105 results about "Expected value of sample information" patented technology

The invention discloses a power generation equipment state fault diagnosis method and system based on artificial intelligence, and the method comprises the steps: actively injecting a mechanical excitation signal of a preset frequency spectrum into a key part according to a physical topological structure of power generation equipment, and carrying out the fusion to generate a time-space-frequency three-dimensional data volume; inputting the three-dimensional data volume into a physical embedded variational auto-encoder, and outputting an equipment state pure feature tensor; inputting the pure feature tensor into a graph space-time causal reasoning network to generate a fault propagation causal graph with probability weight; performing multi-agent diagnosis on the fault propagation causal atlas, and outputting a fault diagnosis report which has a credibility interval and comprises fault positioning and root cause analysis; and mapping the fault diagnosis report to the digital twin of the equipment in real time, and outputting a self-adaptive maintenance strategy sequence which minimizes the expected value of the whole life cycle operation and maintenance cost. According to the embodiment of the invention, the accuracy and anti-interference capability of fault diagnosis can be improved, and the operation and maintenance cost can be effectively reduced.

The invention relates to the technical field of computer data processing, in particular to a multi-model collaborative distillation and dynamic fine tuning model training method and system. The method comprises the steps that data collection is conducted, unlabeled data are input into a model to be made into labeled data, the labeled data and labeled data are integrated into a data set, and then the data set is divided into a training set, a verification set and a test set; inputting the training set into the model, guiding the model to generate extended data by cue words, and training a student model by using an extended training set; the student model is evaluated on the verification set, if the evaluation result shows that the evaluation index of the student model does not reach the expected value, error information is input into the model for analysis and data generation, and then the student model is adjusted until the evaluation index of the student model reaches the expected value; the method has the advantages that multiple large language models are used for collaborative distillation, training data are richer and more diversified, student models are more accurate, and the student models are circularly and dynamically fine-tuned, so that the models are higher in precision and adaptability.

The invention relates to an equipment residual life prediction method based on a physical-data model, and the method comprises the steps: building a degradation model based on a physical degradation mechanism and a standard Wiener process model; obtaining fixed parameters in the degradation model by using a maximum likelihood estimation method in combination with a nonlinear regression method; acquiring an equipment state observation value in real time, and updating random parameters in the degradation model by using a weight optimization particle filter algorithm; and utilizing the degradation model, the fixed parameters in the degradation model and the random parameters in the degradation model to obtain a probability density function of the residual life of the equipment under a random failure threshold value, and performing numerical integration on the probability density function of the residual life of the equipment to obtain an expected value of the residual life of the equipment, and outputting the expected value as the residual life of the equipment. Compared with the prior art, high-precision and high-reliability residual life prediction is realized by fusing a physical mechanism and a data driving method.

The invention relates to the technical field of tunnel engineering automation and intelligent control, and discloses a TBM tunneling parameter intelligent optimization decision-making system based on an LSTM network, and the system comprises a data collection and preprocessing module which obtains external data and outputs a tunneling parameter sequence; the probabilistic tunneling trend prediction module is used for outputting a prediction expected value and prediction uncertainty; the prospective geological precursor sensing module is used for matching and identifying known risks and outputting alarm events; and determining an optimal tunneling mode by the dynamic risk avoidance decision matrix. When the prediction uncertainty is too high, activating the prospective template driven by the uncertainty to excavate a new precursor template and update the template library; meanwhile, the decision-efficiency relevance evaluation and strategy self-optimization engine optimizes the decision rule according to the actual tunneling efficiency. According to the method, decision making is carried out through quantitative risk prediction and fusion of multi-source information, and a double learning closed loop of knowledge discovery and strategy optimization is established, so that the reliability, the adaptability and the long-term efficiency of system decision making are remarkably improved.

The invention discloses a time delay constrained Internet of Vehicles multi-mode cooperative sensing method and system. The method comprises the following steps: collecting point cloud data and image data, and carrying out feature extraction, fusion and space projection; constructing a two-dimensional convolution neural network, and extracting a position-level importance probability graph and a mask used for screening key information; constructing a benefit estimation neural network and a constraint evaluation neural network, and predicting expected values of system benefits and constraint conditions; meanwhile, an overall strategy selection module is established; the motorcade carries out target identification and calculates rewards based on information interaction so as to obtain a perception experience sample; and updating network parameters based on the perception experience sample to obtain a final overall strategy selection module. According to the method, the transmitting power, the transmission channel and the sensing data area of the vehicle are optimized based on reinforcement learning, a strategy correction mechanism is introduced, a risk strategy of high time delay or high packet loss is dynamically corrected, and target identification deviation and decision errors caused by sensing information delay and wrong transmission are avoided.

The invention discloses a straw returning decay promotion effect evaluation method and system based on multi-source data analysis, and relates to the technical field of straw returning. A neural network model with influence factor parameters in the straw decomposition process as input samples and straw quality variation as output samples is constructed, the neural network model is trained, and the number of neurons in a hidden layer of the neural network model is adjusted according to a training result until a prediction result reaches an expected value. The method comprises the following steps: acquiring an incidence relation mapped from an influence factor parameter sample to a straw quality variation sample through a constructed neural network model, acquiring time sequence data of influence factors in a straw decomposition process in a monitoring area, acquiring a straw quality variation when the influence factors change once in the decomposition process according to the incidence relation, and calculating the straw quality variation according to the straw quality variation. Calculating and acquiring an accumulated value of the straw mass variation from the initial moment of the decomposition process to the moment t, and calculating the decomposition rate at the moment t according to the acquired accumulated value of the straw mass variation.

The invention discloses an MIQP problem optimization processing method and system based on an improved random branch strategy, and the method comprises the steps: constructing a graph network model according to a multi-source historical data set, and traversing a path from a power node to each load node in a target power distribution network based on a graph search algorithm, generating a branch strategy candidate set of target power distribution network reconstruction according to the path; reconstructing a target power distribution network in the typical power distribution network operation scene, and calculating expected values of different branch strategies in the branch strategy candidate set in the reconstruction process; inputting the real-time load characteristic data, the real-time line characteristic data and the real-time topological characteristic data into a pre-trained decision tree model to obtain an initial weight parameter of each branch strategy in a target power distribution network reconstruction process; and generating a strategy weight parameter of the branch strategy based on the initial weight parameter and the expected value, and optimizing a solution process of target power distribution network reconstruction according to the strategy weight parameter. According to the method provided by the invention, the efficiency of the reconstruction process is improved.

The invention relates to the field of text semantic analysis, in particular to a special field text semantic element recognition method based on an enhanced CRF layer, and the method comprises the steps: carrying out the sample screening analysis of a sample library, so as to obtain a plurality of training samples; for a single training sample, performing semantic element extraction on the training sample according to a preset strategy function to obtain a tag sequence; calculating an accumulated discount reward corresponding to the tag sequence; obtaining a probability award expected value corresponding to the tag sequence by adopting a preset objective function, performing gradient calculation by adopting a preset gradient formula, and performing parameter updating for the model parameters; determining whether to reselect the label according to a preset exploration strategy; under a preset stopping condition, stopping model training learning; according to the method, the capability of effectively capturing the complex dependency relationship between the tag sequences by the model is improved, so that the accuracy and robustness of the model on a semantic element recognition task are improved, and the capability of processing a long text is improved.

The invention discloses a medical treatment risk detection method, device and equipment and a computer storage medium, and relates to the technical field of data processing. The method comprises the following steps: acquiring medical-seeing communication behavior data of a user; inputting the medical-seeing communication behavior data of the user into a medical-seeing risk detection model, and determining a medical-seeing risk classification result corresponding to the medical-seeing communication behavior data of the user according to a pre-trained target expected value table; the target expected value table comprises a corresponding relation between the doctor-seeing communication behavior data and the doctor-seeing risk classification result, the target expected value table is obtained by training the doctor-seeing communication behavior data and a target user tag, and the target user tag is determined according to a relative distance among a user clustering tag, a user prediction tag and a data point; the user clustering label is obtained by clustering according to the local density and the minimum distance of the user data points, and the user prediction label is obtained through a pre-trained classifier. According to the embodiment of the invention, the optimal decision is made according to the target expected value table, and the risk identification accuracy is improved.

The invention discloses a primary frequency modulation optimization system and method for a new energy station and a medium. The method comprises the following steps: performing principal component analysis on to-be-measured historical data to obtain dimension-reduced historical data; training a plurality of XGBoost models by using a sample set formed by the historical data after dimension reduction, obtaining data to be measured at the current moment, inputting the data to be measured at the current moment into the trained XGBoost models for net load prediction, and obtaining a predicted net load at the next moment; obtaining an expected value and a standard deviation of the predicted net load at the next moment according to the distribution of the predicted net load at the next moment, calculating a net load variable quantity at the next moment according to the expected value of the predicted net load at the next moment and the net load at the current moment, and calculating a frequency deviation according to the net load variable quantity at the next moment and an equivalent inertia constant; and performing frequency modulation control according to the frequency deviation and the standard deviation of the predicted net load at the next moment. According to the invention, high-precision prediction is carried out on the load by using the extreme gradient lifting algorithm, the frequency modulation response speed of the new energy generator set is improved, and the accuracy of frequency regulation is also enhanced.

The invention relates to a multi-source scene-driven power distribution network probabilistic load flow calculation method and system, and the method comprises the steps: obtaining the historical data of 24 hours per day in a complete year of load and photovoltaic, and carrying out the clustering division based on a comprehensive similarity index, thereby obtaining a plurality of optical load typical daily output scene sets and the probability distribution thereof; processing data of each scene set to obtain an expected value and a variance of optical charge power; combining the network topology of the power distribution network, taking an expected value as a node injection amount, and performing probabilistic load flow calculation by adopting a two-point estimation method; and carrying out statistical analysis on a calculation result to obtain related statistical data of branch power and node voltage in each typical scene. According to the method, the typical scene probability distribution set is generated by using historical data, the estimation point selection precision and the calculation result reliability are improved, the calculation times are reduced, the efficiency is improved, the actual power grid operation behavior can be better reflected, decision support is provided for a power grid operator, resource allocation is optimized, the cost is reduced, and renewable energy utilization is promoted.

The invention discloses a water pump operation energy-saving control method and system based on multi-source data analysis, and relates to the technical field of water pump energy saving. The method comprises the steps of obtaining state parameters actually measured at the same moment and corresponding water pump energy consumption value data, eliminating abnormal data points in sampling data, training a neural network by using the sampling data until a square error expected value of a prediction result is smaller than a set expected threshold value, and outputting a mapping relation between the state parameters and water pump energy consumption values. Obtaining the adjustable interval of the active state parameter and the interval of the water pump energy consumption value corresponding to the measured value of the passive state parameter at the current moment by utilizing the mapping relation, performing anomaly detection on each water pump energy consumption value in the interval by taking lambda as a detection step length, and obtaining the reasonable interval of the water pump energy consumption value at the current moment; and the active state parameter value corresponding to the reasonable interval upper limit of the water pump energy consumption value at the current moment is taken as the optimal state value of the state parameter, and the active state parameter is adjusted.

The invention relates to the technical field of insurance data processing, in particular to a customer resource data system and analysis method in the insurance industry, and the system comprises a data simplification module which is used for calculating expected values and variances of hidden variables and observation variables of each insurance record based on the insurance records of insurance customer data, the method comprises the following steps: comparing principal component vectors of insurance records, screening the vectors, calculating cosine similarity between the vectors, and integrating the vectors with high similarity to generate a principal component matrix; according to the invention, through deep analysis and optimization processing of the insurance customer data, the data precision and the decision reliability are improved. Through calculation of expected values and variances of hidden variables and observation variables in the insurance records, screening and high-similarity integration are carried out on principal component vectors, a simpler principal component matrix is formed, and the accuracy of data is enhanced.

The embodiment of the invention discloses a data reconciliation method and device based on anomaly detection, belongs to the technical field of intelligent reconciliation, and can improve the efficiency and accuracy of data reconciliation. Comprising the steps of obtaining source data of right and interest data to be reconciliated, and calculating a reconciliation result corresponding to the right and interest data according to a preset reconciliation algorithm; inputting the reconciliation result into a pre-trained anomaly detection model, and determining an abnormal value of the reconciliation result according to an expected value of a path length of a first isolated forest in the anomaly detection model and reconciliation factors of the reconciliation result; and detecting an account checking result based on the abnormal value, when the account checking result is abnormal, detecting source data of the right and interest data by using a pre-trained abnormal detection model to obtain a detection result, and performing right and interest data account checking based on the detection result, the detection result comprising normal information or error information of the source data.

The application relates to the technical field of automatic testing, and discloses an interface assertion method, a system, a terminal and a storage medium. The method comprises the following steps: when an automatic testing process corresponding to a target business scenario is executed for the first time, if it is detected that a target interface call return state is consistent with an expectation, valid data returned from the target interface is acquired as expected response data; based on a preset business process association rule, corresponding data records are extracted from a plurality of database tables related to the target business scenario as expected business data; the expected response data and the expected business data are integrated into an expected data set, and after the expected data set is associated with a use case identifier, the expected data set is stored in an expected value database to form a data snapshot; when a testing process of a same business scenario as the target business scenario is executed subsequently, actual response data and actual business data are acquired, and the actual response data and the actual business data are compared with corresponding data snapshots in the expected value database to generate an interface automatic assertion result.

The invention relates to a Bayesian online detection method and device for an electric spark perforation machining penetration state, and the method comprises the steps: calculating an expected value of posterior probability distribution of key parameters, and dynamically updating a probability statistical model constructed by an original signal in an electric spark perforation machining process; based on the dynamically updated probability statistical model, modeling and calculating the duration of the current processing stage; and on the basis of the duration of the machining stage, the current machining stage of the electric spark perforation is quantitatively identified, and the moment when the machining stage is converted into the machining penetration stage from the blind hole machining stage is detected online according to a quantitative result, so that an online detection result of the machining penetration moment is generated. According to the embodiment of the invention, the feeding speed of the tool electrode in the processing process can be sampled as the characteristic signal, the dynamically updated probability statistical model is established to describe the electric spark perforation processing state, and the adaptive detection of the penetration moment with high robustness is realized by identifying the quantized model parameters and identifying the instant penetration state.

The application provides a traffic prediction method and device, electronic equipment and storage medium, wherein the method comprises the following steps: determining historical traffic data and a missing position matrix of the historical traffic data; inputting the historical traffic data and the missing position matrix into a prediction completion model to obtain a prediction result output by the prediction completion model; the prediction completion model is obtained by training an initial model on the basis of weight parameters of a data completion module in the initial model, and the training is performed on the initial model by applying sample data, sample missing data, sample data expected values and a sample missing position matrix; and the weight parameters of the data completion module in the initial model are obtained by pre-training based on the sample data, the sample missing data and the sample missing position matrix. The method provided by the application realizes information interaction between a prediction module and a data completion module in a prediction model, completes end-to-end prediction, improves instantaneity, and improves the accuracy of a prediction result.

The embodiment of the present application provides a screening method and a screening system, the screening method comprises the following steps: obtaining historical flow data of a plurality of target objects from a storage system, and extracting historical values of target variables of the plurality of target objects from the historical flow data; generating a predicted selection probability corresponding to the target objects according to the historical values of the target variables; inputting the predicted selection probability and the historical values of the target variables into an established expectation value model, performing screening analysis on the target objects, and generating expected values of the target variables of the plurality of target objects; and screening the target objects whose expected values of the target variables are greater than a preset threshold value, and storing the corresponding identification information of the target objects into a database. According to the predicted selection probability of the target objects, the expected values of the target variables are obtained by using the expectation value model, and the target objects are screened according to the expected values of the target variables, so that the required target objects can be quickly and accurately screened, and the screening quality is better.

This application provides a reward data analysis method, apparatus, device, storage medium, and program product. The method includes: collecting target data generated by a model executing a game; wherein the target data includes: reward data obtained from the robot's action data under the corresponding state data at each moment in the game, and the mapping relationship between the state data, action data, and reward data; using a neural network model to infer reward data corresponding to the state data and action data; determining a first SHAP value for the state data relative to the expected value and a second SHAP value for the action data relative to the expected value; the first SHAP value characterizes the degree of influence of the state data on the expected value, and the second SHAP value characterizes the degree of influence of the action data on the expected value; adjusting the reward data corresponding to the robot's state data according to the first SHAP value, and adjusting the reward data corresponding to the robot's action data according to the second SHAP value.

The application belongs to the technical field of knowledge graph, and particularly relates to a knowledge reasoning method and system based on an agent dynamic path completion strategy, which extracts entities and the relationships between the entities in a target knowledge graph, and mines rules in the target knowledge graph and confidence score corresponding to the rules; a reinforcement learning agent is constructed, the agent is dynamically guided to complete the path of the knowledge graph according to the current entity state and historical path information through rules, and the total reward of the agent is calculated according to the hit reward when the agent hits a target entity and the rule reward when the reasoning path of the agent conforms to the rules, the strategy network of the agent is trained by maximizing the expected value of the total reward of the agent, and the corresponding knowledge reasoning result is obtained by using the trained agent to perform path reasoning in the knowledge graph for a given target condition to be queried. The application dynamically completes the most likely path in the reasoning process by using the dynamic path completion strategy to obtain a complete reasoning path, and solves the reasoning truncation problem caused by the missing path of a sparse knowledge graph.

This invention provides a blockchain-based data verification method and apparatus, relating to the field of blockchain technology and applicable to the financial technology field or other technical fields. The blockchain-based data verification method includes: receiving data verification information from a client; wherein the data verification information includes data to be tested, a data model, an expected value, and an error value; determining the actual data based on the data to be tested and the data model; determining the verification result based on the actual data, the expected value, and the error value; and sending the verification result to the client. This invention can achieve data validity evaluation while protecting the privacy of the data holder's data.

The application discloses a method and system for strengthening main grid elements considering renewable energy under extreme weather, and mainly relates to the technical field of renewable energy research. The method comprises the following steps: collecting an expected value of system loss load caused by extreme weather, and establishing a first objective function with the minimum expected value; modeling an operation mode of a main grid; establishing a linear programming model for calculating a worst probability distribution of uncertain variables; establishing a three-layer mixed integer linear programming model according to the first objective function and the modeling of the operation mode of the main grid; and solving the three-layer mixed integer linear programming model and the linear programming model of the worst probability distribution and outputting a result. The application has the beneficial effects of solving the problem of difficult information acquisition, reducing the difficulty of information provided by a power grid operator, and improving the efficiency of power transmission grid restoration.

Embodiments of the invention disclose a target system ground state energy calculation method and a related device. The method comprises the steps of calculating a Hamiltonian of a to-be-solved target system and an opinion of an operator in a predefined operator pool; performing de-randomization processing on the peer by using a shadow chromatography method to obtain an optimized peer, the component included in the optimized peer being less than the component included in the unoptimized peer; and determining the ground state energy of the to-be-solved system according to the size of the expected value of the optimized etenant. By adopting the embodiment of the invention, the calculation speed of the ground state energy of the target system is improved, and the consumption of hardware resources is reduced.

One example method includes receiving a Quadratic Unconstrained Binary Optimization (QUBO) problem that comprises a matrix that includes various cells having data. It is then determined that one or more of cells is missing data or has corrupted data. A machine learning (ML) model performs a denoising process that removes random noise from the one or more cells having the missing data or corrupted data. This results in data being imputed to the one or more cells having the missing data or the corrupted data. The imputed data approximates the missing data or approximates an expected value of the corrupted data before the corrupted data was corrupted.

The present invention discloses a self-supervised behavior recognition method and system based on adversarial training. The self-supervised behavior recognition method based on adversarial training includes: extracting video data and inputting it into a standardized flow model; performing attack processing on the video data to obtain an adversarial video; calculating the adversarial loss of the video data and the adversarial video in a learning mode; updating the expected value parameters of the standardized flow model based on the adversarial loss, and judging whether the update is completed; if so, generating adversarial samples based on the standardized flow model and the expected value parameters; and inputting the adversarial samples into comparative learning. The present invention attacks samples based on the adversarial attack of the standardized flow, and also treats the attacked samples as positive samples and adds them to the training of comparative learning for adversarial training, thereby improving the accuracy and robustness of the model.

The invention discloses a cloud model-based tailing dam rainfall stability risk assessment method. The method comprises the following steps of S1, establishing a complete index system; s2, carrying out denoising preprocessing on the monitoring data by adopting a variational mode decomposition (VMD) algorithm; s3, carrying out weight analysis, and determining a combined weight; and S4, constructing a cloud model, determining a standard cloud of the cloud model, and judging a risk level. According to the method, the VMD algorithm is innovatively introduced to carry out denoising reconstruction on the monitoring data, original noisy data does not need to be stored, only effective signal components are reserved, risk level distribution characteristics can be described only through expected value-entropy-hyper-entropy parameterized representation of the cloud model and by using three groups of core parameters, dependence of a traditional method on full-amount historical data is replaced, and the method is high in reliability and high in reliability. The problem that the size of the buffer area is difficult to match the data size is thoroughly solved; under the mechanism, the risk evolution key information can be completely reserved with lower storage cost, and risk misjudgment caused by data loss can be avoided.

The present application belongs to the technical field of health state evaluation and prediction of mechanical transmission system, and proposes a kind of parallel misalignment state mechanical transmission component wear amount probability prediction method.It includes the following steps: step S1: build the three-dimensional model of mechanical transmission component containing parallel misalignment error;Step S2: considering the random uncertainty of load profile, carry out wear simulation analysis based on grid node iteration;Step S3: considering the cognitive uncertainty of the model itself, train the Bayesian long short-term memory network model;Step S4: predict the expected value of wear and its uncertainty interval.The present application improves the description accuracy of traditional finite element method on mechanical transmission component wear behavior, breaks through the limitation of wear amount deterministic numerical prediction, provides high robustness and high reliability of probability prediction result, and provides scientific basis and data support for transmission system condition-based maintenance decision.

The present invention relates to the technical field of computer data processing, and in particular to a training method and system for a multi-model collaborative distillation and dynamic fine-tuning model; the method comprises: collecting data, inputting unlabeled data into a model to make labeled data, integrating the data with labeled data into a data set, and then dividing the data into a training set, a validation set and a test set; inputting the training set into the model, using prompt words to guide the model to generate extended data, and using the extended training set to train a student model; evaluating the student model on the validation set, and if the evaluation result shows that the evaluation index of the student model does not reach the expected value, then inputting error information into the model for analysis and data generation, and then adjusting the student model until the evaluation index of the student model reaches the expected value; the method has the advantages of using multiple large language models for collaborative distillation, making the training data richer and more diverse, and the student model more accurate, and then dynamically fine-tuning the student model through a cycle, so that the model has higher accuracy and stronger adaptability.

There are provided systems and methods for a machine learning model for probabilistic anomaly detection in streaming device data. A server obtains a plurality of features associated with a user interaction between a user device and the server. The server selects a combination of features from the plurality of features, where the combination of features comprises features having a variance of expected values that exceeds a threshold variance. The server selects a prediction engine to process the combination of features with a corresponding non-parametric statistical model of a plurality of non-parametric statistical models and to generate a prediction indicating a likelihood that the combination of features represents an anomaly corresponding to fraudulent activity. The server issues a remedial action to the user device through an application programming interface with a remedial action engine based on the prediction for applying the remedial action on the user interaction.

The embodiment of the invention relates to the field of artificial intelligence, and discloses a model training method, a video generation method, devices, equipment and a medium, and the model training method comprises the steps: obtaining a first segmentation model, a sample video and a reference segmentation point; training the first segmentation model according to the reference segmentation points to generate a second segmentation model; constructing a first score model for endowing segmentation scores for statements in the sample video based on the second segmentation model; training the first score model according to the reference segmentation points to generate a second score model; generating a statement data set according to the training data set; an objective function is constructed, and the output of the objective function is matched with a segmentation score expected value and the updating step length of the second segmentation model; and training the second segmentation model according to the statement data set until the output of the target function is converged to obtain a target segmentation model for segmenting the sample video based on the content theme of the video, so that the model accuracy, the content theme recognition accuracy and the editing effect are improved, and the segmentation point output by the model is closer to the actually required segmentation point.