Daisy chain bidirectional communication multi-dimensional diagnosis and partition degradation adaptive fault-tolerant system and method

Abstract

The application relates to a communication technology field, and relates to a multi-dimensional diagnosis and partition degradation self-adaptive fault-tolerant system and method for daisy chain bidirectional communication, which comprises a main control module embedded with a comprehensive decision engine, a bidirectional communication interface, a sampling module group, and the like. total The application can maintain the maximum range of communication ability after the breakpoint occurs through bidirectional diagnosis and intelligent routing, and improves the availability of the system.

Description

Technical Field

[0001] This invention relates to the field of communication technology, specifically to a multidimensional diagnostic and partition degradation adaptive fault-tolerant system and method for daisy-chain bidirectional communication. Background Technology

[0002] In large-scale energy storage systems, the Battery Management System (BMS) typically employs a daisy-chain topology to connect the main controller (MCU) with multiple analog front-end (AFE) chips to monitor the voltage, temperature, and other statuses of a large number of battery cells. The reliability of the daisy-chain communication network directly affects the safety and availability of the entire energy storage system.

[0003] Currently, existing technologies have the following main drawbacks:

[0004] 1. Conservative and simplistic fault handling strategies: When existing BMS detects a communication verification failure, it typically adopts a "one-size-fits-all" shutdown strategy that immediately cuts off the charging and discharging circuits. While this approach can ensure battery safety, it cannot distinguish between transient interference and permanent faults, easily leading to frequent and unplanned system shutdowns, which seriously affects grid dispatching and economic operation.

[0005] 2. Fragile communication paths and lack of intelligent diagnostic and predictive mechanisms: When a physical break occurs in a traditional unidirectional daisy chain, all analog front-ends (AFEs) after the break will be completely disconnected, creating a blind spot for battery status information and posing a risk of local overcharging or over-discharging. Although some studies have attempted to introduce bidirectional communication to improve network diagnostics, they have failed to deeply integrate communication status with system operating status (such as state of charge (SOC), operating conditions, and fault duration), lacking the ability to predict fault trends and resulting in delayed decision-making and response.

[0006] 3. Crude degraded operation strategy: In communication degraded scenarios, traditional BMS simply limits the overall power without considering the differences in communication reachability within the battery cluster, which can easily lead to safety hazards due to information blind spots.

[0007] In summary, existing technologies for daisy chain communication fault tolerance suffer from problems such as a disconnect between diagnosis and scheduling, and rigid response strategies, making it impossible to achieve high availability of the system while ensuring security. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a multi-dimensional diagnostic and partition degradation adaptive fault-tolerant system and method for daisy-chain bidirectional communication. Through bidirectional diagnostics and intelligent routing, it can maintain maximum communication capability even after a breakpoint occurs, thereby improving the availability of the system.

[0009] On one hand, this invention provides a multi-dimensional diagnostic and partition degradation adaptive fault-tolerant system with daisy-chain bidirectional communication, comprising:

[0010] The main control module is the control core of the system; a comprehensive decision engine is embedded in the main control module to perform bidirectional diagnosis, multi-dimensional state fusion, fault prediction and fault-tolerant decision-making.

[0011] The bidirectional communication interface includes a first communication module and a second communication module located at the beginning and end of the daisy chain, respectively, and is connected to the main control module through a high-speed communication channel to form a bidirectional closed loop;

[0012] A sampling module group, connected in series between the first communication module and the second communication module, is used to collect voltage and temperature data of each battery cell; the sampling module group includes n sampling modules connected in series.

[0013] The main control module performs bidirectional logical address allocation to the sampling module group in parallel through the first and second communication modules, recording the number of successfully allocated sampling modules N1 and N2 respectively; by comparing N1 and N2 with the total number of sampling modules N... total To determine the location of the breakpoint;

[0014] Based on the determination of the breakpoint location, intelligent routing is selected, and corresponding fault handling measures are output.

[0015] On the other hand, this embodiment of the invention also provides a multi-dimensional diagnostic and partition degradation adaptive fault-tolerant method for daisy-chain bidirectional communication, implemented based on the above-mentioned adaptive fault-tolerant system. The adaptive fault-tolerant method includes the following steps:

[0016] S1. The main control module performs bidirectional logical address allocation to the sampling module group in parallel through the first communication module and the second communication module, determines the breakpoint location based on the allocation result, and obtains the number of sampling modules with normal communication.

[0017] S2. Based on the determined breakpoint location and the number of sampling modules for normal communication, the comprehensive decision engine performs intelligent routing selection and outputs the corresponding communication path.

[0018] S3, the integrated decision engine, performs real-time status assessment and multi-dimensional integrated decision-making on the system's communication status and application layer status, and outputs corresponding fault handling measures; among them, the communication status includes the breakpoint location and the number of sampling modules with normal communication; the application layer status includes the battery charge status, power consumption conditions, and fault duration.

[0019] S4, the integrated decision engine judges the severity of the fault based on the breakpoint location and the number of sampling modules with normal communication. Combined with the battery state of charge and power consumption conditions, it judges the safety boundary of the current system through context awareness and further predicts the fault evolution to dynamically upgrade or downgrade fault handling measures.

[0020] S5. Based on the real-time status assessment results of step S3 and the fault evolution prediction results in step S4, perform dynamic decision fusion to dynamically adjust the triggering logic and time response threshold of fault handling measures, and output communication path selection and optimal decision measures based on the fusion results.

[0021] Compared with the prior art, the beneficial effects of the present invention include:

[0022] 1. Improve system availability: Based on parallel address allocation of bidirectional communication interface, accurate diagnosis and intelligent routing selection of daisy chain breakpoints are achieved; through bidirectional diagnosis and intelligent routing, the maximum range of communication capability can still be maintained after a breakpoint occurs, avoiding immediate system shutdown.

[0023] 2. Achieving a balance between safety and operation: Adopting dynamic and forward-looking fault-tolerant decision-making, the comprehensive decision engine integrates multi-dimensional states, real-time communication status, system operation status and lightweight fault prediction to achieve gradient fault-tolerant response, ensuring safety while maintaining operation to the maximum extent.

[0024] 3. Enhance system resilience: Introduce a fault evolution prediction mechanism, use a sliding time window to extract communication quality trends, and output fault stability and risk probability through a lightweight prediction model, thereby identifying fault trends in advance, achieving proactive fault-tolerant scheduling, and optimizing decision-making.

[0025] 4. Optimize degraded operation capability: Through trusted partition energy scheduling, trusted zones and isolated zones are dynamically divided based on breakpoint diagnosis results in communication degradation scenarios. Energy scheduling is performed only based on the state of trusted zones, so that the available battery capacity can still be used safely and efficiently under incomplete information conditions, significantly extending the system's degraded operation time and ensuring the safe operation of the system. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the system hardware architecture according to an embodiment of the present invention;

[0027] Figure 2 This is a flowchart illustrating the bidirectional communication and multidimensional decision diagnosis process in an embodiment of the present invention.

[0028] Figure 3 This is a flowchart illustrating the implementation of the integrated decision engine in an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram of the trusted partition energy scheduling strategy process in an embodiment of the present invention.

[0030] Specific implementation methods

[0031] To make the technical solution of the present invention clearer, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments, but the implementation of the present invention is not limited thereto.

[0032] The abbreviations and key terms involved in this invention are described in detail below:

[0033] BMS: Battery Management System;

[0034] MCU: Main Control Unit;

[0035] AFE: Analog Front End;

[0036] SOC: State of Charge.

[0037] SOH: State of Health;

[0038] EMS: Energy Management System;

[0039] PTZ: Primary Trusted Zone;

[0040] STZ: Secondary Trusted Zone;

[0041] IZ: Isolated Zone;

[0042] EWMA: Exponential Weighted Moving Average.

[0043] LSTM: Long Short-Term Memory network. Example

[0044] This embodiment provides a daisy-chain bidirectional communication multidimensional diagnostic and partition degradation adaptive fault-tolerant system, the architecture of which is as follows: Figure 1 As shown, it mainly includes the following components:

[0045] The main control module (MCU) serves as the control core of the system and is used to run the fault-tolerant algorithm of this invention. A comprehensive decision engine is embedded in the main control module in software form to perform bidirectional diagnosis, multi-dimensional state fusion, fault prediction, and fault-tolerant decision-making.

[0046] The bidirectional communication interface includes a first communication module C1 and a second communication module C2, which are located at the beginning and end of the daisy chain, respectively, and are connected to the main control module MCU through a high-speed communication channel (e.g., SPI) to form a bidirectional closed loop.

[0047] Sampling module group AFE: Connected in series between the first communication module C1 and the second communication module C2, it is used to collect voltage and temperature data of each battery cell. In this embodiment, the sampling module group includes n sampling modules connected in series, namely sampling modules AFE1-AFE. n .

[0048] The adaptive fault-tolerant system in this embodiment employs a bidirectional breakpoint diagnosis and intelligent routing method. See also... Figure 2 , Figure 3 When the system powers on or performs periodic checks, the main control module (MCU) performs bidirectional logical address allocation to the sampling module group AFE in parallel through the first communication module C1 and the second communication module C2, and records the number of successfully allocated sampling modules N1 and N2 respectively; by comparing N1 and N2 with the total number of sampling modules N... total To accurately determine the breakpoint location:

[0049] If the number of sampling modules N1 successfully allocated through the first communication module C1 is equal to the total number of sampling modules N... total They are equal, that is, N1 = N total If N2 = 0, then the breakpoint is determined to be located between the second communication module C2 and the daisy chain;

[0050] If the number of sampling modules N2 successfully allocated through the second communication module C2 is equal to the total number of sampling modules N... total They are equal, that is, N1 = 0, N2 = N. total If so, the breakpoint is determined to be located between the first communication module C1 and the daisy chain;

[0051] If the sum of the number of successfully allocated sampling modules N1 and the number of sampling modules N2 equals the total number of sampling modules N... total That is, N1 + N2 = N total Then the breakpoint is determined to be located between the N1th sampling module and the Nth sampling module in the daisy chain. 1+1 Between sampling modules;

[0052] If the sum of the number of successfully allocated sampling module groups N1 and the number of sampling module groups N2 is less than the total number of sampling module groups N... total That is, N1 + N2 < Ntotal If so, it is determined that there are at least two breakpoints on the daisy chain.

[0053] Based on the determination of the breakpoint location, the system performs intelligent routing selection and automatically selects the communication interface that does not pass through the breakpoint and has the shortest path (i.e., selects the first communication module C1 or the second communication module C2) to communicate with the target sampling module, thereby maximizing the communication range.

[0054] This embodiment also provides a multi-dimensional diagnostic and partition degradation adaptive fault-tolerant method for daisy-chain bidirectional communication. The specific process of its adaptive fault tolerance includes the following steps:

[0055] S1. The main control module MCU performs bidirectional logical address allocation to the sampling module group AFE in parallel through the first communication module C1 and the second communication module C2. Based on the allocation result, the breakpoint position is determined and the number of sampling modules with normal communication is obtained.

[0056] The main control module (MCU) will use the number of sampling modules N1 successfully allocated through the first communication module C1, the number of sampling modules N2 successfully allocated through the second communication module C2, and the total number of sampling modules N to determine the total number of sampling modules N. total By comparing the data, the exact location of the breakpoint (P) and the number of sampling modules in normal communication (N) can be determined to identify whether there are single or multiple breakpoints.

[0057] S2. Based on the determined breakpoint location and the number of sampling modules for normal communication, the integrated decision engine embedded in the main control module MCU performs intelligent routing selection and outputs the corresponding communication path.

[0058] In this step, the system automatically selects the communication interface with the optimal path without any breakpoints (i.e., the first communication module C1 or the second communication module C2) through the comprehensive decision engine to communicate with the target sampling module and exchange data, so as to maximize the communication range and ensure that as much cell status information as possible can be collected even in fault environments.

[0059] Specifically, in this embodiment, the integrated decision engine performs multi-dimensional integrated decision-making based on the diagnosed breakpoint location and the number of sampling modules with normal communication, thereby achieving intelligent routing selection. Intelligent routing selection mainly involves communication path selection, with the selection result being either communication with the target sampling module through the first communication module C1 or communication with the target sampling module through the second communication module C2, denoted as R ∈ {through C1, through C2}.

[0060] S3, the integrated decision engine, performs real-time status assessment and multi-dimensional comprehensive decision-making on the system's communication and application layer states, and outputs corresponding fault handling measures (also known as fault tolerance strategies). The communication state includes the breakpoint location and the number of sampling modules with normal communication; the application layer state includes the battery state of charge (SOC), power consumption conditions (charging / discharging / standby), and fault duration (T).

[0061] In the multi-dimensional integrated decision-making process, the integrated decision engine receives communication state vectors and application layer state vectors, and outputs fault handling measures. The communication state vector C = {P, N}, where P is the breakpoint location and N is the number of sampling modules with normal communication; the application layer state vector S = {SOC, G, T}, where SOC is the battery state of charge, G is the power consumption condition (specifically charging, discharging, or standby), and T is the fault duration.

[0062] The fault handling measures include cutting off the charging and discharging of the corresponding sampling module, limiting the power of the corresponding sampling module, and keeping the corresponding sampling module running, denoted as A ∈ {M1, M2, M3}, where the risk level is: M1 (cut off charging and discharging) > M2 (limit power) > M3 (keep the status quo).

[0063] S4, the integrated decision engine determines the severity of the fault based on the breakpoint location and the number of sampling modules with normal communication. Combining the battery state of charge (SOC) and power consumption condition (G), it determines the safety boundary of the current system through context awareness and further predicts the fault evolution to dynamically upgrade or downgrade fault handling measures.

[0064] The mechanism for predicting fault evolution is as follows: Based on historical data of the judgment period, extract time-series features such as communication quality volatility, decay rate of effective node number (i.e., number of sampling modules in normal communication), and growth slope of fault duration T; introduce a lightweight fault prediction model (such as EWMA or minimal LSTM unit) to perform time-series modeling of communication quality volatility, decay rate of effective node number, and growth slope of fault duration, output prediction results, and use them to dynamically adjust the trigger threshold of fault handling measures, thereby realizing the leap from post-event response to pre-event prediction.

[0065] The prediction results include a fault stability score Φ and the probability P_risk of entering a high-risk state in the future. The fault stability score Φ∈[0,1], where Φ→0 indicates high instability and potential deterioration.

[0066] S5. Based on the real-time status assessment results of step S3 and the fault evolution prediction results in step S4, perform dynamic decision fusion to dynamically adjust the triggering logic and time response threshold of fault handling measures, and output communication path selection (R) and optimal decision measures (A) based on the fusion results.

[0067] The optimal decision measures include: cutting off charging and discharging (M1), limiting power / degrading operation (M2), and maintaining the status quo (M3).

[0068] The dynamic decision fusion process adopted in this embodiment is as follows: if the fault stability score Φ>0.8 and the probability P_risk<5% in the prediction result, the transient tolerance window is relaxed. Even if the fault duration T>the first preset period t1, the tolerance window of M3 (maintain the status quo) can still be extended; if the fault stability score Φ<0.4 and the probability P_risk>30%, the time response threshold is tightened. Even if the current fault duration T<the second preset period t2, M2 (limit power) is activated in advance; if the battery state of charge SOC is in the critical range (<10% or>90%) and the probability P_risk>20%, M2 is directly locked, triggering degraded operation.

[0069] S6. In the power-limited degraded operation mode, communication reachability analysis is performed based on the breakpoint location, and the battery cluster is dynamically divided into a primary trusted zone, a secondary trusted zone, and an isolation zone; different energy scheduling strategies are implemented in different zones.

[0070] like Figure 4 This embodiment employs a trusted partitioned energy scheduling strategy. In the power-limited degraded operation mode (M2), the system determines communication reachability based on breakpoint locations and dynamically divides battery clusters into: a primary trusted zone (PTZ), which is a continuous AFE group with optimal communication quality; a secondary trusted zone (STZ), which is a sampling module area with normal but suboptimal communication quality; and an isolation zone (IZ), which is a sampling module area completely disconnected. Furthermore, the system reports the available power of the EMS and calculates the remaining capacity based on SOCtrust, limiting charging and discharging according to the historical state of the isolation zone to achieve energy scheduling.

[0071] The energy scheduling strategy includes: re-estimating the SOC (SOCtrust) based on cell data from the main trusted zone PTZ and the secondary trusted zone STZ, and reconstructing the trusted SOC; dynamically prohibiting charging or limiting discharging according to the location of the isolation zone IZ; and dynamically tightening the voltage and temperature safety thresholds of the main trusted zone PTZ to compensate for the uncertainty caused by missing information.

[0072] In addition, the protection threshold of the individual unit voltage within the trusted region can be dynamically adjusted to compensate for the uncertainty caused by missing information and tighten the safety boundary of system operation.

[0073] The technical solution of this embodiment can be replaced by the following equivalent substitutions: the bidirectional communication interface can be replaced by other communication protocols such as CAN and UART instead of SPI; the fault prediction model can be replaced by timing prediction algorithms such as ARIMA and lightweight Transformer instead of LSTM; in the trusted partitioning strategy, the selection of PTZ can be determined based on multiple indicators such as communication delay and signal quality, rather than solely based on the communication success rate.

[0074] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A multidimensional diagnostic and partition degradation adaptive fault-tolerant system with daisy-chain bidirectional communication, characterized in that, include: The main control module is the control core of the system; a comprehensive decision engine is embedded in the main control module to perform bidirectional diagnosis, multi-dimensional state fusion, fault prediction and fault-tolerant decision-making. The bidirectional communication interface includes a first communication module and a second communication module located at the beginning and end of the daisy chain, respectively, and is connected to the main control module through a high-speed communication channel to form a bidirectional closed loop; A sampling module group, connected in series between the first communication module and the second communication module, is used to collect voltage and temperature data of each battery cell; the sampling module group includes n sampling modules connected in series. The main control module performs bidirectional logical address allocation to the sampling module group in parallel through the first and second communication modules, recording the number of successfully allocated sampling modules N1 and N2 respectively; by comparing N1 and N2 with the total number of sampling modules N... total To determine the location of the breakpoint; Based on the determination of the breakpoint location, intelligent routing is selected, and corresponding fault handling measures are output.

2. The adaptive fault-tolerant system according to claim 1, characterized in that, The process for determining the breakpoint location is as follows: If the number of sampling modules N1 successfully allocated through the first communication module is equal to the total number of sampling modules N... total If they are equal, then the breakpoint is determined to be located between the second communication module and the daisy chain; If the number of sampling modules N2 successfully allocated through the second communication module is equal to the total number of sampling modules N... total If they are equal, then the breakpoint is determined to be located between the first communication module and the daisy chain; If the sum of the number of successfully allocated sampling modules N1 and the number of sampling modules N2 equals the total number of sampling modules N... total Then the breakpoint is determined to be located between the N1th sampling module and the Nth sampling module in the daisy chain. 1+1 Between sampling modules; If the sum of the number of successfully allocated sampling module groups N1 and the number of sampling module groups N2 is less than the total number of sampling module groups N... total If so, it is determined that there are at least two breakpoints on the daisy chain.

3. A multidimensional diagnostic and partitioned degradation adaptive fault-tolerant method for daisy-chain bidirectional communication, implemented based on the adaptive fault-tolerant system described in claim 1 or 2, characterized in that... The adaptive fault-tolerant method includes the following steps: S1. The main control module performs bidirectional logical address allocation to the sampling module group in parallel through the first communication module and the second communication module, determines the breakpoint location based on the allocation result, and obtains the number of sampling modules with normal communication. S2. Based on the determined breakpoint location and the number of sampling modules for normal communication, the comprehensive decision engine performs intelligent routing selection and outputs the corresponding communication path. S3, the integrated decision engine, performs real-time status assessment and multi-dimensional integrated decision-making on the system's communication status and application layer status, and outputs corresponding fault handling measures; among them, the communication status includes the breakpoint location and the number of sampling modules with normal communication; the application layer status includes the battery charge status, power consumption conditions, and fault duration. S4, the integrated decision engine judges the severity of the fault based on the breakpoint location and the number of sampling modules with normal communication. Combined with the battery state of charge and power consumption conditions, it judges the safety boundary of the current system through context awareness and further predicts the fault evolution to dynamically upgrade or downgrade fault handling measures. S5. Based on the real-time status assessment results of step S3 and the fault evolution prediction results in step S4, perform dynamic decision fusion to dynamically adjust the triggering logic and time response threshold of fault handling measures, and output communication path selection and optimal decision measures based on the fusion results.

4. The adaptive fault-tolerant method according to claim 3, characterized in that, It also includes the following steps: S6. In the power-limited degraded operation mode, communication reachability analysis is performed based on the breakpoint location, and the battery cluster is dynamically divided into a primary trusted zone, a secondary trusted zone, and an isolation zone; different energy scheduling strategies are implemented in different zones.

5. The adaptive fault-tolerant method according to claim 3, characterized in that, The result of intelligent routing selection in step S2 is: communicating with the target sampling module through the first communication module or communicating with the target sampling module through the second communication module.

6. The adaptive fault-tolerant method according to claim 3, characterized in that, The fault handling measures in step S3 include cutting off the charging and discharging of the corresponding sampling module, limiting the power of the corresponding sampling module, and keeping the corresponding sampling module running.

7. The adaptive fault-tolerant method according to claim 3, characterized in that, The mechanism for predicting fault evolution in step S4 is as follows: Based on historical data of the judgment period, the communication quality fluctuation rate, the attenuation rate of the number of sampling modules in normal communication, and the growth slope of the fault duration are extracted; a lightweight fault prediction model is introduced to perform time-series modeling of the communication quality fluctuation rate, the attenuation rate of the number of sampling modules in normal communication, and the growth slope of the fault duration, and the prediction results are output for dynamically adjusting the trigger threshold of fault handling measures.

8. The adaptive fault-tolerant method according to claim 3, characterized in that, The prediction results for fault evolution in step S4 include the fault stability score Φ and the probability of entering a high-risk state in the future, P_risk. The optimal decision measures in step S5 include: cutting off charging and discharging M1, limiting power / degrading operation M2, and maintaining the status quo M3. The dynamic decision fusion process is as follows: if the fault stability score Φ>0.8 and the probability P_risk<5% in the prediction result, the transient tolerance window is relaxed; if the fault stability score Φ<0.4 and the probability P_risk>30%, the time response threshold is tightened; if the battery state of charge is in the critical range and the probability P_risk>20%, M2 is directly locked, triggering degrading operation.

9. The adaptive fault-tolerant method according to claim 4, characterized in that, In step S6, the primary trusted region is the group of continuous sampling modules with the best communication quality; the secondary trusted region is the area of ​​sampling modules with normal communication but suboptimal quality; and the isolation region is the area of ​​sampling modules that are completely disconnected.

10. The adaptive fault-tolerant method according to claim 4, characterized in that, The energy scheduling strategy in step S6 includes: re-estimating the battery state of charge based on the cell data of the main trusted region and the secondary trusted region, and reconstructing the trusted battery state of charge; dynamically prohibiting charging or limiting discharging according to the location of the isolation region; and dynamically tightening the voltage and temperature safety thresholds of the main trusted region to compensate for the uncertainty caused by the lack of information.

Images