Intelligent driving controller time synchronization system and method

Abstract

The invention discloses an intelligent driving controller time synchronization system and method, and aims to solve the problems of low time synchronization precision, timestamp jumping, time shuttling and the like caused by dependence on whole vehicle GPS time in the prior art. According to the system, the master control module (MCU) is combined with the whole vehicle CAN bus time and the locally stored historical time to generate stable reference time, high-precision time synchronization of receiving modules (SOC, Radar and the like) is achieved through an internal self-operation counter, it is ensured that an intelligent driving application layer obtains continuous and consistent timestamps, and the stability and reliability of the system are improved.

Description

Technical Field

[0001] The present invention relates to the field of automobile intelligent driving technology, and in particular to an intelligent driving controller time synchronization system and method. Background Art

[0002] In the intelligent driving system of automobiles, the intelligent driving controller needs to synchronize time with multiple sensors (such as cameras, radars, etc.) to ensure that the perception data collected by the sensors are consistent in the time dimension and provide reliable input for the upper-level application algorithms. In the scenario where the vehicle only has CAN bus timing but no Ethernet timing, the intelligent driving controller will use the vehicle's GPS time as the synchronization benchmark. In actual applications, due to the poor stability of GPS time: the GPS time accuracy of the entire vehicle is at the second level, and the initialization and calibration time of the GPS device at the beginning of power-on is relatively long, it is easy to cause the sent GPS time to be invalid or jump. This will cause the intelligent driving controller that uses the vehicle's GPS time as the time synchronization benchmark to be unable to quickly and accurately obtain the valid GPS time, resulting in the timestamp sent to the sensor jumping, and then the perception data sent back by the sensor has an abnormal timestamp, causing problems such as upper-level application errors and function exit.

[0003] If the intelligent driving controller's chip MCU, SOC and external sensors (such as radar) use different time sources, inconsistent time bases may occur, leading to data fusion errors, which in turn may cause time asynchrony among multiple chips.

[0004] Therefore, a solution is needed that can decouple the GPS time dependency of the entire vehicle and ensure stable time synchronization of multiple chips. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems of time jump and time asynchrony caused by relying on the GPS time of the entire vehicle in the prior art, and to provide a time synchronization system for an intelligent driving controller.

[0006] The technical solution of the present invention is a time synchronization system for an intelligent driving controller, comprising:

[0007] The time source is used to send UTC time messages in Tb seconds;

[0008] The time synchronization master is an MCU, equipped with NVM and RAM memories. The MCU receives Tb and stores the effective UTC time of the vehicle when it is powered off in T seconds in NVM. After powering on, the MCU generates a reference time of T0 nanoseconds based on Tb or T and stores it in RAM. At the same time, the MCU generates an internal self-running cumulative counter value of Δt nanoseconds.

[0009] The time synchronization receiving end is SOC and Radar, which respectively receive the synchronization time sent by MCU and generate the sensing data timestamp;

[0010] The time usage end is an application layer, which is used to receive and use the perception data timestamp.

[0011] By adopting this technical solution, the entire vehicle's GPS time dependency can be effectively decoupled, ensuring that the intelligent driving controller can quickly acquire a stable time reference even when GPS time is invalid or jumps. Furthermore, by using the MCU to generate an internal self-running cumulative counter value, this solution ensures stable time synchronization across multiple chips, avoiding data fusion errors caused by inconsistent time references. Furthermore, this time synchronization system has a simple structure and is easy to implement, making it widely applicable to various intelligent driving controllers, improving the stability and reliability of intelligent driving systems.

[0012] The present invention is further configured such that the MCU generates T0 in the following manner: if Tb reception times out or is invalid, T0 is the nanosecond conversion value of T in the NVM; if Tb is valid and received on time, T0 is the nanosecond conversion value of the current Tb.

[0013] With this further configuration, if the signal is unstable or initialization is incomplete, and the MCU cannot receive a valid Tb in time, it will automatically read the UTC time T saved when the vehicle was last powered off from the NVM and convert it to nanoseconds as T0, ensuring the continuity of the time base. Once the signal is restored, the MCU can receive and convert Tb to T0 in real time, achieving an instant time base update. This design not only improves the flexibility of the time synchronization system but also further enhances the adaptability of the intelligent driving controller in complex environments.

[0014] The present invention is further configured as follows: the SOC communicates with the MCU, the SOC is used to receive T0 and Δt sent by the MCU, and generate a synchronization time of Tt nanoseconds, and the calculation formula of Tt is Tt=T0+Δt.

[0015] With this further configuration, the SOC can accurately calculate the synchronization time Tt by receiving the reference time T0 and time deviation Δt sent by the MCU. The calculation formula is simple and clear: Tt = T0 + Δt. This configuration not only optimizes the time synchronization process but also significantly improves its accuracy. After receiving the signal, the intelligent driving controller can quickly and accurately synchronize time, thereby ensuring time consistency across all components of the intelligent driving system.

[0016] The present invention is further configured such that the timestamp of the perception data generated by the SOC is Δt1 nanoseconds, and the calculation formula of Δt1 is Δt1=Tt-T0.

[0017] With the above configuration, the SOC can generate a timestamp Δt1 for the sensed data by recording the difference between the synchronization time Tt and the reference time T0. This timestamp Δt1 represents the time interval from signal recovery, when the MCU completes the time reference update, to when the SOC receives the synchronization time and completes the data sensing.

[0018] The present invention is further configured as follows: Radar is used to receive Δt sent by the MCU, and obtain the radar sensing data timestamp as Δt2 nanoseconds based on Δt.

[0019] With this further configuration, the Radar component receives the time offset Δt sent by the MCU and calculates the timestamp Δt2 of the radar-sensed data. This timestamp Δt2 accurately reflects the interval between the Radar receiving the time offset Δt sent by the MCU and completing the radar data sensing. This configuration also enables the Radar component in the intelligent driving controller system to achieve time synchronization with the SOC, further enhancing the overall time consistency and data synchronization of the intelligent driving system.

[0020] The purpose of the present invention is to provide a time synchronization method for an intelligent driving controller.

[0021] The technical solution of the present invention is a time synchronization method for an intelligent driving controller, comprising the following steps:

[0022] (1) The time source sends the UTC time message to the MCU via the CAN bus in Tb seconds;

[0023] (2) MCU records the effective UTC time when the vehicle is powered off as T seconds and stores it in NVM;

[0024] (3) MCU judges the reception timeout of Tb, and converts T or Tb into nanosecond time according to the judgment result, defines it as T0 nanoseconds and stores it in RAM;

[0025] (4) The MCU sends the internal self-running cumulative counter value Δt nanoseconds to the Radar via the CAN message. The Radar assigns Δt to the radar perception data and sends it back to the MCU via the CAN, and then obtains the radar perception data timestamp Δt2;

[0026] (5) MCU sends T0 and Δt to SOC, SOC generates Tt and calculates the timestamp of the sensed data as Δt1 nanoseconds;

[0027] (6) The application layer receives and uses Δt, Δt1, and Δt2.

[0028] With the above further configuration, this method uses the MCU as the core of time synchronization, not only using the UTC time Tb seconds as the benchmark, but also relying on the UTC time T seconds recorded when the vehicle is powered off as a backup time source, thereby ensuring the continuity and reliability of time synchronization.

[0029] In step (3), the MCU performs a timeout check on the received Tb. This mechanism effectively avoids time synchronization errors caused by network delays or data loss. By converting T or Tb into nanosecond time T0 and storing it in RAM, the MCU can update and maintain an accurate time base in real time.

[0030] In steps (4) and (5), the interaction between the MCU, Radar, and SOC further enhances the accuracy of time synchronization. The MCU sends the cumulative counter value Δt nanoseconds to the Radar, which assigns it to the radar perception data and transmits it back to the MCU, while obtaining the timestamp Δt2 of the radar perception data. The MCU then sends T0 and Δt to the SOC, which generates Tt and calculates the timestamp Δt1 nanoseconds of the perception data. This series of operations ensures that each timestamp can be accurately recorded and transmitted from the time source to the application layer.

[0031] Ultimately, when the application layer receives and uses Δt, Δt1, and Δt2, the system can achieve consistency in the timestamps of perception data between components, providing an accurate time reference for intelligent driving decision-making and control.

[0032] Preferably, the judgment method in step (3) is: if Tb reception times out or is invalid, then T0 is the nanosecond time of T conversion in NVM; if Tb reception does not time out and is valid, then T0 is the nanosecond time of current Tb conversion.

[0033] Preferably, in step (5), the calculation formula of Tt is Tt=T0+Δt, and the calculation formula of Δt1 is Δt1=Tt-T0. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of a specific embodiment of the present invention;

[0035] Figure 2 for Figure 1 A partial enlarged view of part A;

[0036] Figure 3 for Figure 1 A partial enlarged view of part B;

[0037] Figure 4 for Figure 1 A partial enlarged view of part C. DETAILED DESCRIPTION

[0038] The following will clearly and completely describe the technical solutions in this embodiment with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0039] like Figure 1-4 As shown, a time synchronization system for an intelligent driving controller of the present invention includes:

[0040] The time source is used to send UTC time messages in Tb seconds;

[0041] The time synchronization master is an MCU, equipped with NVM and RAM memories. The MCU is used to receive Tb and store the effective UTC time of the vehicle power-off in NVM in T seconds. After power-on, the MCU generates a reference time of T0 nanoseconds based on Tb or T and stores it in RAM (the MCU generates T0 in the following way: if Tb reception times out or is invalid, T0 is the nanosecond conversion value of T in NVM; if Tb is valid and received on time, T0 is the nanosecond conversion value of the current Tb). At the same time, the MCU generates an internal self-running cumulative counter value of Δt nanoseconds.

[0042] The time synchronization receiving end is SOC and Radar, which respectively receive the synchronization time sent by MCU and generate the sensing data timestamp;

[0043] The time usage end is an application layer, which is used to receive and use the perception data timestamp.

[0044] Specifically, the SOC communicates with the MCU via the SPI communication protocol. The SOC receives T0 and Δt from the MCU and generates a synchronization time of Tt nanoseconds. Tt is calculated as Tt = T0 + Δt. The SOC generates a timestamp of Δt1 nanoseconds for the sensed data. Δt1 is calculated as Δt1 = Tt - T0.

[0045] Specifically, Radar communicates with the MCU through the CAN bus. Radar is used to receive the Δt sent by the MCU and obtain the radar perception data timestamp of Δt2 nanoseconds based on Δt.

[0046] A time synchronization method for an intelligent driving controller, characterized by comprising the following steps:

[0047] (1) The time source sends a UTC time message of Tb seconds to the MCU via the CAN bus; (2) The MCU records the effective UTC time of the vehicle power-off as T seconds and stores it in the NVM; (3) The MCU determines the reception timeout of Tb and converts T or Tb into nanosecond time based on the judgment result, which is defined as T0 nanoseconds and stored in RAM;

[0048] (4) The MCU sends the internal self-running cumulative counter value Δt nanoseconds to the Radar via the CAN message. The Radar assigns Δt to the radar perception data and sends it back to the MCU via the CAN, and then obtains the radar perception data timestamp Δt2;

[0049] (5) MCU sends T0 and Δt to SOC, SOC generates Tt and calculates the timestamp of the sensed data as Δt1 nanoseconds;

[0050] (6) The application layer receives and uses Δt, Δt1, and Δt2.

[0051] In this process, GPS time can be converted to UTC time. The three timestamps Δt, Δt1, and Δt2 each carry different time information, and together constitute the core of the intelligent driving controller time synchronization system. Δt represents the time interval from the time the MCU generates the reference time T0 to the time the MCU sends the cumulative counter value of the time deviation to the Radar. Δt1 reflects the time difference between the synchronization time Tt received by the SOC and the reference time T0, and represents the time interval for the SOC to complete data perception. Δt2 accurately records the time interval from the time the Radar receives the time deviation Δt sent by the MCU to the time it completes radar data perception. At the application layer, the system can evaluate and optimize the performance of the intelligent driving controller time synchronization system based on the values ​​of these three timestamps.

Claims

1. A time synchronization system for an intelligent driving controller, characterized in that: include: The time source is used to send UTC time messages in Tb seconds; The time synchronization master is an MCU, equipped with NVM and RAM memories. The MCU receives Tb and stores the effective UTC time of the vehicle when it is powered off in T seconds in NVM. After powering on, the MCU generates a reference time of T0 nanoseconds based on Tb or T and stores it in RAM. At the same time, the MCU generates an internal self-running cumulative counter value of Δt nanoseconds. The time synchronization receiving end is SOC and Radar, which respectively receive the synchronization time sent by MCU and generate the sensing data timestamp; The time usage end is an application layer, which is used to receive and use the perception data timestamp.

2. The system according to claim 1, wherein: The MCU generates T0 in the following manner: if Tb reception times out or is invalid, T0 is the nanosecond conversion value of T in NVM; if Tb is valid and received on time, T0 is the nanosecond conversion value of the current Tb.

3. The system according to claim 1, wherein: The SOC communicates with the MCU. The SOC is used to receive T0 and Δt sent by the MCU and generate a synchronization time of Tt nanoseconds. The calculation formula of Tt is Tt=T0+Δt.

4. The system according to claim 3, characterized in that The timestamp of the perception data generated by the SOC is Δt1 nanoseconds, and the calculation formula of Δt1 is Δt1 = Tt-T0.

5. The system according to claim 1, wherein: Radar communicates with the MCU through the CAN bus. Radar is used to receive the Δt sent by the MCU and obtain the radar perception data timestamp of Δt2 nanoseconds based on Δt.

6. A time synchronization method for an intelligent driving controller, characterized in that: The following steps are involved: (1) The time source sends a UTC time message of Tb seconds to the MCU via the CAN bus; (2) The MCU records the effective UTC time of the vehicle power-off as T seconds and stores it in the NVM; (3) The MCU determines the reception timeout of Tb and converts T or Tb into nanosecond time based on the judgment result, which is defined as T0 nanoseconds and stored in RAM; (4) The MCU sends the internal self-running cumulative counter value Δt nanoseconds to the Radar via the CAN message. The Radar assigns Δt to the radar perception data and sends it back to the MCU via the CAN, and then obtains the radar perception data timestamp Δt2; (5) MCU sends T0 and Δt to SOC, SOC generates Tt and calculates the timestamp of the sensed data as Δt1 nanoseconds; (6) The application layer receives and uses Δt, Δt1, and Δt2.

7. The method according to claim 6, characterized in that The judgment method in step (3) is: if Tb reception times out or is invalid, then T0 is the nanosecond time of T conversion in NVM; If Tb is accepted and valid without timeout, T0 is the nanosecond time of the current Tb conversion.

8. The method according to claim 6, characterized in that In step (5), the calculation formula of Tt is Tt=T0+Δt, and the calculation formula of Δt1 is Δt1=Tt-T0.

Images