43 results about "Active power flow" patented technology

The invention discloses a new energy electric power robustness control method, system and device for preventing chain overload of a power transmission section and a medium. The method comprises the following steps: obtaining the active power flow size and the operation state of each branch in the power transmission section; calculating the comprehensive sensitivity of all the controllable nodes to each branch in the power transmission section, and generating a composite priority index based on the calculated comprehensive sensitivity, the classification of the nodes and the new energy fluctuation risk; performing priority ranking on the generator nodes to generate a node sequence; in combination with a reverse equivalent pairing principle, selecting an adjustment node pair, and calculating an adjustment amount of node output by considering an adjustment amount dynamic margin constraint of new energy output uncertainty; recalculating the active power flow and the operation state of each branch in the section; and judging whether the section power flow and the margin meet a termination condition or not to realize optimal control. According to the method, the operation robustness of the new energy power system is improved, and the overload of the transmission section can be quickly and efficiently eliminated.

The invention relates to the technical field of power dispatching, in particular to an energy storage combined frequency modulation control method and system in a source-network-load-storage integrated regional power grid, and the method comprises the steps: dynamically calculating the adjustable capacity of each resource, dynamically aggregating the total adjustable capacity of the system, sensing the multi-dimensional frequency situation and calculating the adjusting power, and selecting a frequency modulation mode and generating a power distribution strategy. And generating a splitting and isolated network frequency modulation stability control strategy and performing system grid-connected resynchronous operation. The method is suitable for guiding power supply, energy storage and load all-resource collaborative frequency modulation strategy generation in a regional power grid. For consistency of frequency deviation, frequency change rate and active power flow direction, multi-dimensional frequency situation awareness is constructed, power distribution strategies in different modes are generated based on a total adjustable capacity interval of a system, and a frequency stability control and grid-connected resynchronization strategy after splitting is given when the frequency and the adjustable interval exceed limits. Accurate matching of all-resource collaborative frequency modulation and strategies in a regional power grid is realized, and the frequency modulation response speed of the system is improved.

The invention discloses a power grid switching operation anti-misoperation verification method and device based on real-time topology, and relates to the technical field of power grid anti-misoperation, and the method comprises the following steps: collecting the phasor data of target equipment and the real-time active power flow direction of a power grid, constructing a real-time directed topological graph, and extracting a to-be-reconstructed sub-graph associated with an operation instruction of the target equipment; converting the target equipment operation instruction into an impedance correction term, constructing an estimated impedance matrix based on the to-be-reconstructed sub-graph, and calculating a corresponding power flow transfer vector; performing double-layer constraint decoupling on the topological structure of the to-be-reconstructed sub-graph to generate a source flow path chain set; and generating a verification result based on the power flow transfer vector, the source flow path chain set and a preset anti-error rule. The method is used for solving the problems of electrical rehearsal real-time lagging and complex topology anti-error logic confusion in a complex power grid environment.

The present invention belongs to the technical field of power electronic converters and discloses a medium-voltage distribution network flexible interconnection switch topology and control method. The medium-voltage distribution network flexible interconnection switch topology provided by the present invention includes three connected reactors L1, three connected reactors L3, two chain-type STATCOMs, two three-phase isolation transformers, two three-phase LC filters, two three-phase inverter circuits, and a DC port. Back-to-back three-phase converters are introduced at the ends of the chain-type STATCOMs of two feeders. Through the back-to-back three-phase converters, active power flow mutualization between the two medium-voltage chain-type STATCOMs, that is, between two different feeders, can be achieved. The present invention can achieve DC capacitor voltage balancing of the chain-type STATCOM submodules of the two feeders. Compared with the BTB-VSC commonly used in existing medium-voltage distribution network flexible interconnection demonstration projects, this solution has lower cost, simpler topology, easier control, and does not need to consider the active and reactive power decoupling issues between interconnected feeders, making it easy to promote and apply on a large scale.

The invention discloses a large-scale distributed photovoltaic power generation scene simplified clustering method and device suitable for annual mode calculation of a power system. The method and device are suitable for annual mode calculation of the power system. The method comprises the steps of establishing a sensitivity model between photovoltaic output and line active power flow change, quantifying influence degrees of different photovoltaic scenes on power transmission line power flow, and constructing a disturbance quantity characteristic matrix of a distributed photovoltaic output scene on the power transmission line active power flow; and clustering the large-scale distributed photovoltaic output scenes by taking the disturbance quantity feature vectors as samples. According to the method, typical photovoltaic output scenes with representativeness are constructed, so that the number of redundant scenes is reduced, and typical output characteristics which have the greatest influence on safe operation of the system are reserved, and therefore, the simulation efficiency and the result reliability are improved; depth coupling of photovoltaic scene clustering and power system physical characteristics is realized, and more accurate data support is provided for system power flow analysis, load transfer research, scheduling decision and the like.

Provided is a power conversion device disposed at an end of a high-voltage DC power transmission system connected to a renewable energy power source, the power conversion device making it possible to configure a multi-pole high-voltage DC power transmission system by connecting a plurality of power conversion devices in parallel to the same renewable energy power source. The power conversion device includes a power converter and a control device. The power converter converts AC power supplied by the renewable energy power source to DC power and outputs the converted DC power to a DC power transmission line. The control device controls the AC voltage outputted by the power converter to the renewable-energy-power-source side. The control device has a voltage control unit and a current control unit. The voltage control unit generates a voltage command value for controlling the amplitude and the phase of the AC voltage on the basis of at least the power value of active power flowing in from the renewable energy power source. The current control unit corrects the voltage command value via current feedback for reducing at least a DC current component of a lateral-flow current flowing between the current control unit and the power converter of another pole on the renewable-energy-power-source side.

The invention provides a carbon footprint tracking method and system based on electric vehicle path planning, and belongs to the technical field of path planning and carbon footprint tracking, and the method comprises the steps: obtaining the road network planning reference data of a target region, and inputting the data into a preset road network planning model, so as to enable the road network planning model to generate a corresponding road network planning result; according to the road network planning result and the real-time operation data of the power system of the target area, performing power flow planning on the power system through a preset power flow planning model, and determining the active power flow distribution of the power system and the actual active power of each generator; and according to the carbon emission intensity of each generator, the actual active power of each generator and the active power flow distribution of the power system, calculating the node carbon potential of each power grid node. The problem that in the prior art, carbon emission accounting is disjointed with the dynamic behavior of an electric vehicle user can be solved.

The invention discloses a series-parallel hybrid five-port energy router capable of realizing transformer area interconnection and a control method of the series-parallel hybrid five-port energy router. The five-port energy router comprises four low-voltage alternating current ports AC1, AC2, AC3 and AC4, a direct current port DC, two parallel converters PCM1 and PCM3, and two series converters SCM2 and SCM4. AC ports of the PCM1 and the PM3 are respectively connected with the AC1 and the AC3, and AC ports of the SCM2 and the SCM4 are respectively connected with the AC2 and the AC4 through series transformers; dC ports of the PCM1, the SCM2, the PCM3 and the SCM4 are connected together to form a common DC bus; dC is led out from the common DC bus. The five-port energy router supports parallel connection-series connection and parallel connection-parallel connection power flow mutual aid modes, DC bus voltage is stabilized through the PCM, reactive power and active power are controlled through the PCM, voltage amplitude and phase control active power flow are adjusted through the SCM, and the required configuration capacity is small. According to the invention, the flexible interconnection of the low-voltage transformer area can be realized, the dynamic capacity increase of the transformer area and the fault rapid transfer are realized, and the power supply reliability and the distributed power supply acceptance capability are improved.

The invention discloses a carbon reduction strategy optimization method, system and device and a storage medium, and the method comprises the steps: obtaining active power flow data and the carbon emission intensity of a power supply point, and calculating an electric carbon factor of each load enterprise node according to the active power flow data and the carbon emission intensity of the power supply point; identifying the carbon emission characteristics of each load enterprise node according to the electrical carbon factors, and generating an enterprise portrait corresponding to the carbon emission characteristics; and based on the enterprise portrait and the production process and energy consumption elasticity of each load enterprise, generating a differentiated carbon reduction strategy, realizing optimization of the carbon reduction strategy, and providing the differentiated carbon reduction strategy for power consumption subjects with different positions and different characteristics in the park.

The present application discloses a carbon emission assessment method, device, storage medium and product, and relates to the field of carbon emission technology. The present application is based on node power parameters, node heat parameters, heating carbon emission intensity of various sources in the park, and power supply carbon emission intensity of various sources in the park. For any node, the heat loss of the branches between each node, the heat transfer amount of the branches between nodes, the actual heat consumption corresponding to the node, and the heating carbon emission intensity of various sources in the park are calculated; based on the power load of each node, the active power flow between each node, and the power supply carbon emission intensity of various sources in the park, the power consumption carbon emission intensity of the node is calculated, thereby effectively transmitting the carbon emission responsibility of electricity and thermal resources to specific nodes, helping users corresponding to the nodes understand the cleanliness of energy use, and providing basic signals for users' low-carbon response.

The application discloses a single-core independent phase-shifting transformer and a control method thereof. The single-core independent phase-shifting transformer comprises an excitation winding, a source-side voltage regulation winding, a load-side voltage regulation winding and an on-load voltage regulation switch. The excitation winding is a primary side of the single-core independent phase-shifting transformer and is connected in a triangle mode. The source-side voltage regulation winding and the load-side voltage regulation winding are secondary sides of the single-core independent phase-shifting transformer and are directly connected in series in a circuit. The source-side voltage regulation winding and the load-side voltage regulation winding are both provided with the on-load voltage regulation switch. The source-side voltage regulation winding and the load-side voltage regulation winding are divided into several sections according to phase-shifting angles, that is, the phase-shifting angle of each section is changed by the same value, and the voltage of each section is changed by different values. The application can independently adjust the phase and effective value of the circuit voltage, independently adjust the active power flow and the reactive power flow of the circuit, and further optimizes the adjustment performance of the phase-shifting transformer.

The invention discloses a power system carbon emission flow calculation method and system, electronic equipment and a medium, and relates to the technical field of power system carbon emission flow calculation. The method comprises the following steps: calculating an extended incidence matrix according to active power transmitted by each branch, power injected to a corresponding node by an upstream node of each node and power injected to a generator set at each node; determining a unit power injection distribution matrix of the power generation side of the power system, the carbon emission intensity of each generator set and a unit injection carbon flow rate matrix according to the generator set injection power at each node; determining a system load distribution matrix according to the active power of the load at each node; and calculating the carbon emission flow of the power system according to the matrix, the active power flowing through the nodes, the carbon emission intensity and the active power flowing through each branch. According to the invention, the accuracy of the calculation result of the carbon emission flow of the power system can be improved.

The invention discloses a two-stage reactive power flow generation method considering the switching times of a reactive power compensation device. According to the method, a two-stage solving strategy is adopted to solve the reactive power flow generation problem considering the switching frequency limitation of the parallel reactive power compensation device, and in the first stage, the minimum branch active power flow deviation and the voltage square difference between the two ends of a branch serve as a target function; an optimization model containing constraints such as the switching frequency limitation of the discrete reactive power compensation device is constructed, linear processing is carried out, and rapid convergence of the alternating current power flow is achieved by adjusting the switching state of the reactive power compensation device; and in the second stage, fixing the switching state obtained in the first stage, and further adjusting the active output, the reactive output and the terminal voltage of the generator set by taking the minimization of the active adjustment amount of the generator set as a target function, so that the system meets the AC feasibility requirement. According to the method, the solving complexity of the reactive power flow generation problem can be remarkably reduced, the solving efficiency is improved, and the rapid generation and feasibility requirements of the reactive power flow of the power system are met.

The invention provides a carbon management evaluation method and system based on edge calculation. The method comprises the following steps: combining the carbon injection rate of each power distribution side edge node with the active power flow distribution of each line of the power distribution network, and generating a carbon flow vector of each line; generating a carbon emission feature vector based on the carbon flow vector, the power grid topological data and the equipment current waveform parameter of the user side edge node; generating a carbon emission characteristic divergence rate based on the phase space parameter of the power grid frequency change rate of the power distribution side edge node; when the carbon emission characteristic divergence rate exceeds a stability threshold value, generating carbon flow optimization regulation and control data based on the dynamic control signal through an edge control node; and based on the carbon flow net injection rate and the fusion state vector of each power distribution network node, positioning an abnormal user side edge node and generating an exception handling instruction, and the method can solve the problems that carbon emission accounting is distorted, low-carbon and safety cooperation is difficult to realize, and exception judgment has deviation.

The invention provides a method for calculating power grid active loss caused by automatic voltage control based on sensitivity, and belongs to the technical field of power system voltage control. The method comprises the following steps: when a current automatic voltage control cycle arrives, obtaining a reactive power equipment set influencing the active power loss of the power grid and calculating the sensitivity of reactive power output of each reactive power equipment in the set to the active power flow loss of the power grid; calculating the active power flow loss of the power grid caused by the reactive power equipment in combination with the reactive power output variable quantity of the reactive power equipment in the current automatic voltage control period; and based on the power grid active power flow loss caused by the reactive equipment, performing integration by adopting a rectangular method to obtain the power grid active power loss caused by the reactive equipment. According to the method, the quantitative calculation result of the active loss of the power grid caused by automatic voltage control after the automatic voltage control system is put into closed-loop operation can be obtained online in real time, and safe and efficient operation of the power grid is facilitated.

This invention provides a method for managing voltage exceedance in distributed photovoltaic (PV) systems by optimizing the active and reactive power ratio. When the grid connection point voltage exceeds the limit, this application abandons the maximum power point tracking (MPPT) mode and performs the following steps: calculating the reactive and active power ratios that the PV system should inject into the grid connection point after the voltage exceedance, ensuring that the active power injected into the distribution network bus is maximized when the grid connection point voltage does not exceed the limit; calculating active and reactive power commands using the calculated reactive and reactive power ratios; obtaining current reference values ​​on the dq axis in a dq coordinate system with equal amplitude coordinate transformation based on the active and reactive power commands; and adjusting the active and reactive power injected into the grid connection point based on the current reference values ​​on the dq coordinate axis to regulate the grid connection point voltage. This invention ensures maximum active power flowing to the grid connection point while maintaining the grid connection point voltage within the specified limits.

The application relates to a method and device for scanning a predicted fault applied to an element of a power system, a computer device, a storage medium and a computer program product, and relates to the technical field of power grid safety monitoring. The application can shorten the time of detailed power flow calculation and improve the efficiency of the predicted fault scanning. The method comprises the following steps: obtaining a node admittance matrix and an active power flow total increment of an element to be detected under a predicted fault condition; obtaining a voltage phase angle increment corresponding to each bus in the element to be detected according to the node admittance matrix and the active power flow total increment; obtaining an active power flow increment of a current branch according to a first voltage phase angle increment, a second voltage phase angle increment and a current reactance value of the current branch for each branch in the element to be detected; determining a weight coefficient and a power flow limit value of the current branch; and inputting the active power flow, the weight coefficient and the power flow limit value into a preset severity index operation model to obtain a predicted fault severity index when the current branch is disconnected.

The application discloses a method and system for improving structural controllability of a power system, and belongs to the field of structural controllability optimization of a power system, which comprises the following steps: taking a generator, a transformer substation and a power consumption unit as a node, taking a power transmission line as an edge, the direction of the edge being consistent with the active power flow direction of the power transmission line, establishing an original network model of the power system, deleting the incoming edges of the generator nodes in the original network model to obtain a power network, solving the maximum matching of the power network, merging the edges with the same nodes in the maximum matching to obtain a matching path set and a matching circle set, marking the paths with the generator nodes as starting points in the matching path set as control paths, marking the remaining paths as uncontrollable paths, adding a directed edge between the terminal point of the control path and the starting point of the uncontrollable path, and adding a power transmission line between the nodes corresponding to the added edges. The application makes the power system structurally controllable only when the generator nodes are used as driving nodes, and effectively improves the structural controllability of the power system.

The application discloses a kind of carbon flow monitoring and accounting method based on Internet of Things, specifically comprising the following steps: S1.collection power system tidal information;S2.carries out flow calculation to power network, obtains the active power flow matrix F of entire power network;S3.carbon flow density is calculated to the grid in stages, gradually obtains the carbon flow data of entire power network;S4.through the carbon flow data obtained carbon flow matrix C, and provide final visual result;S5.collection carbon emission data;S6.create carbon flow neural network model;S7.train carbon flow neural network model;S8.utilize carbon flow neural network model to the carbon emission of entire system is accounted for.This application mainly utilizes Internet of Things data acquisition technology and data analysis visualization technology, for power system manager directly shows carbon flow in entire power network The flow path of, help to realize the low carbonization and green of entire system.

This application relates to a location method, program product, and related equipment for a grid-type static var generator (SVR). The method includes: acquiring multiple sets of candidate schemes, each set including several location nodes; calculating the total power supply capacity connected to each location node, obtaining the annual maximum active power flow and values ​​of multiple preset indicators for each location node; determining the weight coefficients of each preset indicator in the corresponding location node based on the annual maximum active power flow and total power supply capacity; calculating the location base number of the corresponding location node based on the preset indicator values ​​and weight coefficients; selecting several location nodes sequentially in descending order of the location base number as candidate nodes; calculating the average location base number of the corresponding candidate schemes; and selecting the candidate node in the candidate scheme with the highest average location base number as the final location target. This invention, by connecting the grid-type SVR to the location target, can improve grid stability to meet grid operation requirements.

An electrified railway flexible hub traction substation framework is characterized in that an alpha power supply arm and a beta power supply arm are connected to a three-phase power grid through a traction transformer, and two alternating current ends of a back-to-back converter are respectively connected to the alpha power supply arm and the beta power supply arm; the power distribution system bus is connected to a railway power distribution network through a first grid-connected transformer, and the power distribution system load is connected to the power distribution system bus through a power distribution transformer; the system further comprises an energy feedback system. The energy feedback system comprises a bidirectional DC / AC converter and a second grid-connected transformer. The direct current end of the bidirectional DC / AC converter is connected to the direct current end of the back-to-back converter, and the alternating current end of the bidirectional DC / AC converter is connected to the power distribution system bus through the second grid-connected transformer. The beneficial effects of the utility model lie in that the traction power supply system and the railway power distribution network are connected through the energy feedback system, mutual transfer of active power flow of the traction power supply system and the railway power distribution network can be realized, and the operation energy efficiency of railway power supply is improved.

The invention discloses a method, a system and equipment for calculating the influence degree of grid-connected equipment on transient frequency stability of a power system, and a medium. The method comprises the following steps: determining a network equation, node frequency and active power data of the grid-connected equipment during a transient period through primary time domain simulation; for each time point, calculating the sensitivity of the active power of the grid-connected equipment to the active power of a branch connected with an inflow node according to a network equation; according to the node frequency, the active power of the grid-connected equipment and the sensitivity, calculating the influence degree of the grid-connected equipment on the transient frequency stability of the node; and summarizing the transient frequency stability influence degrees of the grid-connected equipment on all the nodes to obtain the transient frequency stability influence degree of the grid-connected equipment on the power system. According to the method, the influence of all the grid-connected devices on the transient frequency stability of the power system in different time periods can be evaluated only through one-time simulation, a decision basis is provided for optimization control of the grid-connected devices, and the calculation efficiency is improved.

The application discloses an active power flow adjusting method, device and equipment for an AC-DC distribution network, which is characterized in that: the actual active power vector of a converter station is calculated by using the rated voltage of a DC system and the parameters of the converter station, the connection node of the DC system converter station is equivalent to an AC load node, and the actual active power vector of an AC system node is solved, so that the steps of the whole adjusting process are simple, the data does not need to be iteratively processed, the solving speed is fast, and the adjusting efficiency of the active power flow of the AC-DC distribution network is improved.

The application discloses an environmental protection gas ring main unit fault isolation and recovery control method for a power distribution system and relates to the technical field of power distribution automation.The fault isolation method comprises the following steps of S01: collecting three-phase voltage and current instantaneous values of a current feeder interval access point by an intelligent terminal of the feeder interval, determining a disturbance direction by calculating a power disturbance direction parameter, and determining that the disturbance active power flows from a bus to a line if the disturbance active power flows from the bus to the line.Through synchronous collection of three-phase voltage and current instantaneous values, extraction of positive sequence components, calculation of a power disturbance direction parameter based on power changes before and after the disturbance, and determination of whether disturbance energy flows from a bus to a line or from a line to a bus, the method eliminates, in principle, the problem of healthy branch mis-isolation caused by overcurrent discrimination failure due to distributed power reverse current, realizes accurate identification and selective removal of a fault section of a power distribution network containing a distributed power, and significantly improves the accuracy of fault isolation and power supply reliability.

The application discloses a power grid switching operation anti-misoperation checking method and device based on real-time topology, relates to the technical field of power grid anti-misoperation, and comprises the following steps: collecting phasor data of a target device and real-time active power flow direction of a power grid, constructing a real-time directed topology graph, and extracting a to-be-reconstructed subgraph associated with a target device operation instruction; converting the target device operation instruction into an impedance correction term, constructing an estimated impedance matrix based on the to-be-reconstructed subgraph, and calculating a corresponding power flow transfer vector; performing double-layer constraint decoupling on the topology structure of the to-be-reconstructed subgraph to generate a source-flow path chain set; and generating a checking result based on the power flow transfer vector, the source-flow path chain set and a preset anti-misoperation rule. The application is used to solve the problems of real-time lag of electrical pre-rehearsal and confusion of anti-misoperation logic in a complex power grid environment.

The present invention belongs to the fields of power grid operation and artificial intelligence technology, and discloses a power grid flow control method and system based on an artificial intelligence neural network. The method comprises: obtaining flow data of a target power grid; taking the amplitude and phase angle of the voltage at all nodes of the target power grid, the active and reactive power flows of the line, and a preset change in the active power flow of the line to be controlled as input; introducing the data into a pre-trained artificial intelligence neural network model of the line to be controlled for identification to obtain the voltage difference between the two ends of the line to be controlled; and controlling the voltage difference between the first and last nodes of the line to be controlled to be equal to the voltage difference between the two ends. The present invention associates the power adjustment value of the line with the voltage difference between the two ends of the line through the artificial intelligence neural network model. When the line flow control scheme is activated, the flow optimization on the line is achieved by controlling the node voltage.

The invention relates to the technical field of new energy, in particular to a time sequence production simulation method based on refined alternating current power flow. Optimizing an objective function in the time sequence production simulation model based on the direct current power flow according to the influence of the line loss on the operation of the power system; power balance constraint, Ohm law constraint, voltage out-of-limit constraint, current out-of-limit constraint, line active power flow constraint, biomass unit operation constraint and second-order cone constraint in the time sequence production simulation model based on the direct current power flow are optimized, and relaxation operation is performed on the second-order cone constraint; constructing a time sequence production simulation model based on the direct current power flow; and dividing the year-round time period into a plurality of sub-time periods, and adding the divided sub-time periods according to the solving results of the time sequence production simulation models of the sub-time periods to obtain the operation result of the time sequence production simulation with the fixed time length. Compared with a traditional time sequence production simulation model, the effectiveness and accuracy of the time sequence production simulation model are improved.

The application provides a power distribution network power flow calculation method, a power distribution network power flow calculation device and a power distribution network power flow calculation system. The method comprises the following steps: establishing a power distribution network power flow equation; in the case that any node in the power distribution network is disconnected, a first equation is established according to the voltage phase angle, the reactance and the initial power flow of each target branch, the target branch being an affected branch after the node is disconnected; a ratio of the power flow increment of the target branch to the initial power flow of the node is determined as a distribution factor; a second equation is established according to a vector composed of the distribution factors and an active power flow vector of the affected branch when the power distribution network is in a steady state; the power flow increment of the affected branch is calculated by simultaneously solving the power distribution network power flow equation, the first equation and the second equation; and the power distribution of the affected power equipment in the power distribution network is adjusted according to the power flow increment of the affected branch. The method solves the problem that the affected power equipment cannot be adjusted in time due to the low efficiency of the power flow calculation in the prior art.

The application discloses an active power distribution network power flow optimization method based on a large language model and belongs to the field of power system optimization. The method first realizes optimal power flow problem modeling according to the topology and equipment parameter data of the active power distribution network to be subjected to power flow optimization; then designs a structured prompt word through the optimal power flow problem, drives a large language model to generate an initial population of diversified heuristic algorithms in zero samples; further, a thinking chain evolution strategy is adopted to take the calculation error as a judgment standard, guide another large language model to simulate selection, crossover and mutation operations to realize iterative optimization of the heuristic algorithm, and based on the heuristic algorithm with the minimum calculation error, the required output of each distributed energy source of the active power distribution network is obtained, and the power flow optimization of the active power distribution network is completed. The application solves the core problem of low design efficiency and high dependence on expert experience of the heuristic algorithm in the complex optimal power flow problem of the active power distribution network, and can realize high-quality algorithm automatic generation for a self-defined problem scene.

The invention relates to a power transaction outsourcing electric quantity prediction method and device, and belongs to the technical field of power prediction, and the method comprises the following steps: obtaining node net injection power, outsourcing electric quantity and line resistance and reactance parameters, and constructing a net injection power vector and a node feature matrix; forming a node admittance matrix based on reactance, and calculating active power flow in combination with a power vector; a network loss quadratic matrix is constructed based on resistance and reactance, a causal influence coefficient vector is solved, and then a causal attention matrix is generated in combination with active power flow. And inputting the matrix and a node feature matrix into a graph convolutional network to obtain node causal embedding and converge into a graph-level embedding vector. And inputting the graph-level embedded vector and historical data of the purchased electric quantity into a long-short-term memory network, and outputting a predicted value of the purchased electric quantity at the next moment. According to the method, the causal attention mechanism is introduced, the causal influence between the nodes is quantified, the model focuses on a key information source, and the prediction accuracy is remarkably improved.