472 results about "Probability model" patented technology

Even when there is little measurement data of abnormal signals, it is an object to manage the abnormality of the signals. 【Means for solving the problem】 The abnormality management device 1 uses, as teacher data, normal data indicating a normal frequency spectrum among the frequency spectra of a plurality of signals, for each frequency component intensity, and the parameter of a probability model that outputs the posterior probability that each frequency component intensity is normal is learned by maximum likelihood estimation. And a derivation unit 12 configured to derive a probability distribution of abnormal data indicating a frequency spectrum including a frequency component with an abnormal intensity, based on the posterior probability estimated by the learned probability model, the probability distribution of the normal data, and the prior probability of being normal.

The invention discloses an available energy estimation method and system for a prefabricated cabin type energy storage system, and belongs to the field of energy storage systems.The method comprises the steps that probability distribution of environmental parameters is constructed; historical operation data of equipment is collected, a probability relation between the environment temperature and equipment energy consumption is established, and an equipment energy consumption confidence interval is generated according to the probability relation; determining a confidence interval of the SOH by combining the battery aging experiment data and the real-time monitoring error; inputting the probability distribution of the environmental parameters, the equipment energy consumption confidence interval and the battery health state confidence interval into a probability model to generate a probability interval of available energy; and the optimal confidence level is selected in combination with the power grid dispatching requirement and the risk tolerance, and the available energy of the prefabricated cabin type energy storage system is obtained on the basis of the optimal confidence level according to the probability interval of the available energy. According to the method, confidence analysis is introduced, environmental parameters, equipment energy consumption and uncertainty of battery attenuation are quantified through a probability model, a probability interval of available energy is generated, and a basis is provided for scheduling decision making.

The invention discloses a distributed photovoltaic power prediction method, system and device based on a Gaussian mixture model and a medium, and belongs to the technical field of photovoltaic power prediction.The distributed photovoltaic power prediction method comprises the steps that a time sequence vector is collected, principal component analysis is carried out on the time sequence vector, low-dimensional feature representation is obtained, and a power feature vector of each photovoltaic power station is formed; performing clustering analysis based on a probability model on the power feature vector to generate a plurality of photovoltaic power station clusters; for each cluster, acquiring meteorological input data through a set data source priority rule and a completion mechanism; and constructing a neural network power prediction model based on the accumulated power data in the cluster and the corresponding meteorological features, and outputting a future power generation power prediction value of the photovoltaic power station in the corresponding cluster. According to the invention, N photovoltaic power stations in a region are divided into M clusters through a GMM clustering method, so that the design is simplified; and the power prediction of the whole area is realized.

It aims to appropriately manage the bias of the signal processing amount. 【Solution means】 The abnormality management device 1 uses, as teacher data, normal data indicating the number of normal packets distributed to each device for each time zone, and learns, by maximum likelihood estimation, the parameters of a probability model that outputs the posterior probability that the number of packets distributed to each device for each time zone is normal. A learning unit 11, the posterior probability for each device for each time zone estimated by the learned probability model, the probability distribution of the normal data for each device estimated based on the normal data, and the prior probability of being normal. A derivation unit 12 that derives the probability distribution of abnormal data indicating the number of abnormal packets distributed to each device in the time zone, and the probability distribution of the normal data and the probability distribution of the abnormal data for each device for each time zone are equal. And a setting unit 13 that sets, as a threshold value for abnormality determination of the number of packets distributed to each device, the number of packets corresponding to the value of the probability distribution that becomes the value.

The invention relates to the technical field of electronic component detection, in particular to an IC (integrated circuit) carrier plate detection method based on surface state image extraction, which comprises the following steps: acquiring a gray image, analyzing structural parameters, extracting gradient features, detecting boundary disturbance, integrating the image, calculating an abnormal score, identifying a defect position area, extracting features and outputting an identification result. According to the invention, by analyzing the structure parameters of the bonding pad in the gray level image, calculating the edge line segment, the center coordinate and the spacing, and constructing the two-dimensional coordinate system, the regional positioning reference is enabled to have geometric consistency, the coordinate mapping is combined with the gradient direction change frequency and the continuous aggregation point, the boundary disturbance identification precision is improved, and the image division is executed based on the disturbance region. According to the method, non-functional region mixing is effectively avoided, a clustering and probability model is introduced after region gray level statistics, a deviation scoring mechanism is constructed, gray level feature abnormity is accurately recognized, the discrimination capability of small-amplitude and low-contrast defects is improved, and the selectivity and target focusing performance of feature detection are enhanced.

The invention discloses a tunnel rockburst type prediction method and system based on multi-dimensional mechanism fusion and a medium. Relates to the technical field of tunnel rockburst. Dynamically acquiring multi-dimensional basic information in a tunnel construction process; inverting first proportions of different fracture modes of the micro-seismic event according to the micro-seismic information fusion moment tensor, and performing energy evolution on the acoustic emission information to obtain second proportions of different fracture modes of each acoustic emission event; the dynamic failure weights of different fracture modes are comprehensively obtained; spatial clustering analysis is carried out based on the dynamic damage weight and the multi-dimensional basic information, rockburst type probability models of different rockburst types are constructed, and the probabilities of different rockburst types are predicted; according to the scheme, the rockburst type probability model is constructed in combination with real-time multi-dimensional basic information, dynamic judgment and prediction of rockburst types are achieved, and technical guarantee is provided for deep tunnel construction safety.

It aims to more reliably manage abnormal communication. 【Solution means】 The abnormality management device 1 is configured such that a first learning unit 12 treats each observed value of a normal presence time series as discrete values independent of each other, and estimates probability parameters of a probability model representing the normal presence time after quantization based on the appearance frequency of each observed value; and a second learning unit 13 configured to fix the generator parameters of a generator 131 that generates pseudo-normal data sufficiently deviated from the distribution of true normal data, with each observed value of the normal presence time series being true normal data, and update only the discriminator parameters of a discriminator 132 that discriminates between true normal data and pseudo-normal data in a direction to maximize the discrimination accuracy, based on the probability parameters of the normal presence time estimated by the first learning unit 12.

The purpose is to easily manage abnormal program operation. [Solution] The abnormality management device 1 includes a first learning unit 11 that uses normal data indicating processing resource usage data in which the processing resource usage corresponding to each process ID is normal as training data from among a plurality of processing resource usage data each including the usage of processing resources used in executing a process corresponding to each process ID, and learns parameters of a probability model that outputs a posterior probability that the processing resource usage corresponding to each process ID is normal by maximum likelihood estimation, and a derivation unit 12 that derives a probability distribution of abnormal data indicating processing resource usage data including abnormal processing resource usage based on the posterior probability estimated by the learned probability model, the probability distribution of the normal data, and the prior probability of normality.

The invention relates to the technical field of hydrological data processing, in particular to a water conservancy intelligent internet-of-things sensing method and system, and the method comprises the following steps: combining statistical data into a feature set based on hydrological monitoring data including flow, water level and rainfall information, judging whether the data accords with normal distribution, and generating a data distribution feature vector set. According to the method, the change trend of the data is captured through multi-level data distribution characteristic analysis, and the monitoring precision of the data exception is enhanced. In the prediction process, new data and weight adjustment are gradually fused, so that the model can adapt to environmental changes in real time, and the hydrological state in a future short period is accurately predicted. Besides, abnormal data can be identified in real time based on analysis of quantiles, potential problems can be quickly found, and state division of different levels can be carried out according to abnormal degrees. In sampling scheduling, the priority of data acquisition is adjusted according to a probability model of state transition, so that monitoring resources are efficiently configured.

The invention relates to the technical field of data processing, in particular to a cloud database data compression method based on an adaptive quantization algorithm. The method comprises the following steps of: 1, regarding to-be-compressed original data in a cloud database as a matrix, dividing the data into a plurality of non-overlapped data blocks according to predefined statistical characteristics, and calculating a mean value and a variance of each data block; 2, defining the complexity of each data block, and calculating the self-adaptive quantization step length of each data block according to the mean value and variance of each data block; step 3, performing non-uniform adaptive quantization on each data item in each data block to obtain a quantized value; and 4, establishing a probability model for a quantization result of each data block, calculating a coding length, obtaining a compression ratio, and carrying out entropy coding. The quantization step size of the data block is dynamically adjusted to be matched with the local characteristic of the data, so that the quantization error is reduced, and the data recovery precision is improved.

The invention discloses an optimization system and method for participation of an electric vehicle cluster in an electric power standby market, and is applied to the field of an electric power auxiliary service market, and the system comprises a market uncertainty modeling module which is used for constructing a joint uncertainty model of market price and standby calling probability, wherein the model comprises a probability distribution model and a conditional probability model; the risk quantitative evaluation module is used for evaluating risk exposure degrees of different decision-making schemes, including establishing a multi-period risk accumulation model, and comprehensively evaluating balance indexes of expected income, risk openness and opportunity cost; the robust optimization decision-making module constructs a decision-making model according to the market uncertainty modeling module and the risk quantitative evaluation module, and generates an optimal strategy that the electric vehicle cluster participates in the standby market; the adaptive learning optimization module is used for optimizing the decision model through reinforcement learning; according to the invention, risk-controllable revenue maximization can be realized, so that the electric vehicle cluster can efficiently participate in the electric power standby market.

The invention discloses an automatic control gradient LED lamp energy-saving illumination method, which comprises the following steps: S1, asynchronous sampling is carried out through an illumination sensor A and an illumination sensor B, ambient light change data is collected, a sampling timestamp is recorded, a time difference weight is obtained based on the timestamp, data fusion is carried out through the time difference weight, and ambient light intensity data is obtained; s2, performing data calibration on the ambient light intensity data through a state space model, and performing optimization through a weighted moving average method to obtain an ambient light intensity change rate; and S3, acquiring a light change condition based on an environment illumination intensity change rate, adjusting a brightness duty ratio through a probability model based on Bayesian reasoning to output an initial control variable, performing optimization based on the initial control variable in combination with a particle filtering algorithm to obtain a final control scheme, and performing optimization through the probability model and the algorithm to obtain a final control scheme. A relatively accurate brightness adjustment scheme can still be given, and the intelligent level of lamp control is improved.

The invention relates to the technical field of data processing, and provides a missing data interpolation method and system based on a generative adversarial network. The method comprises the following steps: clustering a missing data matrix to obtain a clustering cluster containing a cluster label; based on the clustering cluster, performing classification prediction on a data feature vector corresponding to the missing data matrix through a logistic regression algorithm to obtain a cluster label prediction model; performing probability distribution modeling on the data feature vector through a Gaussian mixture model to obtain a probability model; training a generative adversarial network framework based on the cluster label prediction model and the probability model to obtain an interpolation model; and interpolating data to be interpolated through the interpolation model to obtain an interpolation data matrix. According to the invention, the interpolation precision and stability of nonlinear data are improved.

The invention discloses a statistical eye pattern generation method and device and a storage medium. The method comprises the following steps: inputting a step signal at a transmitting end of a channel, and determining a step response function based on the step signal; introducing a dithering signal at the transmitting end, and determining a dithering probability model based on the dithering signal; based on a jitter probability model, determining a jitter value set and a jitter probability set; determining a state transition probability matrix between the first moment and the second moment in the input step signal; based on the state transition probability matrix and the step response function, establishing an equivalent response distribution kernel for the target sampling moment; performing convolution operation on the voltage probability distribution at the previous sampling moment of the target sampling moment and the equivalent response distribution kernel, determining the voltage probability distribution at the target sampling moment, and determining the voltage probability distribution at multiple sampling moments; and generating a statistical eye pattern based on the voltage probability distribution at the plurality of sampling moments. The method has the technical effect of assisting in improving the credibility of simulation evaluation.

The invention discloses a nuclear magnetic resonance free induction decay signal reconstruction method based on a diffusion probability model, and relates to the field of magnetic resonance spectrums.The method comprises the steps that a simulation data set is constructed based on a mathematical model of free induction decay signals, and the simulation data set is divided into a training set, a verification set and a test set; sampling each data set by adopting a Poisson gap non-uniform sampling strategy, and simulating an actual under-sampling condition; constructing a diffusion probability model, gradually adding Gaussian noise through a forward diffusion module to generate a noise-containing sample, and recovering an original signal from a noise signal through a reverse denoising module; and inputting test data into the optimal model for multiple times to obtain a plurality of reconstruction results. According to the method, the reconstruction method based on the diffusion probability model is constructed, training optimization is carried out through Poisson gap non-uniform sampling and simulation signal data sets, and high-precision reconstruction of the under-sampling nuclear magnetic resonance free induction decay signals is achieved.

The invention discloses a human body behavior prediction method and system, and the method comprises the steps: carrying out the skeleton point sequence extraction of a real-time behavior video image of a target person, and obtaining a joint point coordinate set and a skeleton motion sequence; determining a spatio-temporal feature sequence of the joint point coordinate set, and determining a skeleton motion sequence to perform spatio-temporal attention coding to obtain global spatio-temporal dynamic features; fusing the action probability distributions corresponding to the spatial-temporal feature sequence and the global spatial-temporal dynamic features to obtain short-time action probability distribution data; candidate action screening is carried out on the short-time action probability distribution data, and candidate action comprehensive features are obtained; performing feature coding on the historical action sequence of the target person to obtain a historical context vector, and splicing the historical context vector with the candidate action comprehensive features to obtain a fusion feature; and inputting the fusion features into a probability model for intention probability evaluation to obtain a behavior prediction result of the target person. According to the method, the accuracy of behavior prediction is improved.

The invention discloses a unified representation and dynamic knowledge reasoning method and device for an intelligent cluster system, and the method comprises the steps: determining a decyclization task sub-graph based on a composite task state space model of the intelligent cluster system through the priori information; constructing a real-time data sub-graph by using a real-time data feedback mechanism, and constructing a traceability mode sub-graph based on the decyclization task sub-graph; candidate traceability sub-graphs with the highest matching similarity with the traceability mode sub-graphs are determined from the real-time data sub-graphs to serve as traceability sub-graphs; constructing a reasoning path set based on the decyclization task subgraph, and calculating a probability value of each reasoning path in the reasoning path set by using a path probability model constructed based on a Bayesian conditional probability chain rule; and screening out the reasoning paths of which the probability values are smaller than a preset probability value in the reasoning path set by utilizing a dual path pruning strategy, and adding the reasoning paths containing abnormal nodes in the reasoning path set into an alarm path set so as to provide support for state prediction and risk management of the intelligent cluster system.

Methods and systems use an additional determination to generate predictions of future data processing load predictions. Specifically, the methods and systems generate additional predictions based on other data sets and data calculation methodologies (e.g., data with non-homogeneous time dependencies). The methods and systems then use these additional predictions to determine whether or not the prediction of the real-time average of data processing loads is an outlier. Thus, the methods and systems generate a time-specific data processing prediction.

The invention relates to the technical field of intelligent security and protection, in particular to a public area video monitoring system and method, and the system comprises an intelligent collection and recognition module, a feature fusion module, a heterogeneous trajectory prediction module, a dynamic risk field module, a hierarchical response module and a collaborative warning module. The feature fusion module integrates video data and real-time weather and traffic flow information, and constructs a multi-dimensional environment state vector, the heterogeneous trajectory prediction module adopts Social-LSTM, TransMotion and random field probability models to process motion characteristic differences of three types of targets respectively, the prediction precision is significantly improved, and the dynamic risk field module calculates the trajectory intersection probability, so that the prediction efficiency is improved. According to the method, the environmental dynamic factors and the object danger level are weighted and integrated, dynamic quantification of space-time risks is achieved, the problem of prediction errors caused by motion characteristic differences is solved, the defect of risk misjudgment influenced by environmental factor changes is overcome, and accuracy and practicability are higher.

The invention relates to the technical field of vehicle intelligent management, and discloses a vehicle state monitoring method and system based on deep learning, and the method comprises the steps: analyzing the physical dependence relation among vibration abnormality detection, loosening positioning and fatigue evaluation monitoring tasks based on structural mechanics constraints, and carrying out the monitoring of the vibration abnormality detection, the loosening positioning and the fatigue evaluation; generating a task dependence directed graph; inputting the multi-modal monitoring data matrix into a shared encoder of a multi-task learning network to generate a unified representation vector; respectively inputting the unified representation vector into a vibration anomaly detection decoder, a loose positioning decoder and a fatigue evaluation decoder, and outputting an initial prediction result of each task; and establishing a conditional probability model between task outputs based on the constraint relationship defined by the task dependent directed graph, and generating a consistency diagnosis result meeting physical constraints through maximum posteriori estimation. The technical problems of one-sided diagnosis results, mutual contradiction and low calculation efficiency are solved.

The invention discloses a few-photon non-vision field imaging method based on adaptive windowing, and the method comprises the steps: collecting photon signals reflected from a hidden target through a non-coaxial two-dimensional scanning system, building a probability model of signal photons and noise photons, and analyzing the distribution characteristics of the signal photons and the noise photons in a time domain; the method comprises the following steps: combining adjacent pixels into pixel blocks by using space-time correlation, determining the adaptive window width of each pixel block through a matched filtering method, and separating signal photons from noise photons by applying windowing operation on a time domain; and filling the windowed transient data by using full-variable regularization constraint, and solving a reconstruction problem by using an alternating direction multiplier method to obtain a target reconstruction result. According to the method, the detection efficiency of sparse photons is remarkably improved through the space-time related pixel blocks and the matched filtering technology, and target reconstruction under the extremely low photon level is achieved.

The invention provides a multi-granularity text pedestrian search method based on attribute learning and feature screening, and the method comprises the steps: carrying out the coarse-granularity feature learning of an image-text pair, filtering noise data through a probability model, and optimizing the cross-modal alignment through a self-supervision cost function, so as to obtain a global feature; tokens of the image and the text are screened based on the self-attention weight output by the encoder, and medium-granularity feature representation is generated; performing random masking on pedestrian attribute vocabularies conforming to adjective-noun combination in the text, fusing features and predicting masked vocabularies through a multi-modal interaction encoder, and realizing fine-grained attribute alignment; and combining the coarse-grained loss, the medium-grained loss and the fine-grained loss, iteratively training the model, storing an optimal model according to verification accuracy, and outputting a pedestrian search result.

The invention discloses a Lidar point cloud registration optimization method based on geometric feature reliability, and the method comprises the steps: carrying out the preprocessing and initial alignment of a laser radar scanning point cloud, extracting the features of a maximum principal curvature and a minimum principal curvature based on the local curvature of the point cloud, dividing a point region into two types of geometric features of angular points and plane points according to the threshold value of the maximum principal curvature, and carrying out the registration of the angular points and the plane points. On the basis, fitting quality factors including fitting errors, local curvatures and spectral entropies of the linear features and the plane features are calculated respectively, then the three factors are fused into a unified reliability weight based on a Bayesian probability model, and finally the feature reliability weight is introduced into an optimization objective function of iterative nearest point registration to execute weighted ICP registration. And outputting positioning and attitude determination results. According to the method, the reliability of geometric features is quantitatively evaluated, and a weighted optimization framework is constructed, so that the point cloud registration precision and robustness are remarkably improved, and the problem that a traditional ICP algorithm is sensitive to unreliable features in feature degradation or high-dynamic scenes is effectively solved.

The invention discloses a transformer substation fault detection method and related equipment, and the method comprises the steps: obtaining the real-time data of a transformer substation secondary system, forming a scene prediction model through a deep learning model and a dynamic Bayesian network, and determining a fault detection result in combination with a digital twin system. The deep learning model can deeply mine multi-source real-time data features, accurately recognize fault information in a complex scene, make up for the problem of insufficient information mining of a traditional data driving method, and improve the accuracy of fault detection and positioning. The dynamic Bayesian network carries out probability modeling on the dynamic change of the system state, and the problems that a traditional method is difficult to adapt to complex working conditions and cannot accurately simulate system evolution are solved. Meanwhile, the digital twin system realizes deep fusion analysis of multi-source data, constructs an intelligent processing closed loop from data acquisition to maintenance decision, remarkably improves the timeliness, accuracy and intelligent level of fault detection, and provides reliable guarantee for stable operation of a transformer substation.

The invention discloses an underwater three-dimensional reconstruction and sonar pose joint optimization method based on a multidirectional sonar, and belongs to the technical field of image reconstruction, and the method comprises the steps: obtaining and preprocessing multi-modal sensor data, and converting coordinates into a world coordinate system; dead reckoning is carried out through the IMU and the DVL to obtain a rough pose; constructing and optimizing an implicit occupancy probability model; screening effective frame pairs; calculating point cloud loss; a total loss function is constructed, the pose and the model parameters are jointly optimized, and a reconstruction result and an optimization track are output; according to the underwater three-dimensional reconstruction and sonar pose joint optimization method based on the multidirectional sonar, the reconstruction problem caused by lack of features and unstable sensor data in a turbid water area is effectively solved.

The invention discloses a large-scale MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) joint active user detection and channel estimation method. The invention provides a probability model for modeling a user beam domain channel based on channel support and channel values for a large-scale MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) unlicensed random access system. Under the framework of the Bethe free energy theory, the problem of large-scale MIMO-OFDM joint active user detection and channel estimation is converted into the problem of Bethe free energy minimization. The Bethe free energy minimization problem is solved through a Lagrange multiplier method to obtain a related mixed message passing algorithm, and joint active user detection and channel estimation of the large-scale MIMO-OFDM unlicensed random access system are realized by using the algorithm. According to the method, the accuracy of joint active user detection and channel estimation of the large-scale MIMO-OFDM authorization-free random access system can be effectively improved.

A method performed by at least one processor of a video decoder includes receiving a coded video bitstream including at least one picture and one or more syntax elements encoded in accordance with multi-hypothesis arithmetic coding. The method further includes decoding each syntax element from the one or more syntax elements based on the multi-hypothesis arithmetic coding. The method further includes selecting a probability update rate from a plurality of probability update rates based on a predetermined condition, the plurality of probability update rates including a first probability update rate that is higher than a second probability update rate. The method further includes updating at least one probability model utilized in the multi-hypothesis arithmetic coding based on the selected probability update rate. The method further includes decoding at least one block in the at least one picture based on the decoded one or more syntax elements.

The invention provides a steel plate surface defect image generation method based on a de-noising diffusion probability model and a visual Transform, relates to the technical field of industrial surface defect detection, and aims to solve the problems of low quality of generated samples, lack of detail richness and insufficient global dependency relationship modeling capability in the conventional defect image generation method. The performance improvement of a defect detection algorithm is limited; according to the generation method provided by the invention, by introducing the ViT model, the capability of capturing the global dependency relationship in the image by the model is remarkably enhanced, the limitation of the traditional DDPM in modeling the global structure of the image is overcome, the diversity and quality of the generated image are effectively improved, and the method is suitable for further improving the modeling precision of a complex defect structure. A channel-space self-attention module is designed, and the CSSAM of the module can dynamically adjust the attention between different space regions and channels, so that the details of complex defects are finely simulated, and the authenticity and high fidelity of a generated image on the aspects of a global structure and local textures are ensured.

A power consumption prediction and management method and system based on a solid state disk, a medium and a product relate to the field of solid state disks. The method comprises the following steps: predicting the workload intensity of a host in a first preset time window based on an I / O (Input / Output) instruction stream to obtain an expected host workload curve, and determining an expected host idle time period; calculating an expected host power consumption value in the expected host idle time period; extracting internal state parameters of the solid state disk according to the historical operation log data; establishing a background task triggering probability model, and calculating the triggering probability of starting the high-power-consumption background task by the solid state disk in a first preset time window; performing power consumption conflict judgment according to the expected host power consumption value and the triggering probability, and generating a comprehensive predicted power consumption value; and based on the comprehensive predicted power consumption value, determining an optimal target power consumption state, and instructing the solid state disk to be switched to the optimal target power consumption state. By implementing the technical scheme provided by the invention, the stability of state switching of the solid state disk is improved.

The invention provides a big data compliance risk control method and system, and relates to the technical field of big data processing and risk control. By obtaining the original business data and adding the metadata information for packaging, standardization and traceability of the data source are realized. Then, the metadata information is utilized to carry out preliminary matching in the data mode dictionary, and the potential mode of the data can be quickly identified. The data structure and the field type are inferred through the probability model, and the inference confidence is output, so that the problems of analysis errors and label noise caused by nonstandard data format and update lag of the analyzer in the prior art are effectively solved.