5029 results about "Training set" patented technology

The invention discloses a maritime accident prediction method and device based on interpretable integrated machine learning, and relates to the technical field of maritime affair safety risk analysis, and the method comprises the steps: obtaining accident investigation data, carrying out the preprocessing, balancing the data through a ten-fold layered oversampling method, and carrying out the cross verification training, and determining a performance optimal model by using the test set and carrying out interpretable analysis to explain the influence of the characteristics on the accident prediction result. By constructing a closed-loop'data processing-model optimization-explanation output 'process and adopting SMOTE oversampling and ten-fold layered cross validation training and a heterogeneous base model ensemble learning strategy, the processing capacity of the data imbalance problem of accident categories is improved, the data leakage problem of oversampling is avoided, and the possible bias of a single model is overcome. The interpretability analysis of the model prediction result can quantitatively display the contribution degree of each feature to prediction globally and locally, reveal the nonlinear relationship and interaction effect between the features, and provide transparent interpretation of model decision.

The invention discloses a lithium battery residual life prediction method and system and terminal equipment, and relates to the technical field of lithium battery health management. The method comprises the following steps: receiving a battery capacity attenuation sequence as an original input sequence, detecting and filtering abnormal data by adopting a 3 sigma criterion, and carrying out noise suppression processing on the battery capacity attenuation sequence through a Dropout mask; and carrying out normalization processing on the preprocessed battery capacity attenuation sequence, dividing the battery capacity attenuation sequence into a training set, a verification set and a test set through a sliding window algorithm, and generating a time sequence characteristic matrix and a corresponding residual service life label. According to the method, a neural network structure fusing trend prior perception and dynamic attention regulation is constructed, a multi-scale capacity modeling strategy is introduced to separate a degradation trend, fluctuation disturbance and high-frequency noise, and compared with a traditional time sequence neural network or a single attention model, pseudo fluctuation characteristics caused by capacity regeneration can be more effectively recognized, and the method is more efficient and more reliable. And the judgment accuracy of the model in a complex degradation scene is improved.

The invention provides a sequential network traffic prediction method and system based on swarm intelligence parameter optimization, and relates to the technical field of network traffic prediction. The method comprises the following steps: acquiring indexes such as throughput packet loss rate and round-trip delay of a target link by using a network probe, and performing deletion filling normalization and multi-scale decomposition to obtain a standardized traffic sequence; calculating information entropy, constructing a traffic complexity feature vector, and dividing a training set and a verification set; constructing a hybrid depth prediction model composed of a one-dimensional convolutional network and a gating cycle unit, and establishing a hyper-parameter search space; using particle swarm optimization and entropy-driven inertia weight adjustment and mutation probability mapping to reconstruct a speed and position updating strategy, and iteratively outputting a global optimal hyper-parameter; and generating a benchmark prediction result according to full-amount training, extracting a residual error, training a nonlinear residual error compensation model to carry out superposition correction and reverse normalization, obtaining a final flow prediction result, and improving prediction precision and generalization ability.

The invention provides an underground water pollutant concentration prediction method and system based on machine learning, and relates to the technical field of underground water pollutant concentration prediction.The method comprises the steps that historical data, hydrogeological parameters, meteorological data, human activity data and geochemical parameters of underground water pollutant concentration of a target area are preprocessed; dividing a training set, a verification set and a test set; constructing a preset resolution feature set based on a geochemical mechanism; selecting an adaptive machine learning model according to data characteristics and coupling a physical mechanism; performing hyper-parameter tuning by adopting Bayesian optimization, and supplementing small sample data in combination with transfer learning to complete model training; predicting the underground water pollutant concentration of the target area by using the trained model, and outputting a pollutant concentration prediction result with an uncertainty interval; the invention provides a technical scheme for predicting the concentration of underground water pollutants, which is efficient, accurate and high in adaptability.

The invention discloses a distributed training scheduling and communication optimization method and system of a multi-modal large model on a domestic computing power platform. The method comprises the following steps: virtualizing a heterogeneous computing unit of a preset platform into a virtual device pool, and fusing first-order gradient of a multi-modal sample and Hessian matrix information based on quantitative perception training to generate a sample sensitivity grading atlas; virtual device pool attributes and the sensitivity grading atlas are used as input, an optimal hybrid parallel configuration scheme is automatically generated through a configuration search algorithm, and a parallel combination mode, resource mapping and a high-sensitivity sample scheduling strategy are defined; a distributed training code of an integrated communication optimization strategy is automatically generated according to a configuration scheme, pipeline parallel communication and data parallel gradient synchronization constraint are executed in a topology adjacent equipment subset, and a hierarchical aggregation mechanism is adopted; and dynamically screening a core training set and scheduling a calculation task to complete distributed training. According to the method, efficient cooperative training of the multi-modal large model on the domestic computing power platform is realized.

The invention relates to the cross technical field of computer vision and civil engineering, and discloses a deep learning-based bridge disease image segmentation method, which comprises the following steps of: establishing an image data set containing crack and spalling diseases and performing online enhancement; constructing a segmentation network model comprising a frequency dynamic convolution encoder branch, an edge enhancement Transform encoder branch, a gating cooperation unit, a decoder and a depth supervision module; training the model by using a weighted mixed loss function; and inputting the test set to obtain a final segmentation mask. Self-adaptive fusion of local texture features and global context information is realized through a dual-encoder architecture and a gating cooperation mechanism; a frequency dynamic convolution and edge enhancement module is utilized to enhance the anti-noise capability and micro-disease perception under a complex background; and in combination with a category weighting strategy, the problem of pixel category imbalance is effectively solved, and high-precision automatic segmentation of concrete bridge diseases is realized.

The invention discloses a multi-source heterogeneous anomaly detection method based on time correlation, and belongs to the technical field of water diversion engineering, and the method comprises the steps: S1, obtaining multi-source sensor time sequence data in multi-source heterogeneous data, carrying out the preprocessing of the multi-source sensor time sequence data, and dividing a training set and a test set; s2, constructing a double-branch depth feature extraction network model, training by adopting the training set, and testing through the test set to obtain a trained double-branch depth feature extraction network model; the double-branch depth feature extraction network model comprises a double-branch unit, an attention feature fusion unit, a classifier unit and an output layer unit which are connected in sequence; and S3, inputting to-be-detected data into the trained double-branch depth feature extraction network model, and finally outputting an anomaly diagnosis result. A double-branch depth feature extraction network is constructed, adaptive fusion is realized through an attention mechanism, and the problem of poor modal adaptability of heterogeneous data is solved.

The invention discloses a command and control system resource trend prediction method based on fusion of long and short time sequence characteristics. The method comprises the following steps: acquiring a public power load or similar time sequence monitoring data set, and preprocessing the data in the data set; a deep learning network model based on a TCN-Transformer hybrid model is constructed, a TCN model and a Transformer model are adopted for parallel computing to achieve feature extraction, the TCN model extracts short-term information, the Transformer model extracts long-term features, then fusion features are obtained through a cross attention mechanism and multi-layer perceptron (MLP) weighting, and finally prediction output is generated through full connection layer mapping. Taking data in the training set as input, training the constructed TCN-Transform hybrid model, and continuously optimizing the model until convergence meets a set requirement; and performing prediction by using the trained network model. According to the method, the TCN-Transform hybrid model is constructed, so that local fine-grained features are reserved, the global time trend is effectively captured, and the accuracy of command decision making is improved.

The invention provides a multi-source data fused SAR sea surface significant wave height deep learning inversion method, and belongs to the technical field of remote sensing ocean, and the method specifically comprises the steps: preprocessing buoy data, and obtaining sea wave significant wave height data observed by a buoy; acquiring dual-polarization single-view complex SAR data collected in an interference wide-width mode; preprocessing the SAR data, and taking significant wave height data obtained by buoy observation as a label of the SAR data; acquiring auxiliary data corresponding to the SAR data, wherein the auxiliary data comprises wind speed, wind direction data, rainfall data and OCN data; and constructing a multi-source data-fused SAR significant wave high-depth learning inversion model, training the multi-source data-fused SAR significant wave high-depth learning inversion model, and testing and verifying the trained model by using the test set and the verification set to obtain a final significant wave height inversion result. According to the method, the SAR data inversion capability of the model is improved.

Disclosed in the present invention is an artificial-intelligence-based performance prediction processing method for a carbon-fiber carbonization process. The method comprises: preprocessing experimental data under test, so as to obtain said experimental data that has been subjected to data cleaning; then, using a sliding window processing method to slide on time series data, extracting data within a window at each position and using the extracted data as an input sample, and determining an input feature and an output variable feature of each input sample, so as to convert the time series data into a plurality of experimental data samples under test in the format of a target model input; performing random data set division on said plurality of experimental data samples, so as to obtain some training sets and some test sets; and constructing a target model, and inputting said experimental data samples into the target model. The target model can implement a relatively accurate mechanical-performance prediction for a carbon-fiber-precursor carbonization process, and the model has an optimal performance in all aspects and has a relatively good generalization capability.

The invention provides a reinforcement corrosion electrochemical parameter inversion method based on LSTM (Long Short Term Memory) time sequence prediction, which comprises the following steps: S1, acquiring electrochemical time sequence data in a reinforcement corrosion process through an electrochemical workstation to form an original reinforcement corrosion electrochemical time sequence data set; s2, preprocessing is carried out to obtain a training set, a verification set, a test set and normalization coefficients of all parameters; s3, constructing and training an LSTM time sequence prediction model; s4, constructing and calibrating a steel bar corrosion electrochemical parameter forward modeling model; and S5, constructing an inversion framework fusing a particle swarm optimization algorithm, a simulated annealing algorithm and an Adam optimization algorithm, forming closed-loop cooperation by the particle swarm optimization algorithm, the simulated annealing algorithm and the Adam optimization algorithm so as to minimize an error between a target electrochemical response parameter and a theoretical electrochemical response parameter, and outputting an inversion result. According to the method, through organic combination of time sequence prediction and multi-algorithm cooperation, the problems that a traditional inversion method is low in precision and poor in stability are solved, and a reliable technical means is provided for reinforced concrete structure health monitoring.

The invention discloses an optical cable micro-disturbance identification method based on physical simulation and self-supervised time sequence decoupling, and relates to the technical field of optical cable identification, and the method comprises the steps: constructing a physical digital twin simulator, and generating a high-fidelity training set; constructing a deep learning model, wherein the deep learning model adopts a lightweight time sequence decoupling network; training the model by adopting a staged training strategy, and sequentially carrying out self-supervised noise distribution pre-training, simulation supervised training and spectral domain physical consistency fine tuning operation; inputting DAS time sequence data collected in real time into the trained model, and outputting the data as an optical cable identity ID and a physical position; the lightweight time sequence decoupling network comprises a physical guide preprocessing module, a lightweight U-Net separation module, a sparse gating module and an intelligent parallel decoding module. Through the technical means of simulation-driven data generation, staged training strategies and the like, the defects of the prior art in the aspects of reducing the data cost, improving the detection capability in a low SNR environment, realizing multi-source blind source separation and the like are overcome.

The invention relates to the technical field of image target detection, and discloses an unmanned aerial vehicle aerial image small target detection model construction method comprising the following steps: preparing an aerial image data set, preprocessing the aerial image data set, and generating an aerial image sample set; the method comprises the following steps: establishing a basic model on the basis of a YOLOv8 network model, removing a P5 detection layer in a head network Head of the basic model, introducing a P2 detection layer, replacing a specified position of a Conv module in a backbone network Backbone by adopting an ACMConv feature enhancement module, replacing a specified position of a C2f module in a neck network Neck by adopting a C2fMixStructure mixed structure module, and constructing an improved model by adopting a lightweight enhanced detection head structure; and dividing the aerial image sample set into a training set, a verification set and a test set in proportion, training and verifying the improved model, and generating an unmanned aerial vehicle aerial image small target detection model based on AMLP-YOLOv8. According to the invention, the method has higher perception capability when extracting fine target features, and improves the stability and robustness of small target detection of the aerial image of the unmanned aerial vehicle.

The invention discloses a tunnel or mine water gushing space-time prediction method and system coupled with a hydrodynamic numerical model, and the method comprises the steps: outputting multi-source data based on an identified and verified underground water numerical model, complementing the missing of measured data, quantifying the difference between the permeability characteristics of a fault and a normal stratum, and coupling the difference to a data system, and tunnel or mine excavation space data are merged. And constructing an LSTM-isolated forest-K neighbor regression coupling model, and configuring a multifunctional module to realize multi-scene data co-training. The preprocessed multivariate time series data is divided into a training set and a test set, hidden features are extracted through a coupling model, anomaly detection results are fused, a residual error correction model is synchronously trained, and hyper-parameters and weights are adaptively optimized according to multi-engineering prediction error feedback. And based on the trained coupling model, carrying out synchronous water gushing space-time prediction by adopting a window rolling strategy, and outputting prediction data meeting engineering precision in combination with residual correction. And reliable technical support is provided for safety prevention and control of engineering construction.

The invention relates to the technical field of computer vision, in particular to a space-time consistency data generation method for visual target tracking. The method comprises the following steps: firstly, training a path generator on a target tracking training set, learning a motion law of a target in a time sequence by using optical flow estimation and conditional variation coding technologies, and generating a target motion track conforming to physical constraints; and then, based on the generated target trajectory, introducing a space-time consistency attention mechanism to guide a text-video generation model, and under the condition of keeping basic model parameter freezing, constraining the position, scale and continuity of a target in a generation frame through an attention network, thereby synthesizing a video frame sequence with real motion features. According to the method, target tracking video data with real motion characteristics and high time sequence consistency is generated, and the robustness of the model to complex motion, illumination change and shielding conditions can be improved in different scenes.

The invention discloses a wind power prediction method based on multi-source domain deep transfer learning, and relates to the field of new energy power prediction.The method comprises the steps that the data distribution difference between a multi-source domain and a target domain is reduced through an Euclidean alignment method, and the maximum mean value difference between the domains after optimization is obtained; setting a migration weight factor for each source domain based on the maximum mean value difference, and constructing a weighted migration training set; constructing a multi-source domain deep migration learning model, and carrying out migration training on the model based on a weighted migration training set; and finely tuning the pre-training model according to a small amount of data of the target domain to obtain a target domain wind power prediction model. The wind power data distribution difference between the multi-source domain and the target domain is reduced through the Euclidean alignment method, the knowledge migration efficiency is improved, and the negative migration risk is reduced; setting a migration weight factor for each source domain based on inter-domain MMD, so that the model preferentially learns high-correlation source domain wind power characteristics; and the multi-source domain deep migration learning model is fused into a dynamic weight module, so that low-efficiency migration is avoided.

The invention provides a TLF-GV signal correlation earthquake magnitude prediction method and system, and a medium, belongs to the technical field of earthquake prediction, and aims at solving the problems that a traditional magnitude prediction model lacks physical constraints, TLF-GV features are not fully utilized, the interpretability is poor, and cooperation with a preorder technology is lacked. Comprising the following steps: extracting a multi-source feature vector of a TLF-GV signal containing a gravity component absolute amplitude, a vibration component relative amplitude and a signal phase; based on the G-R law, constructing the positive correlation physical characteristics log10 (GVCIpeak) and BF * Tanomal of the abnormal intensity and the magnitude of the TLF-GV signal; weighting the features according to a Bayesian factor BF value; fitting according to historical samples to obtain a physical empirical formula of positive correlation of GVCI peak logarithm and magnitude as a physical constraint, and constructing a loss function by taking the physical constraint as a regularization item and combining with a mean square error of a magnitude prediction task; and constructing an XGBoost model, and training according to the training set to obtain a prediction model for earthquake magnitude prediction. According to the method, quantitative prediction of earthquake magnitude is realized based on multi-source features of TLF-GV signals in combination with a physically constrained XGBoost regression model.

Various embodiments of the present disclosure provide prompt engineering and text quality assessment techniques for improving generative text outputs. The techniques include identifying a training cluster for an input document, generating a candidate prompt for a generative machine learning model based on the training cluster and a prompt template, providing the candidate prompt to the generative machine learning model to receive at least a portion of a candidate document, generating a plurality of quality metrics for the candidate prompt based on the candidate document, and selecting the candidate prompt from a plurality of candidate prompts based on the plurality of quality metrics.

The invention provides a drug-disease association prediction method and system, computer equipment and a medium, and belongs to the technical field of computers. The method comprises the following steps: constructing a drug-protein-disease heterogeneous network, and extracting a plurality of element path sub-graphs; inputting the meta-path sub-graph into a multi-scale diffusion graph convolution module, executing learnable multi-step graph diffusion on the basis of graph convolution, synchronously capturing local adjacency and high-order topological information, and generating node embedding; and performing dynamic weighted fusion by utilizing meta-path attention to obtain unified representation. In order to relieve imbalance of positive and negative samples, implementing difficult negative sampling in the embedding space, and constructing a balance training set with the positive samples; medicine-disease features are spliced, a regularization XGBoost classifier is trained, and unknown correlation accurate prediction is achieved. By adopting the method, the drug-disease association prediction precision and efficiency are improved, multi-scale topology and priori knowledge are fused, and a powerful calculation tool is provided for drug relocation.

The invention relates to the technical field of satellite hydrological data processing, and particularly discloses a GRACE data super-resolution network space downscaling method fusing geographic information and environmental variables, which comprises the following steps: S1, acquiring original resolution GRACE data and original GLDAS data of a research area, and preprocessing the data; s2, dividing the data obtained by preprocessing in the step S1 into a training set and a test set, and training the GRACE data space downscaling model by using the training set to obtain a trained discriminator and a trained generator; and S3, inputting the GRACE low-resolution data in the test set and the high-resolution environment variable at the moment corresponding to the data into the generator trained in the step S2, and finally obtaining a downscaled high-resolution GRACE image. The method not only can be used for dynamic monitoring of regional scale underground water reserves and flood risk assessment, but also can be expanded and applied to scenes such as agricultural drought monitoring and ecological hydrological process simulation.

The invention requests to protect an unmanned aerial vehicle light-weight small target detection method based on a YOLO-GLL neural network, and the method comprises the following main steps: S1, constructing a YOLO-GLL target detection model based on YOLOv8s, and inputting a training set image to complete model training; s2, extracting different scale feature maps of the image through a backbone network containing a partially grouped multi-scale convolution module; s3, performing cross-level fusion on the feature map by using the lightweight multi-scale fusion feature pyramid neck network; s4, executing target classification and bounding box regression by adopting a lightweight shared detail enhancement detection head to obtain a lightweight high-precision model; and S5, inputting a to-be-detected unmanned aerial vehicle image, and realizing small target detection and identification in combination with the Soft-NMS and Shape-IoU post-processing strategies. The method has the advantages that (1) the detection precision is improved; (2) the parameter quantity of the model is reduced, and the deployment of resource-constrained equipment is adapted; and (3) a post-processing strategy is improved, the problem of missing detection of small targets in a dense shielding scene is relieved, and the robustness of a complex scene is enhanced.

A density perception image defogging method in combination with physical prior comprises the following steps: S1, acquiring a fog-containing image, marking a fog area and a fog-affected area, constructing a data set D1 through size unification, normalization and multi-dimensional data enhancement, and dividing the data set D1 into a training set and a verification set according to a proportion; s2, a density perception defogging network is constructed, the density perception defogging network comprises an encoder sub-network, a middle layer sub-network and a decoder sub-network, core modules are a gridding atmospheric light attention module and a space detail enhancement module, and a GAAB integrates a multi-scale density perception convolution module, a grid perception atmospheric attention module and a feature fusion module; fog density features can be accurately captured, and physical prior can be fused; s3, inputting the data set into a defogging network to generate a multi-scale feature map; s4, carrying out training optimization on the network by adopting a combined loss function of L1 loss and perception loss; s5, evaluating the network performance through the verification set; according to the invention, adaptive defogging in different fog concentration scenes is realized, and the definition of the defogged image is effectively improved.

The invention relates to the technical field of process fault diagnosis. The invention discloses an industrial process fault diagnosis method and device, equipment and a medium, and the method comprises the steps: obtaining historical operation data, and dividing the historical operation data into a training set and a test set; a double-constraint model is constructed, the double-constraint model is trained according to data in the training set, and constraint conditions of the double-constraint model comprise hard sparse neighborhood constraint and manifold structure regularization keeping constraint; testing the trained double-constraint model according to data in the test set; and carrying out industrial process fault diagnosis based on the tested double-constraint model. According to the method, extremely strict constraint conditions are constructed through hard sparse neighborhood constraint and manifold structure keeping regularization constraint in a double constraint model, the problem of small fault detection in a complex industrial process is solved, and high sensitivity and low false alarm rate in fault detection are ensured.

The invention relates to the technical field of new energy power system instability risk assessment, in particular to a new energy power system frequency instability risk assessment method based on a heterogeneous graph attention network. The method comprises the following steps: constructing a wind power-photovoltaic heterogeneous graph model according to a power grid topology; performing combined sampling on different operation modes and anticipated disturbances, and determining corresponding input feature vectors; calculating a frequency stability index value label of the sample set; expanding the training set through an active learning iteration process, and carrying out model training; and inputting the collected operation data into the trained heterogeneous graph attention model, outputting a frequency stability index value, and evaluating the system frequency instability risk in combination with the risk matrix. By adopting the frequency instability risk assessment method for the new energy power system based on the heterogeneous graph attention network, the problem of low efficiency of risk assessment in a high-dimensional uncertain scene is solved, and the heterogeneous graph attention network can reflect the influence of different types of devices at different positions and disturbance types on the dynamic frequency of the system; and the accuracy of frequency instability risk assessment is improved.

The invention discloses a stacking-based steel hot-rolled product mechanical property prediction method, and relates to the technical field of steel product quality prediction. The method comprises the following steps: constructing a steel hot rolling actual production data set and carrying out missing value interpolation on the data set; performing normalization processing on the complete data set, and dividing a part of data from the normalized data set as a training set; establishing a stacking ensemble learning model based on a regression chain; taking the yield strength, the tensile strength and the elongation in the training set as training labels, taking other data types as training features, and training the learning model by using the training set to obtain a steel hot-rolled product mechanical property prediction model; and the steel hot-rolled product mechanical property prediction model is practically applied to perform real-time prediction on the mechanical property of the steel hot-rolled product. According to the method, the coupling relation among the multiple target variables can be effectively processed, and the accuracy and stability of mechanical property prediction of the steel hot-rolled product can be effectively improved.

The invention discloses an interpretable load prediction method based on knowledge distillation and a time self-attention additive neural network, and the method comprises the steps: collecting the historical load and meteorological data of a power grid as the input characteristics of a model, carrying out the detection of a data quartile abnormal value, dividing the data quartile abnormal value into a training set, a test set and a verification set, and carrying out the detection of the data quartile abnormal value; standardization and abnormal value filling are carried out through Z-shaped orthogonalization and linear filling, and finally, a tensor form meeting the model input requirement is converted through a sliding window; designing a knowledge distillation'teacher-student 'framework based on multiple scales and multiple cycles; constructing a time self-attention additive neural network TSA-NAM as a student model; calculating a shape function representing the contribution degree and the characteristic value in the sub-network to obtain the interpretability of the characteristic dimension; exporting the attention weight of the time self-attention module to obtain the interpretability of the time dimension; performing simulation verification; according to the method, high reliability and high precision are guaranteed, and meanwhile, multi-dimensional interpretability is brought to power load prediction.

The invention provides an SAR ship wake detection method based on multidirectional perception convolution and frequency-space fusion attention, and the method comprises the following steps: S1, collecting an SAR image with a ship wake target, arranging the SAR image into a data set, and marking the ship target in the data set; s2, converting the format of the data set labeled in S1 into a YOLO format, and dividing the data set into a training set, a verification set and a test set; s3, on the basis of YOLOv8, constructing an SAR ship wake initial detection model based on multi-direction perception convolution and frequency-space fusion attention; s4, training and verifying the initial detection model by using the training set and the verification set to obtain a final detection model; and S5, detecting the wake target by using the final detection model generated in the step S4, and outputting a detection result. The method has excellent detection precision and robustness for the ship wake target in the SAR image.

The multi-field coupling collaborative modeling and parameter autonomous intelligent decision-making method for the roller type quenching process comprises the steps that four key process parameters including the nozzle flow, the water ratio, the roller gap distance and the plate passing speed are collected and subjected to normalization preprocessing, and a training set and a test set are divided; the four key process parameters of the training set serve as input, the water-cooling convective heat transfer coefficient and the phase change plasticity coefficient serve as output, a BP neural network is trained, and the workpiece surface water-cooling convective heat transfer coefficient and the material internal phase change plasticity coefficient are predicted through the trained BP neural network; a multi-field coupling model composed of a temperature field, a structure phase change field and a stress-strain field is solved, and quantitative calculation of quenching core quality indexes is achieved; by taking water consumption minimization as a target, constructing a nonlinear optimization model under the condition that multiple core quality index constraints of the plate temperature, the outlet hardness and the plate shape are met; and performing global optimization on the optimization model by adopting a differential evolution algorithm, and autonomously deciding an optimal process parameter set value.

The invention discloses a hybrid sequential network power consumption prediction method for reinforcement learning dynamic calibration, and the method comprises the steps: collecting multivariable power consumption time sequence data, completing the data preprocessing through resampling, feature engineering, normalization and sliding window technologies, generating a supervised learning sample set, and dividing the supervised learning sample set into a training set, a verification set and a test set; constructing a hybrid prediction model comprising a dynamic capture module, a long-term dependence modeling module, a regression prediction module and a reinforcement learning dynamic fine tuning module; training and optimizing by adopting a staged training strategy to obtain a hybrid prediction model; multivariable power consumption time sequence data are collected in real time and preprocessed, the preprocessed data serve as input, real-time prediction of future total consumption is conducted through the mixed prediction model, and a final prediction result after dynamic fine adjustment is output. According to the method, accurate and efficient prediction of the power grid load can be realized, and a reliable technical solution can be provided for power system scheduling optimization, demand side management, market transaction and other scenes.

The invention relates to the technical field of automatic driving end-to-end perception, in particular to an end-to-end automatic driving long tail recognition method based on comparative learning pre-training, and the method comprises the steps: firstly generating synthetic image data with long tail distribution characteristics through a conditional diffusion model; a fine-grained scene classifier is adopted to carry out systematic arrangement and semantic annotation on the generated samples, and a structured multi-modal image-text alignment data set is constructed; and finally, fusing the enhanced data set with the original training set, and optimizing a vision-language joint embedding space through a multi-task contrast loss function to realize parameter updating of the pre-training model. According to the method, a closed-loop optimization mechanism of a generative data enhancement and contrast learning framework is creatively established, the problem of data scarcity in a long-tail distribution scene is effectively relieved, and the cross-modal representation capability and downstream task generalization performance of the model on low resource categories are remarkably improved.