Energy storage data non-inductive access system and method based on RPC proxy and distributed storage

Abstract

The invention discloses an energy storage data non-inductive access system and method based on RPC proxy and distributed storage. The system comprises a plurality of slave cabinets and a master cabinet. The slave cabinets store time sequence data and self information including cabinet identifiers, cabinet websites and database access ports in a local database, and periodically report the time sequence data through multicast; the master cabinet does not persistently store slave cabinet data, and a slave cabinet discovery module of the master cabinet monitors multicast and automatically generates and maintains a dynamic slave cabinet information list containing cabinet identifiers of the slave cabinets and database connection addresses; and the energy storage application layer initiates an SQL access request by calling a unified RPC interface of the main cabinet, an embedded database access proxy engine analyzes the SQL and queries a list, establishes connection with a target slave cabinet database, forwards and executes, and returns a result through the RPC. According to the invention, distributed storage and transparent unified non-inductive access of data are realized, capacity expansion of a main cabinet disk is avoided fundamentally, application development is simplified, and plug-and-play expansion of the system is supported.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage system data management, in particular to an energy storage data non-inductive access system and method based on RPC (Remote Procedure Call) agent and distributed storage. BACKGROUND

[0002] The existing industrial and commercial energy storage system is usually composed of multiple (such as 2 to 8) energy storage cabinets, one of which is used as a master cabinet, and the rest are used as slave cabinets. Each slave cabinet needs to independently collect real-time running information of physical devices such as BMS (Battery Management System), PCS (Power Conversion System) and the like in the cabinet, and parse and convert them into time series data. The time series data needs to be stored for subsequent processing and reading.

[0003] The traditional data storage scheme adopts centralized storage, that is, all time series data obtained from the slave cabinets are stored in the central database of the master cabinet. This scheme has significant defects: the storage disk capacity of the master cabinet becomes the bottleneck of the system scale, and as the number of slave cabinets increases, the storage disk capacity requirement of the master cabinet is uncertain and continuously increasing, frequent disk expansion is required, resulting in high hardware cost of the master cabinet and poor system scalability. In addition, centralized storage also brings single-point performance bottleneck and single-point failure risk, and once the central database of the master cabinet fails, it will affect the data access of the entire system.

[0004] To solve the technical problems caused by centralized storage, the existing technology adopts a distributed storage architecture, that is, after each slave cabinet parses and converts the real-time running information of the physical devices in the cabinet into time series data, it is stored in the local database. However, under this architecture, the common implementation method requires the application layer program to directly connect to the local database of each slave cabinet. This means that the application layer directly faces the slave cabinets, and the application layer developers must know the network topology, address and authentication information of all slave cabinets, resulting in complex application layer programming, tedious implementation, and high coupling with the underlying data. In addition, when the number of slave cabinets changes dynamically, the application layer needs to frequently modify the configuration and connection logic, making maintenance extremely difficult. SUMMARY

[0005] The technical problem to be solved by the present application is to provide an energy storage data non-inductive access system and method based on RPC agent and distributed storage, which can provide the energy storage application layer with unified, transparent and elastically expandable non-inductive access capability for distributed data in the entire network without relying on centralized storage of the master cabinet.

[0006] The technical solution adopted by the present application to solve the above technical problems is: In a first aspect, the present invention provides a seamless data access system for energy storage based on RPC proxy and distributed storage, comprising: Multiple slave cabinets, each slave cabinet is equipped with: A local database is used to store at least the time-series data acquired by this cabinet and the cabinet's own information, which includes at least the cabinet identifier, the cabinet's URL, and the database access port. The multicast module is used to periodically send notification messages containing at least its own information to the network via multicast communication; One main cabinet, equipped with: The slave cabinet discovery module is used to listen to the multicast communication to discover online slave cabinets. Based on the cabinet URL and database access port in the notification message, it generates the corresponding database connection address and maintains a dynamic slave cabinet information list that records the cabinet identifier and database connection address of online slave cabinets. The RPC proxy module provides at least one RPC interface and has an embedded database access proxy engine. The energy storage application layer is used to initiate data access requests containing SQL statements by calling the RPC interface; The database access proxy engine is configured as follows: Receive data access requests and parse SQL statements to extract cabinet identifiers; Based on this, query the dynamic slave cabinet information list and determine the database connection address of the target slave cabinet; Based on the database connection address, establish a connection with the target cabinet's local database and forward the SQL statement for execution; Receive the execution result and return it via the RPC interface.

[0007] In some embodiments, each of the slave cabinets is further provided with a data acquisition module for collecting real-time operating information of the physical devices inside the cabinet; the time-series data is generated based on the real-time operating information.

[0008] In some embodiments, the multicast module is configured to periodically send the notification message to a preset UDP multicast address based on the UDP multicast protocol; the notification message is in JSON format, and its content includes at least the cabinet identifier, the cabinet URL, and the database access port.

[0009] In some embodiments, parsing the SQL statement specifically includes: parsing the table name prefix in the SQL statement, the table name prefix containing the cabinet identifier; and the database access agent engine establishing a connection with the local database of the target cabinet via JDBC or ODBC protocol.

[0010] In some embodiments, the master cabinet is further provided with: a data acquisition module, a local database, and a multicast module that are logically identical to those of the slave cabinet.

[0011] Secondly, the present invention provides a method for seamless access to energy storage data based on RPC proxy and distributed storage, applied to the aforementioned system, the method comprising the following steps: S1. The multicast module of the slave cabinet periodically sends notification messages containing at least its own information to the network through multicast communication. The own information includes at least the cabinet identifier, the cabinet URL, and the database access port. S2. The slave cabinet discovery module of the master cabinet listens to the multicast communication to discover online slave cabinets. Based on the cabinet URL and database access port in the notification message, it generates the corresponding database connection address and maintains a dynamic slave cabinet information list that records the cabinet identifier and database connection address of online slave cabinets. S3. The energy storage application layer initiates a data access request containing SQL statements by calling the RPC interface of the RPC proxy module of the main cabinet; S4. The database access proxy engine in the RPC proxy module receives the data access request, parses the SQL statement to extract the cabinet identifier, and queries the dynamic slave cabinet information list accordingly to determine the database connection address of the target slave cabinet. S5. The database access proxy engine establishes a connection with the local database of the target slave cabinet according to the database connection address, forwards the SQL statement for execution, and returns the execution result through the RPC interface after receiving the execution result.

[0012] In some embodiments, the method further includes: The slave cabinet collects real-time operating information of the physical devices inside the cabinet through its data acquisition module, and generates time-series data based on the real-time operating information, which is then stored in the slave cabinet's local database.

[0013] In some embodiments, the notification message is a JSON format message sent based on the UDP multicast protocol, and its content includes at least the cabinet identifier, the cabinet URL, and the database access port.

[0014] In some embodiments, parsing the SQL statement to extract the cabinet identifier of the target cabinet includes: parsing the table name prefix in the SQL statement, which contains the cabinet identifier, i.e., extracting the cabinet identifier of the target cabinet; the database access agent engine establishes a connection with the local database of the target cabinet through JDBC or ODBC protocol.

[0015] In some embodiments, the method further includes: When a new slave cabinet is connected to the system, step S1 is executed automatically; The above step S2 is executed automatically, adding the cabinet identifier and database connection address of the newly added slave cabinet to the dynamic slave cabinet information list; Subsequently, the energy storage application layer can access its time-series data through an SQL statement containing the cabinet identifier of the newly added slave cabinet without modifying the configuration.

[0016] Compared with the prior art, the advantages of the present invention are as follows: 1) Completely solves the problem of main cabinet storage disk expansion: This invention sets up a local database in the slave cabinets, which is the key to realizing distributed data storage and avoiding centralized storage in the main cabinet. Adopting a distributed data storage architecture, time-series data is stored in the local databases of their respective slave cabinets, while the main cabinet does not centrally or persistently store the time-series data of the slave cabinets, but only acts as a proxy and coordination center. Therefore, when the system scales up (adding slave cabinets), it will not put any pressure on the storage disk capacity of the main cabinet, and the main cabinet does not need to expand its storage disk, greatly reducing system hardware costs and operational complexity.

[0017] 2) Achieving seamless access with unified logic: The RPC proxy module in the main cabinet of this invention is the only entry point for the energy storage application layer. The database access proxy engine within the RPC proxy module can parse SQL statements, query dynamic slave cabinet information lists, establish connections with the local database of the target slave cabinet, and forward SQL statements. This allows the energy storage application layer to access the time-series data of the target slave cabinet simply by knowing the URL of the main cabinet, without needing to know the details of the underlying distributed database. It can access the time-series data of the target slave cabinet using only standard SQL statements and a single RPC interface. The access experience is no different from operating a single local database, achieving seamless access and making the programming and implementation of the energy storage application layer simple and convenient.

[0018] 3) System Scalability: The slave cabinet discovery module in the master cabinet of this invention automatically listens for multicast communication, generates database connection addresses, and maintains a dynamic slave cabinet information list, realizing automatic registration of slave cabinets. That is, when a new slave cabinet comes online, it can be automatically recognized by the master cabinet. When the system is expanded, only slave cabinets need to be added. No hardware upgrades or complex software configurations are required for the master cabinet, realizing plug-and-play horizontal expansion.

[0019] 4) System Robustness and Decoupling: In this invention, the master cabinet acts as a unified agent, achieving complete decoupling between the energy storage application layer and the underlying data storage. Even if a single slave cabinet fails, it will not affect the energy storage application layer's access to data from other slave cabinets. The master cabinet has a single responsibility, resulting in high system stability and maintainability.

[0020] 5) Standardized technology and low cost: By adopting standard SQL statements and mature RPC technology, the development threshold of the energy storage application layer is low and the technology migration cost is small, which is conducive to the promotion and application of the technology.

[0021] 6) The database access agent engine in the master cabinet of this invention forwards SQL statements to the local database of the target slave cabinet for execution. The physical I / O pressure of data storage and reading is entirely borne by the target slave cabinet, and the master cabinet only acts as an agent to forward instructions and results.

[0022] 7) In this invention, the master cabinet can be switched to the role of slave cabinet by configuration when necessary, which greatly improves the utilization rate of system hardware resources and deployment flexibility, reduces spare parts costs, and provides a physical basis for high availability deployment of the system (such as master cabinet failure switching). Attached Figure Description

[0023] Figure 1 This is a diagram showing the connection relationships between the modules in the system of this invention; Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0024] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0025] Firstly, this embodiment provides a seamless data access system for energy storage based on RPC proxy and distributed storage, such as... Figure 1 As shown, it includes: Multiple slave cabinets 1, each slave cabinet 1 is equipped with: Local database 11 is used to independently store at least the time-series data acquired by this cabinet and the cabinet's own information. The information includes at least the cabinet identifier, the cabinet URL, and the database access port.

[0026] Multicast module 12 is used to periodically (e.g., every 30 seconds) send notification messages containing at least its own information to the network via multicast communication.

[0027] One master cabinet 2, which does not persistently store the timing data of each slave cabinet 1, and is configured with: The slave cabinet discovery module 21 is used to listen to multicast communication to discover all online slave cabinets 1. Then, based on the cabinet URL and database access port in the notification message, it generates the corresponding database connection address and maintains a dynamic slave cabinet information list (cabinet_info_list) 13 containing the cabinet identifier and database connection address of all online slave cabinets 1. Since the online status of slave cabinet 1 changes, the dynamic slave cabinet information list 13 is also updated accordingly.

[0028] RPC proxy module 22 provides at least one RPC interface 221 and has an embedded database access proxy engine 222.

[0029] The energy storage application layer 3 is used to initiate a data access request containing SQL statements to the database access agent engine 222 by calling the RPC interface 221 in order to access the local database 11 of the slave cabinet 1.

[0030] The database access proxy engine 222 is configured to: receive data access requests and parse SQL statements to extract the cabinet identifier of the target slave cabinet, that is, to extract the cabinet identifier from the SQL statement, and the slave cabinet 1 corresponding to the cabinet identifier is taken as the target slave cabinet; query the dynamic slave cabinet information list 13 according to the cabinet identifier of the target slave cabinet to determine the database connection address of the target slave cabinet; establish a connection with the local database 11 of the target slave cabinet according to the database connection address, and forward the SQL statement to the local database 11 of the target slave cabinet for execution, and the SQL statement remains unchanged during the forwarding process; receive the execution result and return it to the energy storage application layer 3 through the RPC interface 221.

[0031] In practice, the main cabinet 2 and all slave cabinets 1 are usually deployed in the same industrial site. The main cabinet 2 and slave cabinets 1 are connected via Ethernet to form a physically or logically independent dedicated local area network as a cabinet networking network to ensure the reliability of communication and network isolation.

[0032] In some embodiments, each slave cabinet 1 is further provided with a data acquisition module 14, which is used to collect real-time operating information of the physical devices inside the cabinet; the time-series data is generated based on the real-time operating information, and the generated time-series data is stored in the local database 11.

[0033] In one specific embodiment of the present invention, the data acquisition module 14 of cabinet 1 is responsible for acquiring raw real-time operating information from the physical devices within the cabinet, such as total voltage, total current, battery pack temperature, and state of charge (SOC) from the BMS, and operating mode (charging / discharging / standby), switch status, and output power from the PCS. Using conventional time-series data processing methods in the art, the aforementioned sets of real-time operating information are converted into standard time-series data. These time-series data processing methods include, but are not limited to, adding high-precision timestamps, format standardization, and validity verification. After this step, the real-time operating information is organized into structured data with time-series markers and stored in the local database 11.

[0034] In some embodiments, multicast communication is preferably implemented via the UDP multicast protocol. The multicast module 12 is configured to: read the self-information of cabinet 1 from the local database 11; and periodically send notification messages to a preset UDP multicast address and port through the cabinet network based on the UDP (User Datagram Protocol) multicast protocol; the notification messages are in JSON format, and their content includes at least the cabinet identifier (cabinet_id), cabinet URL (ip), database type (db_type), and database access port (port).

[0035] In one specific embodiment of the present invention, slave cabinet 1 connects to the cabinet network via its Ethernet port. Multicast module 12 is configured to send a heartbeat message every 30 seconds via the cabinet network to the UDP multicast address and port "239.255.255.250:9999" using the UDP multicast protocol. The message is in JSON format, with an example content as follows: {"cabinet_id":"03","ip":"192.168.1.103","db_type":"taos","port":"6041","status":"online"}. The propagation range of the UDP multicast address is limited to the cabinet network. Slave cabinet discovery module 21 of master cabinet 2 listens to the same UDP multicast address and port, thereby achieving automatic discovery and status maintenance of all online slave cabinets within the cabinet network.

[0036] In some embodiments, parsing the SQL statement specifically includes: parsing the table name prefix in the SQL statement, which contains the cabinet identifier; and the database access agent engine 222 establishing a connection with the local database 11 of the target cabinet via JDBC or ODBC protocol.

[0037] As an example, suppose the energy storage application layer 3 sends a data access request to the database access agent engine 222 as SELECT voltage, current FROM cabinet_03_bms WHERE ts>'2024-06-01'. The database access agent engine 222 parses the table name prefix "cabinet_03" in the SQL statement. This table name prefix contains the cabinet identifier 03. It queries the dynamic slave cabinet information list 13, finds the entry with the cabinet identifier 03, and determines the database connection address of the target slave cabinet.

[0038] The table name in the SQL statement above has a specific format: cabinet_XX_table.

[0039] In some embodiments, the master cabinet 2 is further provided with: a data acquisition module, a local database, and a multicast module that are logically identical to those of the slave cabinet 1.

[0040] In addition to running the program of the energy storage application layer 3, the main cabinet 2 also takes into account the data collection of the physical devices inside the main cabinet 2. Through software configuration, the main cabinet 2 can switch between two modes: running as the main cabinet 2 and running as the slave cabinet 1. When configured to run as the slave cabinet 1, its data acquisition module, local database and multicast module are activated, while the slave cabinet discovery module 21 and RPC agent module 22 stop service.

[0041] As an example, both slave cabinet 1 and master cabinet 2 use embedded industrial computers. Each slave cabinet 1 is equivalent to a data storage and notification unit, and multiple slave cabinets 1 distribute data storage. Master cabinet 2 is equivalent to an agent and coordination center, without a centralized storage database. The RPC agent module 22 is implemented using the gRPC framework, defining a bGate service and a Query method. The RPC interface 221 is a gRPC interface, which is the unified access entry point for the energy storage application layer 3. The program of the energy storage application layer 3 runs on the master cabinet 2, such as an energy management program. The database access agent engine 222 serializes the execution result into JSON format and then returns it to the energy storage application layer 3 through the RPC interface 221.

[0042] Secondly, this embodiment provides a method for seamless access to energy storage data based on RPC proxy and distributed storage, which is applied to the aforementioned system, such as... Figure 2 As shown, the method includes the following steps: S1. The multicast module of the cabinet periodically sends notification messages containing at least its own information to the network through multicast communication. The information includes at least the cabinet identifier, the cabinet URL, and the database access port.

[0043] S2. The slave cabinet discovery module of the master cabinet listens for multicast communication to discover all online slave cabinets. Based on the cabinet URL and database access port in the notification message, it generates the corresponding database connection address and maintains a dynamic slave cabinet information list containing the cabinet identifier and database connection address of all online slave cabinets.

[0044] Steps S1 and S2 constitute a multicast and discovery process.

[0045] S3. The energy storage application layer initiates a data access request containing SQL statements by calling the RPC interface of the main cabinet's RPC proxy module.

[0046] The database access proxy engine within the S4 RPC proxy module receives data access requests, parses SQL statements to extract the cabinet identifier of the target cabinet, and queries the dynamic cabinet information list based on the cabinet identifier to determine the database connection address of the target cabinet.

[0047] S5: The database access proxy engine establishes a connection with the local database of the target slave cabinet based on the database connection address, forwards the SQL statement to the local database of the target slave cabinet for execution, and returns the execution result returned by the local database of the target slave cabinet to the energy storage application layer through the RPC interface.

[0048] The master cabinet does not persistently store time-series data from the slave cabinet's local database at any step.

[0049] In some embodiments, the method further includes: The slave cabinet collects real-time operating information of the physical devices inside the cabinet through its data acquisition module, and generates time-series data based on the real-time operating information, which is then stored in the slave cabinet's local database.

[0050] In some embodiments, the notification message is a JSON format message sent based on the UDP multicast protocol, and its content includes at least the cabinet identifier, cabinet URL, database type, and database access port.

[0051] In some embodiments, parsing an SQL statement to extract the cabinet identifier of the target slave cabinet includes: parsing the table name prefix in the SQL statement, which contains the cabinet identifier, i.e., extracting the cabinet identifier of the target slave cabinet; and the database access agent engine establishing a connection with the local database of the target slave cabinet through the JDBC or ODBC protocol.

[0052] In some embodiments, the method further includes: when a new slave cabinet is connected to the system, automatically executing step S1; automatically executing step S2, adding the cabinet identifier and database connection address of the new slave cabinet to the dynamic slave cabinet information list; thereafter, the energy storage application layer can access its time series data through an SQL statement containing the cabinet identifier of the new slave cabinet without modifying the configuration.

Claims

1. A system for non-sensing access of energy storage data based on RPC agent and distributed storage, characterized in that, The system comprises: a plurality of slave cabinets, each of which is provided with: a local database for storing at least time-series data acquired by the cabinet and self-information of the cabinet, the self-information at least including cabinet identification, cabinet URL and database access port; a multicast module for periodically sending a notification message at least including the self-information to the network through multicast communication; a master cabinet provided with: a slave cabinet discovery module for listening to the multicast communication to discover online slave cabinets, generating a corresponding database connection address according to the cabinet URL and the database access port in the notification message, and maintaining a dynamic slave cabinet information list recording the cabinet identification and the database connection address of the online slave cabinets; an RPC proxy module providing at least one RPC interface, which is embedded with a database access proxy engine; an energy storage application layer for initiating a data access request including a SQL statement by calling the RPC interface; the database access proxy engine is configured to: receive the data access request and parse the SQL statement to extract the cabinet identification therein; query the dynamic slave cabinet information list according to the cabinet identification to determine the database connection address of the target slave cabinet; establish a connection with the local database of the target slave cabinet according to the database connection address and forward the SQL statement for execution; receive the execution result and return it through the RPC interface. 2.The RPC agent and distributed storage based energy storage data access system of claim 1, wherein, Each of the slave cabinets is further provided with a data acquisition module for acquiring real-time running information of physical devices in the cabinet; and the time-series data is generated based on the real-time running information. 3.The RPC agent and distributed storage based energy storage data non-sensing access system according to claim 1, characterized in that, The multicast module is configured to periodically send the notification message to a preset UDP multicast address based on the UDP multicast protocol. The notification message is in JSON format, and the message content at least includes the cabinet identification, the cabinet URL and the database access port. 4.The RPC agent and distributed storage based energy storage data access system of claim 1, wherein, The parsing of the SQL statement specifically includes parsing a table name prefix in the SQL statement, the table name prefix including the cabinet identification; and the database access proxy engine establishes a connection with the local database of the target slave cabinet through the JDBC or ODBC protocol.

5. The RPC agent and distributed storage based energy storage data access system of claim 2, wherein, The master cabinet is further provided with a data acquisition module, a local database and a multicast module which are logically identical to those of the slave cabinets.

6. A method for non-inductive access of energy storage data based on RPC agent and distributed storage, characterized in that, The method applied to the system of any one of claims 1-5 comprises the following steps: S1. The multicast module of the slave cabinet periodically sends a notification message at least including its own information to the network through multicast communication, the own information at least including cabinet identification, cabinet URL and database access port; S2. The slave cabinet discovery module of the master cabinet listens to the multicast communication to discover online slave cabinets, generates a corresponding database connection address according to the cabinet URL and the database access port in the notification message, and maintains a dynamic slave cabinet information list recording the cabinet identification and the database connection address of the online slave cabinets; S3. The energy storage application layer initiates a data access request including a SQL statement by calling the RPC interface of the RPC proxy module of the master cabinet; and S4. The database access proxy engine receives the data access request, parses the SQL statement to extract the cabinet identification therein, queries the dynamic slave cabinet information list according to the cabinet identification to determine the database connection address of the target slave cabinet, establishes a connection with the local database of the target slave cabinet according to the database connection address and forwards the SQL statement for execution, receives the execution result and returns it through the RPC interface. S4, the database access agent engine in the RPC agent module receives the data access request, parses the SQL statement to extract the cabinet identifier therein, and queries the dynamic slave cabinet information list according to the cabinet identifier to determine the database connection address of the target slave cabinet; S5, the database access agent engine establishes a connection with the local database of the target slave cabinet according to the database connection address, forwards the SQL statement for execution, receives the execution result, and returns the execution result through the RPC interface.

7. The method of claim 6, wherein, The method further comprises: The slave cabinet collects real-time running information of physical equipment in the cabinet through a data collection module of the slave cabinet, generates time series data based on the real-time running information, and stores the time series data in a local database of the slave cabinet. 8.The method of claim 6, wherein, The notification message is a JSON format message sent based on a UDP multicast protocol, and the content of the notification message at least includes the cabinet identifier, the cabinet URL, and the database access port. 9.The method of claim 6, wherein, The parsing of the SQL statement to extract the cabinet identifier of the target slave cabinet includes: parsing a table name prefix in the SQL statement, the table name prefix including the cabinet identifier, so as to extract the cabinet identifier of the target slave cabinet; and the database access agent engine establishes a connection with the local database of the target slave cabinet through a JDBC or ODBC protocol.

10. The method of claim 6, wherein, The method further comprises: When a new slave cabinet accesses the system, the step S1 is automatically executed; The step S2 is automatically executed to add the cabinet identifier and the database connection address of the new slave cabinet to the dynamic slave cabinet information list; Thereafter, the energy storage application layer does not need to modify the configuration, and can access the time series data of the new slave cabinet through a SQL statement including the cabinet identifier of the new slave cabinet.

Images