Vehicle-mounted network node management method based on AUTOSAR architecture

By dividing the main network layer and subnet layer in the AUTOSAR architecture and improving the state machine for fault detection, the real-time and compatibility problems of AUTOSAR network management during cross-domain are solved, the fault detection efficiency is improved, and the reliability of the vehicle network is ensured. sex.

CN118945081APending Publication Date: 2024-11-12CHONGQING UNIV OF POSTS & TELECOMM
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411174284.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Currently, AUTOSAR network management is difficult to meet the real-time and compatibility requirements of network management packets during cross-domain network management, and the fault detection time is too long, which makes it easy to misjudgment the node status, affecting the normal operation of the automotive intelligent system.

Method used

Adopting the on-vehicle network node management method based on AUTOSAR architecture, the entire vehicle network segment is divided into the main network layer and the subnet layer, cross-domain transmission is realized through the AUTOSAR standard Ethernet NM PDU, and the state machine is improved to perform fault detection in the subnet layer. , ensure real-time and reliable communication of network nodes. Specific measures include the design of cross-domain transmission and fault detection strategies, and the fault judgment is made by proxy node monitoring and mapping network management messages, calculating indicators such as message arrival time difference.

Benefits of technology

It improves the real-time and compatibility of network management messages, shortens the fault detection time, enhances the network reliability and fault detection efficiency, and avoids misjudgment of node status.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118945081A_ABST
    Figure CN118945081A_ABST
Patent Text Reader

Abstract

The invention relates to a vehicle-mounted network node management method based on an AUTOSAR architecture, and belongs to the technical field of automobile intelligence. According to the method, a whole vehicle network segment is divided into a main network layer and a sub-network layer according to a domain centralized electronic and electrical architecture. The main network layer is formed by mutually connecting a plurality of domain controllers, and cross-domain transmission of network management messages is realized by using an Ethernet NM PDU (Network Management Protocol Data Unit) of an AUTOSAR standard; the sub-network layer is formed by connecting domain controllers and nodes subordinate to the domain controllers, and fault detection is carried out on network nodes by improving an AUTOSAR network management state machine and adding a fault detection state in the sub-network layer, so that real-time reliable communication of the vehicle-mounted network nodes is ensured. According to the invention, the compatibility, real-time performance and fault detection efficiency of vehicle-mounted network management are effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of automobile intelligent technology, relates to vehicle network management technology, and specifically relates to a vehicle network node management method based on AUTOSAR architecture. Background Art

[0002] Since the 1990s, with the development of intelligent, convenient and networked automobiles, the number of on-board electronic control units (ECUs) and on-board communication data have increased significantly. In order to ensure the real-time and reliable communication of the vehicle control network communication system, the Automotive Open System Architecture (AUTOSAR) has proposed a network management (NM) function with functions such as state distributed monitoring and collaborative sleep to meet the growing data communication needs of smart cars.

[0003] At present, cars have integrated automatic driving functions such as fully automatic parking and emergency braking. This intelligent driving system collects data through sensors (such as ultrasonic radar and cameras) and transmits it to the intelligent control system through the network. The system then issues instructions to coordinate the acceleration, braking and steering of the vehicle to achieve automatic driving. In this process, data is constantly transmitted between Ethernet and CAN, and the demand for cross-domain network management increases. In order to ensure the normal operation of the automatic driving system, the real-time requirements for fault detection of each ECU are also significantly improved. Therefore, it is crucial to study network management in this scenario.

[0004] However, the current AUTOSAR network management has the following shortcomings: (1) When managing cross-domain networks, the compatibility and real-time performance of network management messages are difficult to meet requirements; (2) AUTOSAR official fault detection is implemented through the node's own send / receive error timeout, which takes too long, and when the node is in the ready sleep state, it will not send network management messages, which can easily mistake the faulty node for a normal sleep node. These shortcomings are particularly prominent in the environment where the degree of automotive intelligence is constantly deepening. Summary of the invention

[0005] In view of this, the object of the present invention is to provide a vehicle network node management method based on the AUTOSAR architecture, to improve the real-time performance and compatibility of network management messages, and to reduce fault detection time.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A vehicle network node management method based on the AUTOSAR architecture, the method divides the vehicle network segment into a main network layer and a subnet layer according to the domain centralized electronic and electrical architecture. The main network layer is composed of multiple domain controllers connected to each other, and the cross-domain transmission of network management messages is realized by using the AUTOSAR standard Ethernet NM PDU; the subnet layer is composed of each domain controller connected to the nodes under the domain controller, and the AUTOSAR network management state machine is improved in the subnet layer, and the fault detection state is added to perform fault detection on the network node, so as to ensure the real-time and reliable communication of the vehicle network nodes.

[0008] Among them, cross-domain transmission includes data exchange between Ethernet and CAN bus. Ways to achieve cross-domain transmission include:

[0009] 1) Each domain controller is used as a proxy node to periodically monitor the network management messages sent in the subnet layer, and map the received network management messages into network management messages in a unified format; specifically, the received network management messages are divided into network management messages and application messages, and the network management messages and application messages are mapped into network management messages in a unified format;

[0010] 2) The mapped messages are encapsulated into NM Packets, and the NMPackets are sent to the destination domain controller via the Ethernet AUTOSAR UDP NM mechanism;

[0011] 3) After receiving NM Packets, the destination domain controller filters irrelevant network management messages and sends relevant network management messages to the corresponding nodes in the subnet layer.

[0012] Further, the fault detection includes performing fault detection in a NOS state and performing fault detection in a RSS state.

[0013] Furthermore, fault detection in the NOS state includes: each node monitors all nodes in the network segment and other network segments, and is monitored by other nodes; each node receives network management messages from other nodes, records the NM cycle of other nodes, and calculates the corresponding value; if a node calculates the If the value is greater than the preset threshold value Φ, it will be reported to the application layer Node failure; application layer notification The node enters the fault detection state. The node sends an NM PDU message to try to communicate with the bus and detect whether its own message is sent successfully. If the detection and sending fail, the application layer reports its own fault.

[0014] in, The value is calculated by the following process:

[0015] 1) The node stores the network management messages received nearly k times. Each time a message arrives, the arrival time is stored in the sampling window. If the sampling window reaches the storage limit, the oldest data is deleted. The time difference between two adjacent messages is recorded as Δt, and the mean μ and variance σ of the k-times sample time difference are calculated. 2 ;

[0016] 2) Calculate the probability P of receiving the next message within t time slices after receiving the previous message later (t):

[0017]

[0018] 3) Calculate the current time value:

[0019]

[0020] Where, T last Indicates the time when the last message was received, t now Indicates the current time.

[0021] Furthermore, fault detection in the RSS state includes: the node sends a frame of NM PDU message and enters the fault detection state just before entering the PBM state after the RSS state; the other nodes determine whether they have received the NM PDU message, and if not, report to the application layer that they have failed; the application layer performs statistics to determine the faulty node based on the fault alarm information.

[0022] Among them, the application layer performs statistical judgment on the fault nodes based on the fault alarm information, including:

[0023] 1) If a node A reports the sending failure of all other nodes to the application layer, it is determined that node A has a receiving failure;

[0024] 2) If a certain node B is reported as having a transmission failure by all other nodes, it is determined that the node B has a transmission failure;

[0025] 3) If a certain node A only reports the sending failure of several nodes and is not reported as such by all other nodes except itself, while a certain node C is reported as such by some nodes including node A, it is determined that a sending failure occurs at node C.

[0026] Furthermore, the active wake-up bit in the NM PDU message used for fault detection is 0, which will not keep the entire network in an awake state, thereby not affecting the coordinated sleep of the cluster nodes.

[0027] The beneficial effects of the present invention are as follows: the present invention solves the problem of poor compatibility and real-time performance of network management messages of current vehicle-mounted network management nodes during cross-domain network management, solves the problem of long fault detection time, and when the node is in a ready sleep state, the network management message will not be sent, and the faulty node will be mistaken for a normal sleep node, thereby improving the reliability of the network and the efficiency of fault detection.

[0028] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0030] Figure 1 This is a schematic diagram of the operation of the cross-domain network management mechanism;

[0031] Figure 2 For the improved AUTOSAR state machine;

[0032] Figure 3 This is a schematic diagram of the fault handling process when the node is in NOS state;

[0033] Figure 4 The figure is a schematic diagram of the troubleshooting process when the node is in the RSS state. DETAILED DESCRIPTION

[0034] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0035] Explanation of relevant abbreviations:

[0036] NM PDU: Network Management Protocol Data Unit, Network Management Protocol Data Unit

[0037] UDP NM: User Datagram Protocol Network Management, the main purpose of which is to coordinate the transition of the network between NOS state and BSM state

[0038] ECU: Electronic Control Unit

[0039] BSM: Bus Sleep Mode, bus sleep state, each node enters this state after the vehicle is powered on

[0040] RMS: Repeat Message State, a node enters this state after receiving a wake-up request

[0041] NOS: Normal Operation State, normal sending state, enter this state after the RMS timer ends

[0042] RSS: Ready Sleep State, the node enters this state after the wake-up source is lost

[0043] PBM: Pre Bus-Sleep Mode, ready to sleep state, the node enters this state after exiting the fault detection state, and enters the BSM state after exiting this state

[0044] The present invention proposes a vehicle network node management method based on the AUTOSAR architecture, which divides the vehicle network segment into a main network layer and a subnet layer according to the domain centralized electronic and electrical architecture. In the main network layer, the cross-domain transmission of network management messages is realized by using the AUTOSAR standard Ethernet NM DPU, and in the subnet layer, the AUTOSAR network management state machine is improved and the fault detection state is added to ensure the real-time and reliable communication of the vehicle network nodes.

[0045] The specific contents of this method are as follows:

[0046] 1. Hierarchical network architecture design for the entire vehicle network segment

[0047] The vehicle network is divided into a main network layer and a sub-network layer according to the domain-centralized electronic and electrical architecture. The main network layer is formed by multiple domain controllers connected to each other, while the sub-network layer is formed by the nodes under each domain controller connected to the corresponding domain controller, such as Figure 1 shown.

[0048] 2. Unified network management message format design

[0049] In order to enhance the compatibility of subnet network management, the network management messages of different protocols of multiple subnets are mapped into unified network management messages according to the AUTOSAR network management standard, depending on whether they send network management messages or application messages, to achieve mutual mapping between different network management strategies between domains. The network management message of each subnet is encapsulated in an NM Packet and sent uniformly by the Ethernet of the domain controller. The mapping relationship between network management states of different domains is shown in the following subscript 1:

[0050] Table 1

[0051]

[0052]

[0053] 3. Implementation of cross-domain network management

[0054] (1) Proxy node detection: The domain controller node acts as a proxy node, periodically monitoring the network management messages sent in the subnet and mapping the received network management messages into unified network management messages.

[0055] (2) Message encapsulation and transmission: The mapped messages are encapsulated into NM Packets inside the domain controller and sent to the destination domain controller through the AUTOSAR UDP NM mechanism of the backbone network.

[0056] (3) Message filtering and distribution: The destination domain controller filters irrelevant network management messages and sends relevant network management messages to the corresponding ECU nodes, thus achieving the unification and simplification of cross-domain network management.

[0057] 4. Implementation of network management between domain nodes.

[0058] Each domain controller not only transmits the network management message information of the subnet through NM Packets, but also includes its own status information in the NM Packets, which is located in the first two bytes of the message. The communication between the nodes under the domain controller follows the standard AUTOSAR network management strategy to ensure the compatibility and efficiency of cross-domain network management.

[0059] In this way, the main network layer only retains the simplest network management logic, that is, the domain controller in Ethernet independently and periodically receives and sends NM Packets. Through the coordination of the main network layer NM and subnet layer NM mechanisms, the global monitoring of the vehicle network communication status is achieved.

[0060] 5. Fault detection function implementation

[0061] A fault detection state (Fault Check, each node enters this state when a fault occurs, and re-enters the RMS state to join the network after the fault is repaired successfully) is added to the standard AUTOSAR state machine to implement fault detection.

[0062] In the improved AUTOSAR architecture, the entry / exit conditions of each state are as follows:

[0063] 1) Power-on initialization.

[0064] 2) The node receives the wake-up request, enters the RMS state and sends an NM PDU to wake up the other nodes.

[0065] 3) After receiving the NM PDU, the remaining nodes enter the RMS state, the T_REPEAT_MESSAGE timer countdown ends, and they enter the NOS state to perform the normal sending mode of the network management message.

[0066] 4) The node receives a duplicate message request.

[0067] 5) Network release, such as the local node needs to sleep.

[0068] 6) Receive an NM PDU without the Repeat Message bit.

[0069] 7) Receive NM PDU with Repeat Message bit.

[0070] 8) Network release.

[0071] 9)T_NM_TIMEROUT timer times out.

[0072] 10) Receive the network wake-up request again.

[0073] 11)T_WAIT_BUS_SLEEP timer times out.

[0074] 12) Receive other node failure notification in NOS state.

[0075] 13) Repaired successfully from the fault detection state (Fault Check).

[0076] 14) Receive other node failure notifications in RMS state.

[0077] 15) After entering the RSS state, it automatically enters the Fault Check and performs a self-test before sleeping.

[0078] 16) When the fault judgment conditions are met, report to the application layer.

[0079] 17) The application layer attempts to restart or switch to a redundant system for repair.

[0080] 6. Fault detection strategy design

[0081] (1) Based on historical messages Value calculation

[0082] ① By collecting normal network management message information, it is found that due to the delay of network management messages, their distribution roughly conforms to the normal distribution. The normal distribution assumption based on historical messages can be used for fault judgment.

[0083] The time when the message arrives is stored in the sampling window. The network management messages for the last k times are stored. Each time a message arrives, the arrival time is stored in the sampling window. If the sampling window is full, the oldest data is deleted. The time difference between the two messages before and after is Δt. The average μ and variance σ of these k samples are calculated. 2 :

[0084]

[0085] ② Based on these k samples, calculate the distribution of message arrival time intervals. The fault detection assumes that the distribution of the message arrival time follows a normal distribution, P later (t) represents the probability of receiving the next message within t time slices after receiving the previous message, which is calculated by the sample mean μ and variance σ 2 , calculate P later (t):

[0086]

[0087] ③ Use normal distribution to calculate the current Value. last Indicates the time when the last message was received, t now Indicates the current time. last ,t now and P later Substitute (t) into the following formula to calculate value:

[0088]

[0089] Take the arrival interval data of 30 groups of health messages shown in Table 2 as an example:

[0090] Table 2

[0091]

[0092] Calculate the threshold value φ under the healthy message state (calculated in advance by the above formula). According to the above formula, the average value μ=50.06333 and the standard deviation σ=0.611126 of these 30 groups of messages are calculated. According to different confidence intervals, the appropriate Φ value is selected, as shown in Table 3 below:

[0093] Table 3

[0094]

[0095] The transmission period of the message is 50ms. Analysis of the healthy network management period message results shows that the maximum period error is 1.12ms. Φ=2 can be selected as the fault detection threshold.

[0096] (2) Fault detection strategy design in NOS state

[0097] In the NOS state, the domain controller acts as a proxy node and periodically forwards network management messages required by each node to the target node. Each node monitors all nodes in its own network segment and other network segments, and is monitored by other nodes. Figure 3 As shown in Figure 1, the node receives network management messages from other nodes, records the NM cycles of other nodes, and calculates the corresponding NM cycles of the remaining nodes. If a node a calculates another node b's If the value is greater than the preset threshold value Φ, the node b fault is reported and reported to the application layer. After receiving the application layer fault notification, node b enters the fault detection state, sends an NMPDU message with the active wake-up bit as 0 and the fault bit as 1, attempts to communicate with the bus, and detects whether its own message is sent successfully. If the detection fails, the node fault is reported. The application layer then repairs it by resetting the ECU or switching the redundant system, repeating the detection process until the message is sent successfully, and re-enters the RMS state to join the bus communication.

[0098] (3) Fault detection strategy design in RSS state

[0099] After the vehicle is powered off or the wake-up source is lost, the node enters the RSS state from the NOS state. When the node is in the RSS state, it cannot send NM PDU; if the node fails at this time, it also cannot send messages. For other nodes, the external performance of this node is no different, so the fault condition of this node cannot be identified.

[0100] Taking this situation into consideration, before a node enters the PBM state after the RSS state, it sends a frame of NM PDU message with the fault detection bit set to 1 to allow it to enter the fault detection state. The remaining nodes do not receive the fault detection NM PDU from a certain node in the Fault Check state, indicating that the sending node is faulty or the receiving node is faulty. A receiving node cannot make a final judgment. As long as the node does not receive the fault detection NM PDU, it considers that it has failed and reports it to the application layer. The application layer determines the faulty node based on the fault alarm information. The active wake-up bit in the fault detection NM PDU is 0, which will not affect the collaborative sleep of the cluster nodes. After sending the fault detection message, the node enters the PBM state. After the waiting bus sleep timer WaitBus Sleep times out, it enters the BSM state and the bus goes into sleep.

[0101] The application layer uses the statistical analysis method shown in Table 4 to perform fault diagnosis.

[0102] Table 4

[0103]

[0104] In addition, in this embodiment, the Ethernet UDP NM message format is set as shown in Table 5 below:

[0105] Table 5

[0106]

[0107] In Table 5, Byte 0 is used to indicate the source address Source Domain Identifier of the domain controller node, Byte 1 is the CBV control bit, which is designed with reference to NM PDU, and in the subsequent Byte bits, every 8 bytes are used to indicate the communication status of the subnet node.

[0108] In this embodiment, the format of the NM PDU message is set as shown in Table 6 below:

[0109] Table 6

[0110]

[0111] In Table 6, Bit 0-7 is the node source node identifier; Bit 8 is the Repeat Message Bit; Bit 12 is the node actively waking up the network or passively waking up the network; Bit 15 is the fault detection bit; the design of other bits refers to the original AUTOSAR standard.

[0112] In this embodiment, the time parameters of the network management strategy are improved, as shown in the following Table 7:

[0113] Table 7

[0114]

[0115] In Table 7, the T_REPEAT_MESSAGE timer is used for RMS state countdown. After the countdown ends, the node enters NOS state. The T_WAIT_BUS_SLEEP timer is used for waiting for bus sleep timer. After the countdown ends, the node enters BSM state. The T_NM_Immediate_Time timer is the sending cycle timer of the network management message in the repeat message sending state. The function of the T_NM_TIMEOUT timer is to start this timer as long as the node enters the network mode. When this timer expires, the node will enter the pre-sleep mode.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A vehicle network node management method based on AUTOSAR architecture, characterized in that: The method divides the vehicle network segment into a main network layer and a subnet layer according to the domain centralized electronic and electrical architecture; the main network layer is composed of multiple domain controllers connected to each other, and the cross-domain transmission of network management messages is realized by using the Ethernet NM PDU of the AUTOSAR standard; the subnet layer is composed of each domain controller connected with the nodes under the domain controller, and the AUTOSAR network management state machine is improved in the subnet layer, and the fault detection state is added to perform fault detection on the network node, so as to ensure the real-time and reliable communication of the vehicle network nodes.

2. The vehicle network node management method according to claim 1, characterized in that: The cross-domain transmission includes data exchange between Ethernet and CAN bus.

3. The vehicle network node management method according to claim 2, characterized in that: The method of realizing cross-domain transmission includes: using each domain controller as a proxy node, periodically monitoring the network management messages sent in the subnet layer, and mapping the received network management messages into network management messages in a unified format; The mapped messages are encapsulated into NM Packets, which are sent to the destination domain controller via the Ethernet AUTOSAR UDP NM mechanism; After receiving NM Packets, the destination domain controller filters irrelevant network management messages and sends relevant network management messages to the corresponding nodes in the subnet layer.

4. The vehicle network node management method according to claim 1, characterized in that: The fault detection includes fault detection in a NOS state and fault detection in a RSS state.

5. The vehicle network node management method according to claim 4, characterized in that: Fault detection in NOS state includes: each node monitors all nodes in the network segment and other network segments, and is monitored by other nodes; each node receives network management messages from other nodes, records the NM cycle of other nodes, and calculates the corresponding value; if a node calculates the If the value is greater than the preset threshold value Φ, it will be reported to the application layer Node failure; Application layer notification The node enters the fault detection state. The node sends an NM PDU message to try to communicate with the bus and detect whether its own message is sent successfully. If the detection and sending fail, the application layer reports its own fault.

6. The vehicle network node management method according to claim 5, characterized in that: The The value is calculated by the following process: 1) The node stores the network management messages received nearly k times. Each time a message arrives, the arrival time is stored in the sampling window. If the sampling window reaches the storage limit, the oldest data is deleted. The time difference between two adjacent messages is recorded as Δt, and the mean μ and variance σ of the k-times sample time difference are calculated. 2 ; 2) Calculate the probability P of receiving the next message within t time slices after receiving the previous message later (t): 3) Calculate the current time value: Where, T last Indicates the time when the last message was received, t now Indicates the current time.

7. The vehicle network node management method according to claim 4, characterized in that: Fault detection in the RSS state includes: the node sends a frame of NM PDU message and enters the fault detection state just before entering the PBM state after the RSS state; the other nodes determine whether they have received the NM PDU message. If not, they report the fault to the application layer; the application layer performs statistics based on the fault alarm information to determine the faulty node.

8. The vehicle network node management method according to claim 7, characterized in that: The application layer statistically determines the faulty node according to the fault alarm information, including: 1) If a node A reports the sending failure of all other nodes to the application layer, it is determined that node A has a receiving failure; 2) If a certain node B is reported as having a transmission failure by all other nodes, it is determined that the node B has a transmission failure; 3) If a certain node A only reports the sending failure of several nodes and is not reported as such by all other nodes except itself, while a certain node C is reported as such by some nodes including node A, it is determined that a sending failure occurs at node C.

9. The vehicle network node management method according to claim 5 or 7, characterized in that: The active wake-up bit in the NMPDU message used for fault detection is 0.

Citation Information

Cited By

  • Multi-domain time synchronization system based on AUTOSAR AP, medium and vehicle

    CN121664345A