Large-scale multi-node high-throughput data communication system of energy storage battery

Abstract

The invention relates to the field of energy storage battery communication management, and particularly discloses a large-scale multi-node high-throughput data communication system of an energy storage battery, which comprises an equipment layer, a communication layer and a management scheduling layer, the management of the equipment layer on the battery comprises an IP mode and grouping management based on a battery state; a broadband industrial bus of the communication layer is used for data transmission among a BMU, a BCU and a BAU in the BMS and data interaction between the BMS and an EMS; the distributed domain bus is used for dividing the BMS into different domains, and cluster communication between BMS nodes is achieved in each domain. Cluster management is realized by using a distributed domain bus, the problems of complex wiring and difficult maintenance of traditional networking are solved, unified scheduling of energy storage devices of different manufacturers is realized through a self-defined domain address strategy, and addressing efficiency and distributed management and control capability of large-scale nodes are optimized.

Description

Technical Field

[0001] The present invention relates to the field of energy storage battery communication management, and in particular to a large-scale multi-node high-throughput data communication system for energy storage batteries. Background Art

[0002] Energy storage technology has developed rapidly in recent years, playing an increasingly important role in the interaction between renewable energy and the power grid. With the large-scale application of new energy technologies (such as photovoltaics and wind power), the communication needs of energy storage systems are also growing. The increase in energy storage capacity and the richness of energy storage types have also made the demand for efficient and stable data communication systems more urgent. These communication systems need to support key functions such as battery management and energy monitoring, while ensuring high throughput and low latency in large-scale energy storage systems.

[0003] Currently, energy storage system communication technologies mainly include the following methods: Battery Management System (EMS): monitors the battery status through protocols such as the CAN bus.

[0004] Power Monitoring System (PCS): Uses wireless sensor networks or fiber optic communications to monitor the energy output and efficiency of the energy storage system in real time.

[0005] Battery Management System (BMS): Accurately controls and protects the battery status through low-power interfaces such as I2C or SPI.

[0006] This traditional energy storage communication technology has obvious low bandwidth and poor real-time performance, and cannot meet the communication requirements of large-scale, high-throughput energy storage systems. For the broadband data transmission needs of real-time monitoring of energy storage systems and large-scale energy storage data (millions of kilowatts of energy storage), traditional bus and Ethernet communications have difficulty meeting the communication bandwidth and real-time requirements. In addition, for the future deployment of intelligent energy storage strategies, it cannot meet the clustered distributed management and control needs of energy storage equipment from different manufacturers.

[0007] The development of energy storage systems faces the need for large-capacity, multi-node clustered management. The integration of more energy storage nodes places higher demands on communication bandwidth and real-time performance. A large-scale energy storage communication system combining distributed and centralized management should be implemented to fully optimize energy efficiency and balanced management. In traditional energy storage systems, each subsystem uses industrial switches or bridges to implement a star network architecture. This networking requires a large number of cables and connectors, making it complex and difficult to maintain. Field-level communication uses narrowband communication technology, which has poor real-time performance and low field communication bandwidth, making it impossible to transmit analysis and calculation data from more intelligent terminals. Furthermore, energy storage systems composed of large nodes are difficult to address, making it difficult to achieve refined management of a cluster or pack of batteries.

[0008] Among the relevant existing technical solutions, the large-scale distributed energy storage communication system based on Ethernet has the advantages of high bandwidth, strong real-time performance, mature technology, support for multiple networking features and strong anti-interference ability. However, it has the disadvantages of requiring switches, resulting in high wiring costs, complex construction, and difficult maintenance. Wiring is difficult under limited space and environment conditions, and clock synchronization requires a separate PTP protocol and a complex protocol stack, resulting in high development and maintenance costs. The large-scale distributed energy storage communication system based on RS485 / CAN serial communication technology has good compatibility, is easy to connect to traditional energy storage system terminal equipment, can achieve full access to equipment data, and has mature and stable applications. However, it has the disadvantages of low bandwidth and poor real-time performance, which are not conducive to the intelligent unmanned management needs of large-scale energy storage systems.

[0009] With the popularization of energy storage applications, the scale of energy storage systems is increasing, and the nodes and data communication volume in energy storage systems are also increasing. Traditional energy storage systems cannot meet the data communication needs of an increasing number of smart terminal devices, nor can they meet the communication resources required for large-scale node computing in energy storage systems. There are significant gaps and risks in real-time performance and bandwidth. In order to meet the data communication needs of such large-scale, multi-node, high-throughput energy storage systems, this application is proposed to achieve high-bandwidth and high-real-time communication effects for the entire energy storage system. Summary of the Invention

[0010] The purpose of the present invention is to overcome the defects of the prior art and provide a large-scale multi-node high-throughput data communication system for energy storage batteries to solve the problems raised in the above background technology.

[0011] A large-scale multi-node high-throughput data communication system for energy storage batteries of the present invention includes a device layer, a communication layer, and a management and scheduling layer; The device layer includes multiple battery packs, each equipped with a BMS. Each BMS includes a BMU, BCU, and BAU. The device layer manages batteries using IP and group management based on battery status. The communication layer comprises a wideband industrial bus, a distributed domain bus, an IP address management unit, a routing unit, and a whole-network time synchronization unit, wherein the wideband industrial bus is used for transmitting data between BMUs, BCUs, and BAUs within the BMS and for data interaction between the BMS and the EMS; the distributed domain bus is used for dividing a plurality of BMSs into different domains and realizing cluster communication between BMS nodes in each domain; the IP address management unit is used for allocating the same network address to each BMU, BCU, and BAU in the system and independently mapping the corresponding global IP address; the routing unit is configured with a two-layer routing and a three-layer routing and is used for performing a data forwarding operation between the BAU and the EMS and realizing data intercommunication between the BMS subsystem and the upper-layer management system; and the whole-network time synchronization unit is used for providing a unified time reference signal for all communication nodes. The management and scheduling layer comprises an EMS, an intelligent analysis unit, and a policy scheduling unit, the EMS is used for uniformly collecting running state data of a plurality of distributed BMSs and performing centralized monitoring, the intelligent analysis unit is used for performing model calculation based on the collected data and performing energy efficiency optimization, fault prediction, and running state evaluation, and the policy scheduling unit is used for issuing a control policy according to an analysis result and performing coordinated operation and task scheduling of a distributed BMS cluster.

[0012] Further, the device layer manages battery grouping based on battery states, specifically, the device layer aggregates running data of the BMUs and the BCUs belonging to the device layer through the BAU, dynamically combines the battery families according to the policy issued by the upper layer and the state parameters of the batteries, and when the policy requires centralized regulation and control of the battery families in a specific state, the BAU identifies the battery families meeting the conditions according to the state parameters and divides them into the same management group.

[0013] Further, the distributed domain bus is used for dividing a plurality of BMSs into different domains, specifically, in the division process, each BMS is assigned a unique domain address identifier according to the domain in which it is located, in the domain address identifier, the high 4 bits of the address are used as the domain identifier, and the low bits of the address are kept unique, and each BMS node within a domain addresses through the domain address identifier.

[0014] Further, each domain bus is provided with a master node, the master node is used for coordinating the communication behaviors of all BMS nodes in the domain, including scheduling data transmission priority, processing node access and exit requests, and monitoring the state of the communication link within the domain.

[0015] Further, the distributed domain bus is configured with a domain control, the set of BAUs and BCUs is centrally managed by the domain control, and the domain control integrates the BMU collected data aggregated by each BAU and the control state of the BCU.

[0016] Further, each BMS node is configured with an independent communication processing module within the domain, which is used to parse, pack and forward data frames from other nodes in the same domain, and dynamically adjust the communication response strategy according to the data type and source address.

[0017] Further, each BMS of the device layer includes two wideband industrial bus interfaces, one of which is a southbound interface for connecting the BMU and the BCU inside the BMS, and the other is a northbound interface for connecting the communication layer and the management and scheduling layer outside the BMS.

[0018] Further, the IP address management unit is internally integrated with an address allocation module, a mapping rule control module and a security policy interface, the address allocation module allocates addresses through static presetting and dynamic registration, the mapping rule control module establishes a mapping table for all allocated addresses, and the security policy interface configures an access control list according to management permissions to limit the management commands and data flow paths of different levels.

[0019] Further, the routing unit is configured with two-layer routing and three-layer routing, the two-layer routing is used for data forwarding between BAU nodes in the same subnet or the same domain, and the data frame is identified and the path is selected through the MAC address or the domain logical identifier, and the three-layer routing is used for cross-domain forwarding of data packets between different subnets, different domains or remote BMSs.

[0020] Further, the network-wide time synchronization unit transmits a unified reference time signal to nodes at all levels through clock synchronization, and adds timestamp information to each data frame generation and control command issuance, and the time synchronization signal is distributed through the wideband bus or the domain bus.

[0021] The beneficial effects of the present application are as follows: By adopting the wideband industrial bus, the communication bandwidth and real-time performance of the energy storage system are significantly improved, the fine collection and control of battery voltage, current and temperature data are realized, high-precision data basis is provided for energy efficiency balancing algorithm, and the service life of the battery is prolonged; the distributed domain bus is used to realize cluster management, solve the problems of complex traditional networking and wiring, and difficult maintenance, and the unified scheduling of energy storage devices of different manufacturers is realized through the self-defined domain address strategy, the addressing efficiency and distributed management and control capability of large-scale nodes are optimized; based on the network-wide time synchronization strategy, the time error in data collection and control in the traditional technology is eliminated, the clock consistency from the collection terminal to the management unit is ensured, and the real-time collaborative processing capability of the system is improved; with the unified IP address management, the unique addressing of all nodes of the energy storage system and the one-network-to-bottom management mode are realized, the battery cluster state grouping management and routing forwarding are supported, the high-throughput data interaction and intelligent strategy deployment requirements of the million-kilowatt energy storage system are met, and the economic value of the system is maximized. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 This is a system architecture diagram of a large-scale multi-node high-throughput data communication system for energy storage batteries of the present invention; Figure 2 This is a system architecture diagram of a large-scale multi-node high-throughput data communication system for energy storage batteries of the present invention; Figure 3 This is a system architecture diagram of a data communication system for energy storage batteries in the prior art; Figure 4 Schematic diagram of node division of a distributed domain bus in an embodiment of the present invention; Figure 5 Schematic diagram of the principle of the IP address management unit in an embodiment of the present invention; Figure 6 Schematic diagram of the principle of the whole network time synchronization unit in an embodiment of the present invention. DETAILED DESCRIPTION

[0023] A large-scale multi-node energy storage battery data communication system, such as Figures 1-2 As shown, it includes device layer, communication layer, and management scheduling layer; refer to Figure 1 As shown in the figure, within each battery pack's BMS, the BMU, BCU, and BAU exchange data via a broadband industrial bus. This bus supports high-bandwidth, low-latency data transmission, ensuring that basic battery cell data collected by the BMU can be quickly uploaded to the BAU and that operational instructions such as charge and discharge, balancing control, and fault protection executed by the BCU can be accurately distributed to the corresponding nodes. During data transmission, all information collected by the BMU and instructions issued by the BCU are timestamped. This timestamp is generated based on a unified time reference provided by the network-wide time synchronization unit. This ensures that data from each node within the device layer remains consistent in time, providing a reliable time reference for subsequent status analysis and control coordination.

[0024] The device layer manages battery packs and battery families in an IP-based manner. The BCU corresponding to each battery pack, the BMU cluster contained in each battery family, and the BAU responsible for aggregation management are all assigned unique network addresses. These addresses are associated with the physical ownership and logical grouping of the battery pack and battery family according to preset rules, so that the specific battery pack or battery family can be directly located through the address, realizing accurate identification and unique addressing of each component of the device layer, making it easier for the upper-level management and scheduling system to directly access the target object through the address.

[0025] Battery status-based group management is implemented at the device level through collaborative control of the BAU. The BAU aggregates operational data from its associated BMUs and BCUs in real time, dynamically grouping its battery clusters based on policies issued by upper layers and battery status parameters such as voltage, current, and temperature. When policies require centralized control of battery clusters in specific states, the BAU identifies eligible battery clusters based on these status parameters, assigns them to the same management group, and implements unified charge-discharge balancing or protection control, ensuring targeted and effective group management.

[0026] The custom data frame format used by the device layer includes a source node identifier, a destination node identifier, a data type identifier, a timestamp, a valid data segment, and a check segment. The source node identifier and the destination node identifier correspond to the IP address information of each node in the device layer. The data type identifier distinguishes between collected data and control instructions. The timestamp ensures the time consistency of the data frame. The check segment is used to detect the integrity of the data transmission process. This format ensures the accuracy, efficiency, and identifiability of data during transmission within the device layer, meeting the communication requirements of high bandwidth and high real-time performance.

[0027] In the domain bus-based management strategy of the communication layer, each domain bus has a master control node. This node serves as the core of intra-domain communication and management, coordinating the communication behavior of all BMS nodes within the domain, including scheduling data transmission priorities, processing node entry and exit requests, and monitoring the status of intra-domain communication links. The master control node aggregates operational data uploaded by each BAU within the domain in real time, uniformly maintaining the node address list and status information within the domain. When cross-domain interaction is required, the master control node serves as the sole interface to connect with the master control nodes of other domains or upper-level management systems, ensuring the orderliness and security of cross-domain data transmission. At the same time, it dynamically adjusts communication resource allocation based on the load within the domain to ensure the real-time transmission of critical data.

[0028] Under the unified processing mechanism of the domain controller, the collection of BAUs and BCUs is centrally managed by the domain controller system. The domain controller integrates the BMU-collected data collected by each BAU with the control status of the BCU. Through pre-set collaborative logic, it achieves centralized data analysis and unified command generation. BAUs do not need to interact with upper layers individually. Instead, they submit data directly to the domain controller, which conducts comprehensive analysis and issues coordinated control commands to the BCU, reducing the latency caused by hierarchical transmission. Simultaneously, the domain controller monitors the operating status of the BAUs and BCUs in real time. If a node experiences an anomaly, it can quickly switch to a backup node or activate a redundant link to ensure uninterrupted control flow, thereby improving the overall system's real-time responsiveness and operational reliability.

[0029] The routing function of the communication layer is implemented by a routing unit, supporting intra-domain layer-2 routing and cross-domain layer-3 routing. Through cooperation with the IP address management unit, the target node is accurately located according to the IP address information of the node, and efficient forwarding of data between different BAUs and between BAU and EMS is realized. For cluster management, the routing function can automatically identify the node addresses of the same management group according to the cluster policy, and batch forward the control instructions or collection requests to the target nodes, reducing the overhead of individual addressing. During the routing process, the broadband industrial bus provides a high-bandwidth and low-latency transmission channel to ensure the rapid interaction of a large amount of data within the cluster, and all data transmitted through the bus are attached with timestamps generated by the network-wide time synchronization unit, so that each piece of data routed has a consistent time reference, ensuring the alignment of data from each node in the cluster in the time dimension, and providing accurate timing reference for state coordination and policy execution in cluster management.

[0030] The energy management server of the management and scheduling layer undertakes the function of global monitoring, establishes stable connection with each domain master node through the broadband industrial bus of the communication layer, and can directly access any node in the distributed BMS through the unique addressing mechanism provided by the IP address management unit, to obtain the battery cell data collected by the BMU, the control state of the BCU and the aggregation information of the BAU in real time. The server integrates the received data in a structured manner, combines the unified timestamp given by the network-wide time synchronization unit, and constructs a dynamic view covering the running state of all battery packs. When voltage abnormalities, temperature overruns and other situations are detected, the server immediately triggers an alarm and associates the physical location information of the corresponding node, supports management personnel to quickly locate the problem source, and automatically saves historical data to form traceable operation records, providing a basis for long-term energy efficiency analysis.

[0031] The intelligent analysis based on large models relies on the global data gathered by the energy management server. Through training of massive historical operation data and real-time collected multi-dimensional parameters, the large model forms a deep understanding of battery characteristics, operation rules and environmental influences. During the analysis process, the model combines the time synchronization data of each node to identify the state correlation of battery packs in different domains, such as comparing the charge-discharge curves of different battery families under the same timestamp to determine the energy efficiency balance state, predicting potential degradation risks using voltage and temperature trends of battery cells, and optimizing dynamic strategies for energy scheduling based on external load fluctuations and battery response characteristics. The large model has the ability to iterate autonomously, continuously correcting the analysis logic according to newly generated operation data, and improving the accuracy of energy efficiency optimization suggestions and fault warnings.

[0032] The policy scheduling and distribution link generates control strategies for different scenarios based on the output of intelligent analysis. The policy scheduling unit divides the policy instructions by domain through the routing function of the communication layer, and transmits them to the master node of the corresponding domain with the help of the domain bus. The master node then distributes them to the relevant BAU and BCU according to the IP address. For cross-domain collaborative strategies, instructions are transmitted between different domains through three-layer routing, relying on the time synchronization mechanism to ensure that each node performs operations within the same time window, such as synchronous charging and discharging adjustments of battery packs in multiple domains. The policy content includes specific control parameters and execution priorities. After receiving the policy, the BAU can directly drive the BCU to execute without secondary analysis, and at the same time, the execution results are fed back to the energy management server in real time to form a scheduling closed loop to ensure the timeliness and accuracy of the policy implementation.

[0033] In a specific embodiment, the equipment layer includes multiple battery packs, each battery pack is equipped with a BMS, and each BMS includes a BMU, a BCU, and a BAU.

[0034] BMS stands for Battery Management System. BMS (Battery Management System) is the core of the entire battery management system and is responsible for monitoring, protecting and controlling the battery pack. BMS ensures that the battery pack operates in a safe, stable and efficient state by monitoring and regulating each battery cell.

[0035] BMU is a battery unit. BMU (Battery Management Unit) is a part of BMS. It is responsible for real-time monitoring and control of individual battery units, reporting collected data to BMS, and accepting instructions from BMS for adjustment.

[0036] BCU is the battery control unit. BCU (Battery Control Unit) is mainly responsible for high-level control of battery management and manages the operation of the entire battery system, including: controlling battery charging and discharging according to the instructions of the BMS; ensuring that the power between each battery cell is balanced to avoid overcharging or over-discharging of individual battery cells; monitoring the overall health of the battery pack and initiating safety protection measures when necessary; communicating and working in collaboration with the BMS.

[0037] BAU (Battery Array Unit) is a component in the BMS that manages multiple battery cells (or battery packs). It is used to: aggregate data from multiple BMUs and BCUs and collaborate with the BMS; when the battery system is large, the BAU ensures stable operation of the system by managing multiple battery cells or battery packs; and adjusts the working status of the battery pack according to demand to ensure the overall efficiency and safety of the system.

[0038] The communication layer includes a broadband industrial bus, a distributed domain bus, an IP address management unit, a routing unit, and a full-network time synchronization unit. The broadband industrial bus is used to transmit data between the BMU, BCU, and BAU within the BMS, and for data interaction between the BMS and the EMS.

[0039] The broadband industrial bus is used to connect the BMU, BCU, and BAU within the BMS, supporting high-bandwidth, high-real-time data transmission. By optimizing the data transmission path and transmission rate, the bus ensures that data acquisition and control instructions from the battery cell to the control unit can be transmitted quickly and accurately. Specifically, the BMU is responsible for real-time collection of basic data such as voltage, current, and temperature of each battery cell, and the collected data is uploaded to the BAU via the broadband industrial bus. The BCU performs advanced control tasks such as battery charging and discharging, balancing control, and fault protection through the broadband industrial bus, and feeds back relevant control data to the BMU and BAU.

[0040] As the aggregation unit of multiple BMUs and BCUs, the BAU not only manages data exchange and status coordination among multiple battery cells, but also, with the help of a broadband industrial bus, transmits aggregated data to the EMS via the northbound bus. This allows the EMS to obtain timely information on the battery system's health status, efficiency data, and energy efficiency balance, and issue precise scheduling and control commands based on this data. The high bandwidth and low latency of the broadband industrial bus enable large-scale energy storage systems to process massive amounts of data, ensuring real-time and efficient operation.

[0041] Furthermore, the broadband industrial bus serves as a data bridge between the BMS and EMS, transmitting important status information, control commands, and alarm data between the two levels. The EMS connects to each BMS via the broadband industrial bus to obtain and monitor the operating status of the entire energy storage system. This bus enables clustered management of the energy storage system, ensuring coordinated operation of each BMS subsystem and precisely controlling the charge, discharge, and health status of each battery pack.

[0042] The broadband industrial bus can significantly improve the efficiency and accuracy of battery management. Compared with traditional CAN bus or RS485 serial communication (such as Figure 3 Compared to conventional industrial buses (as shown in Figure 2), broadband industrial buses offer greater bandwidth and lower latency, supporting greater data traffic and enabling more refined battery management. The advantages of broadband industrial buses are particularly pronounced in large-scale energy storage systems, meeting the demands of large-scale, highly concurrent, and high-precision data communication, ensuring system stability and economic efficiency.

[0043] The distributed domain bus is used to divide multiple BMSs into different domains, realize cluster communication between BMS nodes in each domain, and control data scheduling and unified management between different domains through domain-level master nodes.

[0044] Among them, multiple BMSs are divided into different domains. During the division process, a custom domain address encoding rule is adopted, and the upper 4 bits of the address are used as the domain identifier. For example, 0x01 represents the BMS cluster in area A, and 0x02 represents the BMS cluster in area B. The last 8 bits of the BMS node address in each domain remain unique and are used to distinguish different BMS devices in the same domain. The division of domains supports dynamic adjustment. When new BMS devices are connected, new domain IDs can be assigned according to system planning, or they can be incorporated into existing domains. At the same time, to ensure communication isolation and priority control between different domains, the bus protocol has a built-in domain access rights management mechanism. Only authorized domain-level master nodes can schedule data across domains to prevent illegal access and data conflicts, thereby ensuring system security and stability.

[0045] For the management of distributed energy storage systems, communication and data assembly and scheduling need to be completed between more BMSs. The use of traditional Ethernet switches is complex and difficult to maintain. The use of traditional serial buses, such as CAN buses, has low bandwidth and poor real-time performance. The use of broadband buses improves bandwidth and real-time performance while using broadband bus addressing strategies to achieve cluster management. Custom domains can be used for different energy storage systems to meet management needs. That is, the system uses customized address mapping rules to ensure accurate data transmission and management between different domains. Each BMS is assigned a unique domain address identifier based on the domain it is in. By using the high-order part as the domain identification mark, the system can effectively distinguish and manage nodes in multiple domains. For each node in the BMS, the low-order part of the address remains unique to ensure that nodes in the same domain can be accurately located and communicate through their addresses. Through this addressing strategy, automatic identification and scheduling of different areas or nodes can be achieved, simplifying the system expansion process while ensuring that data transmission in the network is not interfered with. In addition, after adopting this strategy, the system can flexibly adjust the node position within the domain and efficiently manage and schedule data flows between different BMSs to avoid data conflicts or loss, thereby ensuring the stability and efficiency of large-scale energy storage systems. Figure 4 For example, for three BMSs A, B, and C, the EMS can implement cluster management of these three BMSs. The EMS regards the three BMSs as a large energy storage system. The three BMSs belong to different domains, and the corresponding three domain address segments can be 0x0100, 0x0200, and 0x0300 respectively, and the last 8 bits retain the address of each node in the BMS unchanged.

[0046] For the EMS, all three BMSs belong to the distributed energy storage system, so the top four bits of the address are always the same. This allows the EMS to centrally manage and schedule the distributed energy storage system consisting of the three BMSs. Users can also define the domain IDs of the three BMSs based on their location and importance, which is also beneficial for the deployment and implementation of cluster strategies.

[0047] The BMS master always has two broadband bus interfaces. The southbound interface handles the data collection and control of the BMS subsystem, and the northbound interface sends the collected data and receives the scheduling control commands from the EMS or PCS. This type of BMS master forms a north-south bidirectional communication structure by configuring two broadband bus interfaces. The southbound interface is mainly used to connect the BMU and BCU inside the BMS to realize real-time data collection and control instructions for battery cells and battery packs, ensuring the stable execution of key functions such as charge and discharge control, balancing adjustment and status monitoring. The northbound interface is responsible for connecting the communication layer and the management and scheduling layer. It is responsible for uploading the collected refined data to the EMS in a timely manner, and at the same time receiving energy efficiency optimization, operation control and other instructions issued by the EMS or PCS, supporting cross-system collaborative management and scheduling.

[0048] In this architecture, southbound bus communication supports high-speed, high-precision data flow, ensuring that changes in the status of each battery cell are fed back to the BAU within milliseconds, effectively improving the accuracy and responsiveness of battery management. At the same time, northbound communication channels are tightly integrated with the domain bus and IP address management system, allowing the master control unit to directly interact with multiple EMS nodes or intelligent scheduling platforms, facilitating rapid response to distributed control commands and efficient distribution of scheduling instructions.

[0049] In a specific embodiment, to ensure the stability and timeliness of two-way communication, the master control unit integrates an independent data cache and priority scheduling module. This sets differentiated transmission priorities for critical control data and status reporting data, ensuring smooth communication links even under high system load or when multiple tasks are being scheduled concurrently. Each master control unit also dynamically matches the network address of the target node based on its domain number. Using an integrated time synchronization mechanism, it achieves seamless communication between commands and data within and outside the domain, enhancing the coordinated execution capabilities of the entire system.

[0050] In addition, the main control structure supports hot swapping and remote configuration, making the system highly scalable and easy to maintain. Whether deployed in densely populated regional power plants or in multi-point coordinated control applications in distributed microgrids, it can quickly adapt to different deployment scales and network structures, meeting the future demands for intelligent, automated, and highly reliable energy storage communications.

[0051] To enable cluster communication between BMS nodes within each domain, the BAUs of all BMSs within the domain are equipped with broadcast and point-to-point communication capabilities, supporting multiple communication mechanisms such as multicast and unicast to ensure data synchronization and command coordination between nodes. A unified communication scheduling mechanism is pre-set in the domain bus protocol. Each domain sets a master node responsible for scheduling the communication behavior of other BMS sub-nodes, implementing message prioritization, access conflict handling, and automatic fault isolation, ensuring stable system operation in high-concurrency communication scenarios.

[0052] BMS nodes within each domain are addressed using predefined addressing rules. Nodes can communicate with each other based on a policy-triggered task model. When a BMS node detects a specific battery status or environmental change, it can broadcast this information to other nodes within the domain or request coordinated scheduling based on the policy. A synchronous cache mechanism is also implemented within the domain, allowing nodes to share key state variables. This supports cross-node coordinated control and data fusion computing, improving the collaborative efficiency and processing accuracy of cluster communications.

[0053] The intra-domain communication protocol supports a highly reliable transmission mechanism with built-in error detection, automatic retransmission, and congestion control capabilities to adapt to dynamically changing load environments. Each BMS node is configured with an independent communication processing module within the domain. This module is responsible for parsing, packaging, and forwarding data frames from other nodes in the same domain, and dynamically adjusting the communication response strategy based on the data type and source address. This approach enhances the collaborative capabilities between nodes, enabling the formation of a clustered communication system with unified scheduling, information sharing, and status linkage across the entire domain, meeting the core requirements of distributed energy storage systems for large-scale node autonomous collaboration.

[0054] When controlling data scheduling and unified management between different domains through domain-level master control nodes, the system sets each domain's master control node as the sole bridge point for intra-domain and external interactions, ensuring the orderliness of cross-domain communication and the efficiency of command transmission. The domain-level master control node has intelligent scheduling and collaborative control capabilities, able to aggregate data from all BMS nodes within the domain and coordinate the execution of corresponding control tasks by sub-nodes within the domain based on global policy instructions from the EMS. At the same time, the master control node acts as a gateway for cross-domain communication, exchanging key data within the domain with other domains or the management and scheduling layer based on unified communication protocols and routing strategies.

[0055] During implementation, the domain-level master control node has the ability to schedule intra-domain communication, forwarding data and managing load according to task priority, ensuring stable operation of critical services under high communication pressure. This node authenticates and manages cross-domain data interactions through access control policies, preventing unauthorized commands from interfering with local domain operations. Furthermore, it possesses dynamic addressing and path optimization capabilities, intelligently adjusting communication paths based on network status, node load, and other factors, improving overall system efficiency and robustness.

[0056] The domain-level master node also handles inter-domain state synchronization, anomaly communication, and coordinated response, ensuring consistency across domains in policy deployment, energy efficiency management, and fault response. By collecting and distributing data such as time synchronization information, scheduling policies, and status feedback, this node achieves the critical transition from distributed autonomy to network-wide collaboration, supporting unified management and control of the entire system under high-concurrency and complex network structures. This design effectively supports the multi-domain energy storage system's goals of high scalability, high consistency, and strong adaptability.

[0057] The IP address management unit is used to assign a unique network address to each BMU, BCU, and BAU in the system, and establish a correspondence between the address and the global IP address according to the preset address mapping rules to achieve unified addressing and management of all nodes; In a specific embodiment, Figure 5 As shown, each terminal device has the same network address, which is independently mapped to the corresponding global IP address. The domain ID reflects the location of the network ID. For EMS and PCS, data collection and control can be performed on any node in the distributed energy storage system through the IP address.

[0058] The IP address management unit, a core component of the network layer within this system, is tightly coupled with the communication layer to assign unique identifiers and perform address mapping for all energy storage subnodes (BMUs, BCUs, and BAUs). Based on a unified address allocation strategy, this unit establishes a corresponding IP address for each device node. By integrating the system's hierarchical structure and domain partitioning mechanism, the IP address not only ensures network-level reachability but also embeds information about the node's domain, functional type, and topological location. This enables efficient addressing and identification even in complex network topologies. The IP address management unit integrates an address allocation module and a mapping rule control module, supporting both static pre-set and dynamic registration for address allocation, facilitating rapid access to new nodes during system expansion or maintenance. To ensure unified access and control of lower-level nodes by the management and scheduling layer, the IP address management unit establishes a mapping table for all assigned addresses, enabling rapid resolution from logical addresses to physical locations. This mapping mechanism supports multi-level caching and fast querying, ensuring that the EMS or policy scheduling unit can quickly locate the target node and promptly return the results when initiating control commands or status queries. The high bits in the address mapping rules are retained as domain identifiers and type indicators, allowing cross-domain management and scheduling to achieve unified routing and efficient scheduling based on logical addresses even when physical networks are isolated. Furthermore, this unit works in conjunction with the routing unit to implement intra-domain and inter-domain routing path selection and load regulation, improving the flexibility and fault tolerance of the system's communication links.

[0059] The IP address management unit also performs the system's addressing integrity verification and conflict detection tasks. When the network topology is dynamically adjusted or nodes are hot-swapped, it automatically detects potential address duplication or allocation conflicts and performs timely corrections to ensure the stable operation of the entire communication system. The unit also has a built-in security policy interface that can configure access control lists based on management permissions, limiting management commands and data flow paths at different levels to ensure the system's data integrity and the credibility of scheduling commands. Through this unified IP address system, the entire energy storage system achieves seamless integration and unified management and control from physical devices to logical networks, providing a solid network foundation for distributed intelligent energy storage management.

[0060] In a specific embodiment, three-layer routing is implemented in the EMS, and two-layer routing forwarding functions are implemented in the BAU. For each independent BMS subsystem consisting of a complete BAU, BCU and BMU, the IP address of each node in the system can be uniquely addressed and accessed. Specifically, the routing unit is configured with two-layer routing and three-layer routing to perform data forwarding operations between the BAU and the EMS, and to communicate data between each BMS subsystem and the upper-level management system. The routing unit is located at the core of the communication layer, and undertakes the key data forwarding task between the BAU and the EMS. It supports the simultaneous configuration of two-layer and three-layer routing capabilities, thereby realizing flexible communication and efficient collaboration across domains and subsystems. Two-layer routing is used to quickly forward data between BAU nodes in the same subnet or the same domain, and to identify and select data frames and paths through MAC addresses or logical identifiers within the domain, to achieve low-latency and interference-free communication between nodes within the domain, to ensure that data uploaded by devices such as BMU and BCU can be quickly aggregated to the BAU of this domain, and to respond to control strategies in a timely manner through local processing. The three-layer routing is based on the IP address management mechanism, which realizes the cross-domain forwarding of data packets between different subnets, different domains or remote BMSs. The unified policy control command initiated by the EMS can be accurately delivered to the target BAU node through the three-layer network protocol, ensuring the cross-domain implementation and multi-domain collaborative control of the dispatching instructions. At the same time, the routing unit has a dynamic path learning and fault avoidance mechanism, which can dynamically adjust the routing path according to the network link status, thereby improving the robustness and load adaptability of the entire energy storage communication network. In actual operation, the routing unit supports multi-level address resolution and fast forwarding by integrating the domain address mapping mechanism with a unified IP addressing strategy. Without relying on complex switching equipment, it realizes efficient information interaction and precise distribution of task instructions between the energy storage system management and dispatching layer and multiple distributed BMSs, building a communication bridge that supports high throughput, low latency, and multi-node collaboration, and providing basic guarantees for the overall intelligent operation and clustered dispatching of the system.

[0061] The traditional BMS northward and southward are implemented by different communication technologies, cannot achieve network-wide time synchronization, and cannot achieve real-time synchronous acquisition and control. Due to the time asynchronization, the acquired data and the issued control commands cannot be well cooperated, and there is time error or poor time synchronization effect.

[0062] In the embodiment, as shown in Figure 6 The network-wide time synchronization unit is configured to provide a unified time reference signal for all communication nodes in the system, so that the data acquired by each node and the control commands issued have consistent timestamp information. The network-wide time synchronization unit undertakes the key unified clock source distribution task in the system, ensures that all communication nodes have consistent time reference in the data acquisition and control execution process, and avoids the problems of data misplacement or control lag caused by clock drift or delay difference. The unit transmits the unified reference time signal to nodes at all levels, including EMS, BAU, BCU and BMU, etc. through centralized or distributed clock synchronization mechanism, fundamentally eliminates the error that may exist in time layer between the southward acquisition link and the northward scheduling link. The system adds timestamp information when generating each data frame and issuing control commands, so that the data in the entire energy storage network can be aligned in time dimension, thereby supporting high-precision time sequence backtracking and trend prediction when the EMS performs policy optimization, state evaluation and abnormal analysis.

[0063] The time synchronization signal can be distributed through a high-reliability low-delay wideband bus or domain bus, and a time delay compensation mechanism is matched in the propagation process to improve the synchronization accuracy. The synchronization mechanism cooperates with the routing unit and the IP address management unit to dynamically adjust the synchronization path and beat refresh frequency according to the topology structure and node distribution, so that each node can still receive consistent time signals even in a multi-level, multi-domain distributed architecture. The system also supports a time synchronization fault-tolerant mechanism. When the main clock source is abnormal, it can automatically switch to the backup clock source for redundant synchronization to ensure the time sequence continuity and scheduling reliability of the overall system. The unified timestamp system also provides a unified time reference for cross-domain data analysis and scheduling strategy distribution, supports real-time collaboration and control loop between large-scale nodes, and provides a solid guarantee for high-precision energy efficiency optimization and state linkage of the entire system.

[0064] The EMS and the distributed BMS are based on the same clock source signal, realizing network-wide time synchronization. From the acquisition terminal (BMU) to the management unit (BAU) and the EMS, based on the network-wide time synchronization sensitive network, the time synchronization of the entire distributed energy storage system is realized, and more accurate timestamps are provided for the network-wide monitoring and management. This time synchronization strategy will greatly improve the energy efficiency management and battery service life of the entire system.

[0065] The management and scheduling layer includes EMS, intelligent analysis unit and policy scheduling unit. EMS is used to uniformly collect the operating status data of multiple distributed BMSs and conduct centralized monitoring. EMS establishes a high-speed, low-latency communication link with each BMS through a broadband industrial bus to achieve unified collection and real-time monitoring of the operating status of multiple distributed BMS subsystems. In the architectural design, EMS, as the core node of the management and scheduling layer, can accurately locate and access the key control nodes in each BMS subsystem based on the global unique addressing mechanism provided by the IP address management unit. Through the domain division and access rules set in the domain bus protocol, EMS can simultaneously retrieve operating data from different domains, covering basic information such as voltage, current, temperature of battery cells, as well as charge and discharge status, balance status, fault alarm information, etc. of battery packs.

[0066] During data collection, the EMS leverages a network-wide time synchronization mechanism to unify time tags for all incoming data, ensuring consistency and temporal comparability across domains and nodes, thus avoiding misjudgments of status or policy execution failures caused by time errors. The EMS incorporates a multi-threaded data processing module capable of concurrently receiving data streams from multiple BMS subsystems. Through data decoding, caching, and status tagging, it generates a structured data set that can be used by subsequent intelligent analysis and policy scheduling units.

[0067] In the entire system, EMS is not only responsible for receiving and recording operating status data, but also for real-time monitoring of the stability and safety of battery operation. When an abnormal state is detected in a BMS subsystem, such as excessive temperature, voltage imbalance, or communication interruption, EMS can promptly issue control instructions through the established control link to isolate the abnormal node, adjust the load, or switch the strategy to ensure the continuous and stable operation of the entire energy storage system. At the same time, EMS has remote monitoring and logging capabilities, and can connect to the external management platform through an integrated communication interface to achieve centralized remote management of the energy storage system in different physical deployment scenarios. The system structure supports horizontal expansion, that is, when adding new BMS nodes or expanding the domain range, EMS can quickly adapt without reconfiguring the basic communication architecture, ensuring that the energy storage system has good scalability and long-term evolution capabilities.

[0068] The intelligent analysis unit is used to perform model calculations based on the collected data and perform energy efficiency optimization, fault prediction, and operational status assessment. The intelligent analysis unit receives structured data aggregated from the EMS and performs model processing based on unified time synchronization. The system has built-in multi-type state analysis models. Based on the voltage, current, temperature, charge and discharge rate, and health status parameters collected by nodes at all levels of the BMS, it establishes a state assessment model for operational safety, energy efficiency, and system reliability through multi-dimensional feature extraction and association calculation. During the operation of the model, a timing analysis mechanism is introduced to track the trajectory of node state changes. By combining historical operation trajectories with current state parameters, potential abnormal patterns or signs of performance degradation can be accurately identified.

[0069] To optimize energy efficiency, the system builds a dynamic energy efficiency scheduling model based on real-time load demand, remaining power distribution, individual battery balance status, and ambient temperature fluctuations. While maintaining overall battery pack stability, the model maximizes system energy conversion efficiency by controlling charge and discharge rates, adjusting the battery cluster operation sequence, and load distribution. The scheduling logic fully considers the health status and usage cycle of each node, dynamically adjusting operating strategies to extend battery life and reduce system operating costs.

[0070] Fault prediction utilizes a deep state assessment algorithm to generate real-time predictive modeling of key performance parameters. Feature dimension compression and model residual analysis are used to determine if there are fault precursors such as abnormal temperature rise, voltage fluctuation, or communication anomalies. When an abnormality trend is identified, the system promptly provides alerts and recommended solutions to the policy dispatch unit, supporting proactive prevention and control and rapid response.

[0071] The operational status assessment, based on an integrated status scoring mechanism, categorizes the operational status of all BMSs within and across domains. This assessment covers indicators such as node health, load balancing, charge-discharge consistency, communication integrity, and response latency. These indicators are weighted and synthesized to form a comprehensive, global, and local operational status report, providing support for subsequent scheduling strategies.

[0072] The intelligent analysis unit supports continuous model iteration and parameter optimization. During system operation, it adaptively corrects model parameters based on real-time feedback and evaluation results to ensure long-term stability of prediction accuracy and control response capabilities. It also supports data fusion analysis between multi-domain systems, providing a reliable decision-making basis for the optimization of multi-node collaborative scheduling strategies.

[0073] The policy scheduling unit is used to issue control policies based on analysis results and coordinate the operation and task scheduling of distributed BMS clusters. After receiving the operational status assessment results and energy efficiency optimization recommendations from the intelligent analysis unit, the policy scheduling unit generates specific scheduling instructions based on pre-set control logic and policy templates, and issues them to the corresponding BAU nodes based on domain division and node priority. The scheduling unit includes functional modules such as policy parsing, instruction generation, task distribution, status feedback, and fault-tolerant control, enabling full-process control and closed-loop management of policy issuance within a multi-domain, multi-node system architecture.

[0074] During the scheduling process, the system accurately maps each control instruction to the target node based on the unique network address information provided by the IP address management unit, ensuring that each BMS subsystem and its internal BCU and BMU can receive and execute tasks based on their functional role and operating status. The policy scheduling unit supports static policy configuration and dynamic policy updates, adjusting scheduling priorities based on multiple parameters such as operating cycle, load changes, battery health status, and external environment, ensuring priority execution of key nodes and improving the overall system's operational efficiency, safety, and stability.

[0075] In terms of task scheduling, the scheduling unit implements distributed coordinated control using a model of intra-domain centralization and inter-domain collaboration. Intra-domain tasks are dispatched and distributed to each BMS sub-node via the master control node, achieving balanced load distribution and optimized resource utilization. Cross-domain tasks are relayed through the routing unit and coordinated with the execution rhythm using a network-wide time synchronization mechanism to ensure the uniformity of control instructions and the coordination of execution. The scheduling unit also has a task status tracking and result reporting mechanism that can monitor the execution status of instructions in real time, detect task completion, and promptly feedback the execution results to the EMS and intelligent analysis unit as input for the next round of strategy generation.

[0076] To ensure the reliability and real-time nature of scheduling, the strategic scheduling unit deploys redundancy mechanisms and failover strategies within the system architecture. Upon detecting an anomaly in a domain's master node or link, the scheduling unit automatically switches to a backup path or node to continue scheduling tasks, preventing scheduling link interruptions from impacting system operations. Furthermore, the scheduling unit can strategically combine and schedule different battery clusters, enabling targeted control and refined management of certain nodes. This supports the coordinated achievement of multiple scheduling objectives, including energy efficiency optimization, peak-valley regulation, balancing control, thermal management, and life extension.

[0077] By working in collaboration with EMS, intelligent analysis units, and communication layer modules, the strategic scheduling unit builds a highly real-time, reliable, and controllable scheduling and control system, ensuring unified, efficient, and orderly scheduling and management of multi-domain, multi-node distributed BMS clusters in complex operating scenarios, and fully supporting the intelligent operation and flexible response capabilities of large-scale energy storage systems.

[0078] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0079] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A large-scale multi-node high-throughput data communication system for energy storage batteries, characterized in that: Including equipment layer, communication layer, and management and scheduling layer; The device layer includes multiple battery packs, each equipped with a BMS. Each BMS includes a BMU, BCU, and BAU. The device layer manages batteries using IP and group management based on battery status. The communication layer includes a broadband industrial bus, a distributed domain bus, an IP address management unit, a routing unit, and a full-network time synchronization unit. The broadband industrial bus is used to transmit data between the BMU, BCU, and BAU within the BMS, and for data interaction between the BMS and the EMS. The distributed domain bus is used to divide multiple BMSs into different domains and realize cluster communication between BMS nodes in each domain. The IP address management unit is used to assign the same network address to each BMU, BCU, and BAU in the system, and independently map it to the corresponding global IP address. The routing unit configures layer 2 routing and layer 3 routing to perform data forwarding operations between the BAU and the EMS, and to communicate data between each BMS subsystem and the upper-level management system. The full-network time synchronization unit is used to provide a unified time reference signal for all communication nodes. The management and scheduling layer includes EMS, intelligent analysis unit and policy scheduling unit. EMS is used to uniformly collect the operating status data of multiple distributed BMSs and conduct centralized monitoring. The intelligent analysis unit is used to perform model calculations based on the collected data, and perform energy efficiency optimization, fault prediction and operating status evaluation. The policy scheduling unit is used to issue control strategies based on the analysis results, and coordinate the operation and task scheduling of the distributed BMS cluster.

2. The large-scale multi-node high-throughput data communication system of energy storage battery according to claim 1, characterized in that: The device layer manages battery groups based on battery status. Specifically, the device layer aggregates the operating data of the BMU and BCU in real time through the BAU, and dynamically combines the battery families according to the policies issued by the upper layer and the battery status parameters. When the policy requires centralized control of battery families in a specific status, the BAU identifies qualified battery families based on the status parameters and divides them into the same management group.

3. The large-scale multi-node high-throughput data communication system of energy storage battery according to claim 1, characterized in that: The distributed domain bus is used to divide multiple BMSs into different domains. Specifically, during the division process, each BMS is assigned a unique domain address identifier based on the domain it is in. In the domain address identifier, the upper 4 bits of the address serve as the domain identifier, and the lower part of the address remains unique. The BMS nodes within each domain are addressed through the domain address identifier.

4. The large-scale multi-node high-throughput data communication system for energy storage batteries according to claim 1, characterized in that: Each domain bus is equipped with a master node, which is used to coordinate the communication behavior of all BMS nodes in the domain, including scheduling data transmission priorities, processing node access and exit requests, and monitoring the status of communication links within the domain.

5. The large-scale multi-node high-throughput data communication system of energy storage battery according to claim 1, characterized in that: The distributed domain bus is configured with a domain controller. The collection of BAUs and BCUs is centrally managed by the domain controller, which integrates the BMU collection data gathered by each BAU and the control status of the BCU.

6. The large-scale multi-node high-throughput data communication system for energy storage batteries according to claim 1, characterized in that: Each BMS node is configured with an independent communication processing module within the domain. The communication processing module is used to parse, package and forward data frames from other nodes in the same domain, and dynamically adjust the communication response strategy based on the data type and source address.

7. The large-scale multi-node high-throughput data communication system for energy storage batteries according to claim 1, characterized in that: Each BMS in the equipment layer includes two broadband industrial bus interfaces, one of which is used to connect the BMU and BCU inside the BMS, and the other is used to connect the communication layer and the management scheduling layer outside the BMS.

8. The large-scale multi-node high-throughput data communication system for energy storage batteries according to claim 1, characterized in that: The IP address management unit integrates the address allocation module, mapping rule control module, and security policy interface. The address allocation module allocates addresses through static preset and dynamic registration. The mapping rule control module is used to establish a mapping table for all allocated addresses. The security policy interface configures the access control list according to management authority, limiting the management commands and data flow paths of different levels.

9. The large-scale multi-node high-throughput data communication system for energy storage batteries according to claim 1, characterized in that: The routing unit is configured with Layer 2 routing and Layer 3 routing. Layer 2 routing is used to forward data between BAU nodes in the same subnet or domain, and to identify data frames and select paths through MAC addresses or logical identifiers within the domain. Layer 3 routing is used for cross-domain forwarding of data packets between different subnets, different domains or remote BMSs.

10. The large-scale multi-node high-throughput data communication system of energy storage battery according to claim 1, characterized in that: The network-wide time synchronization unit transmits a unified reference time signal to nodes at all levels through clock synchronization. Timestamp information is attached each time a data frame is generated and a control command is issued. The time synchronization signal is distributed through a broadband bus or a domain bus.

Images