263 results about "Power system scheduling" patented technology

The invention discloses a source network load storage oriented regional collaborative scheduling method and system, and relates to the technical field related to power resource scheduling, and the method comprises the steps: traversing target region source network load storage to carry out real-time data collection, and constructing a source network load storage multi-dimensional data set to carry out multi-scale layered optimization; operation constraint conditions are set, limitation is carried out in combination with a power grid topological structure, a cooperative scheduling instruction set is determined, cooperative control feedback is carried out, and regional power scheduling feedback parameters are generated; and multi-channel dynamic correction is carried out, and a cooperative scheduling optimization strategy is generated to carry out energy balance cooperative scheduling on the source network load storage in the target area. The technical problems that in the prior art, intermittent fluctuation of new energy is difficult to consume, the collaboration of all links of source network load storage is poor, the dispatching flexibility of a power system is insufficient, and the resource configuration efficiency is low are solved. The technical effects of optimizing regional energy resource allocation and improving the clean energy consumption level, the power grid operation stability and the source grid load storage cooperative response capability are achieved.

The invention discloses a power grid photovoltaic output and load sequence modeling method, system and device and a storage medium, and the method comprises the steps: comprehensively utilizing the multi-scale feature extraction capability of a time-frequency decomposition technology, the time sequence dependence modeling capability of a long and short-term memory network, and the global hyper-parameter optimization capability of a Bayesian optimization algorithm; and carrying out collaborative modeling and prediction on the photovoltaic output and the power load under a unified framework. By introducing a source load time-delay correlation analysis and probability interval construction mechanism, point prediction results and uncertainty intervals of photovoltaic, load and net load can be output at the same time, and a set of source load integrated prediction system with high prediction precision, strong robustness and reliable interval characterization capability is constructed. The method can improve the precision and reliability of photovoltaic power and load prediction, also can reduce the risk in power system scheduling, optimizes the energy storage configuration strategy, and especially has wide popularization potential and application prospects in the scenes of new energy grid-connected operation, intelligent micro-grid and virtual power plant management and the like.

The invention provides a global adaptive power supply adjustment method and system under the condition of local load increase, and relates to the technical field of power system dispatching. The method comprises the following steps: acquiring regional power grid topology, equipment parameters and real-time operation data; historical power consumption data of classified users in recent three years are collected and preprocessed; generating a reference load curve based on real-time data and a historical rule, calculating a deviation ratio, predicting a load in future 2 hours, and outputting a risk level; the regional energy storage equipment is called to dynamically adjust the discharge power when the risk is medium or low, and the adjacent regional resources are evaluated and cross-regional line switching and power supply support are executed when the risk is high or the energy storage is insufficient. Through layered adjustment and a dynamic response mechanism, rapid identification and accurate control of sudden load increase are realized, power supply stability and resource utilization efficiency are improved, and the method is suitable for local load fluctuation response of scenes such as residential areas and industrial parks.

The invention relates to the technical field of power system dispatching automation control, and discloses an AGC load distribution control method and system based on a multi-region greedy hydropower station, and the method comprises the steps: generating dynamic boundary thresholds of a high-efficiency region, a medium-efficiency region and a low-efficiency region based on a real-time performance curve and a dispatching set value of a hydroelectric generating set; judging whether to start a cross-interval distribution strategy or not by utilizing the generated dynamic boundary threshold value; when the number of the units participating in distribution is lower than a set scale, an enumeration algorithm is automatically switched to carry out global optimization, and generation of an optimal solution under a small-scale unit combination is ensured; calculating a deviation value between the actual total load and a target value based on the obtained cross-interval distribution strategy; and preferentially adjusting the output of the high-efficiency area unit through a dynamic compensation mechanism. The system comprises an operation interval dynamic calibration module, a region selection and load pre-distribution module and a deviation compensation and adaptive adjustment module. The power grid frequency adjusting precision and the overall operation efficiency of the hydroelectric system are remarkably improved.

The invention discloses a hybrid sequential network power consumption prediction method for reinforcement learning dynamic calibration, and the method comprises the steps: collecting multivariable power consumption time sequence data, completing the data preprocessing through resampling, feature engineering, normalization and sliding window technologies, generating a supervised learning sample set, and dividing the supervised learning sample set into a training set, a verification set and a test set; constructing a hybrid prediction model comprising a dynamic capture module, a long-term dependence modeling module, a regression prediction module and a reinforcement learning dynamic fine tuning module; training and optimizing by adopting a staged training strategy to obtain a hybrid prediction model; multivariable power consumption time sequence data are collected in real time and preprocessed, the preprocessed data serve as input, real-time prediction of future total consumption is conducted through the mixed prediction model, and a final prediction result after dynamic fine adjustment is output. According to the method, accurate and efficient prediction of the power grid load can be realized, and a reliable technical solution can be provided for power system scheduling optimization, demand side management, market transaction and other scenes.

The invention relates to the technical field of power system dispatching, in particular to a distributed storage system for source-network-load-storage collaborative dispatching, which comprises a distributed data acquisition module, a regional collaborative sensing module, a distributed energy storage evaluation module, a dispatching instruction generation module and a dispatching instruction distribution module, wherein the distributed data acquisition module is used for acquiring real-time operation data of source, network, load and storage nodes in parallel; the regional collaborative sensing module is used for calculating a real-time power deviation value; the distributed energy storage evaluation module is used for generating an energy storage adjustment instruction set; and the scheduling instruction generation module is used for generating a final refined scheduling instruction. According to the invention, through constructing a cooperative scheduling mechanism of the source network load storage full chain, accurate identification of power deviation and optimal distribution of energy storage resources are realized, and scheduling response efficiency and operation stability of the system are significantly improved.

The invention discloses a new energy consumption-oriented power grid digital twinborn optimization scheduling method and system, and relates to the technical field of power system scheduling, and the method comprises the steps: constructing a digital twinborn model of a power grid; embedding a multi-time-space nested prediction model, establishing new energy output fluctuation characteristics under multiple time scales, performing dynamic evolution correction on each virtual power grid evolution branch according to the new energy output fluctuation characteristics, and establishing an uncertainty evolution trajectory; performing a multi-agent game in the sliding time window; and establishing a scheduling strategy by using a multi-agent game result, and performing optimal scheduling management according to the scheduling strategy. According to the method, the technical problems of low new energy consumption efficiency and high power grid operation safety risk caused by difficulty in accurately coping with the uncertainty of new energy output and lack of a multi-subject benefit cooperation mechanism in new energy consumption scheduling in the prior art are solved, and the purposes of realizing new energy consumption optimization scheduling and improving the power grid operation safety risk are achieved. And the new energy consumption efficiency and the power grid operation safety are improved.

The invention provides an electric power system source network load high-order uncertainty cooperative scheduling method integrating scene generation and schedulable flexibility, and relates to the technical field of electric power system scheduling optimization. Comprising the following steps: generating a typical working condition by using a joint probability model, and forming an uncertainty scene data set through Monte Carlo extension; determining robustness set data according to scheduling feasibility and risk constraints; constructing a unified flexibility model, correcting the boundary of the unified flexibility model, and forming a deliverable flexibility set and a rescheduling strategy; and establishing a source-network-load multi-side collaborative optimization model, and solving an optimal operation scheme in combination with multi-scene disturbance, thereby realizing risk-controllable and flexible coordinated scheduling of multi-type adjustable resources. The problems that in the prior art, source network load high-order uncertainty description of an electric power system is insufficient, deliverable modeling is not uniform, and multi-dimensional cooperative scheduling lacks closed-loop connection are solved.

The invention provides a cogeneration unit peak regulation capability evaluation method, which comprises the following steps: firstly, constructing a thermodynamic system mathematical model, solving system coupling parameters under a deep peak regulation working condition through a multi-layer iterative loop architecture, and generating a multi-working-condition data set which covers a full working condition and meets physical consistency; secondly, deducing a linear qualitative relation equation of the power, the main steam flow and the heat supply steam extraction flow, preprocessing a data set, training a pure data driving model as a performance reference, introducing the linear qualitative relation equation as a physical constraint item into a loss function, and constructing and training a PINN; and finally, the rated boundary condition of the peak regulation capacity is determined, a thermal characteristic curve is obtained through PINN simulation, the peak regulation capacity range and the real-time peak regulation margin under different heat supply requirements are calculated, and the result is transmitted to a power system dispatching platform. According to the method, the all-working-condition peak regulation evaluation precision and reliability can be improved, and the flexibility and universality of the model are enhanced.

The invention discloses a power distribution network scheduling method and device based on approximate dynamic programming and load aggregation and electronic equipment, and belongs to the technical field of power system scheduling, and the method comprises the steps: obtaining parameters such as the line, cost, topology and load prediction of a power distribution network, and the operation constraint of distributed loads; constructing an aggregation feasible region of the adjustable load cluster; establishing an optimal dispatching model of the power distribution network by taking minimization of the operation cost as a target; reconstructing the power distribution network optimization scheduling model into a Markov decision model; and solving the Markov decision model period by period under the aggregation feasible region and the physical constraint of the power grid. And in each time period, constructing a post-decision state value function by using the current state and a preset fixed slope parameter, solving an optimal scheduling scheme of the current time period, updating the state of the next time period, generating an all-day optimal scheduling scheme, and executing the all-day optimal scheduling scheme. By implementing the method and the device, the technical problem of poor practical operability of dispatching of the power distribution network in the prior art can be solved.

The invention discloses a main-distribution-micro-cooperation fault processing method and system, and belongs to the technical field of power system dispatching and fault handling, and the method comprises the steps: carrying out the fault perception and risk analysis of main-distribution-micro-cooperation after a fault is triggered, and generating an event fault event; through grid identification and power-loss monitoring, analyzing a power-loss condition and determining a power-loss influence level; generating an alarm event based on the event fault event and the grid identification and power loss monitoring result; and according to the alarm event and the power loss condition analysis result, generating a main-distribution-micro cooperative processing scheme, and supporting visual layered display. According to the method, the defects of fragmentation, low cooperation efficiency and scattered disposal schemes of main and distribution micro fault disposal in the prior art are overcome, the fault positioning accuracy, provincial and regional cooperation efficiency and the scientificity of the disposal schemes are remarkably improved, and the influence of faults on safe operation of a power grid and power supply of users is effectively reduced; the method is suitable for a main and distribution micro-grid fault handling scene containing distributed resources.

The invention relates to the technical field of power system dispatching and control, in particular to a fire coal-fused salt energy storage dynamic characteristic decoupling method based on multi-time scale control, which comprises the following steps: constructing a multi-physical field coupling model of a coal-fired unit and a fused salt energy storage system; dividing a minute-level adjustment scale and a second-level response scale based on the coupling model; aiming at the minute-level regulation scale, designing a power distribution strategy based on model prediction control, matching a load trend in a power grid, and reserving a power regulation margin for second-level response; aiming at a second-level response scale, designing a compensation strategy based on sliding mode control, and correcting power deviation to track a real-time instruction of a power grid; and associating the minute-level strategy with the second-level strategy through a dynamic weight coefficient to form a closed loop to cooperatively complete dynamic characteristic decoupling. The problem of power supply and demand mismatch of the coal-fired unit and the fused salt energy storage system in coupling operation can be solved, the response speed, the adjustment precision and the operation stability of the system are improved, and technical support is provided for efficient operation of a novel power system.

The invention provides a virtual power plant demand side response and load management method and system, and relates to the technical field of power system scheduling, and the method comprises the steps: obtaining a demand side response request and a historical response track of each distributed energy source; performing time-frequency domain decomposition on the historical response trajectory, extracting a transient component and a steady component, constructing a dynamic response model containing the current output power and the load deviation cumulant, and calculating load deviation distribution through Kalman filtering; a Markov decision process is constructed based on load deviation distribution, an optimal power distribution strategy is solved by adopting a value iteration algorithm, and a response execution scheme is generated; collecting a power change sequence after execution of each distributed energy source, converting the power change sequence into a load recovery demand vector, constructing a recovery capacity constraint cone based on the current available adjustment margin of the power grid, and solving an optimal projection point; and performing hierarchical division according to the optimal projection point to obtain a staged recovery sequence. According to the method, the accuracy of virtual power plant demand response and the coordination of load recovery are improved, and the power grid regulation pressure is reduced.

The invention discloses an industrial virtual power plant flexible resource capacity aggregation method based on artificial intelligence, and belongs to the field of power system scheduling and energy management, and the method comprises the following steps: S1, generating a dynamic model parameter containing a constraint boundary and an output prediction curve; s2, scoring the generated dynamic model parameters containing the constraint boundary and the output prediction curve result, and performing priority ranking according to the scoring result; s3, based on a priority ranking result, generating an optimal flexibility resource aggregation capacity allocation scheme; s4, determining a feasible declaration scheme of the industrial virtual power plant; and S5, dynamically updating the steps S2, S3 and S4 by using the real-time operation data. By the adoption of the industrial virtual power plant flexible resource capacity aggregation method based on artificial intelligence, accurate quantification and efficient aggregation of the industrial virtual power plant flexible resource capacity are achieved by fusing the AI algorithm of multi-dimensional feature extraction and industrial scene constraint modeling.

The invention discloses a power grid flexibility potential prediction and evaluation system based on deep learning, and belongs to the technical field of power system dispatching and artificial intelligence crossing. The system comprises a hardware layer, a software layer and an application layer, wherein the hardware layer realizes data acquisition and calculation support through a multi-source terminal, an edge node and a GPU cluster; in the software layer, the data preprocessing subsystem processes multi-source heterogeneous data by adopting an adaptive feature selection mechanism, the space-time fusion prediction subsystem improves the prediction precision based on GNN-Attention-LSTM dual-mode architecture in combination with transfer learning, and the multi-dimensional intelligent evaluation subsystem dynamically optimizes the index weight through an MLP model. And the reinforcement learning decision subsystem generates a scheduling strategy by using a PPO algorithm. According to the method, the problems of high prediction error, subjective evaluation and disjoint decision in the traditional technology are solved, the short-term prediction MAE is less than or equal to 7.8%, the renewable energy consumption rate is greatly improved, the adjustment cost is reduced, and the method is suitable for power grid dispatching optimization under high-proportion new energy access.

The invention relates to a power system scheduling method based on a wind and light output scene and a virtual energy storage model, and the method comprises the steps: sampling and generating day-ahead wind and light data, and carrying out the scene reduction to obtain a typical wind and light output scene; constructing a virtual energy storage model based on the electric vehicle, and introducing the deviation electric quantity to correct the virtual energy storage model; constructing a day-ahead stage multi-agent collaborative optimization model taking the minimum operation cost of an energy storage operator and the minimum energy consumption cost of the park integrated energy system as optimization objectives; and solving the day-ahead stage multi-subject collaborative optimization model by adopting an improved sparrow search optimization algorithm to obtain an optimal solution, namely an optimal scheduling scheme. A typical wind and light output scene set is constructed, the influence of long-time scale prediction deviation on a scheduling plan is reduced, and the robustness of a power system is enhanced; an electric vehicle cluster is modeled as virtual energy storage, the flexible adjustment potential of the electric vehicle cluster is mined through load-storage dual characteristics, a space-time complementary mechanism is formed with physical energy storage, and the overall energy efficiency of the system is improved.

The invention relates to the field of electric power early warning, in particular to an intelligent control method and system for a hydrogen generator set. A data analysis module calculates a power grid structure integrity index and an electric energy quality stability index based on a power transmission line switch trip number, a relay protection action signal number, a system active power deviation value and a key node voltage out-of-limit value, and further weighting is carried out to obtain a comprehensive characterization value; the operation condition division module divides stable or disturbance operation conditions; the strategy scheduling module adjusts the acquisition interval of the next time period according to the operation condition; and the alarm module judges whether to give out early warning or not according to the comparison of the comprehensive characterization value and an alarm threshold value. According to the method, the monitoring frequency is reduced under the stable working condition to save resources, the monitoring density is improved under the disturbance working condition to enhance the sensing capability, and the intelligent operation and maintenance level and the early warning accuracy of the dispatching service of the power system are improved.

The invention provides a cascade water-light complementary scheduling method for a water-light load multi-uncertainty distribution network system. The method belongs to the field of power system dispatching optimization. A cascade hydropower station power generation model, a photovoltaic power generation model and a distribution network system model are combined to establish a cascade water-light complementary power generation joint optimization scheduling model; converting the cascade water-light complementary power generation joint optimization scheduling model into a two-stage robust optimization model, and performing linearization processing; decomposing the two-stage robust optimization model into a main problem and a sub-problem; loosening the decomposed sub-problem and constructing a new objective function to obtain an infeasible test sub-problem; and iteratively solving the main problem, the sub-problems and the infeasible test sub-problems to obtain a day-ahead scheduling plan. According to the method, multiple uncertainties of natural incoming water and photovoltaic output of the small hydropower station are considered, the power generation flexibility of the cascade hydropower station is excavated to improve the photovoltaic absorption capacity, and efficient solution of the model is realized by providing a linearization method of the hydropower model and a sub-problem feasibility test method.

The invention discloses a load-side resource demand response scheduling method, system, equipment and medium under a master-slave game, and relates to the technical field of power system scheduling and demand side management, and the method comprises the steps: collecting the normal operation data of a park, generating a load and photovoltaic curve, analyzing and dividing price-type and excitation-type loads, and constructing a model for describing the dynamic response of the loads. The method comprises the following steps: establishing an operator-dominated and load-followed master-slave game double-layer optimization model, inputting unit price, carrying out iterative solution to obtain an optimal price, implementing demand response, calculating energy purchase cost, solving lower-layer optimization based on actual data and cost, and determining energy storage output power. According to the method, a complete technical chain from load classification, behavior modeling, master-slave game optimization to hierarchical solution is constructed, so that accurate perception, behavior prediction and cooperative scheduling of multi-load-side resources are realized, the interests of multiple subjects are effectively balanced on the premise of ensuring safe operation of a power grid, and the flexibility of the power distribution network is remarkably improved.

The invention relates to the technical field of power system dispatching, and discloses a power load dispatching optimization method considering dynamic response reliability, which comprises the following steps: collecting and classifying user historical and real-time load data; based on the response proportion, the peak clipping contribution degree and the forgetting factor, establishing a reliability coefficient model for interruptible, transferable and reducible loads; in combination with real-time response performance, a user dynamic response reliability calculation model is constructed, and quantitative indexes are obtained; establishing a scheduling optimization model which takes optimal response reliability and lowest scheduling cost as multiple targets and considers capacity and capability constraints; and finally, solving by adopting a hybrid intelligent algorithm to obtain an optimal scheduling scheme. According to the method, the uncertainty of user response is dynamically quantified, and the reliability index is introduced into a decision, so that economy can be considered on the premise of ensuring scheduling reliability, priority calling of high-reliability resources is realized, and the flexibility and stability of power grid operation are improved.

The invention discloses a meta reinforcement learning method and system oriented to source network load storage cooperation, and belongs to the technical field of power system dispatching optimization. The method comprises the following steps: constructing an optimization objective function containing economic cost and network loss cost, defining various constraints, and converting the constraints into a mixed integer second-order cone optimization model through second-order cone relaxation; determining a reinforcement learning action space, a state space and a reward function, and decoupling into a single-section optimization problem; constructing a source load scene and a topological structure task subset, and combining the source load scene and the topological structure task subset to form multiple subtasks; performing model training based on a graph attention neural network, a Transform architecture and an SAC algorithm; interaction-free state space embedding vector calculation is realized through KL divergence constraint; and selecting two meta-test forms to output an optimization decision result. According to the method, the adaptive capacity of the model to the source load and topological change is improved, the strategy safety and economy are ensured, and the method is suitable for novel power system source-network-load-storage collaborative optimization scheduling.

The invention provides an emergency power supply system multi-source cooperative scheduling optimization method based on deep learning, and relates to the technical field of power system scheduling, and the method comprises the steps: obtaining the real-time operation states and load demand data of a plurality of power supplies; calculating supply and demand power constraints and operation constraints to obtain an initial scheduling sequence; performing wavelet transform decomposition on the real-time operation state according to the time scale, constructing an energy-time characteristic spectrum, calculating a comprehensive complementary coefficient through complementary discrimination space mapping, and generating an optimized weight matrix; correcting the initial scheduling sequence according to the optimized weight matrix to form a cooperative scheduling instruction; constructing a fault evolution path map based on the fault sample set, calculating risk assessment values and generating a priority sequence; and constructing an execution time gradient sequence, and generating and executing a switching control sequence. According to the invention, the response speed of emergency power supply and the accuracy of multi-source cooperative scheduling are improved, and the system fault risk is reduced.

This invention relates to the field of power system dispatch and control technology, specifically to a multi-day optimal dispatch method for compressed air storage (CASP) energy storage systems (CASPs) oriented towards continuous extreme scenarios. This method proposes operational constraints for CASPs with unit-like combinations, characterizing practical operational constraints such as minimum continuous operating time, start-stop limits, and waiting times for charge / discharge state transitions. Furthermore, this method proposes an optimal dispatch framework for CASPs on a weekly timescale, addressing various types of extreme scenarios. In addition, the method employs a Benders decomposition-based algorithm and linearization techniques to solve the proposed large-scale mixed linear integer programming problem. This invention improves the engineering practicality and dispatch decision feasibility of CASP operation models, fully explores the long-term energy storage value of CASP in addressing energy imbalances in continuous extreme scenarios, and meets the large-scale dispatch needs of power systems.

The invention relates to the technical field of power system dispatching, and discloses a supply and demand power balance measuring and calculating method based on flexible adjustment resource optimization. The method comprises the following steps: acquiring real-time load data, predicting renewable energy output, flexibly adjusting resource parameters and adaptively optimizing a parameter structure, performing time reference correction and channel standardization processing, and generating an acquisition preparation packet structure; building a flexible adjustment resource model, modeling an energy storage state boundary and an interruptible load response time period, carrying out optimal configuration and constraint solution of short-time, medium-time and long-time multi-layer rolling windows, and generating an optimal configuration result set with mutual exclusion and cooperation rule annotations; and carrying out supply and demand deviation analysis, prediction correction and resource scheme arrangement to realize self-adaptive balance measurement and calculation of supply and demand power. According to the invention, through multi-time scale rolling optimization and flexible adjustment resource collaborative modeling, dynamic response and accurate matching of supply and demand power calculation are realized, and the stability and adjustment efficiency of power grid operation are improved.

The invention discloses a multi-time scale coupled source-storage collaborative economic dispatching method and system, relates to the field of power system dispatching, and solves the problem of low dispatching economy of a power system dispatching scheme. According to the day-ahead scheduling optimization in the embodiment of the invention, the hidden cost of deep peak regulation and high-frequency charging and discharging can be quantified through the deep peak regulation coal consumption function of the thermal power generating unit and the charging and discharging depth-service life attenuation mapping table of the energy storage equipment; the intra-day rolling optimization dynamically adjusts the energy storage behavior based on the real-time electricity price, so that the scheduling plan can be matched with the market fluctuation, and the energy storage income is improved; by combining the two points, the operation characteristics of the power system under the complex working condition can be reflected more accurately, and the scheduling economy of the finally obtained optimal scheduling scheme is improved.

The application discloses an algorithm electricity collaborative demand response bidding decision method and system based on large model training tasks, and relates to the technical field of power system dispatching. First, the power grid invitation signal is received and analyzed to construct an invitation signal feature set. Then, the state of the intelligent algorithm center is collected, and the computing power shadow price of each large model training task is calculated; the state space and action space are constructed according to the distribution characteristics of the invitation signal feature set and the computing power shadow price, the PPO algorithm is used to solve the optimal bid, and the optimal bidding strategy is output. If the bidding is successful, the adjusted task set is selected in the order from low to high according to the computing power shadow price of each large model training task, and hierarchical control is performed as needed. Through the hierarchical response mechanism and cost consideration, when interruption must be performed, the large model training task that has just completed the "checkpoint saving" is automatically locked and suspended through the computing power shadow price sorting, the impact on the training progress is minimized, and the safety of the model parameters is ensured.

The application relates to the technical field of power system dispatching, and discloses an intra-day forward-looking dispatching fast solving method considering large-scale new energy cluster power generation fluctuation, which comprises the following steps: S1: new energy cluster space-time fluctuation scene generation based on neuron cellular automata, S2: power grid dynamic security domain definition and simplification based on physical information neural network, S3: dispatching fast optimization solving based on model prediction path integral, and S4: dispatching scheme dynamic elasticity and stability evaluation based on the theory of Koopman operator. The new energy cluster space-time fluctuation scene generation method based on neuron cellular automata can effectively generate a space-time scene reflecting the large-scale new energy cluster power generation fluctuation, support uncertainty analysis, has the advantages of high calculation efficiency and scene authenticity, and solves the problem that the traditional statistical model ignores space-time coupling, thereby causing inaccurate scene generation.

The invention relates to the technical field of electric power system operation, in particular to an electric power system dispatching method and device considering natural gas power generation quick start-stop capacity and offshore wind power uncertainty, and the method comprises the steps: obtaining historical prediction data and historical actual data of power generation of an offshore wind turbine generator; generating a prediction error probability distribution table of power generation of the offshore wind turbine generator according to the historical prediction data and the historical actual data, and querying the prediction error probability distribution table according to the actual prediction data of power generation of the offshore wind turbine generator to obtain power generation scene data of the offshore wind turbine generator; the power generation constraint data of the traditional unit and the power generation scene data of the offshore wind turbine generator are input into a dispatching model of the power system, the dispatching model outputs a dispatching strategy of the power system, and at least one of the thermal power unit, the natural gas unit and the offshore wind turbine generator is dispatched based on the dispatching strategy. Therefore, the problems that the probability scene of the generated output cannot be accurately estimated, the offshore wind power is uncertain, the scheduling strategy is inflexible and the like are solved.

This invention discloses a day-ahead power prediction method for a single photovoltaic power plant based on a spatiotemporal graph model, belonging to the field of new energy power prediction technology. The method includes: constructing the power prediction task of a single photovoltaic power plant as a graph learning problem; designing a unified spatiotemporal graph structure containing three types of nodes: historical observation nodes integrating historical meteorological and power data, future meteorological forecast nodes, and future power nodes; explicitly fusing the dynamic correlation between meteorological and power data in the temporal and spatial dimensions through this structure; and introducing a graph attention network and a dynamic multi-head attention mechanism to aggregate node information and extract features to capture nonlinear dependencies across time and space, ultimately achieving high-precision day-ahead power prediction. This invention can effectively model the complex "meteorological-power" spatiotemporal coupling mechanism in photovoltaic output, significantly improving prediction accuracy and stability, and providing a reliable basis for power system dispatch and consumption.

The invention relates to a risk early warning method and device based on new energy output fluctuation and a storage medium, which are applied to the technical field of power system dispatching, and comprise the following steps: obtaining the output deviation level of a generator set under a dispatching output plan through historical data, quantifying behavior deviation and incorporating the behavior deviation into a risk model, the risk assessment result is closer to the physical reality and market environment, and the accuracy and reliability of imbalance risk early warning are remarkably improved; accurate risk early warning is provided before a scheduling plan is executed, so that scheduling personnel can identify potential system imbalance risks in advance and have sufficient time to take preventive control measures, and therefore, occurrence of large-scale imbalance events can be effectively avoided, and safe and stable operation of a power grid can be guaranteed; accurate risk early warning helps to reduce unnecessary standby capacity configuration and emergency control cost, avoids sharp fluctuation of spot market price caused by large-scale output deviation at the same time, and helps to maintain stable and economical and efficient operation of the power market.