373 results about "Load forecasting" patented technology

This application relates to the field of predictive control technology, specifically a distribution network load balancing dispatching method and system based on model predictive control, including the following steps: collecting voltage, current, power, photovoltaic irradiance, and wind speed, normalizing them to form predictive inputs and constructing a collaborative relationship; solving for power distribution and voltage levels based on power balance, energy storage, and line constraints to generate control quantities; correcting the rolling step size based on continuous periodic errors and time-series correlations; adjusting energy storage charging and discharging and adjustable load start-up and shutdown according to the difference between load forecast and baseline power to generate instructions; comparing dispatching and forecasting and adjusting the step size weights with mean square error to generate load balancing results. In this application, load and power supply are collaboratively characterized by unified processing of multi-dimensional electrical and environmental quantities; optimization quantities that can be updated with data changes are generated by combining constraint relationships; the rolling rhythm is calibrated in error evaluation; and energy storage and adjustable loads are adjusted under the drive of power difference to keep power flow and voltage stable, thereby improving load balance.

This application discloses a method, apparatus, equipment, and medium for building load forecasting, belonging to the field of electricity forecasting. The method comprises: collecting first load data of a target building and second load data of several adjacent buildings of the same type corresponding to the target building; inputting the first load data and each of the second load data into a pre-trained black-box adaptation network to obtain input embedding vectors, spatial embedding vectors, and feature embedding vectors, respectively; inputting the input embedding vectors, spatial embedding vectors, and feature embedding vectors into a preset large language model to obtain a load forecast embedding vector representing the load forecasting result; wherein, the large language model is iteratively trained on the black-box adaptation network to avoid access to the internal parameters of the large language model; and determining the load forecasting result of the target building based on the load forecast embedding vector. This application can improve the accuracy of building load forecasting in scenarios with few or zero samples.

The present application relates to the technical field of electric power data processing, and particularly relates to a kind of photovoltaic-storage-direct-flexible system regulation and control method, system, computer device and medium;The method comprises: obtaining the multi-source history and forecast data of photovoltaic-storage-direct-flexible system and pre-processing;By photovoltaic output prediction model, output prediction calculation is carried out, and predicted output curve is obtained, and by load prediction model, load prediction calculation is carried out, and predicted load curve is obtained;By day-ahead optimization model, optimization solution is carried out, and day-ahead optimization plan is obtained;According to the actual output of photovoltaic and the corresponding predicted output, the real-time prediction error of photovoltaic output is calculated, to generate real-time regulation and control instruction and issue to photovoltaic-storage-direct-flexible system.Through such a way, the technical problem of insufficient prediction accuracy in the prior art in the building scene of high penetration rate renewable energy and dynamic load access is solved, and the accuracy, adaptability and overall energy efficiency of photovoltaic-storage-direct-flexible system dispatching operation are improved.

The application provides a heat supply joint control method and system based on heat network system load prediction, and belongs to the technical field of data processing. The method comprises the following steps: performing grid division on a target heat network system heat supply coverage area according to heat supply shunt to obtain a plurality of heat supply areas; performing grade division according to heat supply demand attribute proportion conditions to obtain a first heat supply area and a second heat supply area; obtaining first demand load information; obtaining a historical load information set of the second heat supply area; performing data analysis on a plurality of historical area load information sets to generate a plurality of target feature sets; obtaining real-time feature values of the second heat supply area; inputting the load prediction model to obtain a plurality of second prediction load demands; and performing heat supply joint control on the target heat network system. The application solves the problems of long joint control period and low intelligent degree when heat supply is insufficient in the prior art, and achieves the technical effect of efficiently performing differentiated heat supply in the process of joint control of heat supply units to ensure heat supply.

This invention discloses a method for reconstructing battery swapping station operation data based on adaptive truncation and feature correlation repair, belonging to the field of data reconstruction technology. First, a standardized two-dimensional spatiotemporal discrete model is constructed, and sequence mapping technology is used to align the discrete data stream to the date-time period operation space. Second, a zero-first truncation model based on the continuity of frequency distribution is established to adaptively remove outliers and noise. Then, Latin hypercube sampling and inverse cumulative distribution function transformation algorithms are introduced to fill in missing values ​​while ensuring that the reconstructed data conforms to the original probabilistic characteristics. Finally, the battery swapping frequency-electricity data is dimensionally reconstructed to facilitate the inductive analysis of battery swapping behavior characteristics. Numerical examples demonstrate that the method proposed in this invention effectively restores the statistical usability of the data, fully revealing the bi-peak tidal characteristics of battery swapping load and the distribution law of low-electricity anxiety, providing relatively reliable data support for load forecasting and regulation.

This application relates to a real-time power load forecasting method and system based on a hybrid model. The method includes: acquiring current cycle load data and combining it with historical load data to form a load sequence, and acquiring corresponding weather data; performing timestamp alignment, anomaly processing, and normalization on the load sequence and weather data to obtain a standardized input sequence; determining the order parameters of the linear forecasting model and the set of hyperparameters to be optimized, consisting of the network structure and training parameters of the nonlinear forecasting model, and optimizing them through a combined optimization algorithm to update the training configuration of the two models; outputting the first and second forecast sequences for the next cycle in the current cycle, respectively, and using the second forecast sequence as a trend term to compensate the first forecast sequence to obtain a fused forecast sequence; acquiring the fused forecast sequence for the current cycle from the previous cycle, constructing a residual sequence with the current cycle load data and determining the deviation term, calibrating the current cycle fused forecast sequence online, and outputting the calibrated load forecast result.

The application discloses a kind of based on working condition division concentrator electric load modeling and production plan optimization method.The method first defines normal production, low production operation, three production states and corresponding load coefficients of shutdown maintenance, in combination with typical daily load curve, daily random fluctuation factor and seasonal variation factor, construct multi-layer composite annual load refinement model;Again, initial production plan is generated by state coding, and the scattered maintenance day and low production operation day are integrated into continuous period by using centralized scheduling optimization algorithm, to reduce the frequency of production state switching;Finally, the optimized production plan and load model are integrated to generate a 8760-hour refined load sequence throughout the year.The application realizes the deep coupling of production plan and electric load, significantly improves the stability and accuracy of load prediction, provides reliable data support and decision basis for power grid planning, demand side response and enterprise energy efficiency optimization, and is highly adaptable and easy to implement.

The short-term load forecasting method based on SARIMA-random forest combination model comprises the following steps: grouping the original load data by using a sliding window, decomposing the to-be-tested week-before-next day data set of each group to obtain a trend item, a seasonal item and a residual item; establishing a SARIMA model, predicting the trend item to obtain a preliminary prediction result and a residual; clustering weather factors to obtain similar days, grouping to construct a weather-residual data set and establishing a random forest regression model, learning the influence of the weather factors on the residual, and selecting model parameters by using a grid search method; combining the prediction results of the model, and comparing the influence of weather clustering and residual training on the load prediction accuracy. The method can accurately predict the next day load under the condition that the historical load and weather factors of the to-be-tested day are known, and improves the prediction accuracy.

The application provides an air conditioner load prediction method and device, electronic equipment, an air conditioner and a storage medium, and relates to the field of air conditioners. The method comprises the following steps: acquiring actual operation environment data of an air conditioner; wherein the actual operation environment data comprises a space occupancy parameter and a door and window operation parameter, the space occupancy parameter is used for indicating information that an object exists in a space where the air conditioner is located, and the door and window operation parameter is used for indicating information that a door and window in the space is opened or closed; encoding the actual operation environment data to obtain a first feature representation; and performing load prediction according to the first feature representation to obtain a predicted load of the air conditioner. Therefore, by capturing the main disturbance factors in the real operation scene of the air conditioner, including the space occupancy parameter reflecting the heat source distribution of objects such as personnel, equipment and pets in the space where the air conditioner is located, and the door and window operation parameter representing the opening and closing behavior or the ventilation intensity of the door and window, and performing load prediction after encoding the above parameters, the accuracy of air conditioner load prediction can be significantly improved.

In the context of an electrical utility system, changing cloud conditions may cause a customer having solar panels to greatly increase or decrease electrical demand in a difficult-to-predict manner. Accordingly, a spinning reserve maintained by an electric utility company must be larger, and is therefore more expensive. In an example, the spinning reserve may be managed by: calculating a stable sequence of forecasts of smoothed real-time consumption, wherein the calculating is based at least in part on smoothed estimates of consumption data. A stable sequence of forecasts of real-time measured load may be calculated by subtracting forecasts of real-time distributed solar photovoltaic (PV) generation data from the stable sequence of forecasts of smoothed real-time consumption. The spinning reserve of the electricity system may be controlled based at least in part on the stable sequence of forecasts of real-time measured load.

The application discloses a power grid load prediction method and system based on deep learning, which comprises the following steps: collecting multi-dimensional load correlation data through an intelligent power grid data fusion calculation platform, cooperatively capturing space-time correlation characteristics through a space-time residual gated recurrent model, enhancing network and strengthening key feature representation through a gated recurrent unit, then inputting a multi-modal load prediction deep model to build feature mapping relationship, generating preliminary prediction results through feature weight distribution, abnormal value elimination and smoothing optimization, and outputting final data after platform verification. The space-time residual gated recurrent model, the gated recurrent unit enhancement network and the multi-modal load prediction deep model all have exclusive feature processing mechanisms, and the processing accuracy is guaranteed through feature classification, gate initialization, mode division, fusion calculation and other subdivided processes. The application effectively improves the comprehensiveness and reliability of load prediction, and is suitable for accurate scheduling and optimized operation of intelligent power grids.

This invention discloses a power system for load forecasting and dynamic scheduling, relating to the field of power system technology. The system includes: constructing an initial simulation environment for the corresponding scheduling cycle in a digital twin model based on grid status and multi-dimensional load forecasting information from power grid, grid, load, and storage; generating a baseline scheduling plan by solving the solution; selecting candidate scheduling strategies with the highest comprehensive evaluation and associating them with grid status and multi-dimensional load forecasting information from power grid, grid, load, and storage, storing them together in a dynamic strategy library; executing the candidate scheduling strategies and collecting actual operating data to calculate the strategy execution deviation; and adaptively correcting the parameters of the digital twin model and the generation rules of the counterfactual scheduling strategy based on the strategy execution deviation. This invention continuously corrects the parameters of the digital twin model based on actual operating data and adaptively adjusts the disturbance strategy, resulting in higher prediction accuracy and simultaneous improvement in economic efficiency and safety margin of the scheduling strategy.

The application discloses a load prediction and adaptive adjustment method of a district heating system, and belongs to the field of intelligent heating and energy management automation. The method constructs a graph model representing the physical topology structure of the heating pipe network, and quantifies the heat transfer attenuation by giving dynamic connection weights to the connection between nodes based on the pipe physical parameters. After obtaining the thermal data and outdoor environment data, the historical data and the graph model are input into the load prediction model to extract the spatial coupling characteristics between each heat exchange station, and the time domain coding characteristics are extracted based on the outdoor environment data. A feedforward compensation mechanism for the thermal inertia of building envelopes is introduced in the process to correct the temperature response lag. The above-mentioned characteristics are fused to generate a heat load prediction value, and the circulating pump frequency or the primary side electric regulating valve opening degree of each heat exchange station is adaptively adjusted accordingly. The application solves the problems that the existing method cannot fuse the spatial coupling relationship of the pipe network and lacks thermal inertia compensation, and improves the prediction accuracy and system energy efficiency.

This invention relates to a coupled clearing method and apparatus for virtual power plants participating in the spot electricity market, applied in the field of electricity market technology. The method includes: calculating the High-Gross Price Ratio (HGRP) as the threshold for economic compensation of DRVPPs, compensating DRVPPs only when the node marginal price is higher than the HGRP, thus preventing the activation of non-economical DRVPPs during low-price periods, preventing low-cost generation resources from being squeezed out of the market, and ensuring the authenticity and effectiveness of the market clearing price signal; incorporating DRVPPs as a controllable and constrained demand-side adjustment variable into the clearing model, avoiding the failure to reflect the adjustment value due to its inclusion in load forecasting, and overcoming the problem of simply treating it as "negative load" and ignoring response constraints, thereby improving the rationality of the clearing results; and adopting a differentiated economic compensation mechanism based on the node marginal price, linking the adjustment revenue of DRVPPs to the marginal electricity cost of the node, achieving precise matching of DRVPP resources with the actual adjustment needs of the system.

This invention belongs to the field of power system analysis and load forecasting technology, and discloses a method and device for forecasting residential electricity consumption. The method couples the population structure subsystem, socio-economic subsystem, and residential electricity consumption system to construct a closed-loop feedback system, forming a system dynamics model. Based on the current initial value of per capita electricity consumption and the calculated rate of change of per capita electricity consumption, the predicted per capita electricity consumption is calculated. This invention can effectively reflect the dynamic evolution of electricity demand under multi-factor coupling conditions, and can provide intuitive and reliable decision-making basis for power grid planning, power load management, and energy policy formulation. It is applicable to medium- and long-term forecasting scenarios of residential electricity consumption at different regional scales.

The application relates to a multi-source covariate irrigation load short-period prediction method, a device and a medium, and relates to the fields of power system load prediction and intelligent analysis of agricultural irrigation energy. The application is to solve the problems that the existing irrigation load short-period prediction method has limited modeling capability for exogenous factors, cannot effectively utilize user difference information, and is prone to lag or drift in prediction. The application performs correlation evaluation on candidate multi-source covariates and historical global irrigation load, and then selects effective covariates according to the evaluation results; extracts a user feature vector based on the effective covariates; constructs a historical input sequence by combining the historical global irrigation load and the effective covariates, and constructs a future known covariate by combining the known future prior information in the prediction interval; constructs input data by combining the user feature vector, the historical input sequence and the future known covariate; inputs the input data into a load prediction model adopting a TimeXer framework, and outputs an irrigation load prediction result in the prediction interval.

This invention relates to a method and system for short-term load forecasting across seasons in the electricity spot market. The method first collects electricity load, meteorological, and temporal characteristic data to construct a multi-dimensional input feature set, which is then divided into a training set and a validation set. Next, variational mode decomposition is used to decompose the original load data into three modal components, corresponding to three types of input feature sets. Then, with the goal of maximizing the validation set determination coefficient, seasonally differentiated hyperparameter optimization is performed using a particle swarm optimization algorithm. Temporal dependency prediction sub-models and multi-feature association prediction sub-models are constructed based on the optimized Long Short-Term Memory Network and Random Forest, respectively. The three types of input feature sets are then input into their respective sub-models, outputting two types of prediction results. Finally, the two types of results are weighted and integrated to obtain the short-term load forecast result. Compared with existing technologies, this invention has advantages such as significantly improved prediction accuracy.

This application relates to the field of energy-saving technology for industrial air compressors, and particularly to a method, apparatus, equipment, and medium for predicting the air load of an air compressor. This application inputs multi-dimensional feature data from multiple historical periods and production data from the period to be predicted into a trained load prediction model to obtain predicted load values. During the training of the load prediction model, feature transformation processing is performed on the actual production data to obtain continuous production intensity characteristics; deviation processing is performed on the actual production data and planned production data to obtain production deviation characteristics; the continuous production intensity characteristics, production deviation characteristics, equipment air consumption data, and time characteristics are input into the load prediction model for double weighting, and load prediction is performed based on the weights after double weighting to obtain training load values; a loss function is determined based on the training load values ​​and continuous production intensity characteristics; the parameters of the load prediction model are adjusted based on the loss function until the training conditions are met, resulting in a trained load prediction model.

The application discloses a power load multi-model integrated prediction method and device, equipment and medium, and relates to the technical field of power load prediction, which comprises the following steps: cleaning and normalizing the historical load time series data, then adopting a strict causal sliding window mechanism to construct multi-scale features, determining the optimal parameters of a tree model, a gated recurrent single-rank regression meta-model and a bidirectional recurrent neural network through a two-stage hyperparameter joint optimization strategy and constructing corresponding models, inputting structured features, a first time series matrix and a second time series matrix for orthogonal division prediction, and finally outputting results through enhanced stacked fusion. The method realizes root-cause avoidance of data leakage, does not depend on exogenous variables, reduces the parameter tuning calculation complexity, realizes model differentiation and complementation, and improves the accuracy and robustness of power load prediction.

This invention discloses a method, apparatus, device, and product for power load forecasting. The method includes: acquiring a dataset; determining clustering results based on the dataset using a scenario segmentation submodule in a power load forecasting model, wherein the clustering results indicate the mapping relationship between features of various set weather events and corresponding load change patterns; correcting the dataset for temperature using a temperature accumulation effect correction submodule in the power load forecasting model to determine the corrected effective temperature; and performing power load forecasting based on the dataset, the effective temperature, and the clustering results using an adaptive decomposition forecasting submodule in the power load forecasting model, outputting the load forecasting result. The technical solution of this invention addresses the problem of insufficient ability of power load forecasting models to represent temperature accumulation effects, improves the forecasting accuracy of power load data, and provides accurate and reliable technical support for power grid dispatching and risk prevention.

This invention discloses a multi-step load forecasting method for integrated energy systems based on Bayesian optimization and anti-overfitting, comprising the following steps: acquiring and preprocessing historical time-series data of the integrated energy system; constructing an initial feature set, using a random forest model to calculate the importance of each feature in the initial feature set, and selecting a core feature subset from the initial feature set according to a preset cumulative importance threshold; defining the hyperparameter search space of the hybrid model; using a Bayesian optimization framework, aiming to minimize the comprehensive loss on the validation set, and determining an optimal hyperparameter combination through iterative search; configuring and constructing the hybrid model according to the optimal hyperparameter combination; independently training the hybrid model for electricity load, cooling load, and heating load respectively, inputting the current time-series features into the prediction model, and outputting multi-step predicted values ​​of each type of load within a specified future time period. This effectively improves the overall accuracy and generalization ability of multi-step load forecasting for the integrated energy system over the next 24 hours.

This invention provides a method and system for ultra-short-term regional load forecasting. The method includes: acquiring historical measured load sequences, historical measured meteorological sequences, and forecast meteorological sequences for all user nodes in a target area; inputting the historical measured load sequences, historical measured meteorological sequences, and forecast meteorological sequences into a pre-trained ultra-short-term regional load forecasting model to obtain the ultra-short-term regional load forecasting result output by the model; wherein, the ultra-short-term regional load forecasting model is obtained by training a multi-level graph attention network. This invention significantly improves the accuracy and reliability of ultra-short-term regional load forecasting and is applicable to real-time scheduling and control scenarios in distributed energy and smart grid environments.

This invention discloses an intelligent operation and control method and system for cold storage and subcooling cold storage. The method includes: acquiring basic parameters and real-time monitoring data of the cold storage; calculating the real-time load and short-term load change trends of the cold storage through a dynamic load prediction model; formulating an operation strategy based on load factor and time-of-use electricity price information, and intelligently selecting normal cooling, combined cooling with cold storage, or combined cooling with cold release and subcooling modes; achieving smooth switching between the three modes by adjusting the operating status of the compressor unit, electronic expansion valve, electric valve, and water pump; and continuously optimizing the load prediction model and control strategy using real-time feedback of the system's operating status. This invention achieves coordinated control of the cold storage and subcooling processes, improving the system's energy efficiency and operational economy under variable load conditions, and is applicable to the energy-saving and optimized operation of various types of cold storage and subcooling cold storage facilities.

The application provides a power supply load prediction method based on a convolutional neural network. The power supply load prediction method based on the convolutional neural network comprises the following steps: S1. Obtain historical load data of a power system in a certain region, and preprocess abnormal data in the historical load data; S2. Analyze and quantify factors influencing the power load, and normalize the corrected load data, and the specific operation is as follows: normalize the load data to [0, 1] by using a normalization formula, so that the load data is in the same order of magnitude, thereby accelerating the convergence speed of the neural network. According to three types of data such as power, temperature and holidays, the method constructs a convolutional neural network power prediction algorithm, improves the mass data processing efficiency in the power prediction process, comprehensively considers the related information such as temperature, and overcomes the problem that the prediction process excessively depends on personal experience.

The application provides a data center cold load prediction method and system based on feature reconstruction, and relates to the technical field of energy consumption management; including obtaining data center cooling system operation data and meteorological data; the cold load time series of the operation data is decomposed through an empirical mode decomposition method to obtain an intrinsic mode function and a residual term; a trend component, a periodic component and a residual component are calculated based on the intrinsic mode function and the residual term; the trend component value and the periodic component value are processed through a pre-constructed LSTM model to obtain a trend component prediction value and a periodic component prediction value; the residual component is processed through a pre-constructed RF model to obtain a residual component prediction value; the trend component prediction value, the periodic component prediction value and the residual component prediction value are linearly superimposed to obtain the final prediction value of the data center cold load time series. The application predicts after feature reconstruction of the cold load data, and improves the accuracy and robustness of the cold load prediction method.

The application discloses a kind of real-time monitoring and optimization method of electric power load based on intelligent algorithm, it is related to electric power system intelligent control technical field, including, acquisition electrical and environmental data, and are preprocessed, current waveform is analyzed using variational autoencoder (VAE) model, while identifying the equipment combination in operation and assessing its health state, generate health degradation index;Device combination and health state are used as input, corrected and generated with confidence interval and failure probability Load forecasting curve;The prediction result is input into an optimizer of built-in physical information neural network (PINN), the optimizer considers the health of equipment, generates power set point and scheduling instruction that meet physical constraints;The optimized instruction is issued to the execution terminal, and the actual execution result and state are collected.The application forms the complete closed loop of perception-decision-execution-evolution, ensures that the system has long-term operation stability and robustness.

This invention discloses a smart park load forecasting method integrating seasonal and solar term features, comprising: generating a feature input vector and a meteorological-solar term deviation vector based on current input data; a trained weight calculation module outputting a weight vector and a bidirectional dynamic hidden state vector based on these two vectors; a trained gating network outputting a fuzzy probability distribution based on the weight vector and the bidirectional dynamic hidden state vector; and a trained fuzzy boundary hybrid expert network outputting park load forecast data for a preset future time period based on the weight vector, the bidirectional dynamic hidden state vector, and the fuzzy probability distribution. The expert network includes multiple trained seasonal and solar term sub-networks, each of which uses the knowledge learned during training to predict park load data. This invention can deeply capture the nonlinear spatiotemporal relationships of multi-source data, significantly improving the prediction accuracy, convergence speed, and robustness of the model under complex and variable climate conditions.

This invention discloses a big data-based energy storage capacity time-sharing leasing control system and method, relating to the field of time-sharing leasing control technology. It involves marking electricity-consuming enterprises within the monitoring range, acquiring historical electricity consumption records of these enterprises based on historical electricity consumption data from a preset monitoring period, and calculating the basic electricity load forecast values ​​for each time period, taking into account historical influencing factors. For enterprises with only a single working label, the corresponding predicted energy storage demand is calculated according to the label type. For enterprises with more than one working label, the maximum value of the predicted demand corresponding to each label is taken as the final demand. For enterprises without labels, the basic load forecast value is used as the final demand. Time-sharing capacity is pre-allocated based on enterprise type priority, and the results are pushed to the enterprise to confirm leasing intentions. This invention, through a priority capacity pre-allocation mechanism based on enterprise importance, prioritizes electricity demand when capacity resources are limited.