Scheduling sharing system and method based on distributed energy storage

Abstract

The invention discloses a scheduling sharing system and method based on distributed energy storage. The system comprises an intelligent management and control platform, a regional scheduling confluence node and a plurality of local scheduling shared energy storage devices. The local energy storage device comprises an integrated energy storage module, an equipment management control module and a communication interaction module. According to the system, information data of a user side is collected, regional resource information is combined, a planning model is constructed through deep reinforcement learning, an optimal charging scheme under time-of-use electricity price is generated with the purpose of minimizing a charging coefficient, and peak load shifting is achieved. The block chain technology and the composite key encryption protocol guarantee secure transmission and storage of data, and the server is responsible for ciphertext management and verification. The scheduling method comprises the steps of resource information acquisition, charging scheme making, optimal scheduling and the like, the decision process is optimized through Markov decision and reward function iteration, the energy storage resource utilization rate is increased, the power grid stability is enhanced, and user data privacy and scheduling efficiency are ensured.

Description

Technical Field

[0001] This invention relates to the field of dispatch and shared energy storage, and in particular to a dispatch and shared system and method based on distributed energy storage. Background Technology

[0002] With the booming development of distributed energy systems, energy storage technology, as a crucial supporting component, is receiving increasing attention. Against the backdrop of energy structure transformation, shared energy storage, as an emerging economic model, is attracting growing interest and application. In recent years, with the diversification of various stakeholders in the energy system and the continuous enrichment of energy production and exchange methods, the advantages of shared energy storage in energy utilization, flexible response, and multi-party coordination have gradually become apparent, making it an attractive and economical solution for the power system.

[0003] The core value of shared energy storage lies not only in its ability to coordinate dynamic resources across power generation, grid, load, and storage, achieving a higher level of flexible allocation and supply-demand balance, but also in its effective support for renewable energy. By rationally configuring energy storage facilities, the intermittency and volatility of renewable energy generation can be mitigated, wind and solar power curtailment can be reduced, and local consumption of renewable energy can be promoted, thereby improving the efficiency of clean energy utilization and the economics of system operation. However, the development of shared energy storage currently faces many practical challenges, mainly manifested in the immaturity of collaborative dispatch mechanisms, weak unified planning and optimized operation capabilities for energy storage resources, resulting in generally low utilization rates of energy storage facilities. In some regions, energy storage projects remain under decentralized management, lacking effective integration and cross-entity collaboration, failing to fully leverage economies of scale and complementary effects.

[0004] Against this backdrop, effectively promoting the coordinated scheduling of shared energy storage and improving the overall utilization efficiency of energy storage has become a key issue in driving high-quality energy development. (Invention Content)

[0005] To address the shortcomings of existing technologies, this invention provides a scheduling and sharing system and method based on distributed energy storage, with the aim of achieving coordinated scheduling and improving energy storage utilization.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A dispatch and sharing system based on distributed energy storage includes: an intelligent management and control platform, a regional dispatch and convergence node, and multiple local dispatch and sharing energy storage devices. The intelligent management and control platform is connected to the regional dispatch and convergence node, and the regional dispatch and convergence node is connected to the multiple local dispatch and sharing energy storage devices. The regional dispatch and convergence node is used to determine information data within the dispatchable unit based on resource data information. The information data includes: initial SOC, expected end SOC, battery capacity, and charging power.

[0008] The locally dispatched shared energy storage device includes an integrated energy storage module, an equipment management and control module, and a communication interaction module. The integrated energy storage module is used to charge or discharge according to the charging scheme. The equipment management and control module is used to determine the charging scheme based on the information data and the user-side information data, using a planning model. The charging scheme is determined based on time-of-use pricing to minimize the charging coefficient for each charging period. The charging scheme is used to transfer peak charging loads to achieve peak shaving and valley filling. The planning model is determined by using deep reinforcement learning methods, iteratively optimizing and solving the user-side demand response based on the user-side information data and the user-side information data with the goal of minimizing the charging coefficient. The communication interaction module is used to acquire user-side information data and transmit it to the equipment management and control module. The user-side information data includes: charging request data, current state of charge data, expected final state of charge data, and minimum acceptable state of charge data.

[0009] The sharing system also includes a server and a blockchain. The blockchain is connected to the communication interaction module and the device management and control module, respectively. The server is connected to the communication interaction module. The communication interaction module uses a key generated based on a composite key exchange protocol to encrypt user-end information data, uploads the generated ciphertext to the server, and submits data to the blockchain to publish transaction information. The composite key exchange protocol is determined based on the Diffie-Hellman protocol and the SHA-512 function.

[0010] The device management and control module uses a chip integrating MCU and / or FPGA, and the communication interaction module uses a wireless protocol based on 5G and / or LoRa to transmit user terminal information data in a data encryption manner.

[0011] The blockchain, based on the data access request, invokes a smart contract to request the server to verify the identity of the control module. Once the verification is successful, the authorized server sends the ciphertext and the encrypted decryption key to the device management control module. The device management control module decrypts the key based on the ciphertext and the encrypted decryption key to obtain the plaintext of the user terminal information data.

[0012] A scheduling and sharing method based on distributed energy storage includes the following steps:

[0013] S1: Obtain resource data information corresponding to each regional station issued by the intelligent management and control platform;

[0014] S2: Based on the resource data information, the regional scheduling convergence node determines the information data within the schedulable unit;

[0015] S3: For any locally scheduled shared energy storage device, obtain user terminal information data based on the communication interaction module;

[0016] S4: Based on the information data and the user terminal information data, determine the charging scheme based on the planning model;

[0017] S5: Based on the integrated energy storage module, charge or discharge according to the charging scheme;

[0018] S6: Based on all charging schemes, the regional scheduling convergence node uses an optimization model to optimize the charging scheduling and obtain an optimized charging scheduling scheme.

[0019] Preferably, the method for determining the planning model described in S4 specifically includes the following steps:

[0020] S41: Based on user-side information data and user-side demand response determined with the goal of minimizing the charging coefficient, Markov decision-making is used to transform the response into a decision at discrete time steps to obtain the transformation result.

[0021] S42: Obtain initial parameter data;

[0022] S43: For each time step, each agent selects action data from the action set based on the current local observations, and determines the current cumulative expected reward based on the reward function:

[0023] S44: Based on the current cumulative expected reward, the action data and the current corresponding local observation are used as local experience and stored in the replay buffer;

[0024] S45: Sample batch data from the playback buffer, update the weight data based on the loss function, and obtain the updated weights;

[0025] S46: Determine the planning model based on updated weights and local experience, and update the global state and local observations according to all selected action data to obtain the charging scheme.

[0026] Preferably, the expression corresponding to the charging coefficient mentioned in S1 is:

[0027] ,

[0028] in, The charging coefficient; For the first End time of each schedulable unit; For the first Start time of each schedulable unit; For the first Average charging power within each schedulable unit; For the first A binary variable at time; For the first Electricity prices at any given time.

[0029] The optimization model described in S6 includes an objective function and constraints. The objective function is a power grid load fluctuation function, and its expression is as follows:

[0030] ,

[0031] in, For distribution network load fluctuations; The scheduling period is divided for the charging and discharging process; For the distribution network in Total load over the time period; This represents the average load of the distribution network during the charging and discharging process. for The moment within a time period; for The time sequence number within the time period.

[0032] Preferably, the constraints include: charging demand constraints, energy storage capacity constraints, grid output power constraints, and remaining load constraints;

[0033] The expression for the charging demand constraint is:

[0034] ,

[0035] The expression for the energy storage capacity constraint is:

[0036] ,

[0037] The expression for the power grid output power constraint is:

[0038] ,

[0039] ,

[0040] The expression for the remaining charge constraint is:

[0041] ,

[0042] in, For the first End time of each schedulable unit; For the first Start time of each schedulable unit; In the first Within the schedulable unit, the first The charging power at any given time; For the first Total energy billed within each schedulable unit; For energy storage capacity; This represents the maximum discharge power. This is the maximum charging power; Predict output power for wind power; Rated wind power; Predicted output power for photovoltaics; Photovoltaic and wind power; This represents the remaining charge.

[0043] Preferably, the method for scheduling shared energy storage further includes encrypting and transmitting user-end information data using a key generated based on a composite key exchange protocol, based on a server and a blockchain.

[0044] This invention provides a dispatch and sharing system and method based on distributed energy storage, which has the following beneficial effects:

[0045] 1. This invention relates to a dispatch and sharing system and method based on distributed energy storage. The system includes an intelligent management and control platform, regional dispatch convergence nodes, and locally dispatched shared energy storage devices. The intelligent management and control platform distributes resource data information corresponding to each regional site. The regional dispatch convergence nodes determine the information data within the dispatchable unit based on the resource data information. The locally dispatched shared energy storage device includes an integrated energy storage module, an equipment management and control module, and a communication interaction module. The equipment management and control module determines a charging scheme based on the information data and user terminal information data, using a planning model. The charging scheme is used to transfer peak charging loads to achieve peak shaving and valley filling. The planning model is determined using a deep reinforcement learning method. The integrated energy storage module charges or discharges according to the charging scheme. The regional dispatch node performs charging scheduling optimization processing based on an optimization model according to all charging schemes to obtain an optimized charging scheduling scheme, which is used to achieve real-time charging scheduling between renewable energy consumption and local energy storage, enabling balanced optimization.

[0046] 2. This invention integrates deep reinforcement learning methods into the determination of charging schemes and employs an optimization model for charging scheduling optimization, enabling coordinated scheduling among intelligent management and control platforms, regional scheduling convergence nodes, and multiple locally scheduled shared energy storage devices. Furthermore, the charging scheme determined based on the planning model is used to transfer peak charging loads to achieve peak shaving and valley filling, thereby improving energy storage utilization. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a structural diagram of a dispatch and sharing system based on distributed energy storage;

[0049] Figure 2 This is a flowchart of a scheduling and sharing method based on distributed energy storage.

[0050] Marked in the image:

[0051] 1-Intelligent management and control platform; 2-Regional dispatch convergence node; 3-Local dispatch shared energy storage device; 31-Integrated energy storage module; 32-Equipment management and control module; 33-Communication interaction module; 4-Server; 5-Blockchain. Detailed Implementation

[0052] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0053] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0054] Example 1

[0055] like Figure 1 As shown, the present invention provides a dispatch and sharing system based on distributed energy storage, including: an intelligent management and control platform 1, a regional dispatch and convergence node 2, and multiple local dispatch and sharing energy storage devices 3. The intelligent management and control platform 1 is connected to the regional dispatch and convergence node 2, and the regional dispatch and convergence node 2 is connected to multiple local dispatch and sharing energy storage devices 3. The regional dispatch and convergence node 2 is used to determine the information data within the dispatchable unit based on resource data information. The information data includes: initial SOC, expected end SOC, battery capacity, and charging power.

[0056] The locally dispatched shared energy storage device includes an integrated energy storage module 31, an equipment management and control module 32, and a communication interaction module 33. The integrated energy storage module 31 is used to charge or discharge according to the charging scheme. The equipment management and control module 32 is used to determine the charging scheme based on information data and user-side information data, using a planning model. The charging scheme is determined based on time-of-use pricing to minimize the charging coefficient for each charging period. The charging scheme is used to transfer peak charging loads to achieve peak shaving and valley filling. The planning model is determined by using deep reinforcement learning methods, iteratively optimizing and solving the user-side demand response based on user-side information data and the user-side information data with the goal of minimizing the charging coefficient. The communication interaction module 33 is used to acquire user-side information data and transmit it to the equipment management and control module 32. The user-side information data includes: charging request data, current state of charge data, expected final state of charge data, and minimum acceptable state of charge data.

[0057] In this embodiment, the local scheduling shared energy storage device 3 includes an integrated energy storage module 31, an equipment management and control module 32, and a communication interaction module 33. The equipment management and control module 32 adopts a chip with integrated MCU and / or FPGA.

[0058] The intelligent management and control platform 1 is used to distribute resource data information corresponding to each regional station; the resource data includes: output data of renewable energy generator sets, storage and release data of energy storage systems, and controllable load data.

[0059] The regional scheduling confluence node 2 is used to determine the information data within the schedulable unit based on resource data information; the information data includes: initial SOC, expected end SOC, battery capacity and charging power; the schedulable unit includes: charging phase and idle phase.

[0060] The local scheduling shared energy storage device 3 includes a communication interaction module 33 for acquiring user terminal information data and transmitting it to the device management and control module 32; the user terminal information data includes charging request data, current state of charge data, expected final state of charge data, and minimum acceptable state of charge data.

[0061] The equipment management and control module 32 is used to determine the charging scheme based on the information data and user terminal information data, and the charging scheme is determined based on the time-of-use electricity price to minimize the charging coefficient of each charging period. The charging scheme is used to transfer the peak charging load to achieve peak shaving and valley filling. The planning model is determined by using deep reinforcement learning method, based on the user terminal information data and the user-side demand response determined based on the user terminal information data with the goal of minimizing the charging coefficient, and then iteratively optimizing the solution.

[0062] The integrated energy storage module 31 is used to charge or discharge according to the charging scheme; the regional dispatching junction node 2 is also used to perform charging scheduling optimization based on the optimization model according to all charging schemes to obtain the optimized charging scheduling scheme; the optimized charging scheduling scheme is used to realize real-time charging scheduling between renewable energy consumption and local energy storage, and realize energy balance optimization.

[0063] Furthermore, the communication interaction module 33 uses a wireless protocol based on 5G and / or LoRa to transmit user terminal information data in a data encryption manner.

[0064] The dispatch and sharing system based on distributed energy storage also includes: server 4 and blockchain 5. Blockchain 5 is connected to communication interaction module 33 and equipment management and control module 32 respectively; server 4 is connected to communication interaction module 33.

[0065] The communication interaction module 33 uses a key generated based on a composite key exchange protocol to encrypt user terminal information data and uploads the generated ciphertext to the server 4. At the same time, it submits data to the blockchain to publish transaction information. The composite key exchange protocol is determined based on the Diffie-Hellman protocol and the SHA-512 function. The key includes a public key and a private key.

[0066] The device management and control module 32 sends a data access request to the blockchain based on the data publishing transaction information. Blockchain 5, in accordance with the data access request, invokes a smart contract to request verification of the control module's identity from server 4. Upon successful verification, server 4 sends the ciphertext and the encryption / decryption key to the device management and control module 32. The device management and control module 32 decrypts the key using the ciphertext and the encryption / decryption key to obtain the plaintext of the user-end information data. Meanwhile, the communication interaction module 33 uses a key generated based on a composite key exchange protocol to encrypt the user-end information data and uploads the generated ciphertext to server 4. Simultaneously, it submits data publishing transaction information to blockchain 5, specifically including:

[0067] Communication interaction module 33 is based on formula calculate value; This is the private key for the communication interaction module 33; This is the private key for the device management and control module 32; It is the original root; This is a secret value shared between the communication interaction module 33 and the device management and control module 32.

[0068] according to The value is calculated from 128 strings; among them, ; ; For the first A string; It is a prime number.

[0069] Divide the key into 128 consecutive 8-bit blocks. ;in, ; ; For the first An 8-bit block.

[0070] SHA-512 function continuously This is applied to the string. , to obtain the string after application ; 128 of 512 bits The strings are concatenated to obtain the encryption key; the encryption key is used to encrypt the user-end information data, and the generated ciphertext is uploaded to server 4. At the same time, the data is submitted to blockchain 5 to publish transaction information.

[0071] Equipment management and control module 32 according to the formula calculate Value; based on The value generates 128 strings. , The encryption key is divided into 512-bit blocks, resulting in 128 strings. Applying the SHA-512 function until the result equals Until then, subtract 1 from the number of times needed to obtain the equal value, and get Value; 128 8-bit values The bits are connected to obtain the decryption key; the key is then decrypted according to the decryption key to obtain the plaintext of the user terminal information data, so as to access the data and realize data sharing.

[0072] Example 2

[0073] This invention provides a scheduling and sharing method based on distributed energy storage, which is implemented using a distributed energy storage-based scheduling and sharing system. For example... Figure 2 As shown, the scheduling and sharing method based on distributed energy storage includes the following steps:

[0074] S1: Obtain resource data information corresponding to each regional station from the intelligent management and control platform 1. The resource data includes: output data of renewable energy generator sets, storage and release data of distributed energy storage systems, and controllable load data.

[0075] S2: Based on the regional scheduling convergence node 2, the information data within the schedulable unit is determined according to the resource data information. The information data includes: initial SOC, expected end SOC, battery capacity, and charging power; the schedulable unit includes: charging phase and idle phase;

[0076] S3: For any locally scheduled shared energy storage device 3, acquire user-end information data based on the communication interaction module 33. The user-end information data includes: charging request data, current state of charge data, expected final state of charge data, and minimum acceptable state of charge data;

[0077] S4: Based on information data and user-side information data, a charging scheme is determined based on a planning model. The charging scheme is determined based on time-of-use pricing to minimize the charging coefficient during each charging period. The charging scheme is used to transfer peak charging loads to achieve peak shaving and valley filling. The planning model is determined by using deep reinforcement learning methods, iteratively optimizing and solving the user-side demand response based on user-side information data and the goal of minimizing the charging coefficient.

[0078] S5: Based on the integrated energy storage module 31, charge or discharge according to the charging scheme.

[0079] S6: Based on the regional scheduling convergence node 2, the charging scheduling is optimized using an optimization model according to all charging schemes to obtain an optimized charging scheduling scheme. The optimized charging scheduling scheme is used to realize real-time charging scheduling between renewable energy consumption and local energy storage, thereby achieving energy balance optimization. The optimization model includes objective function and constraints.

[0080] The optimization model in S6 includes an objective function and constraints. The objective function is the power grid load fluctuation function, and its expression is as follows:

[0081] ,

[0082] in, For distribution network load fluctuations; The scheduling period is divided for the charging and discharging process; For the distribution network in Total load over the time period; This represents the average load of the distribution network during the charging and discharging process. for The moment within a time period; for The time sequence number within the time period.

[0083] The constraints include: charging demand constraints, energy storage capacity constraints, grid output power constraints, and remaining load constraints.

[0084] The expression for the charging demand constraint is:

[0085] ,

[0086] The expression for the energy storage capacity constraint is:

[0087] ,

[0088] The expression for the power grid output constraint is:

[0089] ,

[0090] ,

[0091] The expression for the residual charge constraint is:

[0092] ,

[0093] in, For the first End time of each schedulable unit; For the first Start time of each schedulable unit; In the first Within the schedulable unit, the first The charging power at any given time; For the first Total energy billed within each schedulable unit; For energy storage capacity; This represents the maximum discharge power. This is the maximum charging power; Predict output power for wind power; Rated wind power; Predicted output power for photovoltaics; Photovoltaic and wind power; This represents the remaining charge.

[0094] Furthermore, the method for determining the planning model in S4 specifically includes the following steps:

[0095] S41: Based on user-side information data and the user-side demand response determined with the goal of minimizing the charging coefficient, Markov decision-making is used to transform the response into a decision at discrete time steps, resulting in the transformation result. The transformation result includes: agent, global state, local observations, action set, reward function, and state transition function. The agent represents the user-side demand response participants at the user end; the global state is the user-side demand response state at the user end; local observations include the active power load of nodes after the demand response; the action set is the set of power layer's allocation and management of power flow in the distribution network.

[0096] In this step, the expression for the charging coefficient is:

[0097] ,

[0098] in, The charging coefficient; For the first End time of each schedulable unit; For the first Start time of each schedulable unit; For the first Average charging power within each schedulable unit; For the first A binary variable at time; For the first Electricity prices at any given time.

[0099] S42: Obtain initial parameter data; the initial parameter data includes: weight data, local observations, action set, and reward function; the local observations, action set, and reward function are stored in the replay buffer.

[0100] S43: For each time step, each agent selects action data from the action set based on the current local observations, and determines the current cumulative expected reward based on the reward function:

[0101] S44: Based on the current cumulative expected reward, the action data and the current corresponding local observation are used as local experience and stored in the replay buffer;

[0102] S45: Sample batch data from the playback buffer, update the weight data based on the loss function, and obtain the updated weights;

[0103] S46: Determine the planning model based on updated weights and local experience, and update the global state and local observations according to all selected action data to obtain the charging scheme.

[0104] S42 reward function The expression is:

[0105]

[0106] in, It is a discount factor. This is the attenuation coefficient. For the first An agent at time step Receive a reward.

[0107] The state transition function represents the state at the next time step. The corresponding state transition is determined by the following formula:

[0108]

[0109] State transitions are affected by the current state of the environment. Local observations of all intelligent agents and the Actions before a time step and environmental random parameters The combination of factors has an impact.

[0110] The loss function uses the soft Q value estimated by the value network, and soft updates (soft target networks) are typically used to avoid overestimation.

[0111] For multiple locally dispatched shared energy storage devices, edge computing is used to reduce the latency of multi-user concurrent dispatch, so as to achieve collaborative dispatch among the intelligent management and control platform 1, the regional dispatch convergence node 2, and multiple locally dispatched shared energy storage devices 3.

[0112] The scheduling and sharing method based on distributed energy storage also includes: using a key generated based on a composite key exchange protocol to encrypt and transmit user-end information data based on server 4 and blockchain 5.

[0113] The scheduling and sharing system and method based on distributed energy storage provided by this invention can be applied to different scenarios such as home microgrids, industrial parks, and electric vehicle charging stations. The scheduling, sharing and encrypted transmission provided by this invention comply with standards such as IEEE 1547 (distributed power grid connection) and IEC 61850 (communication protocol).

[0114] Taking electric vehicle charging stations as an example, in order to reduce the impact of electric vehicle charging on the overall power load, especially during peak load periods, this invention formulates a mathematical optimization-based charging (MOC) scheduling scheme as an optimization problem with the goal of minimizing power load changes, and extends it by incorporating the real-time scheduling process of each plug-in electric vehicle.

[0115] Set the scheduling period to The data used for the basic power load curve (excluding electric vehicle charging load) for the entire dispatch cycle is real data from a certain area, including 9,088 households and 670 electric vehicles, with an electric vehicle penetration rate of 7.4%. The power supply capacity is 17,760 kW / day, and the historical maximum daily load factor is 67%.

[0116] Based on charging data samples collected from an electric vehicle big data platform and combined with electric vehicle charging behavior, a time-of-use pricing intelligent charging method is proposed. The importance of identifying idle parking time and its ratio to required charging time is emphasized. By examining the sequence and timing of parking and charging events, it can quantify the potential of electric vehicles to change charging load. Utilizing one hour of idle parking time, up to 0.45 kWh of net charging load can be transferred from peak hours, and 0.54 kWh can be transferred to off-peak hours for slow charging. For fast charging, the transferred load during peak and off-peak hours reaches 0.55 kWh and 0.38 kWh, respectively.

[0117] Based on real charging data from some charging stations in a certain city, a data sample of electric vehicle charging behavior was established. Each data sample includes battery capacity, initial charging state (SOC), end charging state (end SOC), start charging time, end charging time, and parking time.

[0118] Most fast-charging electric vehicles can reach their target SOC within 3 hours and remain charged for more than 1 hour after completion, while most slow-charging electric vehicles can reach their target SOC within 8 hours and remain charged for more than 10 hours after completion.

[0119] The idle time after electric vehicle charging is completed provides an opportunity for charging scheduling. Utilizing this idle time can, to some extent, shift the charging load of electric vehicles and reduce the impact of electric vehicle charging on the power grid.

[0120] Based on charging behavior data samples collected from an electric vehicle big data platform, and combined with electric vehicle charging behavior, a real-time charging scheduling method for electric vehicles that integrates renewable energy and V2G can be implemented. Using real-world data for validation, the overall optimization objective of charging stations is influenced by adjusting the charging and discharging power of individual electric vehicles. This scheduling method reduces distribution network load fluctuation by 3.94%, reduces the average charging cost of electric vehicles by 1.18 CNY, reduces the average real-time energy consumption difference by 0.47%, and improves scheduling efficiency by 30%.

[0121] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dispatch and sharing system based on distributed energy storage, characterized in that, include: The system includes an intelligent management and control platform (1), a regional dispatch convergence node (2), and multiple local dispatch shared energy storage devices (3). The intelligent management and control platform (1) is connected to the regional dispatch convergence node (2), and the regional dispatch convergence node (2) is connected to multiple local dispatch shared energy storage devices (3). The regional scheduling confluence node (2) is used to determine the information data within the schedulable unit based on the resource data information. The information data includes: initial SOC, expected end SOC, battery capacity, and charging power. The local scheduling shared energy storage device includes an integrated energy storage module (31), an equipment management and control module (32), and a communication interaction module (33). The integrated energy storage module (31) is used to charge or discharge according to the charging scheme. The equipment management and control module (32) is used to determine the charging scheme based on the information data and the user terminal information data, using a planning model. The charging scheme is determined based on time-of-use pricing to minimize the charging coefficient for each charging period. The charging scheme is used to transfer peak charging loads to achieve peak shaving and valley filling. The planning model is determined by using deep reinforcement learning methods, iteratively optimizing and solving the user-side demand response based on the user terminal information data and the user-side demand response determined with the goal of minimizing the charging coefficient. The communication interaction module (33) is used to acquire user terminal information data and transmit it to the equipment management and control module. The user terminal information data includes: charging request data, current state of charge data, expected final state of charge data, and minimum acceptable state of charge data.

2. The dispatch and sharing system based on distributed energy storage according to claim 1, characterized in that, The sharing system also includes a server (4) and a blockchain (5). The blockchain (5) is connected to the communication interaction module (33) and the device management and control module (32) respectively. The server (4) is connected to the communication interaction module (33). The communication interaction module (33) uses a key generated based on a composite key exchange protocol to encrypt user terminal information data and uploads the generated ciphertext to the server (4) and submits data release transaction information to the blockchain (5). The composite key exchange protocol is determined based on the Diffie-Hellman protocol and the SHA-512 function.

3. A dispatch and sharing system based on distributed energy storage according to claim 1, characterized in that, The device management and control module (32) uses a chip that integrates MCU and / or FPGA, and the communication interaction module (33) is based on the 5G and / or LoRa wireless protocol and uses data encryption to transmit user terminal information data.

4. A dispatch and sharing system based on distributed energy storage according to claim 2, characterized in that, The blockchain (5) calls the set smart contract to apply to the server (4) for verification of the identity of the control module according to the data access request. When the verification is successful, the authorized server (4) sends the ciphertext and the encrypted decryption key to the device management control module (32). The device management control module (32) decrypts the key according to the ciphertext and the encrypted decryption key to obtain the plaintext of the user terminal information data.

5. A scheduling and sharing method based on distributed energy storage according to any one of claims 1-4, characterized in that, Includes the following steps: S1: Obtain the resource data information corresponding to each regional station issued by the intelligent control platform (1); S2: Based on the regional scheduling convergence node (2), the information data within the schedulable unit is determined according to the resource data information; S3: For any local scheduling shared energy storage device (3), obtain user terminal information data based on the communication interaction module (33); S4: Based on the information data and the user terminal information data, determine the charging scheme based on the planning model; S5: Based on the integrated energy storage module (31), charge or discharge according to the charging scheme; S6: Based on the regional scheduling convergence node (2), according to all the charging schemes, the optimization model is used to perform charging scheduling optimization processing to obtain the optimized charging scheduling scheme.

6. The scheduling and sharing method for energy storage devices according to claim 5, characterized in that, The method for determining the planning model described in S4 specifically includes the following steps: S41: Based on user-side information data and user-side demand response determined with the goal of minimizing the charging coefficient, Markov decision-making is used to transform the response into a decision at discrete time steps to obtain the transformation result. S42: Obtain initial parameter data; S43: For each time step, each agent selects action data from the action set based on the current local observations, and determines the current cumulative expected reward based on the reward function: S44: Based on the current cumulative expected reward, the action data and the current corresponding local observation are used as local experience and stored in the replay buffer; S45: Sample batch data from the playback buffer, update the weight data based on the loss function, and obtain the updated weights; S46: Determine the planning model based on updated weights and local experience, and update the global state and local observations according to all selected action data to obtain the charging scheme.

7. The scheduling and sharing method based on distributed energy storage according to claim 6, characterized in that, The expression corresponding to the charging coefficient mentioned in S41 is: ， in, The charging coefficient; For the first End time of each schedulable unit; For the first Start time of each schedulable unit; For the first Average charging power within each schedulable unit; For the first A binary variable at time; For the first Electricity prices at any given time.

8. The scheduling and sharing method based on distributed energy storage according to claim 5, characterized in that, The optimization model described in S6 includes an objective function and constraints. The objective function is a power grid load fluctuation function, and its expression is as follows: ， in, For distribution network load fluctuations; The scheduling period is divided for the charging and discharging process; For the distribution network in Total load over the time period; This represents the average load of the distribution network during the charging and discharging process. for The moment within a time period; for The time sequence number within the time period.

9. The scheduling and sharing method based on distributed energy storage according to claim 8, characterized in that, The constraints include: charging demand constraints, energy storage capacity constraints, grid output power constraints, and remaining load constraints. The expression for the charging demand constraint is: ， The expression for the energy storage capacity constraint is: , The expression for the power grid output power constraint is: ， ， The expression for the remaining charge constraint is: ， in, For the first End time of each schedulable unit; For the first Start time of each schedulable unit; In the first Within the schedulable unit, the first The charging power at any given time; For the first Total energy billed within each schedulable unit; For energy storage capacity; This represents the maximum discharge power. This is the maximum charging power; Predict output power for wind power; Rated wind power; Predicted output power for photovoltaics; Photovoltaic and wind power; This represents the remaining charge.

10. The scheduling and sharing method based on distributed energy storage according to claim 5, characterized in that, The method for scheduling shared energy storage also includes encrypting and transmitting user terminal information data using a key generated based on a composite key exchange protocol, based on a server (4) and a blockchain (5).

Images