22 results about "Residual correction" patented technology

This invention belongs to the fields of new energy technology and artificial intelligence technology, and provides a high-resolution long-term photovoltaic power prediction method and system based on a dual-branch architecture. By constructing a parallel dual-branch architecture, the baseline branch extracts multi-scale features, and the ramp branch identifies power ramping events. The features from both branches are embedded through multiple channels and then fed into a BiGRU-Enhanced Transformer model with shared weights. By strategically deploying BiGRU layers at the input and between the encoder and decoder, local instantaneous dependencies and global long-term trends are captured simultaneously. A multi-task joint optimization loss function is designed for the ramp branch, using weighted cross-entropy and... L The probability, temporal location, and fluctuation amplitude of slope occurrence are jointly optimized using the L1 norm. Finally, nonlinear residual correction is applied to the outputs of the two branches to achieve the final prediction. This invention effectively overcomes the oversmoothing effect of deep learning models, demonstrating extremely high capture accuracy and robustness in long-term prediction of one month's worth of data at 1-minute resolution.

The application discloses an effective wave height double-stage space-time prediction method and system based on diffusion residual correction, relates to the technical field of sea wave prediction, and inputs historical effective wave height data and corresponding historical wind field data into an effective wave height space-time prediction model for processing to obtain a preliminary effective wave height prediction result, and then inputs the preliminary effective wave height prediction result, the historical effective wave height data and synchronous forecast wind field data into a residual correction module based on a diffusion model to compensate for the preliminary prediction error and obtain a final effective wave height prediction result. The diffusion model based on the EDM architecture is innovatively introduced to finely reconstruct the prediction residual, the basic prediction field, the synchronous wind field dynamic factor and the prediction time limit are taken as multi-dimensional physical strong conditions to inject the diffusion generation process, the application can deterministically restore the high-frequency texture details and nonlinear wave components missed by the first-stage model, to a certain extent, the application can relieve the long-term deviation accumulation, improve the prediction quality in high wave areas and improve the prediction stability.

This invention discloses a method for designing the composition of nickel-containing high-strength material systems based on transfer learning and the barrel theory optimization. The method includes: first, using a recursive feature elimination algorithm to screen key features across source domain benchmark data and target domain experimental data containing sparse components; second, constructing a deep neural network for pre-training on the source domain dataset, extracting nonlinear feature representations between components and performance through hidden layers; third, freezing the weights of the feature extraction layers of the pre-trained model, and for sparse samples in the target domain, introducing Bayesian ridge regression to construct a residual correction model to achieve cross-domain knowledge transfer; and finally, combining multiple metallurgical constraint criteria to perform an evolutionary search within a limited composition space. This invention solves the problem of effectively modeling sparse data in new material systems. By combining deep feature extraction with Bayesian residual correction, it significantly improves the robustness of high-strength material performance prediction and enables precise optimization of industrial-grade composition formulations.

The application discloses a surface roughness prediction method based on a diffusion model and neural structure search, performs quantile normalization on process parameters, and performs Z-score standardization on surface roughness values; synthetic process parameter data meeting physical consistency and distribution fidelity is generated; a balanced, high-variety mixed training set is constructed to alleviate the data scarcity problem in a small sample scene; a Gaussian process-based Bayesian optimization strategy is adopted to automatically search for an optimal prediction network architecture matching the data complexity; then residual correction and structure fine-tuning are performed to improve the generalization ability and prediction robustness of the model under a small sample, and a surface roughness prediction value is output. The generalization ability and robustness of the model are improved, higher-precision surface roughness prediction can be realized, thereby providing a general solution for small-sample, high-dimensional and strong nonlinear process modeling in the field of precision manufacturing, and the surface quality prediction of other precision machining processes can be further popularized.

This invention discloses a method for dynamic path optimization of an inspection robot based on reinforcement learning, specifically including: S1, constructing a set of path structural units; S2, constructing a path structural state vector; S3, constructing a set of structural actions; S4, performing continuous-time mapping on the path structural state vector using a liquid time constant neural network that introduces a structural residual correction time constant; S5, selecting structural actions based on the output of the liquid time constant neural network and reconstructing the path structural units; S6, collecting information on structural cost changes after path structural unit reconstruction and constructing a reward vector; S7, performing perturbation sampling and weighted update of the gate parameter subset of the liquid time constant neural network using a natural evolution strategy that introduces hierarchical noise in structural actions. This invention uses structural-level reinforcement learning for dynamic path optimization.

This invention provides a method and related device for real-time prediction of rock mass strength in tunnel excavation advance drilling. By acquiring drilling parameters during the advance drilling process, a mechanism characteristic representing rock breaking strength and energy consumption is constructed based on these parameters. This mechanism characteristic is then input into a baseline prediction model, a residual correction branch, and a physical consistency branch. A fusion gating network generates a first fusion weight corresponding to the residual correction and a second fusion weight corresponding to the physical correction. Based on the baseline prediction value, the residual correction, the physical correction, and the first and second fusion weights, the final predicted value of the uniaxial compressive strength of the rock mass is determined. This invention uses drilling mechanism characteristics as input to construct a collaborative prediction framework consisting of a baseline prediction model, a residual correction branch, and a physical consistency branch. A dual-gating mechanism enables sample-level adaptive fusion and robust output, thereby simultaneously improving prediction accuracy, physical rationality, and engineering applicability.

The application discloses an NDVI downscaling method based on random forest and nonlinear residual correction, belongs to the field of ecological remote sensing and geoinformation technology, and comprises the following steps: 1, constructing a random forest model; 2, multi-source data preprocessing and feature grid construction, generating a total sample data set for random forest model training and prediction; 3, training the random forest model to obtain an optimal random forest model; 4, NDVI downscaling prediction, applying the trained random forest model to a high-resolution feature variable data set to obtain a high-resolution annual average NDVI prediction result; 5, performing nonlinear residual correction on the optimal random forest model based on GAM; and 6, applying the constructed nonlinear residual correction model to all downscaling NDVI data to obtain a corrected NDVI result. The application adopts the above method, and the spatial consistency, time continuity, overall precision and stability of the downscaling result are improved.

The application discloses a numerical control rotary table time grating position precision self-evolution compensation method combining a mechanism model and reinforcement learning, and comprises the following steps: collecting time grating sensor original signals and rotary table running states; performing mechanism rough compensation through a harmonic error model to obtain rough compensation position values; forming a state vector by combining the rough compensation position values and the running states, inputting the state vector into a deep Q network intelligent agent, and outputting compensation residual correction values; superimposing the rough compensation values and the correction values to obtain final compensation values; calculating a reward according to a deviation from a high-precision reference, updating network parameters, and realizing self-evolution; and providing a corresponding compensation system based on the method. The application adopts a double-loop architecture to combine mechanism knowledge and a data-driven method, realizes online adaptive compensation of the numerical control rotary table time grating position precision through a composite reward function and a priority experience playback mechanism, and has the advantages of high compensation precision, strong adaptability and good training safety.

The application discloses a deep reinforcement learning-based oil and gas pipeline corrosion prediction method and platform, and relates to the technical field of oil and gas pipeline corrosion prediction. The method comprises the following steps: collecting pipeline operation related time series data through a multi-dimensional sensor, reconstructing non-equidistant data through an improved signal processing technology, constructing a deep Q network prediction framework of a time series residual error corrosion state prediction algorithm, determining an optimal model through multi-group super parameter combination iterative training optimization, outputting a normalized corrosion rate estimation value combined with a residual error correction mechanism, and obtaining an actual corrosion rate prediction result through inverse transformation; and performing specific processes such as prediction framework modeling, network structure construction, residual error algorithm embedding, super parameter optimization, test data processing and prediction result correction, and the platform realizes the whole-process closed-loop operation of data acquisition, processing, modeling, prediction and analysis relying on each functional unit. The application significantly improves the accuracy and time series adaptability of corrosion prediction, and provides technical support for the safe operation of oil and gas pipelines.

This invention proposes a coarse prediction-guided conditional residual diffusion model channel prediction method. The method first obtains a coarse prediction of the future CSI based on historical Channel State Information (CSI) sequences. Then, using this coarse prediction as the prior mean of the reverse refinement process, a conditional residual diffusion forward process is constructed, transitioning from the true future CSI value to the noisy state in the coarse prediction neighborhood. Subsequently, a conditional reverse refinement network is trained to estimate residual correction terms and noise removal terms based on the diffusion state, diffusion step index, and relevant conditional information. Finally, a denoising diffusion implicit model (DDIM) is used for a few steps of reverse update to obtain the future CSI prediction result. This invention can recover fine-grained random variation information missed in the coarse prediction result while maintaining the overall evolution trend of the future channel, improving the prediction accuracy of complex time-varying channels. Furthermore, by constraining the reverse process within the coarse prediction neighborhood, it enhances generation stability, reduces the number of reverse sampling steps, and lowers inference latency.

This application relates to the field of artificial intelligence technology, and more particularly to a multi-scale drought index prediction method based on a hybrid model. It models the linear trend and periodic variations of the SPEI drought index time series using a SARIMA model; mines nonlinear features from the SARIMA model residual sequence using a long short-term memory network; introduces a lightweight spatial attention mechanism to adaptively weight the importance of different spatial grid locations; and constructs a dynamic residual weight adjustment mechanism to adaptively adjust the fusion ratio of linear prediction results and nonlinear correction results based on the actual contribution of residual correction to improving prediction accuracy. The aim is to address the problem of how to improve the reliability of drought prediction results using a hybrid model.

The application provides a time series prediction method based on multi-scale feature fusion and residual correction, and relates to the technical field of time series data prediction.The original time series monitoring data is acquired and preprocessed, the complexity index is calculated based on the physical features in the first sliding time window, the causal wavelet decomposition is performed according to the complexity index dynamic parameter, the wavelet features and the original features are spliced to obtain an enhanced feature set;the data set is divided according to the time series and the input sample is adaptively intercepted, the first-order difference of the prediction target is obtained to obtain an increment label, the normalized input time series prediction model is obtained to obtain a difference prediction result, the residual is calculated and the secondary modeling is performed to output a residual compensation result, the two are fused and combined with a reference value to restore the difference to obtain a final prediction result.The application effectively separates the non-stationary noise and eliminates the prediction lag effect, and improves the prediction accuracy and stability.

The application discloses a face key point generation method based on emotion prior and low-dimensional residual correction, relates to the technical field of image data processing, and generates a base face key point sequence irrelevant to emotion based on input text; a layered emotion prompt tensor is generated based on an emotion identifier, the layered emotion prompt tensor is injected into a decoder layer by layer, a decoder hidden sequence with injected emotion prompts is obtained, an emotion prior feature sequence is extracted from the decoder hidden sequence with injected emotion prompts, an average value of a base expression point subset sequence in a time dimension is calculated according to a predefined expression point index set, a geometric context feature is obtained, and the geometric change is constrained to a specific expression area to maintain the stability of other areas; and a predicted expression residual is calculated based on the emotion prior feature sequence, the geometric context feature and the emotion identifier; the application effectively suppresses geometric drift and maintains the stability of non-expression points while enhancing emotion regulation capability.

This invention discloses a hyperspectral / multispectral chlorophyll-a inversion method based on feature-level residual learning. First, a regularized regression model is trained using multispectral broadband features and measured chlorophyll-a ground truth values. The difference between the ground truth and predicted values ​​is used to construct a residual error vector. Then, using this residual as the target, a multivariate dimensionality-reduction regression model is trained using hyperspectral high-frequency derivative features. The multispectral feature matrix of each pixel in the target water body is input into the trained regularized regression model to obtain the baseline predicted value of chlorophyll-a concentration for each pixel. The hyperspectral derivative feature matrix of each pixel is input into the trained multivariate dimensionality-reduction regression model to obtain the high-frequency residual correction value for each pixel. Finally, the two values ​​are summed and post-processed using a hyperbolic tangent soft thresholding function to obtain the final predicted value of chlorophyll-a concentration, and the inversion map is output. This invention achieves a fundamental breakthrough in the multi-source fusion paradigm, possesses strong physical noise resistance, and improves the accuracy of high-concentration chlorophyll inversion.

The application provides a soil thickness inversion method, system and device, and belongs to the technical field of soil thickness calculation. The method comprises the following steps: acquiring environmental factor data and soil thickness measured point data of a target area and constructing a modeling data set; adopting a series connection strategy of spatial trend-nonlinear residual correction, first capturing the spatial non-stationary trend of soil thickness by using a GWR model, then performing nonlinear fitting and correction on the GWR residual by using an XGBoost model, and introducing a regularization term and a subsampling strategy in the training process to perform sparse processing and adaptive screening on the environmental factors; finally, superimposing the trend component and the residual correction component to generate a high-precision soil thickness spatial distribution map. The application couples the spatial modeling capability of GWR and the nonlinear learning advantage of XGBoost, and introduces key environmental factors such as slope relative position index, thereby improving the precision and reliability of soil thickness inversion in complex terrain areas.

The application discloses a kind of online evaluation method of rail repair precision, including steps 1: calculating profile dynamic adaptation DPA;Step 2: calculating disease residual correction DRC;Step 3: calculating surface roughness standard degree SRC;Step 4: calculating polishing quality uniformity GUI;Step 5: calculating rail polishing quality comprehensive evaluation index CGQI, the present application fuses profile dynamic adaptation, disease residual correction, surface roughness standard, polishing uniformity these four dimensions, forms comprehensive evaluation system (CGQI), based on rail basic parameters and the contact pressure deduced based on Hertz contact theory, combined with laser profile measurement deviation, realize the dynamic evaluation of "the greater the contact pressure, the higher the deviation influence weight", it is in line with actual service working condition;Supplement microelement area division, profile alignment reference, deviation limit and other key rules, solve engineering landing pain points;Give consideration to macro profile, residual disease, micro roughness, polishing uniformity, more comprehensive evaluation.

This application relates to the field of artificial intelligence technology, and in particular to a method, apparatus, equipment, storage medium, and computer program product for predicting port container throughput. The method acquires multi-source heterogeneous influencing factor data associated with the port to be predicted and classifies and constructs a multi-dimensional feature set; it performs anomaly identification and missing feature imputation on the feature set to obtain a complete multi-dimensional feature set, and aligns it at the time granularity to obtain a unified granularity multi-dimensional feature dataset; it trains an ensemble tree model based on the dataset, and determines the contribution by combining feature perturbation evaluation of out-of-bag sample errors, selects a target feature subset, and constructs a dimensionality-reduced feature sequence; it inputs the dimensionality-reduced feature sequence into a time-series prediction model to output a throughput prediction sequence. The time-series prediction model includes frequency domain processing, state update parameter generation, multi-branch time-series modeling, and gated fusion. Multi-branch time-series modeling generates long-term, short-term, and residual correction branch outputs, and gated fusion weights and fuses the outputs of each branch to obtain the prediction sequence.

The application relates to the technical field of feature extraction of gravitational waves, and particularly discloses a gravitational wave signal feature extraction and classification method fusing multi-modal learning, wherein main strain data and auxiliary monitoring data are synchronously collected and converted into time-frequency spectra and spectral arrays; deep feature extraction is performed on the main time-frequency spectra to obtain a primary feature tensor; the core lies in that a differentiable forward mapping network is constructed, the shallow layer representation of the auxiliary data and the primary feature is taken as input, and a systematic deviation caused by a physical coupling effect, that is, a coupling response prediction tensor, is actively predicted; the deviation is subtracted from the primary feature through adaptive residual correction based on learnable weights to generate a decoupled feature tensor which removes the coupling artifact, the decoupled feature and high-order abstract features of the auxiliary data are subjected to multi-scale attention fusion and reweighting, and then the classification discrimination is performed through causal constraints to output the category judgment result of the gravitational wave signal; and the application improves the purity of the signal features and the robustness of the classification system.

This application discloses a method for assessing small failure probabilities based on prior constraint integration and hierarchical correction sampling, relating to the field of uncertainty quantification technology. The method includes: first, constructing an integrated surrogate model incorporating multiple types of prior constraints and performing residual correction using high-fidelity anchor samples; then, calculating soft failure weights based on the corrected surrogate model, and using this weights to perform hierarchical search of candidate samples to obtain a failure sample set; next, clustering the failure samples to identify multiple failure sample clusters and constructing a multi-scale hybrid proposal distribution; then, performing effective sample size-driven bridging sampling, adaptively controlling the transition step size to obtain a stable initial failure probability estimate; finally, using inverse probability weighted unbiased correction, performing high-fidelity verification of the bridging samples prioritizing false negative risk and correcting the initial estimate, outputting the final failure probability. Under the condition of strictly limited high-fidelity evaluation times, this method achieves high-precision, high-stability, and statistically unbiased estimation of small failure probabilities.

This invention proposes a coarse prediction-guided conditional residual diffusion model channel prediction method. The method first obtains a coarse prediction of the future CSI based on historical Channel State Information (CSI) sequences. Then, using this coarse prediction as the prior mean of the reverse refinement process, a conditional residual diffusion forward process is constructed, transitioning from the true future CSI value to the noisy state in the coarse prediction neighborhood. Subsequently, a conditional reverse refinement network is trained to estimate residual correction terms and noise removal terms based on the diffusion state, diffusion step index, and relevant conditional information. Finally, a denoising diffusion implicit model (DDIM) is used for a few steps of reverse update to obtain the future CSI prediction result. This invention can recover fine-grained random variation information missed in the coarse prediction result while maintaining the overall evolution trend of the future channel, improving the prediction accuracy of complex time-varying channels. Furthermore, by constraining the reverse process within the coarse prediction neighborhood, it enhances generation stability, reduces the number of reverse sampling steps, and lowers inference latency.

The application discloses a kind of MIMO-OTFS system signal detection methods based on least square minimum residual preprocessing and residual injection, belong to wireless communication technical field.The method includes: obtaining received signal vector, time delay-Doppler domain equivalent channel matrix and noise power;With and as input, least square minimum residual iteration is solved, and initial estimation vector and residual vector are obtained;According to residual energy ratio, injection factor is calculated;Enhanced observation vector and extended channel matrix are constructed;Based on residual correction noise precision initial value;With and as input, statistical quantity is used as priori initial value to execute unit approximate message passing iterative detection.The application accelerates iterative convergence and improves detection precision by high-quality initial estimation and residual injection mechanism, effectively suppresses residual interference in fractional Doppler channel.