60 results about "Bayesian information criterion" patented technology

The invention relates to an Internet of Things time sequence root cause analysis method based on a dynamic causal diagram, and belongs to the technical field of Internet of Things. The method comprises the following steps: embedding a fine-tuning large language model by utilizing an Internet of Things knowledge graph, embedding, splicing and constructing a graph structure through entities and relationships, and combining text word embedding and mask prediction task optimization model; on the basis of knowledge graph subgraph construction, a triple is converted into a natural language to be input into a large model to generate a causal hypothesis; assumptions are converted into causal constraints, dynamic causal graph structure learning is carried out in combination with a Bayesian information criterion scoring function, and conditional probabilities of father nodes and historical values are modeled; performing parameter learning by adopting kernel density estimation, and quantifying time hysteresis among the features; feature probability distribution is predicted based on sliding window observation data, accumulative error contribution is calculated through asymmetric Shapley values, and causal ancestor features are preferentially sorted to determine root causes. Real-time and effective root cause analysis of the Internet of Things system is realized, and the system has the capability of intelligently solving faults.

The invention relates to the technical field of API security protection, and discloses an API anomaly detection method, system and device based on behavior fingerprints during operation and a medium, and the method comprises the following steps: collecting operation log data in an API calling process in real time; preprocessing the log data, extracting multi-dimensional behavior characteristics from the preprocessed data, and constructing an API behavior fingerprint vector; based on the historical behavior fingerprint vector, a Gaussian mixture model is adopted to establish a normal behavior reference model; dynamically determining the optimal clustering number of the Gaussian mixture model through a Bayesian information criterion; and inputting a behavior fingerprint vector called by the API in real time into the Gaussian mixture model, calculating an abnormal score, judging an abnormal behavior, regularly and incrementally updating parameters of the Gaussian mixture model, and re-optimizing the clustering number. According to the method, through dynamic behavior modeling and a closed-loop self-adaptive mechanism, the accuracy of abnormal recognition in a complex environment and the long-term stability of the system are remarkably improved.

The invention relates to the technical field of e-commerce, and relates to a digital shopping mall management SaaS system, and the system comprises the steps: collecting shopping mall call and network events, and generating an embedded vector through optical reserve wavelength division multiplexing nonlinear mapping; under the constraint of the causal structure in the last round, the de-noising diffusion model generates twinborn events, an updated causal graph is formed through incremental Bayesian information criterion learning, and a causal feature tensor is obtained through random walk embedding; a reinforcement learning framework formed by the neural morphological execution network and the GPU evaluation network fuses tensors with inventory, price and promotion business data into a value state, and outputs a price adjustment-discount-replenishment strategy; the strategy is compiled into a WebAssembly micro contract, and the components are subjected to hot replacement without shutdown through a BPF sandbox; generating a zero-knowledge proof for the micro contract and storing the proof on the side chain of the block chain; in the gray stage, profit, inventory and cost feedback is collected and written back to a data link to complete self-adaptive closed loop; according to the invention, millisecond-level strategy iteration, second-level security release and extreme scene robust operation are realized.

The invention discloses a transcritical working medium thermophysical property intelligent prediction method suitable for physical property mutation characteristics, and the method comprises the steps: obtaining the physical property-temperature data of a target working medium under a specified working pressure through a standard physical property database, and employing a piecewise linear regression change point detection algorithm based on residual sum of squares and dynamic monitoring; recognition of inflection points of a working medium physical property-temperature curve and self-adaptive division of temperature intervals are achieved, polynomial order optimization is carried out on each segmented interval based on a Bayesian information criterion to balance fitting precision and complexity, continuity constraints at segmented connection points are applied, and a globally continuous segmented polynomial function is constructed. According to the method, the full-process automatic transcritical working medium physical property-temperature function modeling of data acquisition-inflection point identification-order optimization-function generation can be realized, the subjectivity and limitation of artificial experience segmentation are effectively overcome, and the over-fitting risk is avoided while the model precision is ensured; and reliable support is provided for engineering simulation and thermodynamic system design under the transcritical working condition.

The invention provides a photovoltaic power prediction method and device based on dual-mode hybrid analysis and electronic equipment, and relates to the technical field of power systems and photovoltaic power generation. The method comprises the following steps: acquiring a current photovoltaic power sequence; carrying out adaptive noise complete set empirical mode decomposition on the current photovoltaic power sequence to obtain multiple groups of intrinsic mode functions; based on a preset clustering algorithm and a Bayesian information criterion, performing clustering processing on the plurality of groups of intrinsic mode functions to obtain a plurality of data sub-pools; for each data sub-pool, inputting the plurality of intrinsic mode functions in the data sub-pool into a pre-trained BiLSTM model and a pre-trained SVR model to obtain a first prediction result and a second prediction result; performing fusion processing on the first prediction result and the second prediction result to obtain a target prediction result; and determining a photovoltaic power prediction sequence according to the target prediction results of all the data sub-pools. The method is used for improving the accuracy of photovoltaic power prediction and the reliability of a prediction result.

The embodiment of the invention discloses a GBIC-based multi-electric aircraft starting power generation system DBN model lightweight method, and relates to the technical field of reliability design and modeling of an equipment complex system.The GBIC-based multi-electric aircraft starting power generation system DBN model lightweight method comprises the steps that a DBN model of a starting power generation system is established, nodes in the DBN model correspond to key components and working states of the starting power generation system, and the nodes in the DBN model correspond to the key components and the working states of the starting power generation system; directed edges in the DBN model are used for describing the mutual relation between the nodes, a redundant structure in the DBN model is recognized, lightweight processing is carried out, and the fault state of the starting power generation system is recognized through the DBN model subjected to the lightweight processing. Aiming at the problems of complex reasoning calculation and low efficiency in the use process of the dynamic Bayesian analysis method of the equipment complex polymorphic system, the grey system theory and the Bayesian information criterion are introduced, the lightweight of the dynamic Bayesian network of the complex polymorphic system is realized, the redundant structure is eliminated, and the accuracy of the result is ensured.

The invention discloses an oil reservoir history fitting method based on graph lasso and set smooth multi-data assimilation, and relates to the technical field of oil and gas field development. The method comprises the following steps: firstly, acquiring to-be-optimized model parameters and observation data of a target oil reservoir, constructing an observation error covariance matrix and initializing a model parameter set, performing numerical simulation by utilizing an oil reservoir numerical simulator to obtain prediction data, constructing a joint state matrix, performing dimensionless standardization processing on the joint state matrix to obtain an empirical correlation coefficient matrix, and then, performing optimization on the joint state matrix. And executing a graph lasso algorithm combined with an extended Bayesian information criterion to obtain an optimal dimensionless sparse precision matrix, extracting correlation coefficient sub-blocks required by Kalman updating from the optimal dimensionless sparse precision matrix based on a Scherr's theorem, obtaining a robust data auto-covariance matrix through a reverse reduction physical quantity outline, calculating Kalman gain in combination with an observation error covariance matrix, and calculating a Kalman filter. And the model parameter set is updated until the preset condition is met, the reservoir history fitting model parameter set is output, and the stability and precision of reservoir automatic history fitting are improved.

Facies of a formation are classified from data charactering properties of a portion of the formation as a function of depth, wherein the number of facies is determined automatically in an unsupervised manner without human input. In one embodiment, a layer-based methodology is provided that performs facies classification based on layer-based properties which are determined from well log data obtained from a plurality of different well logging tools. In another embodiment, a depth-based methodology is provided that performs facies classification based on well log data obtained depth-by-depth from a plurality of different well logging tools. The number of facies can be determined automatically without human input, for example using the Bayesian Information Criterion or a method which determines the optimal number of clusters based on the repeatability of the clustering results. In embodiments, the facies classification can be performed using the Gaussian mixture model (GMM) method.

The invention discloses a cascade water-wind-light multi-power system modeling method based on time-space correlation, and the method comprises the steps: building an autoregression integral moving average-generalized autoregression condition heterovariance model of wind power photovoltaic power and cascade hydropower station runoff, building an R-vine, C-vine and D-vine Copula function spatial correlation model through employing a generalized autoregression condition heterovariance value, and carrying out the modeling of a cascade water-wind-light multi-power system. And selecting an optimal spatial correlation model according to a Bayesian information criterion, an akaike information criterion and a logarithm likelihood value goodness of fit index. Based on the optimal spatial correlation model, constructing a scene generation method of Latin hypercube-Copula function sampling, performing scene generation on the wind power photovoltaic power and the runoff volume of the cascade hydropower station, and constructing a scene reduction method of contour coefficient optimization Kmeans clustering to reduce the generated scene. The invention provides a time-space correlation cascade water-wind-light multi-power system modeling method, and aims to analyze water-wind-light multi-energy complementary operation characteristics and provide reference for a scheduling mode among water-wind-light multi-energy.

The invention provides a construction interruption risk reasoning and plan generation method and device. The method comprises the following steps: acquiring multiple types of risk factors influencing construction interruption and a historical data set; taking a Bayesian information criterion as a scoring function, and constructing a directed acyclic graph structure based on the scoring function, a historical data set and a constraint set determined by domain knowledge; the constraint set is used for limiting whether edges between nodes exist or not in the construction process of the directed acyclic graph structure; based on the directed acyclic graph structure and the historical data set, a conditional probability table of the directed acyclic graph structure is determined, and the directed acyclic graph structure and the conditional probability table form a Bayesian network model; determining the risk probability of the target construction interruption event based on a Bayesian network model; and optimizing a preset construction plan combination based on the risk probability, and determining an optimal plan combination. According to the directed acyclic graph structure, objective data and domain knowledge are considered, and the accuracy of subsequent risk reasoning and plan generation is improved.

The invention relates to the technical field of nonferrous metal beneficiation intelligent control, and discloses a lead zinc ore flotation process state evolution modeling method. The method comprises the following steps: collecting multi-source heterogeneous data of a flotation process, preprocessing the multi-source heterogeneous data, then obtaining a plurality of key features, and constructing a feature matrix according to the plurality of key features; constructing a hidden Markov model based on the hidden state of the flotation process, inputting the feature matrix to the hidden Markov model, optimizing parameters of the hidden Markov model through a Baum-Welch algorithm, and screening evolution models in a training process in combination with a Bayesian information criterion; and acquiring real-time multi-source heterogeneous data of the flotation process and acquiring a corresponding feature matrix, inputting the feature matrix into the evolution model, and acquiring an optimal process state sequence in combination with a dimensional bit algorithm. The problem that an existing beneficiation model cannot effectively and accurately predict the flotation process state is solved.

The invention relates to a shield tunneling machine cutterhead torque parameter prediction method. The method comprises the steps that equipment operation data and geological environment data in the tunneling process of a shield tunneling machine are periodically collected; performing data preprocessing on the equipment operation data and the geological environment data to obtain preprocessed data; determining a stability data sequence in the preprocessed data and a difference order corresponding to the stability data sequence through ADF stability test, and obtaining parameters of stability data sequence adaptive optimization based on a Bayesian information criterion to generate an optimal ARIMA model; performing prediction processing on the preprocessed data by adopting an optimal ARIMA model to obtain a predicted value of the shield tunneling machine cutterhead torque parameter; adjusting the tunneling speed of the shield tunneling machine and the rotating speed of the cutter head in the next period according to the predicted value of the torque parameter of the cutter head; and the steps are repeated until the shield tunneling machine is turned off. The construction cost of the shield tunneling machine is reduced, and the engineering efficiency is improved.

The invention discloses a multi-dimensional night urination behavior monitoring system based on rhythm features, and relates to the technical field of biomedical signal processing, and the system specifically comprises a data acquisition module, a rhythm feature extraction module, a urine volume dynamics modeling module, a clustering monitoring module and a real-time feedback module; collecting individual data in real time through IoT equipment and preprocessing the individual data; calculating a nocturnal urination frequency index, a nocturnal urination time concentration ratio, a urination interval rhythm variation coefficient and a deviation index based on the individual data, and performing rhythm feature extraction; performing quadratic polynomial least square fitting on the night accumulated urine volume, and calculating a urine volume acceleration index by using an obtained second derivative to quantify a urine volume generation trend; the method comprises the following steps: constructing and preprocessing a night urine multi-dimensional digital phenotypic vector matrix, fitting a Gaussian mixture model based on an expectation maximization algorithm of a Bayesian information criterion, automatically mapping individuals into four types of subtypes according to cluster center features, and outputting individual subtype labels and confidence coefficients; and obtaining a comprehensive risk score through normalized risk assessment.

The invention relates to the technical field of uncertainty probability modeling, and provides a limit load modeling method and device for a high-temperature adhesive, a computer readable storage medium and electronic device.The method comprises the steps that self-service sampling processing is conducted on an original limit load sample set of the high-temperature adhesive, and N self-service sample sets are obtained; n is an integer greater than 1; fitting each candidate probability distribution model by using each self-service sample set, and calculating a Bayesian information criterion value corresponding to each candidate probability distribution model; under each self-service sample set, selecting an optimal model based on a Bayesian information criterion value; and counting the frequency of each candidate probability distribution model selected as an optimal model under the N self-service sample sets, and determining the candidate probability distribution model with the highest frequency as an optimal fitting distribution model of the limit load data of the high-temperature adhesive. The invention provides a scientific and reliable method for high-temperature adhesive limit load distribution modeling.

The application discloses a non-invasive load decomposition method based on CEEMDAN and FastICA, and belongs to the technical field of load monitoring. The method comprises the following steps: S1, collecting active power of total load and various single loads and preprocessing; S2, constructing a completely adaptive noise ensemble empirical mode decomposition (CEEMDAN) model, and decomposing total load power; S3, estimating the number of sources based on Bayesian information criterion; S4, performing dimension reduction by using maximum information coefficient (MIC); S5, performing load decomposition by using FastICA blind source separation; and S6, evaluating the approximation degree of decomposed signals and source signals. The application realizes load decomposition from the perspective of signal blind source separation, reduces the cumbersome load information feature extraction, and obtains complete load information through decomposition. Compared with deep learning, the model training time is greatly reduced.

The present application discloses a signal jitter separation method, device, medium and equipment, relating to the technical field of digital signal processing. The present application performs jitter separation based on the spectrum separation method, quantifies the serial signal and performs clock recovery, and then resamples using the time interval error data, jitter timestamp data and clock signal. Since it is the best fit for the number of single frequencies, it is necessary to first fit the sequence of the number of multi-tone signals that may be included in the signal, and finally fit the optimal number of single frequencies. Since the optimal model of the Bayesian information criterion is used for fitting, both the goodness of fit of the model and the complexity of the model are considered, avoiding overfitting, ensuring faster calculation while taking into account the accuracy of the calculation, and being more flexible at the same time. After determining the target number of single frequencies, when extracting the mixed single-frequency information, it can greatly reduce the calculation time, more accurately and quickly separate the jitter information, and improve the effect of jitter separation in serial bus communication.

The invention belongs to the technical field of load prediction, and particularly relates to an external load dynamic prediction method. Comprising the steps that S1, different driving working conditions are simulated through a virtual prototype, a dynamic load data set of the multi-crawler walking device is obtained, and the dynamic load data set is a set of dynamic load values of the multi-crawler walking device walking on the ground at each moment and characteristic parameters related to the dynamic loads; s2, performing data preprocessing on the dynamic load data set to obtain an initial structured data set; obtaining an optimal time window length based on a Bayesian information criterion, and obtaining a recurrent neural network prediction model in combination with the initial structured data set; and S3, obtaining a to-be-detected characteristic parameter sequence based on a rolling time window method, and inputting the to-be-detected characteristic parameter sequence into the recurrent neural network prediction model for processing to realize dynamic load online prediction of the multi-crawler walking device. The application range and the calculation efficiency of the dynamic load prediction method are improved to the maximum extent.

The present invention discloses a method and device for predicting photoacoustic imaging array signals for limited field of view sampling. The method includes: detecting the acoustic pressure signals reflected by the tissue of the object to be measured through ultrasonic transducers in a photoacoustic imaging array, converting the acoustic pressure signals into electrical signals, and performing stationarity tests on the electrical signals; selecting the truncation number and the trailing number through the autocorrelation function and the partial autocorrelation function, and then combining the truncation number and the trailing number pairwise to determine the order of the photoacoustic imaging array signal prediction model based on the Bayesian information criterion; determining an autoregressive moving average model according to the order, and then weighting it with an LSTM network to obtain a photoacoustic imaging array signal prediction model; inputting the electrical signals obtained after the stationarity test into the photoacoustic imaging array signal prediction model for prediction, and ending the prediction when the number of photoacoustic signals output by the photoacoustic imaging array signal prediction model reaches the preset number of ultrasonic transducers, so as to obtain the predicted photoacoustic signals.

This invention relates to a method for highlighting spike discharge data of intracranial electroencephalogram (EEG) signals. The method involves collecting background noise data from the patient's brain without neuronal discharges, preprocessing the background noise data, automatically selecting the optimal order of an autoregressive (AR) model using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), estimating the AR model coefficients using the Yul-Walker equation, and constructing a background noise model. The intracranial EEG signals to be processed are then subjected to high-pass filtering. A short-time Fourier transform (STFT) and window-based frame-by-frame processing strategy are used to subtract the spectrum of the signal from the noise. By adjusting the parameters of the spectral subtraction, noise removal and preservation of neuronal signal features are achieved.

The application discloses a kind of ocean element profile classification method based on clustering analysis algorithm, belong to marine data processing technical field, for ocean element profile classification, including obtaining original ocean element profile sample and pre-processing and standardization;Covariance matrix is constructed to standardization profile, and principal component is selected using cumulative variance contribution rate, and feature vector is constructed;Expectation maximization algorithm is used to estimate Gaussian mixture model parameter;Classification result is verified based on bayesian information criterion and physical interpretability;Ocean structure partition or water mass identification is carried out based on classification result.The application combines principal component analysis with Gaussian mixture model, significantly reduces the calculation complexity while retaining the main vertical variation information of the profile, does not need artificial preset label, has strong adaptability, interpretability and generalizability, can objectively and efficiently complete ocean element profile automatic classification, and is suitable for water mass identification, acoustic environment assessment and other various marine applications.

The invention discloses a biomolecule relation modeling method based on semantic consistency hypergraph contrast learning, and aims to model a complex relation between biomolecules by effectively combining a local structure and global semantic information. According to the method, firstly, local representation of nodes is enriched by expanding a message aggregation mechanism guided by a subgraph, and global context information is combined to improve the modeling precision of the biomolecular relationship; secondly, performing hypergraph reconstruction by adopting a double-layer consistency mechanism, and maintaining the consistency of a local structure and semantics at a node level and a hyperedge level through semantics and structure constraints; besides, soft clustering is carried out by using a Gaussian mixture model (GMM) and a Bayesian information criterion (BIC) strategy, hyperedge selection is further optimized, and structural consistency in a reconstruction process is ensured. In order to ensure the consistency of local and global semantics, the method introduces a multi-granularity comparison target in a comparison learning framework, performs training at node, hyperedge and extended subgraph levels, feeds back high-level semantic information to low-level semantic information through a cross-layer feedback mechanism, stabilizes the model and enhances semantic alignment between representations. Experimental results show that the provided framework is excellent in performance in node classification and clustering tasks on multiple reference data sets, and compared with an existing method, local and global semantic relationships can be better captured and aligned, and the hypergraph learning effect is remarkably improved.

The invention discloses an automatic division method and system for a satellite constellation orbital plane, and relates to the technical field of spaceflight measurement and control. The method comprises the following steps: firstly, calculating an orbital plane unit normal vector based on a satellite orbit inclination angle and an ascending node right ascension, and performing first-level spatial pointing clustering by adopting a density clustering algorithm to form an orbital plane group; secondly, in the same orbital plane group, by taking a semi-major axis and an eccentricity rate as characteristics, automatically determining an optimal shell layer number by utilizing a Gaussian mixture model and a Bayesian information criterion, and realizing second-stage orbital energy shell layer division; and finally, in the same shell layer, sorting and interval statistical analysis are performed on the right ascension of the ascending node, recursive subdivision is performed through dynamic threshold segmentation in combination with a range-based dual-condition verification mechanism, and fine division of a third-level orbital plane is completed. The method realizes full-process automation from orbit data to division results, has the advantages of high division precision, strong adaptability, good engineering practicability and the like, and is suitable for operation and maintenance management of large-scale heterogeneous satellite constellations.

The invention discloses an eVTOL collaborative design optimization method based on improved genetic programming, and the method comprises the steps: firstly constructing pneumatic, propulsion and dynamics high-fidelity models of an eVTOL takeoff stage, building an original collaborative optimization problem, and solving the problem through a direct transcription method; secondly, multi-target genetic programming fusing a root mean square error, an akaike information criterion, a predicted residual sum of squares and a Bayesian information criterion is creatively adopted, and an agent model for calculating thrust is constructed to replace a high-fidelity thrust model to solve and obtain optimal design parameters. The error between the optimal solution obtained by the method and the optimal solution obtained by using the high-fidelity model is very small, the calculation cost is remarkably reduced, and multidisciplinary efficient collaborative optimization is realized.

The present invention discloses a method for estimating a time-varying underwater acoustic channel. The method comprises the following steps: (1) using the Fourier transform method to perform time offset compensation on real underwater acoustic data; (2) performing Kronecker product structural decomposition on the real underwater acoustic channel and selecting structural parameters using the Bayesian Information Criterion; and (3) using the sparse enhanced conjugate gradient method based on known structural parameters to online estimate the time-varying channel impulse response and output a multi-pulse underwater acoustic channel estimation result. The method has the advantages of utilizing the potential structured sparsity characteristics of the real underwater acoustic channel to reduce the channel impulse response parameter estimation dimension, effectively suppressing false estimation under low signal-to-noise ratio conditions, and utilizing the adaptive recursive concept to provide good online continuous estimation capability for multi-pulse time-varying underwater acoustic channels and strong noise interference resistance.

A method for estimating the subspace dimension of a random signal belongs to the technical field of array signal processing. The present invention solves the problem of poor robustness of existing methods under conditions of low signal-to-noise ratio and low array received data volume. The method of the present invention calculates the covariance matrix based on the array received data, performs eigenvalue decomposition on the covariance matrix to solve the eigenvalues, derives the probability density of the array received data and the eigenvalues respectively, and then constructs the Bayesian information criterion expression. After calculating the information value corresponding to all possible subspace dimension values, the subspace dimension value corresponding to the maximum information value is used as the signal system space dimension. The method of the present invention can be applied to the technical field of array signal processing.

The invention provides a grain safety toughness influence factor analysis method based on a Bayesian network. Relates to the field of food safety. The method comprises the following steps: screening grain safety toughness influence factors through a Spearman level correlation coefficient method, an adjusted R2 maximum criterion optimal subset and a random forest feature importance evaluation method, and determining indexes through a union set method; a natural breakpoint method is utilized, and quantitative indexes are converted into high, medium and low qualitative indexes by minimizing intra-group variances and maximizing inter-group variances; taking the grain safety toughness as a dependent variable and the final index as an independent variable, constructing a three-layer Bayesian network by using a simulated annealing algorithm, and evaluating and iterating through a Bayesian information criterion scoring function to obtain an optimal structure; on the basis, a reverse reasoning method is used, the probability that the grain safety toughness is high is set, the reverse reasoning probability and change of each index are calculated, and the influence effect is analyzed. Various factors are incorporated, and the influence effect of each factor can be systematically and comprehensively analyzed.

The present application relates to the technical field of image data processing, and more particularly to a concrete vibration quality evaluation method and system based on image analysis. The method comprises: acquiring a gray image of the surface of the newly vibrated concrete; constructing a region adjacency graph and extracting multi-dimensional features; calculating physical entropy weight based on the multi-dimensional features; performing region merging operation; identifying bubble clusters and calculating the quality evaluation index of the concrete surface according to the statistical characteristics of the bubble clusters; and outputting the concrete vibration quality evaluation result according to the quality evaluation index. The present application divides the concrete surface by the watershed algorithm and region adjacency graph, extracts multi-dimensional features, constructs a physical entropy weight model to evaluate the saliency, uses a boundary gray level priority queue and an adaptive multi-threshold intelligent region merging method to solve the problem of over-segmentation, and combines the Gaussian mixture model and the Bayesian information criterion to identify bubble clusters, so as to comprehensively evaluate the concrete vibration quality and improve the objectivity and accuracy.

The application relates to the technical field of non-ferrous metal ore dressing intelligent control, and discloses a lead-zinc ore flotation process state evolution modeling method. The method comprises the following steps: collecting multi-source heterogeneous data of a flotation process, obtaining a plurality of key features after pre-processing the multi-source heterogeneous data, and constructing a feature matrix according to the plurality of key features; constructing a hidden Markov model based on the implied state of the flotation process, inputting the feature matrix into the hidden Markov model, optimizing the parameters of the hidden Markov model through a Baum-Welch algorithm, and screening an evolution model in the training process in combination with a Bayesian information criterion; obtaining real-time multi-source heterogeneous data of the flotation process and obtaining a corresponding feature matrix, inputting the feature matrix into the evolution model, and obtaining an optimal process state sequence in combination with a Viterbi algorithm. The method solves the problem that existing ore dressing models cannot effectively and accurately predict the state of the flotation process.

The invention belongs to the technical field of intelligent load sensing of energy routers, and particularly relates to a non-intrusive load port monitoring method and system for an energy router. The monitoring method provided by the invention comprises the steps of building a BIC-FHMM model, carrying out load decomposition based on the BIC-FHMM and verifying scheme feasibility. The invention provides a Bayesian information criterion method and a Bayesian criterion-based adaptive load state quantity technology, the load state quantity can be automatically determined in an FHMM model training process, the load state does not need to be manually defined, and thus the accuracy of load decomposition is improved; according to the method, the validity of the algorithm is successfully verified by using the British dataset UK-DALE; according to the non-intrusive load monitoring device for the energy router, serial communication with the PC can be carried out by utilizing self-test load power data, and NILM is carried out through a python language, so that the feasibility of a scheme and the effectiveness of an algorithm are verified.

The present invention discloses a weather classification method based on a hybrid neural network model after feature selection and clustering analysis. First, the maximum information coefficient (MIC) is used to select features, and features with an MIC lower than 0.5, that is, features with low correlation with other features, are removed. Then, the Bayesian information criterion (BIC) coefficient is used to estimate the optimal number of clusters k of the clustering model, and the Gaussian mixture model clustering is performed on the data set to screen similar samples. Next, an MLP classification neural network and an MLNN classification neural network are respectively constructed. Finally, the AdaBoost adaptive boosting algorithm in ensemble learning is used to sequentially train the two models and generate a hybrid model of MLP and MLNN neural networks. The present invention combines feature selection, clustering analysis and hybrid neural networks to process weather data, reducing the training time. By training a multi-layer perceptron and a morphological-linear neural network and incorporating the idea of ensemble learning, and applying the AdaBoost adaptive boosting algorithm during model training and hybridization, the accuracy of weather classification is improved.