152 results about "Supply & distribution" patented technology

The invention discloses a distributed power supply distribution optimization method and system based on a virtual power plant, and the method comprises the steps: generating estimated power distribution data according to a power distribution region power characteristic value, climate and holiday and festival information, and generating estimated power distribution data in combination with a target historical power generation record and the like; then judging a peak power consumption period through the estimated power consumption distribution, obtaining the total electric energy demand quantity, and determining the cache capacity of the energy storage equipment according to the total electric energy demand quantity and the estimated power generation distribution data, and when the difference value between the cache capacity and the flow of the transmission bottleneck point of the power distribution network is lower than a preset margin threshold value; and generating a power plant supplementary electric quantity proportion by using the power distribution proportion adjustment parameter and the real-time load data, further adjusting a power plant scheduling frequency and distributed power supply access density proportion, and finally generating a power distribution scheduling scheme of a target power distribution period. The method can effectively balance the load of the power grid, improves the power distribution efficiency, adapts to the access of a distributed power supply and renewable energy sources, and achieves the intelligent dispatching and optimized operation of the power grid.

The invention relates to the technical field of vehicle power supply management, in particular to a vehicle load power supply distribution method, device and equipment and a storage medium, according to the scheme, a vehicle using scene at the current moment is determined by analyzing vehicle operation data, and the load levels of a plurality of electric appliances in use are determined by combining a vehicle using scene-load level mapping table; and a power supply distribution scheme is formulated, on-demand power supply for each electricity load of the vehicle is realized, and meanwhile, a user is allowed to customize load grade setting of a vehicle using scene. According to the scheme, through reasonable distribution of power supply, the power supply proportion of each load can be adjusted in real time according to the vehicle use scene on the premise of ensuring normal operation of key functions of the vehicle, so that the operation efficiency and stability of the vehicle are improved, it can be ensured that each electricity load obtains proper power supply in different scenes, the energy utilization efficiency is improved, and the energy consumption is reduced. The personalized electricity demand of the user is met, the driving experience of the user is improved, and the intelligent level of the vehicle is improved.

The invention discloses an inspection intelligent analysis method and system for power marketing, and the method comprises the steps: building a mapping relation between power data and business data in a target marketing region according to the power supply distribution information of the target marketing region, and obtaining a fusion data set; performing semantic mapping on the fused data set to obtain a power marketing data view; obtaining a rule between every two pieces of data in the electricity marketing data view to identify an abnormal service; determining target business data corresponding to the abnormal business, and obtaining an abnormal business analysis result based on a pre-constructed marketing inspection full-business model; generating a problem tracking work order according to the abnormal service analysis result, and matching a corresponding work order processing flow to obtain a problem processing work order; extracting an abnormal service processing opinion in the problem processing work order to predict a rectification effect of an abnormal service; and an inspection analysis report of the abnormal service is generated based on the rectification effect and the abnormal service processing opinion, so that the efficiency and accuracy of inspection analysis are improved.

The invention discloses an electric power material efficient dispatching and logistics optimization method for an emergency scene, and belongs to the field of power grid material dispatching and transportation. According to the material dispatching requirements of the power system under different disaster levels, a hierarchical response mechanism is provided, and the hierarchical response mechanism comprises a single-bin quick response mode, a multi-bin collaborative supplement mode and a whole-network centralized support mode. A mathematical optimization model including supply, transfer and demand nodes and transportation paths is established, and improved optimization algorithms such as Sine mapping population initialization, differential variation disturbance and elitist retention, dynamic reverse learning and the like are introduced to improve the solving efficiency and the diversity of solutions. Under the assumption that node positions, material requirements, vehicle parameters and the like are known, the model realizes full-process optimization of supply distribution, transportation paths and resource scheduling. The method can effectively improve the emergency material dispatching efficiency, reduces the transportation cost and the material shortage risk, and has a good practical application value.

The invention relates to the technical field of communication power supply energy distribution, and discloses a power supply distribution method and device based on a communication power supply system, and the method achieves the precise cooperation of multiple targets through constructing a multi-dimensional target reward function based on the equipment energy consumption cost, the balance of remaining electric quantity and the attenuation of the service life of a battery, breaks the limitation of a single target, and improves the power supply efficiency. And performing reinforcement learning training on the initial power supply distribution model by using a historical operation data set, so that the model gradually learns optimal power supply strategies in different scenes, and finally generating a target power supply distribution model capable of autonomously optimizing decisions. The adaptive capacity of a communication power supply system to a complex and changeable communication network environment is remarkably improved, and the problem of power supply redundancy caused by a static distribution strategy is effectively avoided while the power supply reliability is ensured.

The invention relates to a disaster prevention emergency material distribution method and device, equipment and a medium. The method comprises the following steps: acquiring multi-source heterogeneous data of a disaster area; performing standardization and spatial fusion processing on the multi-source heterogeneous data to obtain standardized spatial grid data; based on the standardized space grid data, analyzing the material demand urgency degree of each space grid to obtain demand urgency degree data; generating initial material distribution scheme data according to the demand urgency degree data and preset material inventory data; and based on the initial material distribution scheme data, in combination with preset transportation network data, carrying out scheme adjustment through a predefined multi-objective optimization rule, and generating a material distribution scheme. By adopting the method, the problems of unreasonable distribution and low decision-making efficiency caused by data splitting and dependence on experience in a traditional mode are solved, the integration from disaster assessment to transportation scheduling is realized, the decision is quickly optimized, and the overall efficiency of emergency response is improved.

The invention relates to the technical field of data center computing power and electric power resource cooperative control, in particular to an electric power dynamic monitoring system based on a knowledge graph. Comprising a heterogeneous data spin mapping unit used for collecting current waveform data and voltage data of an intelligent power distribution unit; the topological coupling force field construction unit is used for acquiring physical topological data and logical topological data of the data center and generating a coupling strength matrix defining association weights among nodes; the dynamic potential energy evolution and gradient analysis unit is used for generating a potential energy gradient vector for indicating a regulation and control direction; the potential energy guiding regulation and control decoding unit is used for receiving the potential energy gradient vector; monitoring the modulus length of the potential energy gradient vector; and after the closed loop feedback logic and the hardware control signal are executed, the physical operation states of the intelligent power supply distribution unit and the server calculation unit are changed. According to the method, the mathematical abstraction ability and the quantization accuracy of the stability monitoring of the complex data center system are remarkably improved.

The present invention discloses a photovoltaic four-in-one fusion terminal sensing system and method, which relates to the technical field of sensing control. The present invention includes a low-voltage distributed photovoltaic substation line topology formation module, a sensing analysis module, a power supply distribution analysis module, and a remote control module; the low-voltage distributed photovoltaic substation line topology formation module is used to form the low-voltage distributed photovoltaic substation line topology of the target area; the sensing analysis module is used to predict the photovoltaic power supply risk coefficient of the low-voltage distributed photovoltaic substation; the power supply distribution analysis module is used to analyze the power supply distribution situation of the low-voltage distributed photovoltaic substation. The present invention takes the real-time photovoltaic power supply unabsorbed coefficient of the low-voltage distributed photovoltaic substation as the real-time electricity consumption characteristic value of the substation users, simplifies the comparison and processing process of data, and is beneficial to ensuring the soft control and rigid control of the power supply distribution situation of the substation before the power supply equipment in the substation fails, and improves the sensing effect of the photovoltaic four-in-one fusion terminal.

The invention relates to the technical field of power supply and distribution monitoring, in particular to a power supply and distribution system performance health state dynamic monitoring system which comprises a state acquisition module, a power supply switching module, a power supply monitoring module, a switching matching module and a state evaluation module. The method comprises the following steps: combining power supply equipment as power supply nodes into a power supply distribution network, and checking switching behavior characteristics of a current power supply node according to operation information of the power supply nodes and operation time and switching times of the power supply nodes; the similarity during switching of different power supply modes is checked according to the voltage deviation and the load deviation of the operation information, and updating of the operation information of each power supply node is completed; starting from switching behavior characteristics, describing a switching mode of a current power supply node under power supply mode switching, describing a switching matching network of each power supply node according to a switching sequence of the switching modes, and finally screening an optimal configuration scheme corresponding to each power supply node; the efficiency of anomaly detection and the reliability of switching of the power supply and distribution system are realized.

An electrical control box for superconducting magnet transportation and a device for transporting a superconducting magnet, relating to the field of superconducting magnet transportation, are configured to transport the cryogenic superconducting magnet. The electrical control box for superconducting magnet transportation is provided to address the problem that rated load capacity requirement of the transport apparatus is too high while transporting the superconducting magnet. A function of one input to two outputs is implemented through a power supply distribution module to supply power to a helium compressor and a refrigerator respectively to align with the common practice of transport personnel. In addition, a delay module is arranged in a power supply output connected to the helium compressor to provide power to the helium compressor later than the refrigerator, i.e., the start-up of helium compressor and the refrigerator are asynchronous.

The invention discloses a meteorological measurement method, device and system, relates to the technical field of communication, and aims to solve the problems that meteorological services provided by meteorological detection equipment such as weather radars, rain gauges and meteorological satellites are unbalanced in supply distribution and large in updating time delay, and the surface precipitation cannot be accurately determined. The method comprises the steps that a receiving end receives a reference signal transmitted through a wireless channel, the reference signal is measured to obtain meteorological information, and then the receiving end reports the meteorological information to core network equipment. Correspondingly, the core network equipment receives the meteorological information reported by the receiving end, determines meteorological characteristic parameters based on the meteorological information, and further provides meteorological services corresponding to the meteorological characteristic parameters. Wherein the meteorological information is used for representing the meteorological characteristics of the wireless channel in the transmission process, and / or the channel state of the wireless channel under the to-be-measured weather. The scheme of the invention can be widely applied to the fields of communication technology, artificial intelligence, Internet of Vehicles, smart home networking and the like.

The application relates to a power grid power supply distribution box with a closed intelligent power connection and a closed power connection method thereof, which comprises a driving part installed on a second box door, a gear cylinder arranged on the driving part, and a lifting part installed on the second box door through a guide block and connected with the gear cylinder. When the second box door is opened, a door magnetic switch automatically disconnects a circuit, thereby triggering the conduction of a tripping circuit, which makes a circuit breaker trip and cut off the power supply in the distribution box. In this way, all branch switches in the distribution box are disconnected in a certain order under the action of a step-by-step locking structure, so that safe power-off can be realized when the power is connected. The step-by-step power-off operation design can effectively avoid large-scale power cut-off and reduce the damage of equipment caused by current impact or voltage fluctuation.

The invention discloses a power supply control method and device for a power supply vehicle, and the method comprises the steps: determining a power parameter of an input power supply and a power utilization parameter corresponding to an externally-connected power utilization device through the power supply vehicle when the externally-connected input power supply is detected, and judging whether the power utilization parameter is matched with the power parameter through a control unit through the power supply vehicle, when it is judged that the power utilization parameters are matched with the power supply parameters, a power supply matched with the power supply parameters is output to the electric equipment through a power supply distribution unit, and when it is judged that the power utilization parameters are not matched with the power supply parameters, the input power supply is adjusted to be a target power supply matched with the power utilization parameters. And the target power supply is output to the electric equipment through the power supply distribution unit. Therefore, the technical problem that the power supply used by the electric engineering machinery is single can be solved, and the application scene of the electric engineering machinery is widened, so that the generality of the power supply vehicle is improved, and the earthwork operation efficiency of the electric engineering machinery is improved.

Disclosed herein are a meteorological measurement method and apparatus, and a system, relating to the technical field of communications, intended to solve the problems of uneven supply distribution, large update delays, and an inability to accurately determine surface precipitation in a meteorological service provided by a meteorological detection device such as a weather radar, a rain gauge, and a meteorological satellite. The method comprises: a receiving end receiving a reference signal transmitted by means of a wireless channel, measuring the reference signal to obtain meteorological information, and the receiving end then reporting the meteorological information to a core network device; correspondingly, the core network device receiving the meteorological information reported by the receiving end, determining a meteorological characteristic parameter on the basis of the meteorological information, and further providing a meteorological service corresponding to the meteorological characteristic parameter. The meteorological information is used for characterizing a meteorological characteristic of the wireless channel in a transmission process and / or a channel state of the wireless channel in a meteorology to be measured. The solution of the present application can be widely applied to the technical fields of communication, artificial intelligence, vehicle networking, intelligent home networking, and the like.

The invention relates to the technical field of power distribution and protection, and provides an intelligent power distribution system in order to solve the technical problems that an existing system cannot accurately control on-off of all branches, cannot independently measure voltage and current of all the branches and is incomplete in protection function. Comprising a main power supply access module, a shunt processing module, a parameter measurement module, a main control module and a communication module, all the modules are connected through wires and data lines, power supply distribution of multiple paths of electric equipment is cooperatively achieved, meanwhile, electric power parameters of all branches are accurately monitored, a comprehensive protection function is provided, safe and stable operation of the electric equipment is guaranteed, and the service life of the electric equipment is prolonged. And the reliability and the intelligent level of power use are improved.

The invention relates to the technical field of intelligent power grid management and optimization, in particular to a data mining system and method based on big data in an intelligent power grid, and the system comprises a data collection module which is used for collecting the power utilization data of a user in real time; the data abnormity analysis module is used for judging the abnormal condition of the real-time power consumption data of the user; the energy supply analysis module is used for evaluating the energy supply condition; the energy consumption analysis module is used for judging the load density of the electric quantity area and the importance and demand of the power supply area; the balance analysis module is used for judging the balance state condition of the power grid in real time; the imbalance decision module is used for judging the supply and demand conditions, the electric quantity area and the power supply area in the abnormal imbalance state and outputting a decision; and the decision output module is used for drawing a power supply distribution condition table and outputting dispatching instructions and emergency measure suggestions to the power generation equipment. The operation efficiency of the power grid is improved, and the stability and reliability of power supply are ensured.

The invention belongs to the technical field of textile automatic control, and provides a blank pipe tray feeding, supplying and distributing system for tray type automatic spooling equipment, which sets a simulation operation period, and performs stability analysis on passing interval time of trays on a path A and a path B in the divided simulation operation period; judging whether the tray interval time on the path A and the path B is stable or not, respectively acquiring a representative value of the tray interval time A and a representative value of the tray interval time B according to the stable condition of the tray interval time on the path A and the path B, and performing supply shunting matching degree analysis to determine the matching degree on shunting time of the path A and the path B; the closing time and the releasing time of the anti-blocking device can be reasonably set by determining the matching degree, it is ensured that the tray can smoothly pass through the intersection according to the matching degree condition, the matching degree of the A path and the B path in the shunting time is determined, and the A path and the B path can supply empty pipe trays to the CBF channel according to the reasonable time rhythm.

The utility model relates to an intelligent control system for a sewage treatment system, which comprises a touch screen PLC (Programmable Logic Controller) all-in-one machine unit, and a power supply system unit, a 5G communication system unit and an I / O (Input / Output) interface unit which are mutually connected with the touch screen PLC all-in-one machine unit, the power supply system unit is provided with a power supply conversion module and a power supply distribution element, the 5G communication system unit is internally provided with 5G signal transmission module equipment and a signal transmission module equipment antenna, and the I / O interface unit is provided with I / O signal connection equipment. According to the utility model, the 5G communication system unit integrated in the system is utilized to realize uploading and downloading of system monitoring data, remote modification of system operation parameters and implantation signals in a wireless mode, so that wireless and remote interactive transmission of the data is realized, and remote monitoring in different places is realized.

The invention discloses an urban rail transit power supply control system and method, and the method comprises the following steps: S1, collecting the operation data of a rail transit power supply system, and carrying out the preprocessing; s2, constructing an improved MuZero reinforcement learning model, and optimizing a rail transit power supply control strategy through Monte Carlo tree search; s3, modeling a rail transit power supply resource allocation process by adopting a game theory, and calculating optimal power supply resource allocation by solving Nash equilibrium; s4, adjusting the power supply control scheduling of the rail transit power supply system based on the rail transit power supply control strategy and the optimal power supply resource allocation; and S5, updating a rail transit power supply control strategy by using the improved MuZero reinforcement learning model, and optimizing power supply resource allocation. The improved MuZero reinforcement learning and game theory method is adopted, rail transit power supply control scheduling is optimized, and the advantages of balanced power supply distribution, efficient energy recovery and intelligent self-adaptive adjustment of the system are achieved.

The present invention provides a grid planning method for grid connection of a photovoltaic power station, belonging to the technical field of grid operation and maintenance, and including steps such as historical estimation, preliminary planning, actual measurement, and re-planning. By first performing short-term actual measurement on the target photovoltaic power station, the present invention can directly measure the relationship curve between the light intensity and its actual power supply from the results after grid connection, and rely on this relationship curve to estimate the future all-day power supply distribution of the target photovoltaic power station according to weather forecast data. Then, according to the requirements of the grid, the power supply supplemented by the photovoltaic power station to the grid is pre-planned and adjusted, so that there can be more time for response, regulation, correction, and remedy, thereby avoiding installing complex electrical equipment on each photovoltaic power station for adjustment, which is beneficial to reducing the equipment costs of photovoltaic power generation and grid operation and maintenance, is also more convenient for the management of maintenance personnel, and is also beneficial to improving the safety and stability of grid operation.

The invention belongs to the technical field of distribution boxes, and particularly relates to a distribution box for electrical engineering automatic power supply, which comprises a box body, a servo motor, a bird repelling mechanism and a heat dissipation mechanism, heat dissipation holes are formed in the left side and the right side of the box body, the servo motor is arranged in the box body, and an output shaft is in transmission connection with a transverse rotating rod; the bird repelling mechanism comprises a box body, a vertical rotating rod and two swing assemblies, the box body is fixed above the box body, the vertical rotating rod is rotationally arranged in the box body, the bottom end of the vertical rotating rod is in transmission connection with the transverse rotating rod, a cam is fixed to the vertical rotating rod, each swing assembly comprises a supporting plate, a sliding block and a swing rod, the sliding block is slidably connected to the supporting plate, and the swing rod is arranged on the sliding block. A roller is arranged on one side of the supporting plate, a compression spring set and a rack are arranged on the other side of the supporting plate, a gear is meshed below the rack, one end of a swing rod is fixed on a gear shaft, and a swing block is arranged at the other end of the swing rod. And the heat dissipation mechanism is arranged at the heat dissipation holes of the box body. The power distribution box has a bird repelling function and a heat dissipation function, and normal use and operation of the power distribution box are effectively guaranteed.

The application relates to the technical field of power distribution cabinets, and particularly discloses a power failure uninterrupted self-throw type power distribution cabinet, which comprises a mounting mechanism arranged at the top of the power distribution cabinet and a pressure relief mechanism arranged at the top of the power distribution cabinet through the mounting mechanism. The high-temperature and high-pressure gas is directly contacted with the baffle through the pressure relief mechanism, the baffle is driven to rotate, the high-temperature and high-pressure gas is released, the stability of the high-temperature and high-pressure gas release is improved, the baffle does not shield the channel for upward discharge of the high-temperature and high-pressure gas after being turned over, the pressure relief efficiency is further improved, the pressure relief mechanism is conveniently mounted and dismounted through the mounting mechanism, the situation that the pressure relief mechanism cannot be dismounted or is difficult to dismount after the power distribution cabinet is faulty or burns due to the traditional screw fixing mode is avoided, the pressure relief mechanism is conveniently recycled, and actual dismounting and mounting are facilitated.

System and method for powering electrical consumers of an electric or hybrid vehicle. The invention relates to a power supply system (1) for electrical consumers of an electric vehicle comprising: - a high-voltage battery (8), - two DC-DC converters (4, 6) supplying an on-board network (50), - an energy storage unit (5) connected to the output of the converters (4, 6) and capable of supplying the on-board network (50) in standby mode, - power lines (L2, L5, L6, L7, L8, L10) supplying distribution boxes (12, 13, 14), - the distribution boxes (12, 13, 14), which include electronic switching components (t1, t2, t3, t4, t5, t6) connected to electrical consumers, the power supply system (1) further comprising a coupling device (2, 3) connected to the output of the converters (4, 6). and to the power lines,and comprising at least one electronic switching component(s). (Figure 1)

The invention discloses a cloud computing cabinet power distribution method and device based on AI optimization and a medium, and relates to the technical field of computer power management, and the method comprises the steps: constructing a power demand prediction feature set according to an electric power state sequence, executing forward reasoning through a trained recurrent neural network model, and generating a power demand prediction value set; performing power supply capability constraint verification on the power demand predicted value set based on a cabinet power supply connection model, and performing evaluation and screening in combination with a power matching degree, power fluctuation and a power supply safety margin to form a power supply power distribution scheme selection set; and mapping the power supply power distribution scheme selected set into a power supply distribution parameter set based on the port-path mapping relationship, and performing state updating in combination with power deviation evaluation to generate a cloud computing cabinet power distribution state set. The effect of guaranteeing power supply safety and stability in a complex industrial cloud computing environment is achieved, and the energy efficiency utilization rate is improved while power supply safety and stability are guaranteed.

The present application relates to the field of fuel cell vehicles, in particular to an energy management method for a vehicle hybrid power system.The present application provides an energy management method for a vehicle hybrid power system, the vehicle hybrid power system comprising a fuel cell and a power battery, the energy management method comprising the following steps: determining the working mode of the vehicle, including a static mode and a dynamic mode; in response to the vehicle being in the static mode, adopting a first energy management strategy, including determining the fuel cell output power based on the power battery SOC and determining the power supply distribution between the power battery and the fuel cell for the vehicle demand power; and in response to the vehicle being in the dynamic mode, adopting a second energy management strategy, including determining the fuel cell output power based on the vehicle acceleration mode and the power battery SOC, and cooperatively supplying the vehicle demand power by the fuel cell and the power battery. Not only the energy distribution efficiency is improved, but also the vehicle endurance mileage is improved, and the driving experience is improved.

A method of protecting a power transformer (101) in a power supply distribution grid (100) is described. The method comprises monitoring low frequency electrical signals in a circuit coupled to the power transformer (101) to provide a first detection signal (211); monitoring electromagnetic radiation to provide a second detection signal (212); in the event that the first detection signal (211) exceeds a first threshold, operating a protection switch (204a) connected to the transformer (101) to break the circuit and to reconnect the circuit in response to the first detection signal dropping below the first threshold; in the event that the second detection signal exceeds a second threshold, operating the protection switch (204a) to break the circuit at least until the first detection signal has dropped below the first threshold.

A system for diagnosing a power supply distribution circuit that distributes power supplied from a DC-DC converter and an auxiliary battery to a plurality of loads, the system comprising: a plurality of switches provided between the DC-DC converter and the auxiliary battery and the loads; a switching unit that switches a connection state of the switches; and a diagnosis unit that diagnoses the power supply distribution circuit based on the connection state of the switches and a change in voltage and current in the power supply distribution circuit.

The invention belongs to the technical field of power distribution cabinets, and particularly relates to a power supply branching structure for equipment in a cabinet. Comprising a power distribution cabinet structure, a fixing structure is arranged on the inner side of the power distribution cabinet structure, a component mounting structure and a deconcentrator mounting structure are arranged on the front side of the fixing structure, and a deconcentrator structure is arranged on the front side of the deconcentrator mounting structure. Through the deconcentrator mounting structure and the deconcentrator structure, the wiring of the component body can be subjected to effective branching processing by using the deconcentrator structure, the wiring of each component body can be ensured to be subjected to independent branching processing, and the deconcentrator structure and the fixing structure can be quickly disassembled and assembled by matching with the deconcentrator mounting structure; therefore, the situation that all wire harnesses are mixed together is avoided, when a certain independent component body needs to be maintained, a corresponding cable can be easily distinguished, and identification difficulty caused by cable disorder is avoided.

The utility model discloses a multi-channel direct-current power supply distribution module, and relates to the technical field of power transmission and distribution, the multi-channel direct-current power supply distribution module comprises a control branch and a plurality of power supply branches, each power supply branch comprises an input isolation circuit, a power conversion circuit and a rectification filter circuit which are connected in sequence, and the input end of the input isolation circuit is used for being connected with a direct-current power supply; the output end of the rectification filter circuit is used for being connected with industrial equipment, the control branch comprises a single chip microcomputer, an output isolation optocoupler and a driving circuit which are connected in sequence, and the output end of the driving circuit is connected with the control end of each power conversion circuit. The input isolation circuit is arranged at the access end of the direct-current power supply to isolate interference among different direct-current power supplies in the same channel, and meanwhile, the single-chip microcomputer and the driving circuit are arranged in the control branch to control the output power of each channel under the power demand data input by a user, so that the user experience is improved. And the power requirements of different industrial equipment are met.

The invention discloses a graph reinforcement learning-based distributed power supply-containing power distribution network fault recovery method, and relates to the field of power system optimization scheduling, in particular to a graph reinforcement learning-based distributed power supply-containing power distribution network fault recovery method. The method comprises the following steps: firstly, establishing a power distribution system topological graph model, constructing a multi-agent decision process, and realizing agent and environment interaction on the basis of power flow analysis and topological verification; secondly, abstracting the power distribution network into graph data, extracting topology and electrical characteristic information by utilizing a graph neural network, and realizing a dynamic recovery decision by combining a deep Q network; and meanwhile, a priority experience playback and sparse reward allocation mechanism is designed, so that the training efficiency and the strategy convergence are improved. According to the method, key load recovery and network reconstruction can be quickly realized under complex topological change and constraint conditions, and the toughness of the power distribution network is remarkably improved.