219 results about "Short-term memory" patented technology

The invention discloses a power transmission line fault diagnosis and operation and maintenance scheduling method and system based on networking learning, and relates to the technical field of power transmission line fault diagnosis operation and maintenance scheduling. Related data is extracted to construct a high-risk equipment area and a visual high-risk area thermodynamic diagram, a visual risk grading diagram is constructed in combination with electrical quantity data, and meanwhile, an intelligent recognition storage network and a fault type classification recognition model are constructed in combination with a convolutional neural network-long and short-term memory network hybrid model; the model is optimized through networking learning and an attention mechanism, maintenance teams and resources are autonomously allocated in combination with an operation and maintenance management system, then autonomous optimization and closed-loop operation are achieved, full-process coverage of fault sensing, intelligent decision making and efficient response is achieved, the response time after a line fault occurs is remarkably shortened, and the maintenance efficiency is improved. And the fault handling and operation maintenance capabilities of the power grid system are comprehensively enhanced.

The embodiment of the invention provides an elderly emotion recognition method and device based on multi-modal perception, and the method and device achieve the optimization and enhancement of the signal quality through innovatively constructing a multi-modal data preprocessing mechanism and integrating the facial expression, voice and posture features. And designing a personalized feature mapping model based on historical emotion expression data, and establishing an adaptive feature fusion strategy for intelligent matching in combination with a cross-modal attention network. A hierarchical time sequence classification mechanism is introduced, dynamic modeling of the emotional development trend is realized through a long-short term memory network, and accurate prediction of the emotional state is supported. According to the method, the defects of the traditional technology in the aspects of multi-modal processing, personalized modeling, time sequence analysis and the like are effectively overcome, and the accuracy and reliability of sentiment recognition of the old people are remarkably improved.

The invention belongs to the technical field of ship fuel consumption prediction and big data analysis, and relates to an operating ship fuel consumption prediction method based on a deep learning model. The method comprises the following steps: 1, synchronously acquiring multi-source data from a ship automatic identification system, a cabin monitoring system and an ERA5 meteorological database, and preprocessing the multi-source data; step 2, inputting the preprocessed data into a pre-trained CNN-BiLSTM-Attention model to carry out ship fuel consumption prediction; wherein the CNN-BiLSTM-Attention model is composed of a convolutional neural network, a bidirectional long and short term memory network and a time attention module; the model adopts a composite loss function based on navigational speed-power-fuel oil physical constraint. The method is more excellent in the aspects of prediction precision, robustness and fitting effect. Compared with the prior art, the method has higher accuracy and reliability in the aspect of ship fuel consumption prediction.

The invention relates to the field of power system operation and maintenance, in particular to a power transformation equipment operation and maintenance risk online assessment method and system. A power transformation equipment operation and maintenance risk online evaluation system comprises a data acquisition module, a weight configuration module, a sequence risk evaluation module, a collaborative decision game module and a disposal scheme output module. According to the method, a multi-source data fusion and dynamic threshold mechanism is introduced, real-time state quantity, historical maintenance records, operation modes and external weather information are uniformly mapped to a convolution-long and short-term memory network, key features are adaptively amplified in a feature weighting layer, redundant features are weakened, and collaborative recognition of short-term fluctuation and long-term degradation is achieved; compared with a traditional fixed threshold value or single monitoring quantity model, the method can keep sensitive and steady risk early warning capacity under the complex working conditions of severe weather, heavy load operation and the like, the false alarm rate and the missing report rate are greatly reduced, potential faults are locked in advance, and sudden power failure events are avoided.

The invention provides a chemical ship cabin intelligent coating leakage positioning and early warning control method and system, and relates to the technical field of intelligent monitoring, and the method comprises the steps: collecting data through a multi-class sensor array, constructing a parameter distribution map, extracting features through a convolutional neural network, carrying out the time sequence analysis through a long and short term memory network, and carrying out the early warning. The variational auto-encoder constructs a leakage feature fingerprint database, collects data in real time, then carries out feature extraction and fusion processing, matches the data with the fingerprint database, determines a leakage position and a leakage rate, divides a monitoring area, and executes prevention and control. The leakage monitoring precision and the response speed can be improved, and the chemical accident risk is reduced.

The invention belongs to the field of industrial thermal engineering and intelligent modeling, and discloses a heat exchanger heat exchange coefficient prediction method and system based on a physical information neural network. The method comprises the following steps: acquiring multi-dimensional operation data through a signal acquisition system, cleaning abnormal and blank values, standardizing, and segmenting into time sequence samples by adopting a sliding window method; a double-layer physical information long-short-term memory network is constructed, and a time sequence feature and a physical equation residual error are combined to generate a space-time fusion feature matrix. And a composite loss function including data loss, physical equation loss and physical consistency loss is designed, physical and data driving influences are balanced through hyper-parameter tuning, and accurate prediction of the heat exchange coefficient is achieved based on a gradient descent optimization model. The method combines field physical laws and data features, improves the reliability and physical interpretability of prediction, and is suitable for operation optimization of the heat exchanger of the desulfurization wastewater treatment system of the thermal power plant.

The invention discloses a new energy access region power grid balance scheduling method based on artificial intelligence, and relates to the technical field of power grid scheduling, and the method comprises the steps: data collection and preprocessing: obtaining and processing power grid multi-source data and micro-scale meteorological data; meteorological feature coding: extracting dynamic meteorological features by using a long short-term memory network; constructing a dynamic space-time hypergraph and embedding nodes, and generating node dynamic embedding in combination with a graph convolutional network; based on power grid state prediction and pre-fault analysis of causal intervention, accurate prediction and fault identification are realized; and generating and executing a pre-fault scheduling strategy, and generating and executing an optimization strategy through reinforcement learning, so that the method can realize the transformation of the power grid from response type recovery to prospective self-healing, improves the toughness, reliability and economy of the power grid in an extreme scene, and is suitable for the power grid balance scheduling of a new energy access region.

The invention relates to the field of electric power systems, in particular to an electric power facility remote intelligent maintenance method and system. The method comprises the following steps: collecting three kinds of heterogeneous monitoring data of vibration spectrum, infrared thermal imaging and partial discharge signals; mapping the data to a three-dimensional feature fusion space, and calculating a mahalanobis distance to generate a fusion feature matrix; extracting spatio-temporal features by using a convolutional long-short-term memory network, and outputting a triple diagnosis result including a fault type, a severity level and an evolution trend; searching a dynamic maintenance scheme in a maintenance strategy knowledge graph based on the diagnosis result; solving an optimal resource scheduling scheme by adopting an improved Hungary algorithm in combination with the geographic topology and the resource state; the maintenance operation is remotely guided through the augmented reality terminal, and real-time verification is carried out; and collecting the maintained data, carrying out residual analysis, and reversely optimizing the knowledge graph. The method realizes full-process intelligent management, improves timeliness and reliability of operation and maintenance of electric power facilities, and is suitable for a remote intelligent maintenance scene of an electric power system.

The invention discloses an oral cavity data monitoring and early warning method and system based on deep learning, and the method comprises the steps: solving a problem that the focus recognition is inaccurate because the image collection process of an oral cavity endoscope of a patient is affected by various environment and equipment parameters; according to the method, synchronous binding of collected images and equipment parameters and intelligent image preprocessing are proposed, historical time sequence images and variation trend features are combined, lesion space and time features are extracted through combination of a lightweight convolutional neural network and a long-short term memory network, feature weighted reasoning is realized by using a context awareness attention fusion network, and the focus space and time features are extracted. According to the method, the model parameters and the risk threshold value can be dynamically adjusted, the focus recognition accuracy and the early warning individuation are improved, and clinical grading treatment and intelligent health management are effectively assisted.

The invention belongs to the technical field of mechanical equipment wear state monitoring. The invention provides a gearbox health state stage identification method and system based on oil characteristics, and the method comprises the steps: carrying out the preprocessing of an obtained oil abrasive particle image, and extracting the oil characteristics according to the preprocessed oil abrasive particle image; according to the oil characteristics and a pre-trained long-short-term memory neural network model, obtaining a gearbox health state classification result; training of a long-short-term memory neural network model: fusing the oil characteristics of each oil abrasive particle image sample to obtain a one-dimensional health factor, and obtaining a health index sequence according to the one-dimensional health factor of each oil abrasive particle image sample; and carrying out breakpoint detection on the health index sequence to obtain three breakpoints, and adding health state labels to all the oil abrasive particle image samples according to the obtained breakpoints. According to the method, accurate division of the wear stages and capture of the evolution trend can be automatically completed, and finally decision judgment exceeding artificial experience is formed.

The invention relates to the technical field of dynamic visual target motion tracking control, in particular to a dynamic visual target motion tracking control method and system based on deep learning, and the method comprises the steps: synchronously collecting continuous multi-frame target scene image data through a visual multi-frame collection module; and performing time sequence association and memory fusion on target features in continuous multi-frame target scene image data through a cross-frame feature memory fusion module, and constructing a target feature model. According to the invention, the current and historical stable features are dynamically fused through the cross-frame feature memory fusion module, time sequence association is realized in combination with the long and short-term memory network, and the problem of slow feature model updating under target deformation and shielding is solved; the deformation-shielding bimodal recognition module accurately recognizes a scene state, provides a basis for the multi-branch Kalman filtering prediction module, enables the multi-branch Kalman filtering prediction module to call a corresponding branch, corrects a prediction equation through a compensation factor, and improves the position prediction accuracy.

The invention relates to the technical field of tunnel deformation prediction, in particular to a BPNN parameter inversion and LSTM tunnel deformation prediction method. On the basis of the orthogonal test design principle, multiple parameter combinations are designed for the M-C constitutive model and the HSS constitutive model, numerical calculation is carried out, and vertical displacement data of the tunnel measuring points are obtained through numerical calculation; constructing a BPNN model, and training the BPNN model by taking the vertical displacement data as input and the corresponding constitutive model parameters as output; inputting field actual measurement tunnel settlement data into the trained BPNN model, and performing inversion to obtain optimal constitutive parameters of M-C and HSS constitutive models; performing forward numerical simulation on the whole process of pipe jacking construction by adopting the optimal constitutive parameters obtained by inversion to obtain a tunnel vertical displacement evolution rule; and inputting a vertical displacement monitoring value of the measuring point of the downlink tunnel based on the long-short term memory neural network to obtain a prediction result of future settlement of the tunnel. The accuracy of the predicted value is improved.

The invention relates to the technical field of digital control, and provides a digital prefabricated pump station and a digital control system.Perception data of liquid level height, pressure difference, pump set flow, blockage displacement and pipeline vibration are obtained through a perception judgment module, abnormal working conditions are comprehensively judged, limitation of parameter judgment is overcome, and a reliable basis is provided for follow-up control; the pump-valve cooperation module dynamically adjusts the opening degree of a valve through multiple parameters, optimizes the expansion amount of flexible connection in combination with pipeline vibration and blocking displacement, and cooperatively adjusts the rotating speed of a pump set based on the liquid level height, the flow of the pump set and the blocking displacement, so that linkage control over the pump, the valve and the pipeline is achieved. The energy efficiency optimization module achieves accurate prediction of the liquid level through long and short term memory network processing, optimizes the valve opening degree and the pump set rotating speed by combining a particle swarm optimization algorithm and taking pump efficiency maximization as a target, dynamically generates an optimized pump stopping liquid level, and meanwhile adjusts network parameters through deviation feedback of a predicted value and an optimized value to form a continuously optimized closed loop.

The invention provides an old well reexamination method and system based on a multi-modal digital rock debris virtual well, and the method comprises the steps: obtaining mineralogical data of an old well rock debris sample, optimizing an original well logging curve of an old well through the combination of Bayesian correction, predicting an undrilled layer of the old well through the constraint of a long-short-term memory network neural model and seismic data, and obtaining an optimal well logging curve; and the potential level of the undrilled layer of the old well is quantitatively evaluated through the potential matrix, so that the problems of deep rock debris data, low review efficiency and difficulty in potential layer identification in traditional old well review are solved.

The invention discloses a direct current charging pile detection system and method applied to the field, and relates to the technical field of power equipment on-line monitoring, and the method comprises the steps: inputting a three-dimensional data set into a Toeplitz cyclic measurement matrix for compressed sampling, reconstructing the three-dimensional data set into a high-resolution signal through employing an improved orthogonal matching pursuit algorithm, and carrying out the detection of the high-resolution signal; analyzing and calculating a local variable coefficient through a sliding window, marking abnormal positions, and integrating the abnormal positions into an abnormal position set; inputting the abnormal position set into a pre-trained long and short term memory prediction model, evaluating the insulation residual life and the fault risk level, and calculating the maximum allowable output current of the DC charging pile in real time; the load impedance is dynamically adjusted and detected through the memristor array, the output current of the direct current charging pile is collected in real time, and synchronous compression wavelet transform is adopted to analyze a harmonic ridge line and extract a harmonic component; the accuracy of fault early warning is remarkably improved, rapid dynamic adjustment of the output current is achieved, and the technical problem of multi-physics field coupling monitoring of the rapid charging pile is effectively solved.

The invention relates to the technical field of infectious disease monitoring and early warning, and discloses an intelligent influenza early warning system based on community multi-modal data fusion. The community-level multi-modal data fusion architecture is constructed, medical health data, environmental data, crowd activity data and network behavior data are integrated, spatial-temporal features are dynamically extracted and fused in combination with a deep learning model, and the problem of community monitoring blind areas caused by a single data source of an existing early warning system is solved; a long short-term memory network and convolutional neural network cascade architecture is utilized to capture a localized propagation rule, and the defect that a region-level prediction model cannot adapt to community heterogeneity is overcome; the risk score is generated in real time, the grading response instruction is triggered, a'monitoring-early warning-intervention 'closed loop is established, the early warning timeliness is remarkably improved, a basic-level response chain scission gap is filled, early prevention and control of flu outbreak are finally achieved, and public health resource consumption is reduced.

The invention discloses an electrical system load prediction method based on a data center, and the method comprises the steps: obtaining electrical parameters, environmental parameters and historical operation data of each subsystem through a data collection module, carrying out the normalization and time sequence reconstruction of the collected data, and forming standardized time sequence data; a combined neural network model comprising a convolutional neural network (CNN), an improved long-short term memory (LSTM) network and an attention mechanism is utilized to extract local fluctuation, mutation, frequency and statistical characteristics, and capture of data long-term dependence and key time sequence information is realized, so that an electrical load prediction value is output, and finally real-time regulation and control of a data center electrical system are realized. The method can effectively improve the prediction precision and response speed, and provides reliable technical support for the full life cycle management of the data center.

The invention discloses a fault prediction and self-repairing method and device, electronic equipment and a storage medium, and relates to the technical field of computers. According to the method, the function of dynamically monitoring various data of processor hardware for the target node can be realized, the dynamically monitored hardware state data is input, the fault probability is predicted by a method of weighting and combining key indexes through a bidirectional long-short-term memory model and an attention mechanism, the possible faults are intelligently predicted, and the fault prediction efficiency is improved. And when the target node has a test fault in advance, the target node is repaired in a gradual load reduction mode, so that the effects of real-time monitoring, accurate prediction and rapid self-regulation are achieved. Manual intervention is reduced, the intelligent decision-making capability is achieved, operation and maintenance automation and intelligentization are achieved, resource self-adaptive repairing can be integrated, the self-adaptive capability is improved, and the complex scene fault sensing capability is improved.

The invention relates to the technical field of power distribution network fault monitoring, in particular to a fault recognition method and system based on an intelligent fusion terminal, and the method comprises the steps: obtaining an instantaneous multi-dimensional electrical data set at each moment, and constructing an input sample with the current moment as an end point, inputting the input sample into the trained long-short-term memory model to calculate a first fault probability of the input sample; inputting the input sample into a trained optimal fuzzy clustering model to calculate a second fault probability of the input sample; and distributing respective weights for an output result of the long and short term memory model and an output result of the optimal fuzzy clustering model, carrying out weighted fusion on the first fault probability and the second fault probability to obtain a comprehensive fault probability of the input sample, and judging whether the power distribution network has a fault according to the comprehensive fault probability. According to the invention, through multi-source information fusion, model collaborative optimization and dynamic weight distribution, the fault identification precision and response speed of the power distribution network under complex conditions are effectively improved.

The invention discloses an oil chromatography data trend classification method and system based on depth feature enhancement and an attention mechanism, and the method comprises the steps: carrying out numeralization conversion, deletion detection and grouping trend calculation on oil chromatography original gas component data, and generating a basic feature vector; executing multi-scale sliding statistics, change rate and subsequence feature enhancement, and calculating comprehensive similarity and attention weight based on a template library to generate a weighted similarity vector; splicing the enhanced feature and the weighted similarity vector into a time sequence input sequence, and outputting an oil chromatogram trend classification result after attention expansion and long and short term memory network processing. According to the method, structured processing and basic trend extraction of data are realized, adaptive matching and weighted aggregation of historical operation modes are realized, and a multi-dimensional dependency relationship and time sequence dynamic change are captured, so that accurate classification of oil chromatogram trends is realized.

The invention discloses an agent construction method and device based on a large model, a storage medium and equipment, and belongs to the technical field of artificial intelligence. Selecting a dialogue scene according to the current dialogue information and the scene description information; obtaining tool calling data from an intelligent agent corresponding to the dialogue scene; generating a short-term memory context according to the historical dialogue information, generating a tool context according to tool calling data, generating a file context according to a file, generating a long-term memory context according to the current dialogue information and the historical dialogue information, and forming a scene context; generating a candidate tool list sorted according to preferences according to the current dialogue information or the tool calling data; generating a system prompt word according to the scene context and the candidate tool list; and optimizing the intelligent agent according to the system cue word and the candidate tool list to obtain the intelligent agent based on user preference and context awareness. The intention and preference of the user can be recognized, answering is carried out according to the intention and preference of the user, and the accuracy is improved.

The invention discloses an intelligent bird strike prevention and control method and system based on multi-source data, and relates to the field of aviation safety, in particular to the intelligent bird strike prevention and control method based on the multi-source data. The intelligent bird strike prevention and control method based on the multi-source data comprises the steps that aircraft position data are collected; acquiring bird position data; preprocessing the data; constructing a hybrid space-time convolutional neural network, a long-short-term memory network and a physical constraint aircraft trajectory prediction model; constructing a bird trajectory prediction model of a mixed long and short-term memory network and a graph neural network; constructing a collision detection model combined with a mahalanobis distance; risk grade evaluation; and risk grading response. According to the invention, the problems that the data structure is single and different air defense measures are not taken for different risk levels in the prior art are solved. Therefore, the probability of bird strike accidents is effectively reduced, and the safety performance of air transportation is improved.

The invention belongs to the field of man-machine interaction, and particularly discloses a multi-scene switching method, system, equipment and medium based on a song ordering table, and the method comprises the steps: collecting the behavior data of residence time, song ordering frequency and interaction intensity of a user in a virtual environment through a sensor, and processing the behavior data through a long and short term memory network, obtaining a user dynamic preference vector; matching is carried out according to the user dynamic preference vector and a preset scene type library, if the matching degree is higher than a threshold value, it is judged that the current scene preference is stable, otherwise, it is judged that potential switching intentions exist, and potential switching intention probability distribution is obtained through the judgment; after the probability distribution of the potential switching intention is obtained, a support vector machine classifier is adopted to train association features among multi-scene data; the invention aims to solve the problems of resource loading delay and unsmooth interface switching caused by rapid change of user scene preference in a virtual environment in the prior art.

The invention discloses an incomplete multi-mode dialogue emotion recognition method and system based on a speaker and time sequence information joint graph network, and the method comprises the steps: obtaining the deep features of a text mode, a voice mode and a visual mode in a dialogue through a feature extraction module, and guaranteeing the high expression capability of the features through a pre-training model; the random mode missing simulation module effectively simulates the data incomplete condition in a real scene, and the robustness of the model is improved; a bi-directional long-short term memory network (Bi-LSTM) is combined with a time sequence diagram network (TGNN) to capture context and time dynamic characteristics of a dialogue, and meanwhile, an interaction relationship between speakers is modeled through a speaker influence matrix, so that joint modeling of a time sequence and speaker information is realized; the deep features are further extracted through the graph convolutional network, and the emotion discrimination of the features is enhanced; finally, the modal reconstruction and emotion classification module significantly improves the accuracy and robustness of incomplete multi-modal dialogue emotion recognition through reconstruction of missing modals and multi-class emotion prediction, and is suitable for man-machine interaction and emotion analysis application in a complex real scene.

The invention discloses an intelligent marketing terminal power data communication security protection method and system, and belongs to the field of power data security, and the method comprises the steps: collecting electromagnetic noise spectrum data through a built-in environment electromagnetic sensing module of a terminal, extracting electromagnetic characteristic factors, and constructing a correlation model of the electromagnetic characteristic factors and the complexity of an elliptic curve cryptosystem; building a terminal twinborn mirror image, predicting a communication instruction sequence and calculating a deviation coefficient in combination with a long-short-term memory neural network; classifying and screening the power data collected by the terminal, desensitizing privacy data, and then fusing the privacy data into the hash fragment of the unique identifier of the terminal equipment to generate a desensitized power data packet containing the terminal identity identifier; a data abstract is generated according to encryption parameters determined by the association model, the access authority is verified, a data acquisition timestamp is converted into a binary coding sequence, the data abstract is embedded, and an encrypted data packet with a time sequence traceability watermark is generated and uploaded to the block chain for evidence storage; according to the invention, full-process security management and control of power data are realized, and the communication security and the data credibility are improved.

The invention relates to the technical field of florescence prediction methods, in particular to a florescence prediction method based on a GOD model, which comprises the following steps: acquiring historical daily average temperature data of a target area, and preprocessing the historical daily average temperature data; based on the preprocessed historical daily average temperature data, constructing and training a long and short-term memory network model for predicting day-by-day temperature in a future time period, and outputting a day-by-day temperature sequence; on the basis of the predicted day-by-day temperature sequence, a growth day model is adopted to calculate a daily growth day and an accumulated growth day, and threshold values of an initial flowering stage and an end flowering stage are determined; the florescence length is obtained by subtracting the last florescence from the early florescence, and temperature data predicted by combining the GOD model with the LSTM model is adopted, so that high-precision phenological prediction is realized, and reliable support is provided for flower appreciation tourism and agricultural planning; the method can improve the prediction precision of the flowering phase.

The invention discloses a linkage regulation and control method and system for a vacuum environment, and belongs to the technical field of vacuum environment regulation and control, and the method specifically comprises the steps: collecting time sequence data of a crystal growth environment in real time, constructing a temperature-vacuum-growth stage three-dimensional mapping relation, and dynamically correcting a vacuum threshold value by combining a furnace body material thermal expansion deformation compensation algorithm. Generating a multi-parameter coupling vacuum regulation and control model; generating a dynamic vacuum threshold curve of each growth stage based on a vacuum regulation and control model, and outputting a cooperative control signal through real-time deviation calculation; a three-level linkage control strategy is adopted to maintain the vacuum degree stable, a long-short-term memory network is utilized to predict the change trend of the vacuum degree, and a standby vacuum unit is activated in advance; a distributed optical fiber strain sensor is used for detecting leakage sound wave characteristics of a sealing surface, positioning a failure point and driving a magnetic compensation plug to perform dynamic plugging, and meanwhile, leakage information is fed back to a vacuum regulation and control model, so that closed-loop optimization is realized; the vacuum degree control precision and the equipment reliability in the crystal growth process are improved.

The invention discloses a charging equipment self-diagnosis method and system based on multi-source heterogeneous data space-time fusion in the technical field of charging equipment diagnosis. The method comprises the following steps: performing data preprocessing on an obtained charging pile data set to obtain a processed data set; according to the processed data set, constructing a charging pile real-time topology network based on a dynamic graph convolutional network, and extracting a multi-dimensional spatial feature matrix; according to the processed data set, extracting a multi-dimensional time feature matrix based on a lightweight long-short term memory network; performing feature fusion based on an attention feature fusion algorithm according to the multi-dimensional spatial feature matrix and the multi-dimensional time feature matrix to obtain a spatial-temporal feature matrix; and according to the spatial-temporal characteristic matrix, realizing multi-scale anomaly detection based on a variational auto-encoder algorithm after adversarial training enhancement and sliding window dynamic statistics. According to the method, the blindness and resource consumption of troubleshooting can be effectively reduced, and real-time data support is provided for equipment update decision and maintenance resource allocation.

The invention discloses a rainfall downscaling method and system based on a deep learning network model fusing rainfall priori knowledge, and the method comprises the steps: firstly collecting the topographic data and low-resolution day-by-day rainfall data of a target region, and taking the data as input data; a short-term high-resolution precipitation field generated in a mesoscale weather forecast WRF mode is used as training truth value data; according to the method, the function of accurately downscaling the rainfall data in combination with the convolutional neural network and the long and short term memory network is realized, the spatial-temporal correlation of rainfall is fully considered in the downscaling process, and meanwhile, a likelihood function combined with coupled censored data, Box-Cox conversion and time variation variance Gaussian distribution is adopted as a rainfall loss function; the method not only can represent zero expansibility, skewness and heterovariance characteristics of rainfall, but also can improve the rainfall downscaling precision and quantify the uncertainty of rainfall downscaling, and is suitable for wide popularization and use.

The invention belongs to the field of artificial intelligence and contact type measurement, and discloses a blood oxygen saturation degree measuring method based on a smart phone camera. The method comprises the following steps: acquiring fingertip video data by using a smart phone under the condition of no external equipment, extracting a pulse photoplethysmography (PPG) signal from the fingertip video data, and constructing a multi-dimensional feature set through feature extraction and calculation; furthermore, the invention provides a two-channel deep learning model fusing a convolutional neural network and a long-short term memory network, the multi-dimensional feature set and the preprocessed PPG signal are processed in parallel, the advantages of the two networks in the aspects of feature extraction and time sequence modeling are fully exerted, and the accuracy of feature extraction and time sequence modeling is improved. And the sufficiency of the extracted features and the generalization ability of the model are ensured, so that the accurate estimation of the oxyhemoglobin saturation is realized. The research provides a feasible scheme for health self-examination in a family environment.