65 results about "Probabilistic forecasting" patented technology

The invention belongs to the technical field of electric power, and provides a user load increase and probability boundary prediction method, which comprises the following steps of: firstly, realizing unified projection and deep fusion of multi-source heterogeneous data by adopting a topological adaptive space-time fusion mechanism; secondly, physical constraint comparative learning and hypergraph clustering are utilized to generate a load-sensitive topology mode according with a power grid rule; a CPT-Mama model is constructed by injecting physical rules, meteorological codes and lightweight NTK features, and spatial correction is performed by means of a graph attention network to improve the precision of load increment prediction, so that efficient prediction of the multi-modal load increment is realized; and finally, innovatively introducing a net rack spectral domain built-in operator, reconstructing a residual error and quantifying uncertainty, and accurately describing a probability boundary of load increase. According to the method, the medium and long term load prediction precision is improved, the crossing from point prediction to probability prediction is realized, and reliable technical support is provided for power grid dispatching, risk prevention and control and planning decision.

Computer-implemented methods optimising routes in a transport network for a geographical region are disclosed. Methods comprise obtaining an analysis of the geographical region; generating probabilistic predictions, checking whether at least a subset of the generated probabilistic predictions have previously been solved by a trained machine learning model; based on at least a first subset of the generated probabilistic predictions having previously been solved, retrieving the trained machine learning model previously used to solve the at least a first subset of the generated probabilistic predictions and executing the trained machine learning model to determine a first plurality of suggested routes for the transport network; based on at least a second subset of the generated probabilistic predictions not having previously been solved, training a new machine learning model to solve the at least a second subset of generated probabilistic predictions to determine a second plurality of suggested routes for the transport network.

The invention discloses a marine traffic flow probability prediction method based on multi-source AIS-meteorological data fusion, and belongs to the field of marine traffic safety management, and the method comprises the following steps: S1, obtaining a standardized multi-source data set; s2, fusing the dynamic statistical characteristics of the AIS ship and the meteorological environment characteristics through a ship motion transfer function, and constructing a space-time patch considering a short-term ship motion trend and a long-term sea condition trend; s3, based on the space-time patch, fusing patch Transform time sequence modeling and adaptive dynamic graph space correlation learning to obtain a space-time fusion feature; and S4, outputting the traffic flow state prediction value and the probability distribution of the traffic flow key indexes. The marine traffic flow probability prediction method based on multi-source AIS-meteorological data fusion breaks through the limitation that a traditional method is single in data dimension, insufficient in space-time correlation capture and lack of probability confidence, and high-precision probability prediction of the marine traffic flow under the complex sea condition is achieved.

The invention discloses an odor traceability management and control method based on multi-source data fusion and related equipment. The method comprises the following steps: acquiring weather forecast and monitoring data of a target area, preset emission intensity of a pollution source and real-time odor concentration; inputting the real-time monitoring data and the emission intensity into an odor complaint probability prediction model to obtain a theoretical risk value, and calculating a real-time deviation factor between the theoretical risk value and an actual risk value obtained based on the real-time concentration; obtaining a first complaint risk probability value of a future time period by using weather forecast data and emission intensity, and performing weighted correction on the first complaint risk probability value based on the real-time deviation factor to obtain a second complaint risk probability value; and when the second complaint risk probability value exceeds a risk threshold value, reversely calculating target emission intensity capable of enabling the risk to reach the standard according to the model and the real-time deviation factor, and generating a production working condition adjustment instruction according to the target emission intensity. According to the method, the problem that a static prediction model cannot accurately describe a complex and changeable environment, so that mistaken report and missing report of stink traceability management and control are frequent is solved.

The application discloses a flood forecasting method fusing a physical mechanism and a conditional denoising diffusion model, and comprises the following steps: obtaining multi-source hydrological data of a basin to be forecasted; driving a preset hydrological model to execute multi-stage sequential parameter optimization to obtain a basic forecasting sequence and an initial residual sequence; extracting multi-dimensional conditional variables based on the hydrological data and the initial residual sequence, and using a multi-dimensional evaluation criterion to perform dimension reduction screening to obtain a fusion conditional vector; inputting the fusion conditional vector as a guide condition into a pre-trained conditional denoising diffusion model, performing iterative denoising sampling through a multi-scale denoising operator to generate a forecasting residual probability sequence set; and coupling the basic forecasting sequence and the forecasting residual probability sequence set to obtain deterministic forecasting and probabilistic forecasting results of the basin to be forecasted. The application realizes a leap from single-point forecasting to probabilistic distribution deduction, can accurately obtain a flood deviation boundary under a changing hydrological environmental condition, and enhances the reliability of disaster prevention and mitigation decision-making.

To provide a congestion probability prediction device capable of facilitating power generation control of power generation equipment with non-firm type connection while especially considering the uncertainties.SOLUTION: A disclosed congestion probability prediction device for predicting the congestion of a power system includes: a weather forecast data input unit that acquires weather forecast data regarding the weather a predetermined number of days after the current date; a power system data input unit that acquires demand data regarding electricity demand a specified number of days after the current date and that acquires system data regarding the power system after the predetermined number of days after the current date; a system congestion probability prediction unit that predicts congestion points and congestion probability data of the power system the specified number of days after the current date.SELECTED DRAWING: Figure 1

The application discloses a photovoltaic ultra-short-term probability prediction method based on data decomposition and reconstruction, solves the problems of low prediction accuracy, weak generalization ability, poor prediction interval calibration, insufficient stability across data sets and low calculation efficiency of existing methods, and cannot meet the 5-minute ultra-short-term photovoltaic probability prediction engineering requirements. The method first collects and pre-processes photovoltaic power and meteorological data, uses the ACEEMDAN algorithm to decompose the photovoltaic power into fixed high-frequency and low-frequency dual components through four stages and regularizes; then the corresponding features of the three types of data are extracted through CNN, iTransformer and BiLSTM respectively, and the unified feature representation is obtained after the multi-head attention mechanism fusion, and then the EQN network is input; finally, the improved EQN loss function containing the width penalty term is used to train the model to convergence, and the probability prediction result is output. The method performs excellently in the deterministic and probability evaluation indexes, has strong stability across data sets, and can be directly applied to power grid dispatching and photovoltaic power generation system operation management.

A method and system for probabilistic solar generation forecasting under rapidly changing weather conditions integrate copula theory with an extreme gradient tree boosting (XGBoost) classifier to enhance forecast accuracy. Historical weather data is partitioned into meteorological clusters, and bivariate copulas analyze spatiotemporal correlations to select optimal features. Multivariable Vine and Gaussian copulas model variable dependencies, with an XGBoost classifier dynamically selecting the optimal copula based on real-time weather conditions. Synthetic weather data, generated using the selected copula, captures uncertainties and is applied to a trained XGBoost regression tree to produce probabilistic forecasts. The method achieves up to 60% higher accuracy than conventional models under non-sunny conditions, leveraging Gaussian Kernel Density Estimation and Huber loss for robustness. The system supports real-time grid operations, offering reliable solar power predictions for diverse weather scenarios, validated with real-world data from multiple global locations.

This application discloses a multi-objective power grid dispatching decision-making method and system, relating to the field of power grid dispatching technology. The method includes: constructing a normal dispatching objective based on minimum load fluctuation; constructing an abnormal weather dispatching objective based on minimum dispatching quantity and minimum dispatching loss; constructing a device abnormality dispatching objective based on minimum switching loss; constructing a multi-objective optimization function by combining the normal dispatching objective, the abnormal weather dispatching objective, and the device abnormality dispatching objective with initial priority weights; obtaining probabilistic prediction values ​​based on regional historical load data, regional real-time load data, regional historical weather data, regional real-time weather data, regional historical device maintenance data, and regional real-time device data; obtaining the dispatching weights for each dispatching objective based on the probabilistic prediction values; and updating the multi-objective optimization function with the dispatching weights to update the multi-objective power grid dispatching decision. The beneficial effects of this application are: balancing scenario switching flexibility and power dispatching accuracy when facing complex real-world scenarios.

The invention discloses a Bayesian model averaging-based power prediction model evaluation method and a related device, and belongs to the technical field of power system prediction. The method comprises the following steps: constructing a candidate model set comprising a plurality of power prediction models; secondly, historical data are used for training all the models, the posterior model probability of each model is calculated based on the Bayesian theorem, the posterior model probability serves as the scientific weight of the model, and the goodness of fit and complexity of the model are considered in the weight at the same time; and finally, for a new prediction input, performing weighted average on the prediction distribution of each candidate model by taking the posterior probability as the weight to generate comprehensive probability prediction distribution. According to the method, the advantages and disadvantages of each candidate model are scientifically evaluated through the posterior probability, and a comprehensive and probabilistic prediction result is finally generated, so that the robustness and reliability of prediction are improved, and richer decision information is provided for power grid dispatching.

The invention discloses a day-ahead electricity price prediction method and device based on real-time market demand prediction. The method comprises the following steps: acquiring a market game expected index and traditional market data in a historical transaction period; constructing a fusion feature matrix based on the market game expected indexes in the historical transaction period and the traditional market data; matching the fusion feature matrix with preset historical day-ahead electricity price data to generate a training data set; training the integrated probability prediction model according to the training data set to obtain a target integrated probability prediction model; acquiring a market game expected index and traditional market data of a to-be-predicted period to construct a to-be-predicted feature matrix; inputting the to-be-predicted feature matrix into a target integrated probability prediction model to obtain a prediction processing result; and according to the prediction processing result, generating a day-ahead electricity price single-point prediction value and a day-ahead electricity price prediction interval. By implementing the scheme, more comprehensive and reliable price signals and risk quantification information can be provided for the operation optimization decision of the power system.

The invention belongs to the technical field of post-earthquake building falling object accumulation range probability prediction models and urban road network blocking vulnerability estimation, and discloses a post-earthquake building ruin distribution range probability prediction model and an urban road network blocking vulnerability estimation method. The building falling object accumulation range prediction model proposed based on the Bayesian theory considers multiple damage states of the building under the earthquake action, including building complete collapse and non-structural member damage; posterior probability distribution of unknown parameters of a prediction model can be obtained based on the Bayesian theory, and accordingly cognitive uncertainty from the model parameters is quantified in post-earthquake road blocking vulnerability evaluation. The method not only helps to improve the accuracy of post-earthquake road congestion risk assessment, but also helps to reasonably plan a post-disaster rescue evacuation path.

The invention discloses a new energy output probability prediction method and device and electronic equipment, which are applied to a probability prediction model and comprise a feature extraction module, a feature fusion module, a probability prediction module and a result output module. The method comprises the following steps: acquiring day multi-source data; extracting meteorological element features of the weather forecast data in a first preset time period through a feature extraction module, extracting image features of the remote sensing image data, and extracting time sequence features of the power data; fusing the meteorological element features, the image features and the time sequence features through a feature fusion module to obtain target fusion feature data; processing the target fusion feature data with the time sequence through a probability prediction module to obtain probability distribution of new energy power generation power; and analyzing the probability distribution through a result output module, and outputting a prediction result. According to the method, the complete probability distribution of the future output potential can be directly output, and a more comprehensive decision basis is provided for market participants.

The application provides a new energy power system power imbalance quantitative evaluation method, device and equipment, by considering the random volatility of uncertain power output and load demand and the mutual influence therebetween, a joint probability prediction model for different prediction time scales of a future to-be-evaluated period is constructed, and by using the joint probability prediction model and the working state information of the uncertain power and load in the time dimension provided by the operation scheduling plan, a probability distribution of the net difference of the uncertain power in the to-be-evaluated period is generated, the net difference of the deterministic power and load is considered, and the power imbalance distribution of the whole power system is accurately predicted, so as to provide accurate data support for the scheduling and operation of the power system.

The invention discloses a micro-grid operation random optimization method based on source load probability prediction. The method comprises the following steps: firstly, performing energy-sensitive self-organizing segmentation on source-load historical time sequence data, and extracting morphological fingerprint features; and adopting an improved affinity propagation clustering algorithm fused with a power system operation constraint penalty mechanism to identify a typical operation mode. Secondly, a quantile regression model based on a gated pulse neural P system is established in each mode for probability prediction, and a probability scene library with weights is generated; and finally, constructing a two-stage stochastic optimization model with the goal of minimizing the expected operation cost, and solving by adopting a Benders decomposition algorithm to obtain a fixed equipment plan and a flexible operation strategy. According to the method, through refined mode recognition and probability modeling, on the basis of fully considering the uncertainty of the source load, robust optimization of economic operation of the micro-grid is realized, and the expected operation cost of the system is effectively reduced.

The application discloses a building-traffic energy coupling relationship modeling analysis method and system based on GNN, relates to the technical field of energy management, performs cross-domain energy coupling relationship modeling on a target energy area, obtains a space-time heterogeneous graph network, performs hierarchical prediction to obtain a predicted load sequence, and imports a probability prediction model to determine a probabilistic prediction result; the heterogeneous graph structure and the probabilistic prediction result are subjected to coding and decoding reconstruction processing, and abnormal prediction data are located through reconstruction error; and the target energy area is managed according to the probabilistic prediction result and the abnormal prediction data, thereby solving the technical problems of difficulty in probabilistic adaptive prediction and efficient abnormal positioning under dynamic modeling of building-traffic coupling relationship in the prior art, being capable of adapting to energy consumption characteristic changes under different time periods and different passenger flow scales, and greatly improving the prediction accuracy and stable analysis capability of building-traffic coupling load.

A neural process prediction method for surface reflection characteristics based on micro-surface model guidance is proposed. Current BRDF representations based on analytical models or deterministic neural networks perform poorly in complex material modeling and sparse sampling scenarios. There is a lack of standardized and accurate probabilistic prediction methods for tasks related to quantifying prediction uncertainty and analyzing and controlling error propagation. This invention forms an initial model by training a network on a completed BRDF dataset and applying physical guidance constraints. Based on the initial model, a probabilistic model of the bidirectional reflection distribution function is formed to create a family of BRDF function probability distribution models. The prediction process for reflection characteristics in unobserved directions is then completed based on the established family of BRDF function probability distribution models.

This invention discloses a multivariate probabilistic prediction method and system for photovoltaic power based on mixed evaluation scores. The method includes: acquiring input information from multiple photovoltaic power plants at the time to be predicted, the input information including historical photovoltaic power generation and numerical weather forecasts; inputting the input information into a deep learning model trained with mixed evaluation scores as the loss function, and outputting the construction parameters of a conditional multivariate Gaussian mixture model; generating weights, means, and covariance matrices of multiple Gaussian components according to the construction parameters through fixed rules, and substituting them into the joint probability density function formula modeled by the conditional multivariate Gaussian mixture model to obtain the joint probability density function of future photovoltaic power generation from multiple power plants, thereby achieving joint probabilistic prediction of power generation from multiple power plants. This invention significantly improves the overall accuracy of multivariate probabilistic prediction and can accurately capture the power correlation characteristics between multiple photovoltaic power plants.

The application belongs to the technical field of post-earthquake building collapse range probability prediction model and urban road network congestion vulnerability estimation, and discloses a post-earthquake building collapse range probability prediction model and a method for estimating the vulnerability of urban road network congestion. The building collapse range prediction model based on the Bayes theory considers various damage states of buildings under the action of earthquakes, including complete collapse of buildings and damage of non-structural components; the posterior probability distribution of unknown parameters of the prediction model can be obtained based on the Bayes theory, and thus the cognitive uncertainty from the model parameters is quantified in the post-earthquake road congestion vulnerability estimation. The application not only helps to improve the accuracy of post-earthquake road congestion risk assessment, but also helps to reasonably plan post-disaster rescue evacuation routes.

The invention relates to a physically constrained photovoltaic power probability prediction method, device and equipment and a medium, and belongs to the technical field of new energy power generation prediction.The method comprises the steps that preprocessed regional operation data is acquired; on the basis of the preprocessed regional operation data, regional constraint characteristics at a preset future moment are calculated, an input characteristic set is generated according to the regional operation data and the preprocessed regional constraint characteristics, and the regional constraint characteristics comprise theoretical maximum power and irradiance uniformity indexes; inputting the input feature set into a preset quantile prediction model to obtain initial prediction results under different confidence degrees; constraining the power range boundary of the initial prediction result through a preset physical constraint and a theoretical maximum power to obtain a prediction optimization result; and performing uncertainty correction on the prediction optimization result based on the irradiance uniformity index, and dynamically adjusting the interval width of the prediction optimization result to obtain a final prediction result. The method has the effect of improving the power generation prediction accuracy.

A quality prediction device includes a probabilistic prediction model generation unit that generates a probabilistic prediction model on the basis of a plurality of training data in which a log relating to a molding operating condition or to a state of a molding machine and a quality value of a molding product corresponding to the log are associated with each other, and a quality prediction unit that calculates a quality index of the molding product from the log using the probabilistic prediction model.

The invention discloses a deterministic and probabilistic wind speed prediction method, system and device and a medium. The method comprises the steps that total prediction uncertainty is decoupled into data uncertainty and model uncertainty; a Monte Carlo discarding algorithm is introduced in a neural network reasoning process to carry out multiple times of forward propagation, the uncertainty of the model is quantitatively evaluated, and a variance estimation value of the uncertainty of the model is obtained; an error prediction model is constructed, an independent network is trained by using a point prediction residual error, and the data uncertainty is quantitatively evaluated, so that a variance estimation value of the data uncertainty is obtained; a comprehensive evaluation index system is obtained by designing a determinacy prediction evaluation index and a probability prediction evaluation index based on a variance estimation value of data and model uncertainty; and a wind speed prediction result is obtained by constructing a multi-resolution neural network and carrying out feature fusion on multi-time scale input. According to the invention, the risk assessment and decision accuracy based on the probability result can be improved, and the wind speed data can be fully utilized.

A reservoir earthquake risk evaluation method based on fusion of a physical mechanism and probability prediction comprises the following steps: step 1, constructing a three-dimensional model of a reservoir, simulating spatio-temporal evolution of a stress field and a pore water pressure field under reservoir water level change, and according to a spatio-temporal evolution result, calculating a reservoir earthquake risk evaluation result; calculating coulomb stress and pore pressure variation of each section of the target fault; a second step of jointly inputting the coulomb stress variation, the pore pressure variation and the tectonic strain cumulant into a Bayesian-machine learning hybrid prediction model based on historical earthquake example training, and obtaining conditional probability density functions of each segment corresponding to different earthquake magnitude grades at different time; and a third step of dynamically determining a multi-mechanism weight according to a real-time water level and a fault stress state, generating a standardized space-time risk index according to weighted fusion probability density, and dividing risk levels according to the standardized space-time risk index to obtain a reservoir earthquake risk evaluation result. Therefore, according to the design, dynamic and quantitative evaluation of the reservoir-induced earthquake risk can be realized.

The invention belongs to the technical field of new energy, and discloses a wind power weather forecast fusion method, a wind power climbing early warning method and a related device, and the prediction method comprises the steps: obtaining future weather forecast data predicted in a plurality of forecast modes of a wind power plant; and according to the optimal fusion weight of each forecasting mode, carrying out weighted fusion on the future weather forecasting data predicted in the plurality of forecasting modes of the wind power plant to obtain the future fused weather forecasting data of the wind power plant. The early warning method comprises the steps of obtaining future fused weather forecast data, future wind power prediction power and current prediction deviation of a wind power plant, and obtaining future wind power prediction power probability distribution of the wind power plant through a preset probability prediction model; and obtaining and generating a wind power climbing early warning result according to the difference of adjacent time windows of future wind power prediction power probability distribution and the sudden change direction cumulative probability. The problems that wind power prediction uncertainty quantization is insufficient, climbing event early warning lags behind and the like are solved, and high-precision weather prediction and climbing early warning in advance are achieved.

Provide a system. 【Solution means】 An information collection means for collecting user operation data, An information analysis means for evaluating the impact on the environment based on the collected information, An improvement plan generation means for providing an optimal countermeasure plan to the user based on the evaluation result, A progress tracking means for monitoring the achievement degree of the environmental policy set by the user, A system including a collaboration means for realizing information exchange with other users, A technical means for analyzing the user's energy consumption data and generating a proposal to promote specific behavior change, A notification means for sending a notification as a reminder to the user for the purpose of achieving energy consumption and environmental load reduction, A response means for providing real-time feedback on behavior change based on the data obtained from the user, A system including an evaluation means for evaluating the generated proposal using a probabilistic prediction model.

The invention discloses a short-term power load probability prediction method based on an improved neural network, and belongs to the technical field of power system load prediction. The method comprises the following steps: acquiring and preprocessing load and related data; the method comprises the following steps: constructing a DCS-CNN-BiLSTM-Attention-QR probability prediction model fusing a convolutional neural network, a bidirectional long short-term memory network, an attention mechanism, quantile regression and kernel density estimation; automatically optimizing key hyper-parameters of the model by utilizing a difference creation search algorithm; carrying out model training by adopting a quantile loss function to obtain load prediction values under different quantiles; and finally, generating a probability density curve of the prediction interval through kernel density estimation. According to the method, deterministic point prediction is expanded into probabilistic interval prediction, the uncertainty of load prediction can be quantified, the model precision and generalization ability are improved through an intelligent optimization algorithm, and richer and more reliable decision information is provided for power system scheduling and risk management.

The invention discloses a runoff probability forecasting method based on an improved deep integration strategy, belongs to the field of runoff probability forecasting, and solves the problem that an existing deep integration strategy can only integrate a single network and arithmetic average distribution weight. The method comprises the following steps: 1, selecting LSTM, GRU and SWM as an integrated sub-network according to an inclusive strategy, and constructing a prediction framework of an input step length 4, an output step length 1 and a hidden layer number 3; 2, a sigma unit is added to an output layer of the sub-network, and softplus is used for activating function constraint, so that uncertainty can be predicted in a quantized mode; 3, taking a negative logarithm likelihood function based on normal distribution as a training scoring rule; 4, after the sub-networks are trained in parallel, a training set is split through three-fold cross validation, an integrated weight is dynamically optimized through a genetic algorithm, and an MNDE model is constructed; the model is superior to a single deep learning model and a traditional machine learning model in certainty, interval, probability prediction and reliability, and is suitable for runoff probability prediction.

This invention discloses a photovoltaic power probability prediction method and system based on the copula function. The method includes: acquiring historical photovoltaic power data and historical meteorological data of centralized and distributed photovoltaic power plants; obtaining different weather types through clustering; obtaining the cumulative distribution of photovoltaic power based on the historical photovoltaic power data under different weather types, and constructing optimal copula function models that quantify and dynamically represent the spatial correlation of centralized and distributed photovoltaic power respectively; obtaining the predicted value of distributed photovoltaic power points based on the acquired photovoltaic power data of centralized photovoltaic power plants and the corresponding optimal copula function models; constructing a conditional probability model based on the quantile regression method, and obtaining the conditional probability prediction value corresponding to the predicted value of distributed photovoltaic power points through the conditional probability model, thereby achieving more accurate distributed photovoltaic power prediction and power probability prediction, and ensuring the reliability of power system operation.

The invention belongs to the technical field of power load prediction, and relates to a regional multi-building power load probability prediction method, system and device based on a weather forecast set. Acquiring multi-building historical load, historical meteorological observation data, a historical weather forecast set and date, holiday and festival information in the target area, and performing preprocessing; in the training stage, historical meteorological observation data is adopted to construct a regional multi-building load prediction model, a Transform is utilized to extract multi-building shared load characteristics, and a pre-training time sequence model Timer is combined to learn a mapping relationship among weather, loads and exogenous variables; in the prediction stage, a target day weather forecast set is input, each forecast curve is used as a weather scene, and a load prediction result is obtained. And determining a reliability weight based on a historical forecast set and meteorological observation data, performing weighted aggregation on a multi-scene prediction result according to the weight, and outputting a regional multi-building power load probability prediction result considering weather forecast uncertainty.