4885 results about "Load forecasting" patented technology

The present application relates to the technical field of load prediction, and provides an electrical load prediction method and device based on spatial-temporal correlation. The method comprises: performing spatial-temporal modeling on the basis of historical electrical load data and historical environment data to obtain a multi-layer spatial-temporal model, wherein each layer model is a directed graph and represents spatial causality in the fixed time, and a directed edge between the layer models represents temporal causality; on the basis of an interpretable spatial-temporal attention transformer, performing feature extraction and priori knowledge fusion on the multi-layer spatial-temporal model to obtain spatial-temporal characteristics of an electrical load, wherein the interpretable spatial-temporal attention transformer comprises a spatial causal attention network, a temporal attention network, and a spatial dependence comparison module; and on the basis of the spatial-temporal characteristics of the electrical load, and the real-time electrical load data and the real-time environment data of a target electrical node, predicting by using a trained STEF-DHNet model to obtain the predicted electrical load data of the target electrical node. The present application can improve the prediction accuracy and the model applicability.

The invention relates to the technical field of power plant data processing, in particular to a virtual power plant load prediction and dynamic adjustment optimization system and method, and the system comprises a data collection module, a preprocessing module, a prediction module, an adjustment module and a verification module. The data acquisition module acquires real-time operation data and power market signals of distributed energy nodes; the preprocessing module performs standardization processing on the data through a quantum space-time alignment and anomaly reconstruction technology, and extracts strong correlation vectors of meteorological features and loads; the prediction module adopts an adaptive noise complete set empirical mode decomposition algorithm to separate a trend term, a periodic term and a residual component of a load sequence, and the adjustment module constructs a multi-target optimization model. Efficient aggregation of distributed resources, high-precision load prediction in a meteorological sudden change scene and cooperation of multi-market dynamic scheduling strategies are realized; and the clean energy consumption capability and the virtual power plant market response efficiency are improved.

The invention discloses a virtual power plant peak regulation optimization scheduling method and system, electronic equipment and a medium, and the method comprises the steps: obtaining multi-source data of distributed energy, energy storage equipment and a load side, and carrying out the cleaning and normalization processing; utilizing an LSTM neural network and an XGBoost model to carry out short-term and medium-and-long-term load prediction respectively; based on the prediction result and the resource state, generating a scheduling optimization strategy through a deep Q network (DQN) algorithm; according to an optimization strategy, distributed energy power generation, energy storage charging and discharging and load side response are dynamically managed, and multi-resource cooperative control is achieved; and through a feedback optimization module, DQN model parameters and strategy updating are automatically adjusted according to monitoring data, and the scheduling effect is optimized. The system comprises a data acquisition module, a data preprocessing module, a load prediction module, a scheduling decision module, a resource control module, and a monitoring and feedback optimization module. The method has the beneficial effects that the load prediction precision is improved, the resource scheduling is optimized, the operation cost is reduced, and the method is suitable for scheduling management of a multi-source energy system.

The invention relates to the technical field of electric power short-term load prediction, in particular to an industrial park short-term load prediction method and system based on a Diffusion-Transform-AR model. Acquiring non-stationary historical load data and meteorological data of the industrial park, and performing missing value processing on the data; performing correlation analysis on the data by using the maximum information coefficient, selecting meteorological data with relatively high correlation, and reprocessing time data; performing time sequence decomposition on the load data by using a seasonal trend decomposition method to obtain trend term, seasonal term and residual term features; constructing a diffusion Transform prediction model, and taking the residual term, the meteorological data and the time data as input prediction residual; constructing an autoregression prediction model and taking the trend term as an input prediction trend; according to the method, the advantages of various models are combined, time sequence features are deeply mined, the prediction precision and the model generalization ability are effectively improved, and the method is suitable for complex industrial park load prediction scenes.

The invention relates to the technical field of power systems, provides an association rule analysis-based power load prediction method and system, and aims to solve the problem of hidden fault response lag caused by lack of an equipment health state and load fluctuation dynamic coupling mechanism in the prior art. And the problem of load prediction compensation deviation caused by insufficient weight quantization of the fault propagation path is solved. The method comprises the following steps: generating equipment state data according to a vibration spectrum and an insulation aging index of power equipment; performing fusion analysis, generating an equipment health degree evaluation index, and establishing a dynamic influence model of the equipment abnormal event on the power grid load fluctuation according to the association rule; identifying potential abnormal equipment, performing logic mapping, and generating a fault propagation path weight; and dynamically adjusting according to the weight to obtain an adjusted power load predicted value. According to the technical scheme provided by the invention, equipment vibration and insulation aging data are fused, a health assessment and fault propagation model is constructed, and load prediction is dynamically corrected to prevent and control power grid risks.

The invention discloses a central air-conditioning system optimization control method oriented to building load prediction, which comprises the following steps: building a building load prediction model based on a long short-term memory network, combining real-time acquired building electrical load, indoor and outdoor temperature, humidity and equipment operation power data to accurately predict building cooling load requirements in a future time period, and on this basis, optimizing a central air-conditioning system. According to the method, time-of-use electricity price information is used for dynamically adjusting a cooling capacity distribution strategy of the central air conditioning system, so that the energy consumption and the operation cost of the system are reduced to the maximum extent while the load requirement is met. Compared with the prior art, the method has the advantages that the problems of insufficient load fluctuation response and insufficient energy consumption optimization are effectively solved, and the operation efficiency and economical efficiency of the central air-conditioning system of the building can be remarkably improved.

The invention discloses a power load prediction method based on multi-source heterogeneous data feature fusion, system equipment and a medium, and relates to the technical field of power system load prediction, and the method comprises the steps: obtaining a power load prediction data set, and carrying out the preprocessing; performing input feature processing on the static data, the graph data and the time series data based on a multi-layer perceptron, a graph convolutional network and a gating loop unit network, and fusing the processed features based on a feature splicing method to obtain comprehensive feature information; based on full-connection neural network construction, taking the comprehensive features as input, inputting training set data into a model, and performing iterative optimization according to a model effect to adjust model parameters and hyper-parameters; and based on the parameter optimization adjustment model, using the evaluated optimal model to predict the power load in the future time period. The method is superior to a traditional method in precision and stability, can better adapt to complex and changeable load prediction scenes, and has wide application prospects and remarkable actual value.

The invention discloses a power system load prediction method based on multivariate data fusion, and relates to the technical field of power systems. The problems that in the prior art, multi-source data fusion processing is complex, timeliness and processing efficiency are insufficient, and under the complex working condition, the response capacity to short-term emergencies is weak are solved. By collecting and preprocessing the multi-source data in real time, various factors influencing load changes can be comprehensively considered, the response capability to complex working conditions and emergencies is enhanced, the time sequence window is dynamically adjusted to improve the data validity and prediction performance, the accuracy and reliability of power system load prediction are improved, and the power system load prediction efficiency is improved. According to the method, multiple models are constructed and trained to obtain an accurate load prediction result, intelligent prediction and management of the load of the power system are realized through accurate load prediction, abnormal identification and processing are performed in combination with an emergency database, stable and reliable load prediction is provided for the power system, and thus power resource configuration is optimized.

The invention discloses a computing power load stabilizing and energy-saving management system based on a flywheel energy storage system, and belongs to the technical field of energy storage and management, and the system comprises a data collection module which is used for collecting and calculating a real-time current harmonic spectrum of a cluster and server rack vibration data to generate an electric energy quality characteristic spectrum; the prediction modeling module is used for establishing a prediction model output computing power demand dynamic distribution matrix based on multi-dimensional space-time correlation; the optimization control module is used for analyzing a peak load phase in the power demand dynamic distribution matrix to generate a dynamic coordination control strategy; and the power execution module is used for executing the dynamic coordination control strategy to adjust the phase parameter of the reverse electromotive force of the flywheel so as to offset the voltage harmonic component of the server bus. Harmonic phase compensation is realized through dynamic reverse electromotive force modulation, the calculation load prediction precision can be effectively improved, and the energy utilization efficiency and the equipment operation reliability are remarkably improved.

The invention provides an energy-saving control method and device for a central air-conditioning system and a medium, and the method comprises the steps: constructing a cooling load prediction model of the central air-conditioning system, and predicting the cooling load per hour of the next day; under the conditions that the hourly cooling load of the next day is met and the power utilization cost is the lowest, the cooling capacity in a fixed time period is distributed; and an energy efficiency model and a function model of the central air-conditioning system are constructed, and the operation strategy of the central air-conditioning system corresponding to the given cooling capacity is optimized with the purpose of minimizing the power consumption cost, so that the comprehensive energy consumption of the central air-conditioning system is minimized. According to the method, the cooling load prediction result is improved by comprehensively considering various cooling load influence factors and effectively capturing the long-term dependency relationship in historical cooling load data, the comprehensive energy consumption minimization and the power consumption cost minimization of the central air-conditioning system serve as targets, the operation strategy is optimized through simulation combination and energy efficiency comparison, and the energy efficiency of the central air-conditioning system is improved. Therefore, the requirement of a target environment for an efficient and energy-saving central air conditioning system is met.

The invention relates to the technical field of refrigeration equipment control, and discloses an artificial intelligence-based refrigeration equipment operation energy-saving control method and equipment, and the method comprises the following steps: collecting multi-dimensional data in a refrigeration environment in real time through a multi-sensor system, the data including environment variable data and equipment operation data; preprocessing the collected multi-sensor data, and generating a comprehensive environment variable through credibility evaluation and data fusion; carrying out load prediction on the current environment variable and the equipment operation state based on a deep learning model, and predicting a future refrigeration load demand; and according to the load prediction result, an optimization objective function is constructed, and an optimal control strategy of the refrigeration equipment is generated. According to the invention, through a deep learning load dynamic prediction technical scheme, a technical effect of high-precision prediction of a future refrigeration load demand is achieved, so that the change trend of an environment variable and an equipment operation state can be accurately captured, and the reliability of load prediction is improved.

The invention discloses a power load prediction method and system based on a meteorological prediction model, and relates to the technical field of load prediction, and the method comprises the following steps: obtaining historical power load data and corresponding multi-dimensional meteorological data, constructing a meteorological coupling effect analysis model, calculating a dynamic coupling weight matrix among the meteorological parameters through an attention mechanism network; performing feature fusion on the coupling weight matrix and the multi-dimensional meteorological data through a gating circulation unit to generate a space-time coupling feature vector; establishing a power load prediction model, and extracting a short-term fluctuation feature and a long-term trend feature of the power load from the space-time coupling feature vector based on a deep residual network architecture and a parallel convolution channel; the dynamic coupling weight matrix is updated based on prediction error back propagation, and adaptive adjustment of the meteorological parameter coupling relation is achieved. According to the method, dynamic fusion of the spatial-temporal characteristics is realized through the cross gating unit, so that the model can perform power load prediction by using spatial-temporal information, and the prediction precision is improved.

The invention relates to the technical field of smart power grids, and discloses a smart power grid power dispatching optimization method, which comprises the following steps of: firstly, acquiring power grid operation data, and processing data missing and noise problems by utilizing federal learning; and constructing a load prediction model through a dynamic time warping algorithm and a specific network. A multi-energy coupling scheduling model and a demand response game model are constructed, and a multi-time scale rolling optimization framework is established. And carrying out sensitivity analysis on scheduling parameters, designing a hierarchical collaborative optimization mechanism, and constructing a robust optimization model to cope with the power flow uncertainty. And integrating a scheduling instruction verification module, and deploying an online incremental learning mechanism. The method can effectively process data, accurately predict load, optimize multi-energy scheduling, guide demand response, deal with uncertainty, verify scheduling instructions and update the model in real time, improves the safety, reliability and economy of smart grid power scheduling, and realizes optimal configuration of power resources.

The invention provides a thermal power generating unit performance optimization control system and method, and the system comprises a data collection module which collects various real-time data in the operation process of a thermal power generating unit; the load prediction module is used for establishing a load prediction model and predicting a load change trend by comprehensively considering a historical load condition, a load change rate, a load change acceleration and an external interference factor; the optimization control module is used for dynamically adjusting fuel, an air door and air quantity according to a control strategy in combination with the load prediction result and the unit operation state; the actions of the regulating valve, the air door actuator and the water supply regulating valve are correspondingly controlled; the monitoring feedback module is used for realizing closed-loop control and self-correction; according to the method, a load prediction model based on the combination of time sequence analysis and a physical mechanism and a control strategy comprehensively considering multi-parameter collaborative optimization are utilized, so that quick response to load change, balance consideration of steam quality and combustion efficiency, self-adaptive stable operation under complex working conditions and improvement of overall performance are achieved.

The invention provides an air conditioner energy-saving control method and system based on a virtual power plant, and the method comprises the steps: determining an air conditioner energy-saving control model corresponding to a plurality of air conditioner devices through obtaining an operation parameter set and real-time power grid load data of the plurality of air conditioner devices in a target region, and carrying out the energy-saving control of the plurality of air conditioner devices according to the real-time power grid load data and load prediction data; and determining a power grid load fluctuation threshold value of the target area, dynamically optimizing the operation parameter set of the plurality of air conditioning equipment through a parameter optimization model, generating an optimized equipment operation parameter subset, and adjusting temperature set values and operation modes of the plurality of air conditioning equipment in a future time period. And generating an air conditioner group control strategy corresponding to a future time period, obtaining operation feedback data, updating weight parameters of the load prediction model according to a deviation value between actual power grid load data and load prediction data, calibrating an optimization rule of the parameter optimization model, and generating a calibrated model. According to the invention, the fine regulation and control capability and the energy coordination efficiency of the air conditioner group can be improved.

The invention relates to the technical field of power system dispatching, in particular to an emerging load power supply potential improvement method based on dynamic prediction, which comprises the following steps of: acquiring emerging load power consumption, meteorological environment and power grid schedulable resource data in a target area through an Internet of Things sensing terminal, and performing two-channel modeling to obtain a new load power supply potential improvement model; a deep space-time network is used to predict a load curve, a model is established to quantify resource regulation potential, a scheduling priority list and a capacity allocation strategy are generated by means of a matching rule base according to load fluctuation and resource evaluation results, an actual scheduling effect is fed back to the prediction model, parameters are corrected through error back propagation, a closed-loop optimization link is formed, and the scheduling efficiency is improved. The method improves load prediction accuracy and resource scheduling adaptability, is suitable for emerging load power supply optimization in a novel power system, and guarantees stable and efficient operation of a power grid.

The invention discloses an integrated energy system operation management method based on load prediction, and belongs to the technical field of energy system management, and the method specifically comprises the steps: collecting operation parameters of energy use equipment, the parameters comprising current waveform characteristics, surface temperature distribution and medium flow change; analyzing a time sequence change rule of the operation parameters, and extracting characteristic indexes related to equipment aging; establishing an energy consumption prediction correction model according to the characteristic indexes, and dynamically adjusting a theoretical energy consumption calculation value of the energy use equipment; inputting the corrected theoretical energy consumption calculation value into an energy distribution optimization model to generate a load distribution instruction of the energy network; after executing the load distribution instruction, comparing the deviation between the actual energy consumption and the correction theoretical value, and updating the parameter weight of the energy consumption prediction correction model; according to the invention, continuous and stable operation and optimal management of the integrated energy system in the equipment aging process are realized.

The invention relates to the technical field of electric power energy storage management, in particular to an independent electrochemical energy storage integrated management system, which comprises a load prediction analysis module, a charge and discharge response control module, a dynamic load threshold adjustment module, an energy storage response optimization module, a PID (Proportion Integration Differentiation) adjustment module and an abnormal state monitoring module. According to the method, the change rate and fluctuation amplitude of the load are adjusted in real time, the power demand change is accurately predicted, the stability and response speed of the system are improved, the fluctuation period and sudden change value of the load are analyzed, the demand trend is accurately predicted, the charging and discharging rate is adjusted, the actual demand is matched, the energy utilization efficiency is improved, the service life of equipment is prolonged, and the response threshold value is dynamically adjusted; the operation time interval is optimized, so that the system is maintained in an optimal state under variable loads, the energy consumption and the operation and maintenance cost are reduced, power output and equipment states are monitored in real time, abnormities are handled in time, safety and stability of the system are ensured, and the overall efficiency and reliability of the power system are remarkably improved.

The invention discloses a central air conditioner energy-saving optimization system based on load prediction, and relates to the technical field of air conditioner energy saving.According to the central air conditioner energy-saving optimization system based on load prediction, a VMDTransformerGRU hybrid deep learning model is introduced, through multi-source data fusion, the precision of cold load prediction is remarkably improved, and the prediction model can recognize the load change trend in advance. According to the improvement, the dynamic response capacity of the system is effectively enhanced, pre-regulation and control can be carried out before load changes, energy waste is reduced, a water chilling unit, a cooling tower and a pump are modeled to be independent intelligent agents, multi-agent deep reinforcement learning (MADRL) is adopted, and all the intelligent agents can autonomously learn and optimize control strategies of the intelligent agents. The intelligent agents cooperate with one another, efficient operation of the whole system is achieved through autonomous decision, and operation conflicts between devices are avoided. In this way, dynamic coordination and joint optimization among all the devices are achieved, invalid energy consumption is reduced, and the overall energy-saving effect and operation stability of the system are remarkably improved.

The invention discloses a power grid dispatching method and system adapting to power system requirements, and relates to the field of industrial big data, and the method comprises the following operation steps: S1, multi-source heterogeneous data collection and edge preprocessing; s2, knowledge graph construction and data fusion; s3, load prediction and renewable energy output prediction based on deep learning; s4, generating a dynamic optimization scheduling strategy; s5, carrying out security and credible execution on the data endowed by the block chain; and S6, real-time monitoring and closed-loop feedback optimization are carried out. According to the power grid scheduling method and system adapting to the power system demand, the scheduling method integrates edge calculation, block chain, deep learning and reinforcement learning, can realize multi-source data real-time processing, dynamic optimization strategy generation and data security and credibility, improves the power grid operation efficiency, stability and renewable energy consumption capability, and improves the power grid scheduling efficiency. And the dynamically optimized scheduling strategy can reduce the operation cost of the power grid, and can reduce carbon emission at the same time.

The embodiment of the invention discloses a park load prediction method and device based on multi-modal data and a medium, and relates to the technical field of micro-grids, and the method comprises the steps: collecting multi-modal micro-grid data through a multi-source heterogeneous sensor network disposed at a power distribution node, the multi-mode micro-grid data comprises electrical measurement data, environment monitoring data and equipment state data; constructing a dynamic space-time incidence matrix according to the multi-modal micro-grid data, and performing feature extraction based on the dynamic space-time incidence matrix through a space-time diagram convolutional network deployed at an edge calculation node to determine a space-time feature vector; and inputting the spatio-temporal feature vector into a time sequence fusion prediction model, outputting a load prediction result, and generating an optical storage cooperative scheduling instruction according to the load prediction result, the load prediction result including a load prediction value and a confidence interval.

The invention discloses a control method and device of a building cold and heat source system self-adaptive to load requirements. The control method comprises the following steps: acquiring historical operation data, physical characteristics and equipment system data of a building, and establishing a building load prediction model; generating the hourly load curve, and predicting the cooling or heating demand of each time period in the future according to the curve; dynamically adjusting the refrigerating or heating load demand, and generating a corresponding cooling or heating curve; and constructing a machine learning algorithm based on an equipment mechanism model, optimizing a control strategy of the cold and heat source system based on the historical operation data of the equipment, and outputting control parameters of the equipment. According to the technical scheme, the load of the building cold and heat source system is intelligently predicted and adjusted, the energy efficiency of the system is improved, invalid energy consumption is reduced, energy utilization is optimized, control parameter optimization is achieved through a machine learning algorithm, efficient operation is ensured, indoor comfort is kept, and the purpose of saving energy is achieved.

The invention discloses a distributed new energy storage optimal configuration method and system for a power distribution network, and the method comprises the following steps: building a load prediction model through employing a probability discretization method according to historical power utilization data and real-time weather data; according to load demand data output by the load prediction model, a multi-target optimization model is constructed by taking maximization of distributed new energy power generation utilization rate, minimization of distributed new energy storage energy loss and minimization of operation cost of energy storage and scheduling as optimization targets; according to the multi-objective optimization model, adjusting weight distribution of the three optimization objectives, gradually optimizing power distribution between new energy power generation and energy storage, and generating an optimal new energy complementation strategy; charging and discharging are adjusted according to the optimal new energy complementation strategy, the health state, the battery temperature, the number of charging and discharging cycles, the environment temperature and the load demand change of energy storage equipment are collected in real time, and new energy power generation and energy storage configuration is optimized by dynamically comparing multiple indexes under different energy storage configurations in combination with power grid frequency fluctuation.

The invention discloses a power grid load prediction and scheduling optimization system based on artificial intelligence, particularly relates to the technical field of power system automation, and solves the technical problems of low power grid load prediction precision, poor scheduling strategy robustness and insufficient source grid load storage coordination in the prior art. Multi-source heterogeneous data space-time alignment is realized by constructing a data acquisition layer based on edge calculation, a load prediction result is generated by adopting an AI prediction module fused by a graph convolutional network and an attention mechanism, and a source-network-load-storage collaborative scheduling scheme is generated through a multi-target risk hedging optimization algorithm. And closed-loop optimization is realized by using digital twinborn pre-check and incremental learning. And finally, the load prediction accuracy, the scheduling decision reliability and the system adaptive capability in the new energy access environment are improved.

The invention discloses an intelligent linkage control management system based on air conditioner application, and relates to the technical field of intelligent control and energy management, and the system comprises a load prediction module which obtains an air conditioner load demand prediction value through an intelligent load prediction model based on preprocessed environment energy consumption data; the optimization scheduling module is used for optimizing the current air conditioner operation mode through the building energy management system based on the air conditioner load demand prediction value, and obtaining an optimized air conditioner operation scheduling instruction; the intelligent adjusting module is used for intelligently adjusting the operation parameters of the air conditioner system based on the optimized air conditioner operation scheduling instruction, adaptively adjusting the energy efficiency of the air conditioner and obtaining operation feedback data; and the fault management module is used for identifying potential fault risks of the air conditioner and formulating fault risk prevention measures through a time sequence analysis method based on the operation feedback data. According to the invention, through innovation of an intelligent load prediction and fault prevention mechanism, the energy utilization efficiency and the system stability are effectively improved.

The invention discloses an operation regulation and control method and system for a distributed new energy photovoltaic power station, and relates to the technical field of new energy, and the method comprises the steps: recognizing a power station, building a monitoring system, collecting and monitoring key equipment data through the system, so as to obtain historical and real-time operation data, and predicting photovoltaic output and load based on the data to obtain predicted values, calculating power deviation, inputting the deviation into an optimal scheduling model to obtain a strategy, and adaptively regulating and controlling the operation state of the power station according to the strategy. The technical problems of unbalanced power generation and power utilization, low operation efficiency, poor stability, high cost and poor power grid compatibility caused by the fact that efficient and self-adaptive optimal scheduling cannot be performed according to accurate photovoltaic output and load prediction in the prior art are solved, accurate balance of power generation and power utilization is achieved, and the power generation efficiency and the power utilization efficiency are improved. The operation efficiency and stability of the power station are remarkably improved, the operation cost is reduced, and the compatibility with a power grid is enhanced.

The invention relates to the technical field of power distribution network operation management, and discloses a power distribution network capacity-increase-free access system based on dynamic coordination. The system comprises a multi-source data acquisition and fusion unit which acquires multi-source data and fuses the multi-source data to generate dynamic power grid state characteristics; the multi-time-scale collaborative optimization architecture unit is used for constructing a layered architecture to realize optimization of different time scales; the space-time load prediction unit is used for predicting load distribution based on the space-time diagram model; and the multi-target dynamic optimization solving unit is used for constructing a model to solve an optimal scheduling instruction. In addition, a layered communication protocol unit is arranged to guarantee communication, and a voltage stability cooperative control unit and a distributed energy storage cooperative control unit are arranged to improve the stability of the power grid. Through cooperation of multiple technologies, capacity-increase-free access of the power distribution network is realized, the operation efficiency is improved, the loss is reduced, the voltage stability is ensured, and challenges brought by distributed energy and electric vehicle charging are effectively dealt with.

The invention provides a machine room control method and system based on communication between an air conditioner terminal and a water chilling unit, and the method comprises the steps: obtaining three-dimensional data and equipment attribute data of a machine room and accessory equipment of the machine room through a server, and building a digital twinborn model of the machine room; the server establishes a dynamic load prediction model and sends the dynamic load prediction model to a first controller of the water chilling unit; the intelligent sensor collects environment data and equipment data; a second controller of the air conditioner terminal obtains the environment data and sends the environment data to the first controller; the first controller predicts the cold load change according to the dynamic load prediction model and the historical data and adjusts the operation strategy of the air conditioning system; the server simulates a machine room environment and an equipment operation condition in real time by using a digital twin model, generates a control strategy in combination with a self-adaptive control strategy after evaluating a result, and sends the control strategy to the first controller; and the first controller dynamically adjusts operation parameters of the air conditioning system according to the strategy. By means of the scheme, efficient, intelligent and accurate real-time monitoring and control over the energy efficiency of the air conditioning system can be achieved.

The invention relates to the technical field of intelligent scheduling and prediction of a power system, and discloses a power grid load prediction system and method based on machine learning, and the system comprises a multi-modal data processing and fusion module which is used for collecting multi-modal data from power grid load data, meteorological data, economic data and social events, cleaning and feature extraction processing are carried out on the multi-modal data, weighted fusion is carried out on different data features through a self-attention mechanism, and standardized feature input is provided for the hybrid prediction model module. Through a multi-modal data processing and fusion module, deep fusion is carried out on multi-modal data of power grid load data, meteorological data, economic data and social events, and a self-attention mechanism is introduced to carry out weighting processing on different data features, so that dynamic weighting is realized, key features are screened, and the accuracy of data processing is improved. The accuracy of load prediction and the comprehensive utilization capability of multi-source information are effectively improved, so that the sensitivity of the prediction model to complex environmental factors is remarkably enhanced.

The invention provides a power demand dynamic prediction method based on a multi-scale hybrid architecture. The power demand dynamic prediction method comprises the following steps: step 1, acquiring multi-source data and preprocessing the multi-source data; step 2, constructing a dynamic feature project; step 3, constructing a hybrid prediction model; 4, executing a staged dynamic training strategy; and 5, verifying the model, and deploying and applying the model. According to the invention, by introducing the temperature sensitivity piecewise nonlinear coding and the holiday multi-dimensional dynamic attenuation model, the prediction precision in an extreme temperature scene is significantly improved, and the high-temperature daily load prediction error and the spring festival holiday error are reduced. In addition, through lightweight design and staged dynamic strategy staged extension of the input length, the training convergence speed is obviously improved.