the invention, the flow field velocity field at the required moment can be obtained based on the event data within a period of time.

The invention discloses a tunnel portal construction risk assessment method and system based on multi-source monitoring data fusion. The method comprises the following steps: performing layered acquisition on monitoring objects in a tunnel portal construction area, obtaining multi-source monitoring data of an environment layer, a geological layer and a construction layer, and adding element attributes such as time, space coordinates and equipment health status; performing credibility correction and time sequence reconstruction based on the monitoring data to obtain a time sequence alignment data set with credibility weighting; performing dynamic weight fusion on the data set, and generating a fusion risk feature vector in real time; inputting the feature vectors into a spatio-temporal evolution model, and outputting collapse, water inrush and settlement risk probabilities in a plurality of time windows in the future in combination with coupling calculation of a spatial sub-model and a time sub-model; and forming a multi-dimensional risk portrait according to a prediction result, performing adaptive correction based on construction site feedback, and updating a monitoring index weight and a model parameter so as to realize dynamic optimization of a subsequent prediction period.

The invention provides a short-term photovoltaic power prediction method and system, computer equipment and a medium, and belongs to the field of photovoltaic power generation output power prediction.The method comprises the steps that short-term photovoltaic power and meteorological working condition data samples are obtained, and a Gaussian mixture model is used for conducting multi-modal clustering processing on the meteorological working condition data samples to obtain membership probability embedded vectors; through a Pearson's correlation coefficient weighting and sliding window mechanism, extracting features from the data sample and the membership probability embedding vector, and constructing a multi-modal time sequence feature tensor; a multi-head attention mechanism in a traditional Transform network is replaced with a class domain fusion self-attention mechanism, and a class domain fusion attention model is formed; inputting a multi-modal time sequence feature tensor to train a class domain fusion attention model to obtain an initial prediction value; and residual error estimation is carried out on the initial prediction value by using a residual error learning error compensation strategy, a short-term photovoltaic power prediction result is output, and the accuracy and robustness of the model are improved.

The invention is suitable for the technical field of earth surface monitoring, and provides an earth surface deformation monitoring method and system based on a time sequence InSAR, and the method comprises the following steps: carrying out the SAR image screening and interferogram generation, and collecting corresponding meteorological data and thermal infrared data; performing space-time adaptive atmospheric phase correction, taking the air pressure vertical gradient and the temperature anomaly as driving factors of atmospheric delay, and separating an atmospheric phase through a space-time weighted model; dynamic deformation modeling is carried out, and deformation is decomposed into linear deformation and nonlinear deformation; the method comprises the following steps: taking a mining area road network as a geometric constraint, unwrapping a coherent region by adopting a minimum cost flow algorithm, converting an unwrapping phase into sight-line-direction deformation, and calculating horizontal and vertical deformation components in combination with InSAR sight-line-direction deformation and a digital elevation model. The method adapts to a complex deformation mechanism of a mining area by capturing linear and nonlinear deformation. Through sight line deformation and DEM geometric projection, vertical and horizontal deformation separation is realized, and the deformation direction is determined.

The invention provides a grouting simulation method and system based on time sequence interaction and self-adaptive grid reconstruction, and relates to the technical field of grouting engineering.The grouting simulation method comprises the steps that monitoring data of the grouting construction process are obtained, the monitoring data are input into a grouting diffusion prediction model, a grouting pressure prediction value is output, and a series of prediction responses are conducted based on the grouting pressure prediction value. Obtaining a grouting pressure predicted value, a boundary point set of the grouting diffusion area, the area and the volume of the grouting diffusion area and a diffusion response result of the pressure distribution diagram; constructing a geometric model of the grouting area based on the diffusion response result, and performing grid unit division and grouting diffusion area mapping on the geometric model; and performing local grid encryption and reconstruction on the grid units according to a density control rule to obtain a structured grid data model, adaptively updating the structured grid data model by using real-time monitoring data, and performing real-time visual grouting optimization control. According to the invention, the high-precision prediction response of grouting is improved.

The invention provides an intelligent data alignment method and system based on time sequence dynamic multi-source embedding mapping. The method belongs to the technical field of multi-modal data fusion and spatio-temporal information processing. The method comprises the following steps: performing time sequence dynamic feature extraction on a multi-source heterogeneous data source to generate a heterogeneous data sequence containing a time dependency relationship; and constructing a time sequence dynamic multi-source embedded manifold space based on a manifold learning theory, mapping a heterogeneous data sequence to a unified evolution geometric structure representation space, and generating embedded manifold data. Through the method, heterogeneous data from various different data sources can be effectively processed, unified mapping is carried out through time sequence dynamic feature extraction and a manifold learning technology, cross-source alignment of the data is achieved, and the method is particularly suitable for a data scene needing to consider a time dependency relationship.

The invention belongs to the technical field of computer vision, deep learning and time-space prediction, and discloses an end-to-end time-space prediction method based on improved three-dimensional rotation position coding, which is suitable for various time-space sequence prediction scenes such as weather, traffic flow and the like. According to the invention, through four key improvements, a position coding mechanism is optimized; three-dimensional coding proportions of time, height and width are dynamically adjusted so as to adapt to different scenes; fusing the absolute time and the relative space position, and strengthening local space-time correlation modeling; the position information directly guides attention calculation, and the fusion with an Attention module is deepened; and a rotation matrix cache mechanism is introduced to reduce redundant calculation. Meanwhile, the model is matched with a Patch embedding layer, an adaptive Transform encoder and an MLP de-wharf, a complete link of'feature embedding-position encoding-space-time fusion-prediction output 'is constructed, and the precision, generalization and reasoning efficiency of space-time prediction are effectively improved.

The invention discloses a PINN-based limited space internal explosion damage parameter prediction method and system. The method comprises the steps of completing data acquisition and constructing a time sequence data set; model construction is completed, and the time sequence modeling capability is enhanced; determining a total loss function and adding boundary condition constraints to enable model prediction to accord with physical laws; firstly optimizing data loss, then introducing a physical residual error, and finally activating a boundary speed suppression loss item and adjusting a learning rate; in combination with a learning rate dynamic scheduling and early stop mechanism, the training efficiency and stability are improved through adaptive residual weighted balance data and physical constraint loss; and predicting parameters such as pressure, temperature and speed of each point in the limited space in multiple time frames by using the trained model, calculating impulse based on a pressure time history, evaluating personnel damage, and completing damage zoning in the limited space. And efficient modeling under a complex boundary condition and a limited space environment is completed.

The invention discloses a vehicle-mounted electrical control unit (ECU) debugging simulation system and method based on digital twinning. The system is composed of a data aggregation and time reference module, a consistency evaluation module, a synchronization and correction module, a semantic communication and scheduling module, an arbitration and security debugging module, and a cross-layer adaptive degradation and execution and iteration module. Channel-level time delay is estimated through a unified time reference, a consistency difference spectrum containing amplitude, time sequence, event and derivative components is constructed, confidence is given, time delay compensation and event-level alignment are calculated, and deterministic and priority scheduling is implemented according to semantic benefits; actions such as parameter recharge, message injection, task rhythm perturbation and environmental stimulation are generated and limited under the constraint of a security invariant set, control right soft handover and gating are realized based on confidence, and grading and degrading are performed when the confidence is low or the constraint is triggered; millisecond-level bidirectional collaborative debugging of twin and physical ECUs is realized, and consistency convergence, safety and bandwidth time arrival are improved.

The invention relates to the field of data processing, in particular to an electric energy meter abnormity monitoring method and system based on digital twinning, and the method comprises the steps: obtaining a historical data time sequence of an electric energy meter, and carrying out the preprocessing; dividing the historical data time sequence of each dimension to obtain a plurality of historical data sets containing historical data time sequence segments of each dimension, and training an LSTM model based on the historical data sets; acquiring a current data time sequence of the electric energy meter and preprocessing the current data time sequence to obtain a current data set, inputting the current data set into the trained LSTM model, outputting predicted values of a plurality of target dimensions at the next moment, obtaining measured values of the plurality of target dimensions at the next moment, calculating a judgment index based on the predicted values, the measured values and a historical data set, and determining the current data time sequence of the electric energy meter. And judging the abnormal condition of the electric energy meter based on the judgment index. According to the method, noise residual errors generated when the self-cognition uncertainty of the model is high can be automatically inhibited, meanwhile, a real state deviation signal during cognition determination is amplified, and the accuracy of anomaly judgment is improved.

The invention provides an intelligent monitoring and early warning method based on multi-source information fusion, and belongs to the technical field of intelligent diagnosis and early warning, and the method comprises the steps: synchronously collecting the multi-modal data of a voltage transformer, the multi-modal data comprising voltage / current waveform, partial discharge signal, temperature and vibration data; preprocessing the multi-modal data, wherein a preprocessing method comprises waveform segmentation, spectrogram generation and scalar normalization; constructing a dynamic weight distribution mechanism: calculating a weight matrix based on the relevance between the time sequence features and the spatial features, and dynamically adjusting scalar weights by combining the physical coupling relationship between the temperature and the vibration to generate feature fusion weights; iteratively updating the weight matrix according to the real-time prediction error; fusing the time sequence features, the spatial features and the scalar features through the dynamic weight distribution mechanism to generate a comprehensive feature vector; and synchronously outputting a fault type classification result and an equipment health score based on the comprehensive feature vector, and calculating a real-time prediction error.

The invention discloses a distributed photovoltaic cluster power prediction method and device based on multi-modal fusion. The method comprises the following steps: acquiring historical photovoltaic data and historical photovoltaic images of a photovoltaic region to be predicted; analyzing a time sequence relationship in the historical photovoltaic data, extracting photovoltaic time sequence characteristics, and giving a first prediction result in combination with the data time sequence prediction model; extracting spatial features in the historical photovoltaic image, reconstructing the historical photovoltaic image, and giving a second prediction result in combination with the image prediction model; and in combination with a preset fusion weight, fusing the first prediction result and the second prediction result to obtain a target prediction power, and completing power prediction of the distributed photovoltaic cluster, thereby effectively capturing the influence of sudden weather events on photovoltaic power generation, improving the accuracy of conventional cloud picture data when coping with complex and changeable cloud layer motion, and improving the prediction efficiency of the distributed photovoltaic cluster. Therefore, the accuracy of power prediction is improved.

The invention relates to the technical field of photovoltaic prediction, in particular to a distributed photovoltaic power prediction method and system based on high-dimensional gridding numerical weather forecast, and the method comprises the steps: carrying out the standardization of the numerical weather forecast data and photovoltaic power historical data of a target region, and achieving the time-space alignment based on a preset grid, generating a gridding data set; utilizing convolution processing to extract local space features, and converting and fusing the local space features into a feature sequence containing space and historical time sequence information at the same time; modeling is carried out through an encoder-decoder architecture, an encoder excavates historical power dependence, and a decoder dynamically couples future meteorological characteristics with historical power through an attention mechanism and outputs a grid-level predicted value; aggregating to obtain a system total power prediction result; by establishing a unified space-time grid, refined alignment of data is realized, cross-space-time dynamic fusion is performed in combination with convolution and an attention mechanism, and prediction precision and stability can be kept in complex weather.

The invention relates to the technical field of new energy power data governance, in particular to a multi-source fusion cleaning and closed-loop quality control method and device for wind power data and a medium, and the method comprises the steps: generating an initial data set through collecting SCADA time sequence data and at least one type of external meteorological data of a wind turbine generator; based on the initial data set, executing physical consistency check, grouping adaptive statistical detection, unsupervised machine learning detection and time sequence residual analysis in parallel, and generating four independent exception judgment vectors; inputting the four abnormal judgment vectors into a preset fusion decision model, and outputting a comprehensive abnormal label; obtaining a repaired data set according to the comprehensive abnormal label; and calculating a multi-dimensional quality score representing data quality based on the repaired data set, and triggering a quality control closed-loop operation according to the multi-dimensional quality score and a comparison test result. Through the arrangement, multi-dimensional fusion detection and self-adaptive intelligent repair can be realized, and integrated cleaning of quality quantitative evaluation and closed-loop verification can be realized.

The invention relates to the technical field of disaster prevention and reduction of a power system, and discloses an icing risk early warning method based on multi-model fusion and residual time sequence characteristic analysis, which comprises the following steps: collecting meteorological data of a line area in real time, removing abnormal values through secondary judgment of a Pauta criterion and a trend, and standardizing; adopting a TEROL algorithm to screen high-weight key features; running SWD-BP, MUL-GRNN and ELM models in parallel, constructing a dynamic weight by combining DSI, an independence weight method and an entropy weight method, and calculating a final meteorological predicted value; generating a prediction residual signal, extracting time domain features such as a mean value and a peak value, and constructing a residual feature matrix through a sliding window; and inputting an LSTM model to process a time sequence dependency relationship, and judging an icing risk level. According to the method, meteorological prediction is optimized through multi-model dynamic fusion, and deviation is analyzed and corrected in combination with residual time sequence characteristics, so that the problem of weak generalization ability of a single model is effectively solved, and the accuracy of icing risk early warning is obviously improved.

The invention provides a pilot manipulation quality real-time analysis method and system in a wind shear environment, and relates to the technical field of aviation flight, and the method comprises the steps: firstly capturing flight manipulation and environment interaction data including flight state feedback data and pilot manipulation execution data; secondly, establishing a response association link set of the manipulation action and the environment change, taking the wind shear environment change as a trigger node, taking pilot manipulation execution as a response node, and taking flight state feedback as an association node; tracing the matching relation between the time sequence track of the control action and the environment change, and outputting a time sequence matching relation set; analyzing the adaptability dynamic change of the manipulation action based on the time sequence fit relationship set evolution, and outputting an adaptability dynamic change result; and finally, generating a manipulation adjustment guide based on the dynamic change result of the adaptability, assisting a pilot in optimizing manipulation actions in real time, and improving the flight safety and manipulation quality.

The invention relates to the technical field of valve leakage detection, and particularly discloses an ultrahigh pressure gas valve leakage detection method and system. According to the method, a unified leakage feature vector is obtained through multi-source acquisition and processing; capturing a time sequence dependency relationship, and determining a micro leakage mode; carrying out abnormal peak detection to obtain a preliminary leakage probability; combining the initial leakage probability with real-time environmental parameters, and dynamically adjusting a judgment standard; comparing the unified leakage feature vectors through a judgment standard, fusing position coordinate triangulation positioning, judging whether micro crack leakage occurs or not, and obtaining leakage position estimation; the historical data sequence is fused, the micro crack development trend is predicted, and the early warning level is determined; and generating a corresponding response signal according to the early warning level, and outputting micro crack leakage early warning. Multi-source data fusion, dynamic threshold adjustment and position prediction can be carried out, the accuracy and robustness of micro crack leakage detection are remarkably improved, and efficient guarantee is provided for safe operation of a pipeline.

The invention discloses a motion recognition method and system based on a redox photoelectric memristor, a terminal and a storage medium, and the method comprises the steps: selecting classical motions based on a human body motion data set, extracting time sequence data, and coding the time sequence data into an optical pulse sequence; constructing a reservoir array composed of a plurality of photoelectric memristors, and expanding an optical pulse sequence signal into a high-dimensional state vector; constructing a supervised training model, solving a weight matrix, and constructing a memristive cross array; and outputting an action classification result through simulation domain operation based on the output current multi-path light current signals in combination with the memristor cross array. According to the method, the motion features can be directly fed into a rear-end classification network for action recognition without depending on a complex digital feature extraction algorithm, so that the transmission and processing overhead of redundant data is fundamentally eliminated; a high-efficiency, low-delay and high-robustness hardware solution is provided for real-time and anti-noise motion recognition in scenes such as intelligent monitoring and man-machine interaction.

The invention relates to the technical field of photovoltaic power generation power prediction, and discloses a data time period missing value interpolation method and system based on deep learning, and the method comprises the steps: obtaining power generation power data and meteorological data, recognizing a data missing time period, and obtaining context time period data; searching photovoltaic power stations in a space area where the target photovoltaic power station is located, and screening to form a photovoltaic candidate power station set; performing feature extraction on the generated power curve by adopting an auto-encoder based on a time sequence convolutional network to obtain a time sequence feature vector; calculating similarity scores of the target photovoltaic power station and the photovoltaic candidate power stations to form a dynamic space reference power station set; carrying out weighted averaging on the actual power value of each reference power station according to the weight in the missing time period; and calculating a theoretical generated power upper limit of the target photovoltaic power station in combination with meteorological data, applying physical constraint to the preliminary completion value, and taking a minimum value as a final completion value. According to the method and the device, high-precision and self-adaptive filling value complement is carried out on the missing data of the time period.

The invention relates to the technical field of intelligent diagnosis, and discloses a micro-grid fault diagnosis method and system based on data driving and unsupervised learning. The method comprises the following steps: collecting current, voltage, temperature and power data of a micro-grid and constructing a time sequence matrix; inputting a time sequence prediction network and a time sequence reconstruction network, and performing parallel processing to obtain a prediction error and a reconstruction error; carrying out weighted fusion on the two errors and constructing a two-dimensional error space to judge normal fluctuation and fault abnormity; and extracting a state variable to generate a dynamic threshold to judge a fault. The false alarm rate and the missing report rate of fault diagnosis are reduced.

The invention provides an effective wave height real-time prediction method and device based on dynamic residual correction, and relates to the field of ocean engineering. The method comprises the following specific steps: acquiring wave height data and performing multi-dimensional feature screening; constructing an integrated filter fusing L1 trend filtering and variational mode decomposition, optimizing parameters by using a sea image optimization algorithm, introducing a causal sliding window to extract features so as to construct a time sequence input tensor, and inputting the time sequence input tensor into a stacked bidirectional long-short-term memory network based on an attention mechanism after noise addition standardization so as to obtain a basic predicted value; calculating a manifold coherent structure, PID dynamics and physical statistical characteristics, and cascading with the basic prediction characteristics to construct a comprehensive element characteristic vector; a LightGBM architecture is constructed, and a prediction residual error is fitted after optimization is carried out through a sea image optimization algorithm; and finally, executing linear reconstruction based on the dynamic safety threshold constraint, and outputting a real-time correction result. According to the method, the error evolution rule is deeply mined by using manifold geometric features, and the real-time precision and robustness of significant wave height prediction are remarkably improved.

The invention relates to an unmanned ship smooth collision avoidance method considering marine environment disturbance, and belongs to the field of unmanned ship collision avoidance. The method comprises the following steps: acquiring observation data of a current ship and surrounding target ships at the current moment, preprocessing the observation data, and storing the preprocessed observation data into a historical state cache; traversing historical cache data of each moment in the historical state cache, performing geometric and kinematic calculation to generate feature vectors of corresponding moments, and stacking the feature vectors of all the moments in sequence according to a time sequence; and inputting a stacking result of the feature vectors at each moment into a strategy network to obtain a collision avoidance control instruction at the current moment, and sending the collision avoidance control instruction to an unmanned ship bottom layer control system for execution. According to the method, a Mama sequence encoder is introduced, ship states at multiple moments are modeled, and time sequence characteristics reflecting historical dynamic change trends are extracted; and an action change rate penalty term is added in the reward function to guide the strategy to output smooth and executable collision avoidance actions, so that decision optimization considering both security and execution stability is realized.