A parameter estimation system and method

Abstract

A parameter estimation system includes multiple control units connected via a first type of bus, and the control units connected to a sensor system and / or an execution system via a second type of bus. Each control unit is equipped with a computational model for calculating target vehicle parameters. First, the control units acquire first operational data from the second type of bus; then, they use their local computational models to perform calculations on the first operational data to obtain a first estimation result and transmit this first estimation result to other control units, and receive second estimation results calculated by other control units based on the first operational data; finally, they obtain the calculated target vehicle parameters based on the first estimation result and the received second estimation result. The system also includes a parameter estimation method, a device for parameter estimation, a chip for parameter estimation, a vehicle, a computer-readable storage medium containing computer instructions, and a computer program product containing instructions. By involving multiple control units in the calculation of target vehicle parameters, the system improves the estimation performance and accuracy of vehicle parameters, and fully utilizes the computational resources of the control units.

Description

A parameter estimation system and method Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a parameter estimation system and method. Background Technology

[0002] Vehicle parameter estimation is crucial for vehicle design and development, performance testing and simulation, safety analysis, and autonomous driving. Accurate vehicle parameters help improve vehicle performance, enhance safety, and reduce energy consumption.

[0003] The current method for estimating vehicle parameters is as follows: the calculation model for general vehicle parameters is deployed in the sensor system, and the processor in the sensor system can perform the estimation of vehicle parameters. However, due to the limited computing power of the processor in the sensor system, it can only execute simple calculation models or calculation models that consume less computing resources, resulting in low accuracy of the estimated vehicle parameters. Summary of the Invention

[0004] This application discloses a parameter estimation system and method, which improves the estimation performance of vehicle parameters, increases the accuracy of the estimated vehicle parameters, and can make full use of the computing power resources of the vehicle's control components, thereby improving resource utilization.

[0005] In a first aspect, this application provides a parameter estimation system, which includes multiple control components. These control components are connected to each other via a first type of bus and connected to a sensor system and / or an execution system via a second type of bus. Each control component has a computational model deployed on it for calculating target vehicle parameters. At least one of the control components performs the following operations: first, acquiring first operating data via the second type of bus; then, performing calculations on the first operating data using a local computational model to obtain a first estimation result and transmitting the first estimation result to other control components; receiving a second estimation result from other control components, the second estimation result being associated with the first operating data; and finally, obtaining the calculated target vehicle parameters based on the first estimation result and the received second estimation result.

[0006] For example, the transmission rate of the first type of bus is greater than that of the second type of bus. Both the first and second type of buses can achieve one-to-many communication and data distribution. In applications, the first type of bus often uses multicast to achieve one-to-many communication and data distribution, while the second type of bus often uses broadcast to achieve the same. Here, the bus does not limit the topology of the connected devices; it only indicates that multiple devices can be connected to the bus.

[0007] Optionally, the first type of bus is an in-vehicle Ethernet bus, and the second type of bus is an in-vehicle bus. For example, the in-vehicle bus includes one or more of a controller area network (CAN) bus, a local interconnect network (LIN) bus, a FlexRay bus, or other signal buses used for transmitting chassis data, power data, etc. In some solutions, the second type of bus can also be an in-vehicle Ethernet bus; in this case, the first type of bus can be a high-speed Ethernet bus, and the second type of bus can be a medium-speed Ethernet bus or a low-speed Ethernet bus.

[0008] Here, the aforementioned computational model can also be referred to as an estimation model, estimation algorithm, etc. For example, the deployment of computational models on multiple control components for calculating target vehicle parameters can be understood as: the computational models deployed on multiple control components are associated with the estimation of target vehicle parameters. Taking the first control component as an example, the association between the computational model deployed on the first control component and the estimation of target vehicle parameters can be: the computational model deployed on the first control component is used to calculate the target vehicle parameters themselves, or the computational model deployed on the first control component is used to calculate the intermediate parameters (i.e., intermediate results) required for the estimation of target vehicle parameters.

[0009] In the above method, by utilizing the broadcast characteristics of the second type of bus, multiple control units can simultaneously acquire the first operating data, achieving data alignment. By deploying computational models on multiple control units, these multiple control units can jointly participate in the estimation of the same target vehicle parameters, making full use of the computing resources of the control units. In addition, data exchange on the second type of bus is stable, and different control units can also achieve rapid exchange of estimation results through the first type of bus, thereby providing more complex and higher accuracy parameter estimation services, improving the accuracy of vehicle parameter estimation, and also contributing to improving parameter estimation efficiency.

[0010] In one possible implementation of the first aspect, the at least one control component is further configured to: transmit the calculation results of the target vehicle parameters to the sensor system and / or the execution system. In some embodiments, the at least one control component is further configured to send the calculation results of the target vehicle parameters to the upper-layer application. Thus, the calculation results of the target vehicle parameters obtained through local arbitration can be sent in a timely manner to one or more of the sensor system, the execution system, and the upper-layer application, facilitating their use of the calculation results of the target vehicle parameters.

[0011] In one possible implementation of the first aspect, the plurality of control components include a first control component and a second control component, wherein the computational model deployed on the first control component is the same as the computational model deployed on the second control component.

[0012] By implementing the above method, the same calculation model is deployed on different control units, achieving model backup. Only when the estimation results obtained by different control units are consistent are they output as the calculation results for the target vehicle parameters. This improves the accuracy of the estimated vehicle parameters and enhances estimation performance. Furthermore, since multiple control units execute the same calculation model, if one control unit malfunctions and fails to output the calculation results for the target vehicle parameters, the calculation results output by other control units can be used. This enhances fault tolerance and the reliability of the estimation results.

[0013] In one possible implementation of the first aspect, the plurality of control components include a first control component and a second control component, wherein the computational model deployed on the first control component is a first computational model, and the computational model deployed on the second control component is a second computational model, and the first computational model and the second computational model are different.

[0014] By implementing the above method and deploying different calculation models on different control components, the capabilities of the hardware can be maximized. This provides multiple calculation methods or approaches for the target vehicle parameters. Compared with the current single calculation method, multiple calculations of vehicle parameters are realized, which not only helps to improve the estimation efficiency of vehicle parameters, but also helps to improve the accuracy and reliability of the estimation.

[0015] In one possible implementation of the first aspect, the operating conditions applicable to the first computational model are different from those applicable to the second computational model.

[0016] By implementing the above methods, different calculation models are applicable to different working conditions, which not only improves the flexibility of deployment and meets the application needs of different scenarios, but also improves the accuracy of vehicle parameter estimation.

[0017] In one possible implementation of the first aspect, the operating conditions applicable to the first calculation model are the same as those applicable to the second calculation model, and at least one control component is specifically used to: arbitrate the first estimation result and the received second estimation result based on preset rules to obtain the calculation result of the target vehicle parameters.

[0018] By implementing the above methods, different calculation models can be applied to the same working conditions, providing multiple calculation methods or approaches for vehicle parameters under these conditions, thereby providing calculation results with different levels of accuracy.

[0019] In one possible implementation of the first aspect, the plurality of control components include a first control component and a second control component, wherein the computational model deployed on the first control component is part of the target computational model, and the computational model deployed on the second control component is another part of the target computational model.

[0020] By implementing the above method, a complex computational model can be broken down into different parts or modules and deployed on multiple control units. Each control unit executes a different part of the same computational model. During deployment, computing resources can be allocated according to the processing capabilities of the control units to support the operation of the complex computational model. This fully utilizes the computing resources of multiple control units, and the multiple control units work together to complete complex calculations to provide more accurate estimation results. It also helps to improve the execution efficiency of complex calculations.

[0021] In one possible implementation of the first aspect, when the vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data for a first cycle, and the aforementioned at least one control component is further configured to: perform calculations using a local calculation model on the first operation data and at least one historical calculation result of the target vehicle parameters to obtain a first estimation result. Wherein, at least one historical calculation result comes from at least one other control component, and the historical calculation result is associated with second operation data, which is the vehicle operation data for a second cycle, and the second cycle is the cycle preceding the first cycle.

[0022] By implementing the above method, when vehicle operation data is acquired periodically, the calculation of the first estimation result may also depend on the historical calculation results of the target vehicle parameters. By referencing the historical calculation results of the target vehicle parameters, the control unit can perform continuous estimation, which is beneficial to improving the accuracy of subsequent estimated vehicle parameters.

[0023] In one possible implementation of the first aspect, the first estimation result is carried via a multicast message, which also carries several of the following information:

[0024] Multicast IP address, which is used to identify the multicast group receiving the first estimation result, the multicast group includes the other control components mentioned above;

[0025] Parameter encoding information, which indicates that the calculated vehicle parameters are those of the target vehicle; and,

[0026] Intermediate process variables are used to indicate intermediate parameters involved in the calculation of the first estimation result.

[0027] Implementing the above method, multicast IP addresses are used to share and output the first estimation result. Parameter encoding information can not only distinguish and label different vehicle parameters, but also differentiate the calculation model used and the control unit performing the calculation. Intermediate process variables are key intermediate variables or parameters in the calculation model, serving as auxiliary values ​​for the obtained estimation result. Carrying this information through multicast messages facilitates the traceability and synchronization of calculation results. Furthermore, the multicast method achieves efficient one-to-many data transmission, addressing low network load requirements in scenarios with limited network bandwidth.

[0028] In one possible implementation of the first aspect, when the transmission of vehicle operation data is periodic, the first operation data is the vehicle operation data for the first period, and the multicast message also carries the following information:

[0029] At least one historical calculation result of the target vehicle parameters, wherein at least one historical calculation result is involved in the calculation of the first estimation result, at least one historical calculation result comes from at least one other controller, and the historical calculation result is associated with second operating data, the second operating data being vehicle operating data for a second period, the second period being the period preceding the first period; and,

[0030] Verification information is used to determine whether the first estimation result calculated in the current period is valid.

[0031] By implementing the above method, the multicast message carries the above information, which makes it easier for the receiving end of the multicast message to trace the first estimation result carried in the multicast message, and can also perform validity verification on the first estimation result through the verification information.

[0032] In one possible implementation of the first aspect, the first estimation result or the second estimation result is an intermediate result obtained during the calculation of the target vehicle parameters or an output result of the target vehicle parameters. That is, the computational model deployed on the control unit may directly calculate the output result of the target vehicle parameters, or it may calculate an intermediate result that participates in the calculation of the target vehicle parameters. In this way, the deployment of the computational model on multiple control units is flexible.

[0033] In one possible implementation of the first aspect, the plurality of control components include a first control component, which is directly connected to the sensor system via a signal line; the at least one control component is specifically used to: acquire first operating data via a second type of bus and a signal line.

[0034] Here, compared to a bus, signal lines are more often used to achieve one-to-one data transmission and communication. A signal line is a physical wire used to transmit a single signal, such as a voltage signal, current signal, or pulse signal.

[0035] The sensor system is directly connected to the control unit via signal lines. Compared to the method where the sensor system is connected to the control unit via the vehicle bus, the data collected by the sensor system can be transmitted to the control unit more timely and frequently. This allows the control unit to perform the estimation of relevant vehicle parameters in a timely manner, which helps to improve the estimation efficiency of vehicle parameters.

[0036] Secondly, this application provides a parameter estimation method applied to a first control component, the first control component being equipped with a computational model for calculating target vehicle parameters. The method includes: firstly, acquiring first operating data via signal lines and / or a second type of bus; then, performing calculations on the first operating data using a local computational model to obtain a first estimation result, and transmitting the first estimation result to other control components via the first type of bus; and receiving a second estimation result from other control components via the first type of bus, the other control components being equipped with computational models for calculating target vehicle parameters, the second estimation result being associated with the first operating data; finally, obtaining the calculated target vehicle parameters based on the first estimation result and the received second estimation result.

[0037] For information on Class 1 buses, Class 2 buses, signal lines, etc., please refer to the description in the corresponding section of Part 1; it will not be repeated here.

[0038] In the above method, the control unit acquires first operating data from the second type of bus, and quickly exchanges estimation results calculated using its local computing model with other control units via the first type of bus. Finally, based on the estimation results calculated locally and the estimation results received from other control units, the calculation result of the target vehicle parameters is obtained. It can be seen that multiple control units participate in the calculation of the same vehicle parameters, fully utilizing the computing resources of the control units, providing more complex parameter estimation services with higher accuracy requirements, and improving the accuracy of vehicle parameter estimation.

[0039] In one possible implementation of the second aspect, the method further includes: transmitting the calculated results of the target vehicle parameters to the sensor system and / or execution system via a second type of bus and / or signal line. In some embodiments, at least one of the aforementioned control components is also used to transmit the calculated results of the target vehicle parameters to an upper-level application.

[0040] By implementing the above method, after obtaining the calculation results of the target vehicle parameters locally, the calculation results of the target vehicle parameters are promptly sent to one or more of the sensor system, execution system, and upper-layer applications so that they can use the calculation results.

[0041] In one possible implementation of the second aspect, the computational model deployed on the first control unit satisfies any of the following conditions as well as the computational models deployed on other control units:

[0042] same;

[0043] Different;

[0044] Partially the same; or,

[0045] Deploy different modules of the same computing model separately.

[0046] For example, taking other control components including the second control component and the third control component as examples, the three cases of being the same, different and partially the same as above are specifically explained. Please refer to (1)-(3) below.

[0047] (1) The same computing model deployed on the first control unit as the computing model deployed on other control units means that the computing model deployed on the first control unit, the computing model deployed on the second control unit, and the computing model deployed on the third control unit are all the same.

[0048] (2) The computing model deployed on the first control unit is different from the computing model deployed on other control units. This means that the computing model deployed on the first control unit is different from the computing model deployed on the second control unit, the computing model deployed on the second control unit is different from the computing model deployed on the third control unit, and the computing model deployed on the first control unit is different from the computing model deployed on the third control unit.

[0049] (3) The computational model deployed on the first control unit is partially the same as the computational model deployed on other control units, which may be: the computational model deployed on the first control unit is the same as the computational model deployed on the second control unit, or the computational model deployed on the first control unit is different from the computational model deployed on the third control unit.

[0050] Implementing the above methods allows for different control units to deploy the same or different computational models. Alternatively, the complex computational model can be broken down into multiple parts or modules and distributed across multiple control units, with each control unit deploying a portion of the complex computational model. This flexible deployment of the computational model across multiple control units offers several advantages. Furthermore, executing the same computational model across multiple control units enhances fault tolerance and the reliability of the estimation results, improving estimation performance. Executing different computational models across different control units provides multiple calculation methods or pathways for the target vehicle parameters, which helps improve the accuracy and reliability of the estimation. Distributing the complex computational model across multiple control units fully utilizes the computing resources of the control units, improving the execution efficiency of the complex computational model and providing more accurate estimation results.

[0051] The effective effects of the technical features in the second aspect below are described in the description of the beneficial effects of the corresponding features in the first aspect above, and will not be repeated here.

[0052] In one possible implementation of the second aspect, when the vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data for a first cycle. The calculation result of the target vehicle parameters is obtained based on the first estimation result and the received second estimation result. This includes: performing calculations based on the first estimation result, the received second estimation result, and at least one historical calculation result of the target vehicle parameters to obtain the first estimation result. Wherein, at least one historical calculation result comes from at least one other control component, and the historical calculation result is associated with the second operation data. The second operation data is the vehicle operation data for a second cycle, and the second cycle is the cycle preceding the first cycle.

[0053] In one possible implementation of the second aspect, the first estimation result is carried via a multicast message, which also carries several of the following information:

[0054] Multicast IP address, which is used to identify the multicast group receiving the first estimation result, the multicast group includes the other control components mentioned above;

[0055] Parameter encoding information, which indicates that the calculated vehicle parameters are those of the target vehicle; and,

[0056] Intermediate process variables are used to indicate intermediate parameters involved in the calculation of the first estimation result.

[0057] In one possible implementation of the second aspect, when the transmission of vehicle operation data is periodic, the first operation data is the vehicle operation data for the first period, and the multicast message also carries the following information:

[0058] At least one historical calculation result of the target vehicle parameters, wherein at least one historical calculation result is involved in the calculation of the first estimation result, at least one historical calculation result comes from at least one other controller, and the historical calculation result is associated with second operating data, the second operating data being vehicle operating data for a second period, the second period being the period preceding the first period; and,

[0059] Verification information is used to determine whether the first estimation result calculated in the current period is valid.

[0060] In one possible implementation of the second aspect, the first estimation result or the second estimation result is an intermediate result obtained during the calculation of the target vehicle parameters or an output result of the target vehicle parameters.

[0061] Thirdly, this application provides an apparatus for parameter estimation. The apparatus is a first control component or included within a first control component, and the first control device is equipped with a calculation model for calculating target vehicle parameters. The apparatus includes: an acquisition unit for acquiring first operating data via signal lines and / or a second type of bus; a processing unit for performing calculations on the first operating data using a local calculation model to obtain a first estimation result; a transmission unit for transmitting the first estimation result to other control components via the first type of bus; the acquisition unit is further configured to receive a second estimation result from other control components via the first type of bus, wherein the other control components are equipped with a calculation model for calculating target vehicle parameters, and the second estimation result is associated with the first operating data; and the processing unit is further configured to obtain a calculation result of the target vehicle parameters based on the first estimation result and the received second estimation result.

[0062] In one possible implementation of the third aspect, the transmitting unit is further configured to: transmit the calculated results of the target vehicle parameters to the sensor system and / or execution system via a second type of bus and / or signal line.

[0063] In one possible implementation of the third aspect, the computational model deployed on the first control unit satisfies any of the following conditions as well as the computational models deployed on other control units:

[0064] same;

[0065] Different;

[0066] Partially the same; or,

[0067] Deploy different modules of the same computing model separately.

[0068] In one possible implementation of the third aspect, when the vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data for a first cycle. The processing unit is specifically configured to: perform calculations based on a first estimation result, a received second estimation result, and historical calculation results of at least one of the target vehicle parameters to obtain the first estimation result. Wherein, at least one historical calculation result comes from at least one other control component, and the historical calculation result is associated with the second operation data, which is the vehicle operation data for a second cycle, and the second cycle is the cycle preceding the first cycle.

[0069] In one possible implementation of the third aspect, the first estimation result is carried via a multicast message, which also carries several of the following information:

[0070] Multicast IP address, which is used to identify the multicast group receiving the first estimation result, the multicast group includes the other control components mentioned above;

[0071] Parameter encoding information, which indicates that the calculated vehicle parameters are those of the target vehicle; and,

[0072] Intermediate process variables are used to indicate intermediate parameters involved in the calculation of the first estimation result.

[0073] In one possible implementation of the third aspect, when the transmission of vehicle operation data is periodic, the first operation data is the vehicle operation data for the first period, and the multicast message also carries the following information:

[0074] At least one historical calculation result of the target vehicle parameters, wherein at least one historical calculation result is involved in the calculation of the first estimation result, at least one historical calculation result comes from at least one other controller, and the historical calculation result is associated with second operating data, the second operating data being vehicle operating data for a second period, the second period being the period preceding the first period; and,

[0075] Verification information is used to determine whether the first estimation result calculated in the current period is valid.

[0076] In one possible implementation of the third aspect, the first or second estimation result is an intermediate result obtained during the calculation of the target vehicle parameters or the output result of the target vehicle parameters.

[0077] Fourthly, this application provides a chip for parameter estimation, the chip including a processor and a memory, wherein the memory is used to store program instructions; the processor calls the program instructions in the memory to cause the chip to execute the method in the second aspect or any possible implementation of the second aspect.

[0078] Fifthly, this application provides a vehicle that includes the parameter estimation system of the first aspect or any possible implementation thereof, or includes the device of the third aspect or any possible implementation thereof, or includes the chip of the fourth aspect.

[0079] In a sixth aspect, this application provides a computer-readable storage medium including computer instructions that, when executed by a processor, implement the method in the second aspect or any possible implementation thereof.

[0080] In a seventh aspect, this application provides a computer program product that, when executed by a processor, implements the methods described in the second aspect or any possible embodiment of the second aspect. The computer program product may, for example, be a software installation package. When the methods provided by any possible design of the second aspect are required, the computer program product can be downloaded and executed on a processor to implement the methods described in the second aspect or any possible embodiment of the second aspect. Attached Figure Description

[0081] Figure 1A is a schematic diagram of a parameter estimation system provided in an embodiment of this application;

[0082] Figure 1B is a schematic diagram of another parameter estimation system provided in an embodiment of this application;

[0083] Figure 2 is a schematic diagram illustrating a specific application of a parameter estimation system provided in an embodiment of this application;

[0084] Figure 3 is a flowchart of a parameter estimation method provided in an embodiment of this application;

[0085] Figure 4A is a schematic diagram of the deployment of a computing model on multiple control components according to an embodiment of this application;

[0086] Figure 4B is a schematic diagram of the deployment of a computing model on multiple control components according to an embodiment of this application;

[0087] Figure 4C is a schematic diagram of the deployment of a computing model on multiple control components according to an embodiment of this application;

[0088] Figure 5 is a schematic diagram of the transmission of estimation results of different control components in a parameter estimation method provided in an embodiment of this application;

[0089] Figure 6 is a schematic diagram of an application scenario provided by an embodiment of this application;

[0090] Figure 7 is a schematic flowchart of a single control component performing parameter estimation according to an embodiment of this application;

[0091] Figure 8 is a schematic diagram of a parameter estimation device provided in an embodiment of this application;

[0092] Figure 9 is a schematic diagram of the structure of a computing device provided in an embodiment of this application. Detailed Implementation

[0093] The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit this application. Terms such as "first," "second," etc., in the description and claims of this application are used to distinguish different objects, not to describe a specific order.

[0094] To facilitate understanding, the relevant terms that may be involved in the embodiments of this application will be introduced below.

[0095] (1) CAN bus

[0096] Controller Area Network (CAN) bus is a shared two-wire serial communication bus that can use twisted-pair cables or optical fibers to transmit signals. It can be used in automotive powertrain systems, chassis, and body electronics.

[0097] (2) LIN bus

[0098] Local Interconnect Network (LIN) bus is a low-cost serial communication bus used to implement distributed electronic system control in automobiles. It is suitable for communication scenarios where network bandwidth, performance, or fault tolerance requirements are not high, primarily serving as an auxiliary function to the CAN bus. It is generally used in non-real-time applications, such as local systems for seat, window, and rearview mirror control.

[0099] (3) FlexRay bus

[0100] The FlexRay bus is a time-deterministic, distributed clock-synchronized, and fault-tolerant bus standard. FlexRay can be applied to systems with high real-time requirements, such as advanced driver assistance systems, chassis control systems, and drive-by-wire systems (e.g., suspension control, shift control, braking control, steering control, etc.).

[0101] In this solution, the in-vehicle bus includes one or more of the aforementioned CAN bus, LIN bus, FlexRay bus, or other signal buses used for transmitting chassis data, power data, etc. The in-vehicle bus can be used to transmit bus status data, sensor status data, actuator status data, etc.

[0102] (4) Vehicle Ethernet

[0103] In vehicular networks, Ethernet is used to meet the requirements of high bandwidth and high speed data transmission, especially in advanced driver assistance systems (ADAS), autonomous driving, infotainment systems, and in-vehicle camera networks.

[0104] The above terms may optionally be used in the embodiments described below.

[0105] This solution provides a parameter estimation system, which offers a method for estimating vehicle parameters based on a two-layer network structure. This system can control multiple control components to participate in the calculation of the same vehicle parameter, making full use of the computing resources of the control components. Even under conditions of limited computing resources, the method provided by this solution can achieve complex parameter estimation, thereby improving the accuracy of vehicle parameter estimation.

[0106] The composition of the parameter estimation system is described below. In this scheme, the parameter estimation system includes multiple control components, which are connected to each other via a first-type bus, and each control component is connected to a sensor system and / or actuator system via a second-type bus. The number of sensor systems and actuator systems connected to each control component is not limited here.

[0107] For example, the transmission rate of the first type of bus is greater than that of the second type of bus. Both the first and second type of buses can achieve one-to-many communication and data distribution. In applications, the first type of bus often uses multicast to achieve one-to-many communication and data distribution, while the second type of bus often uses broadcast to achieve the same. Here, the bus does not limit the topology of the connected devices; it only indicates that multiple devices can be connected to the bus.

[0108] Optionally, the first type of bus is an in-vehicle Ethernet bus, and the second type of bus is an in-vehicle bus. For example, the in-vehicle bus includes one or more of the following: controller area network (CAN) bus, local interconnect network (LIN) bus, FlexRay bus, or other signal buses used to transmit chassis data, power data, etc.

[0109] In some solutions, the second type of bus can also be an in-vehicle Ethernet bus. In this case, the first type of bus can be a high-speed Ethernet bus, and the second type of bus can be a medium-speed Ethernet bus or a low-speed Ethernet bus.

[0110] In some solutions, sensor systems may be directly connected to at least one control unit via signal lines. Compared to a bus, signal lines are more often used for one-to-one data transmission and communication. Here, a signal line is a physical wire used to transmit a single signal, such as a voltage signal, current signal, or pulse signal. By directly connecting the sensor system to the control unit via signal lines, compared to connecting the sensor system to the control unit via an in-vehicle bus, the data collected by the sensor system can be transmitted to the control unit more promptly and frequently. This allows the control unit to perform timely estimations of relevant vehicle parameters, thus improving the estimation performance of vehicle parameters.

[0111] In some solutions, a sensor system may be directly connected to an actuator via a signal line. In this case, the sensor data collected by the sensor system can be transmitted to a Type II bus through the actuator to which the sensor system is connected.

[0112] Here, the parameter estimation system is deployed on the vehicle. Depending on the vehicle's power source, the vehicle can be either a new energy vehicle or a traditional vehicle. Traditional vehicles refer to fuel-powered vehicles, such as gasoline vehicles or diesel vehicles. New energy vehicles can be electric vehicles (EVs), hybrid electric vehicles (HEVs), range-extended electric vehicles (EVs), plug-in hybrid electric vehicles (HEVs), fuel cell vehicles, or other new energy vehicles.

[0113] For example, the control component may be a device or component within the vehicle, such as an Ethernet gateway, a telematics box (TBox), a domain controller, etc., and the component may be a chip, a control unit (e.g., a microcontroller unit (MCU), a microprocessor unit (MPU), etc.), an integrated circuit, etc.

[0114] Here, the Ethernet gateway can be, for example, a vehicle integrated / integration unit (VIU). The VIU manages the electronic control units (ECUs) within its area. The VIU allows nearby sensors, actuators, or ECUs to access the system, providing functions such as power supply, electronic fuses, and I / O port isolation. Additionally, the VIU possesses some or all of the gateway functions, such as protocol conversion, protocol encapsulation and forwarding, and data format conversion. When the VIU integrates the functions of ECUs within some vehicle components, it also has electronic control functions to control the corresponding vehicle components. In some possible embodiments, the VIU also has data processing capabilities across vehicle components, i.e., it can perform calculations on data obtained from actuators of multiple vehicle components.

[0115] For example, a domain controller can be an integrated hardware and software platform supporting intelligent driving, i.e., a vehicle computing platform, such as a mobile data center (MDC); it can also be an integrated hardware and software platform supporting body control and chassis control, such as a vehicle domain controller (VDC); or it can be an integrated hardware and software platform providing in-vehicle multimedia services (such as at least one of head-up display, instrument panel display, and entertainment system), such as a cockpit domain controller (CDC). In some solutions, the domain controller can also be a controller that integrates multiple functions of the aforementioned VDC, MDC, and CDC. As an example, by integrating VDC, MDC, and CDC, a domain controller capable of providing body control functions, autonomous driving control functions, and cockpit control functions can be obtained. In this case, the domain controller can also be referred to as a central computing unit.

[0116] For example, MDC provides services for vehicle components that enable autonomous driving functions, including monocular cameras, binocular cameras, millimeter-wave radar, lidar, ultrasonic radar, etc.; CDC provides services for vehicle components in the cockpit domain, including head-up displays (HUDs), instrument displays, radios, navigation systems, cameras, etc.; VDC provides services for vehicle components in the body domain and chassis domain, including window and door controllers, electric mirrors, air conditioning, central locking, etc.; and chassis components include components in the braking system, steering system, and acceleration system, such as the accelerator pedal.

[0117] Here, the sensor system is used to assist the vehicle in perceiving its surroundings, monitoring its status, and providing feedback information. Examples of sensor systems include cameras, LiDAR, millimeter-wave radar, ultrasonic radar, inertial measurement units (IMUs), global positioning systems (GPS), temperature and humidity sensors, air pressure sensors, wheel speed sensors, tire pressure monitoring systems (TPMS), etc.

[0118] An execution system refers to the hardware and systems that control vehicle behavior based on received control commands. For example, an execution system may be a steering system, braking system, powertrain or drive system (or power drive unit), body control system, powertrain system, wheel control system, battery management system (BMS), thermal management system (TMS), thermal dynamics unit (TDU), automatic parking system, etc.

[0119] Multiple control components are equipped with calculation models for vehicle parameters, and the sensor systems and / or execution systems connected to these control components need to receive and use the estimated vehicle parameters. For example, when the execution system is the aforementioned braking system, the estimated vehicle parameters include the slip rate; when the execution system is the aforementioned steering system, the estimated vehicle parameters include the vehicle's yaw rate; and when the execution system is a drive system, the estimated vehicle parameters include the road adhesion coefficient, torque vector, etc.

[0120] Referring to Figure 1A, which is a schematic diagram of a parameter estimation system provided in an embodiment of this application. In Figure 1A, taking a vehicle-mounted Ethernet bus as the first type of bus and an in-vehicle bus as the second type of bus as an example, it can be seen that the aforementioned multiple control components include at least a first control component and a second control component. The first control component and the second control component communicate via the vehicle-mounted Ethernet, and both the first control component and the second control component communicate with the sensor system and / or the execution system via the in-vehicle bus. Figure 1A is only an example of a parameter estimation system; in some solutions, the parameter estimation system may also include a greater number of control components, sensor systems, and execution systems.

[0121] Referring to Figure 1B, which is a schematic diagram of another parameter estimation system provided in an embodiment of this application. Compared to Figure 1A, it can be seen that in Figure 1B, the sensor system is directly connected to the first control unit via signal lines, while the execution system is still connected to the first and second control units via the vehicle bus. Figure 1B is only an example; in some solutions, the sensor system may also be connected to multiple control units via signal lines, or some sensor systems may be connected to the control units via signal lines while others are connected to the control units via the vehicle bus.

[0122] For example, the sensor system includes multiple wheel speed sensors. When applied to FIG1A, the multiple wheel speed sensors can be connected to at least one control component via an in-vehicle bus; when applied to FIG1B, the multiple wheel speed sensors can be connected to a first control component (e.g., VIU1) via signal lines. In some embodiments, the multiple wheel speed sensors can also be directly connected to an actuation system (e.g., the braking system described above) via signal lines.

[0123] Referring to Figure 2, which is a schematic diagram illustrating a specific application of a parameter estimation system provided in an embodiment of this application, the parameter estimation system shown in Figure 2 includes multiple control components, including the aforementioned VDC, VIU1, and MDC. The sensor systems and / or execution systems connected to these control components include electric power steering (EPS) systems, intelligent braking systems (IBS), R-MCUs, F-MCUs, etc. In Figure 2, VDC and VIU1 are connected via an in-vehicle Ethernet connection, and VIU1 and MDC are connected via an in-vehicle Ethernet connection. VDC, VIU1, and MDC are all connected to the aforementioned EPS, IBS, and other sensor systems via a Chassis CAN bus, and VDC, VIU1, and MDC are all connected to the R-MCU, F-MCU, and other sensor systems via an EP-CAN bus.

[0124] The Chassis CAN bus and EP-CAN bus mentioned above belong to the aforementioned in-vehicle bus systems. The Chassis CAN bus primarily serves chassis and body control, ensuring vehicle stability, handling, and safety. It mainly connects chassis-related equipment, including EPS and IBS, as well as electronic suspension systems, wheel sensors, and anti-lock braking systems (ABS). The EP-CAN bus primarily manages and controls the battery and electric motor, ensuring efficient collaboration between various components of the electric powertrain. It mainly connects electric powertrain-related equipment such as the battery management system (BMS), electric motor controller, inverter, charging system, and power distribution system.

[0125] In this solution, taking IBS as an example, as shown in Figure 2, IBS requires the use of the vehicle parameter slip ratio. As shown in Figure 2, IBS is connected and communicates with the three control components VDC, VIU1, and MDC through the in-vehicle bus. Therefore, a calculation model for slip ratio can be deployed on the three control components VDC, VIU1, and MDC, so that the three control components can jointly estimate the vehicle parameter slip ratio. The estimation process is described in the embodiment shown in Figure 3 below. For the sake of brevity, it will not be repeated here.

[0126] It is understood that Figure 2 is only an example of a specific application of the parameter estimation system and should not be construed as limiting the parameter estimation system.

[0127] The parameter estimation system shown in Figure 1A, Figure 1B, or Figure 2 can be applied to a variety of application scenarios, such as the following: mobile internet (MI), industrial control, self-driving, transportation safety, internet of things (IoT), and smart city.

[0128] The parameter estimation method provided in this solution is described below based on the system architecture shown above.

[0129] Referring to Figure 3, which is a flowchart of a parameter estimation method provided in an embodiment of this application, this method can be applied to the parameter estimation system shown in Figure 1A or Figure 1B above. Specifically, it is applied between multiple control components. An illustrative example is given using multiple control components including a first control component, a second control component, and a third control component, but the number of control components is not limited to three. Here, the connection methods between the control components, the sensor system, the execution system, between the control components and the sensor system, between the control components and the execution system, and between the control components themselves, are described in the corresponding content of the foregoing embodiments and will not be repeated here.

[0130] In Figure 3, the sensor system and / or execution system require receiving and / or using target vehicle parameters. The sensor system is connected to multiple control components via the aforementioned second type of bus and / or signal lines. The execution system is also connected to multiple control components via the second type of bus. These control components are interconnected via the aforementioned first type of bus. A computational model is deployed on each of these control components, and the computational model is used to calculate the target vehicle parameters. Here, the first type of bus, the second type of bus, and the signal lines are described in the preceding embodiments and will not be repeated here. The method shown in the embodiment of Figure 3 includes, but is not limited to, the following steps S301-S305.

[0131] S301: At least one control unit acquires first operating data through a second type of bus.

[0132] Here, the first set of running data is the basic data required for calculating the target vehicle parameters.

[0133] For example, the first operating data includes one or more types of data such as sensor data, actuator feedback data, and bus status data. Specifically, sensor data includes data collected by sensors and information indicating whether the sensors are malfunctioning; actuator feedback data includes feedback on the results of actuators executing control commands and information indicating whether the actuators are currently available; and bus status data includes the feedback status of sensor data and actuator feedback data on the bus.

[0134] In one implementation, the first operational data is transmitted in a broadcast manner on a second type of bus, so multiple control components can synchronously acquire the first operational data. For example, the second type of bus is an in-vehicle bus, including the aforementioned CAN bus, LIN bus, FlexRay bus, etc. In the first operational data, the buses on which different types of data are transmitted can be the same or different.

[0135] For example, when the target vehicle parameter is the tire lateral force, the first operating data includes basic data such as the vehicle's longitudinal acceleration, yaw rate, longitudinal speed, lateral acceleration, and steering wheel angle.

[0136] For example, when the target vehicle parameter is slip ratio, the first operating data includes basic data such as the vehicle's longitudinal speed, tire rolling speed, tire radius, tire angular velocity, road surface conditions (i.e., coefficient of friction), and tire traction or braking force.

[0137] In some schemes, these multiple control components may be connected to multiple sensor systems via a second type of bus. Different control components may acquire first operating data from the same sensor system via the second type of bus, or they may acquire first operating data from different sensor systems via the second type of bus, and the first operating data acquired by different control components may be the same. Here, the number of second type buses on which the control components acquire the first operating data is not limited; it can be one or multiple.

[0138] For example, different control components may acquire the first operating data simultaneously, or there may be a time difference in the time when the first operating data is acquired, but the time difference is within the allowable error range.

[0139] In some possible embodiments, when there are multiple sensor systems, some sensor systems are connected to multiple control components via a second type bus, while other sensor systems are directly connected to multiple control components via signal lines. In this case, the control components acquire first operating data via the second type bus, including: the control components acquire the first operating data via the second type bus and the signal lines. In this scenario, a portion of the first operating data is acquired from the signal lines, and another portion is acquired from the second type bus. Further, the data acquired from the second type bus in the first operating data specifically originates from the connected sensor systems and / or execution systems.

[0140] It is understandable that since the above parameter estimation system is deployed on the vehicle, as the vehicle moves, it needs to periodically collect sensor data, and will also periodically generate the above actuator feedback data and bus status data. Therefore, the vehicle operation data can also be transmitted periodically. Thus, the above first operation data is the vehicle operation data of the first cycle, and the first cycle is the current cycle.

[0141] S302: At least one control unit uses a local calculation model to perform calculations on the received first operating data to obtain a first estimation result, and sends the first estimation result to other control units among the plurality of control units.

[0142] For multiple control units to deploy computational models for calculating target vehicle parameters, it means that the computational model deployed on each control unit is associated with the estimation of the target vehicle parameters. Here, taking the first control unit among multiple control units as an example, the association between the computational model deployed on the first control unit and the estimation of the target vehicle parameters means that: the computational model deployed on the first control unit is used to calculate the target vehicle parameters themselves, or the computational model deployed on the first control unit is used to calculate the intermediate parameters (i.e., intermediate results) required for the estimation of the target vehicle parameters.

[0143] For example, when applied to tire force estimation, if the target vehicle parameter is the tire lateral force, the calculation model deployed on the first control unit can be used to calculate the tire vertical force or the tire longitudinal force. Here, both the tire vertical force and the tire longitudinal force are intermediate parameters required for estimating the tire lateral force. Alternatively, the calculation model deployed on the first control unit can also be used to calculate the tire lateral force.

[0144] In this scheme, the deployment of the computing model on multiple control components can be done in the form of the same model deployment, different model deployment, or distributed model deployment, which makes full use of the computing resources of multiple control components and realizes flexible deployment of the computing model.

[0145] In one implementation, the same computational model is deployed on multiple control components. For example, a first computational model is deployed on the first control component, a second computational model is deployed on the second control component, and a third computational model is deployed on the third control component, wherein the first, second, and third computational models are identical. Referring to Figure 4A, that is, the first computational model is model A, the second computational model is model A, and the third computational model is model A, and model A is used to calculate the target vehicle parameters.

[0146] For example, in a slip ratio estimation scenario, Model A is used to estimate the current slip ratio by measuring the wheel's angular velocity in real time. Deploying the same model (i.e., Model A) on multiple control components to calculate the slip ratio and comparing the results helps achieve higher accuracy. Furthermore, there may be a time delay in the slip ratio calculations by different control components using Model A. Utilizing this delay for interpolation algorithms can achieve a higher data density recording of the slip ratio, enabling upper-layer applications (such as anti-lock braking systems) to detect abnormal vehicle driving conditions more quickly and make timely adjustments.

[0147] This implementation deploys the same computational model on different control units, achieving model backup. Furthermore, the calculation results for the target vehicle parameters are only output when the estimation results obtained by different control units are consistent, which improves the accuracy of the estimated vehicle parameters and enhances estimation performance. In addition, since multiple control units execute the same computational model, if one control unit malfunctions and fails to output the calculation results for the target vehicle parameters, the calculation results output by other control units can be used, thus enhancing fault tolerance and the reliability of the estimation results.

[0148] In another implementation, different computational models are deployed on different control components. For example, a first computational model is deployed on the first control component, a second computational model is deployed on the second control component, and a third computational model is deployed on the third control component. The first computational model differs from the second computational model, the second computational model differs from the third computational model, and the first computational model differs from the third computational model. Referring to Figure 4B, the first computational model is model A, the second computational model is model B, and the third computational model is model C. Models A, B, and C are all used to calculate the target vehicle parameters.

[0149] In other words, based on the differences in processing capabilities between different control components and the different characteristics of various calculation models, the calculation models can be deployed differently on different control components. This means that the same parameter estimation service can execute different calculation models, which is beneficial for forming coverage in different application scenarios and also helps to improve the reliability and accuracy of parameter estimation results.

[0150] In a given application scenario, different computational models may be applicable to different working conditions.

[0151] Taking the calculation of the road adhesion coefficient as an example, Model A is used to calculate the road adhesion coefficient in a dry road environment, Model B is used to calculate the road adhesion coefficient in a wet and slippery road environment, and Model C is used to calculate the road adhesion coefficient under other extreme conditions (such as icy and snowy roads). Here, the different calculation models for different extreme conditions are only an example and do not limit extreme conditions to only dry, wet and slippery, and icy and snowy roads. In some schemes, different calculation models for the road adhesion coefficient can be configured according to the road type, or according to the road type and road condition. Road types include, but are not limited to, asphalt pavement, cement pavement, and special pavement (such as mud and sand pavement, grass, etc.), and road conditions include, but are not limited to, dry, wet and slippery, and icy conditions.

[0152] Using different control components to execute different calculation models to output the calculation results of the road adhesion coefficient under different road surface characteristics is beneficial to improving the reliability and accuracy of the output calculation results.

[0153] For example, Model A is used to calculate the road adhesion coefficient when the vehicle is traveling at low speeds, Model B is used to calculate the road adhesion coefficient when the vehicle is traveling at medium speeds, and Model C is used to calculate the road adhesion coefficient when the vehicle is traveling at high speeds. In this way, by deploying different calculation models, the road adhesion coefficient can be estimated in different speed domains.

[0154] For example, two different calculation models can be used to estimate the road surface adhesion coefficient: Model A is a cause-based model, and Model B is an effect-based model. The cause-based model focuses on analyzing the characteristics of the road surface itself (such as its specific shape and physical features), tire characteristics, and the interaction between the vehicle and the road surface to estimate the road surface adhesion coefficient. The effect-based model estimates the magnitude of the road surface adhesion coefficient by testing the motion response on the vehicle body or wheels caused by the road surface adhesion coefficient or changes in the adhesion coefficient. For example, the effect-based model can be any of the following: an estimation model based on tire local strain-stress, an estimation model based on tire force observers and dynamic models, or an estimation model based on μ-s curves.

[0155] In another application scenario, different calculation models may be applicable to the same working conditions.

[0156] Taking the calculation of the road adhesion coefficient in a dry road environment as an example, Model A, Model B, and Model C all use indirect estimation methods to estimate the road adhesion coefficient. For instance, Model A estimates the road adhesion coefficient in a dry road environment based on vehicle longitudinal dynamics (e.g., measuring acceleration or braking distance); Model B estimates the road adhesion coefficient in a dry road environment based on the difference between tire rotation speed and actual driving speed (i.e., slip ratio); and Model C estimates the road adhesion coefficient in a dry road environment based on data provided by various sensors (e.g., accelerometers, GPS, inertial sensors, etc.) combined with calculation methods (e.g., Kalman filtering, machine learning, etc.). In this case, the accuracy of the estimation result obtained by Model B is greater than that obtained by Model A, and the accuracy of the estimation result obtained by Model C is greater than that obtained by Model B. This shows that for the same vehicle parameter estimation, different calculation models can yield estimation results with different levels of accuracy.

[0157] By implementing this method, different computational models can be deployed on different control components, which can maximize the capabilities of the hardware. Different computational models can be applied to different working conditions, which not only improves the flexibility of deployment and meets the application needs of different scenarios, but also helps to improve the accuracy of estimation.

[0158] In another implementation, the same complex computational model is deployed in a distributed manner across multiple control units. For example, a first computational model is deployed on a first control unit, a second computational model on a second control unit, and a third computational model on a third control unit. The first, second, and third computational models are, respectively, partial models or modules within a target computational model, which is used to calculate the target vehicle parameters. Referring to Figure 4C, the target computational model comprises three parts: model part1, model part2, and model part3. These target computational models are distributed across the first, second, and third control units; that is, the first computational model is model part1, the second computational model is model part2, and the third computational model is model part3.

[0159] Taking the target calculation model for calculating the vehicle parameter of tire lateral force as an example, model part1 deployed on the first control unit is used to calculate tire vertical force, model part2 deployed on the second control unit is used to calculate tire longitudinal force, and model part3 deployed on the third control unit is used to calculate tire lateral force. The calculation of tire longitudinal force is related to tire vertical force, and the calculation of tire lateral force is related to both tire vertical force and tire longitudinal force.

[0160] For example, to further improve the accuracy of vehicle parameter estimation, the target calculation model may be a machine learning model, a deep learning model, or a neural network model. In this case, the target calculation model can also be decomposed and deployed on multiple control components. For example, control component 1 is used to implement data mapping, control component 2 is used to implement data calculation, control component 3 is used to implement data calculation, and control component 4 is used to arbitrate the execution results. This solves the problem of different computing capabilities of a single control component and improves the utilization rate of computing power.

[0161] This implementation method breaks down a complex computational model into different parts or modules and distributes them across multiple control units. Each control unit executes a different part of the same computational model, and during deployment, computing resources can be allocated based on the processing capabilities of each control unit to support the operation of the complex computational model. This fully utilizes the computing resources of multiple control units, and the collaborative completion of complex estimations by multiple control units provides more accurate estimation results. It also improves the execution efficiency of complex calculations. This solves the problem of insufficient computing power and resources of a single control unit, which prevents the operation of complex computational models, thus enhancing the system's computational and load-bearing capacity.

[0162] In some possible embodiments, the computational models deployed on multiple control components may be partially the same or partially different. For example, a first computational model may be deployed on a first control component, a second computational model on a second control component, and a third computational model on a third control component, wherein the first computational model is the same as the second computational model, but the second computational model is different from the third computational model. Such deployment can achieve both differentiated deployment of computational models and multiple computations using the same computational model.

[0163] Taking any one of the multiple control components (e.g., the first control component) as an example, if a first calculation model is deployed on the first control component, then the control component uses its local calculation model to perform calculations on the received first operating data to obtain a first estimation result, including: the first control component uses the first calculation model to perform calculations on the received first operating data to obtain the first estimation result. Further, the first control component sends the first estimation result to other control components among the multiple control components.

[0164] For example, the first control unit sends a first estimation result to other control units among a plurality of control units, including: the first control unit transmits the first estimation result in a multicast manner via a first type of bus, wherein multicast can be broadcast or multicast. When multicast is broadcast, the other control units among the plurality of control units besides the first control unit are members of the same broadcast group; when multicast is multicast, the other control units among the plurality of control units besides the first control unit are members of the same multicast group.

[0165] As an example, when the first control unit transmits the first estimation result via multicast, the first estimation result is carried in a multicast message, wherein the multicast message also includes several of the following information:

[0166] Multicast IP address: The multicast IP address is used to identify the multicast group that received the first estimation result;

[0167] Parameter encoding information: Parameter encoding information is used to indicate that the calculated vehicle parameters are those of the target vehicle; and,

[0168] Intermediate process variables: Intermediate process variables are used to indicate intermediate parameters involved in the calculation of the first estimation result.

[0169] The multicast IP address mentioned above is used to share and output the first estimation result, and the multicast group is used to define the calculation units participating in the vehicle parameter estimation. Different vehicle parameters can be managed through different multicast groups. Parameter encoding information not only distinguishes and labels different vehicle parameters but also differentiates the calculation model used and the control unit performing the calculation. Intermediate process variables are key intermediate variables or key parameters calculated by the calculation model, serving as auxiliary values ​​for the obtained estimation result. For the estimation of the same vehicle parameter, the intermediate process variables used may differ under different operating conditions. It can be understood that in the above multicast message, both the intermediate process variables and the first estimation result are obtained through local calculation performed by the first control unit.

[0170] In some possible embodiments, when the vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data of the first cycle. The first control unit uses a local calculation model to perform calculations on the first operation data to obtain a first estimation result, including: the first control unit uses a local calculation model to perform calculations on the first operation data and at least one historical calculation result of the target vehicle parameters to obtain the first estimation result; wherein the at least one historical calculation result comes from at least one other control unit, the historical calculation result is associated with the second operation data, the second operation data is the vehicle operation data transmitted in the second cycle, and the second cycle is the cycle preceding the first cycle.

[0171] In this case, the multicast that carries the first estimation result is also used to carry the following information:

[0172] At least one historical calculation result of the target vehicle parameters, wherein the aforementioned historical calculation result is used in the calculation of the first estimation result, the aforementioned historical calculation result is derived from at least one other control component, and the historical calculation result is associated with second operating data, the second operating data being vehicle operating data for a second cycle, the second cycle being the cycle preceding the first cycle; and,

[0173] Verification information is used to determine whether the first estimation result calculated in the current period is valid.

[0174] For example, the verification information includes the sequence counter of the computation task, the signature, and cyclic redundancy check (CRC) information.

[0175] Refer to Table 1, which shows the information carried by a multicast message. As shown in Table 1, the multicast message includes the multicast group identifier (e.g., multicast IP address), parameter encoding information, the calculation result of the previous cycle referenced, intermediate process variables, the current calculation result (i.e., the first estimation result mentioned above), and verification information.

[0176] Table 1 Information Carried by Multicast Messages

[0177] Here, the information carried by the multicast messages shown in Table 1 is only an example and does not limit the multicast messages to carrying the content shown in Table 1. The number of calculation results from the previous period referenced in Table 1 is 2, but this scheme does not limit the number of historical calculation results referenced to only 2. In some schemes, the verification information, parameter encoding information, etc., in Table 1 can be optional information.

[0178] S303: At least one control unit receives a second estimation result from other control units, the second estimation result being associated with the first operating data.

[0179] Here, the second estimation result is obtained by other control components performing calculations on the first operating data using their local calculation models. For example, when the other control component is the second control component described above, the local calculation model of the second control component is the second calculation model described above; when the other control component is the third control component described above, the local calculation model of the third control component is the third calculation model described above.

[0180] It is understandable that the terms "first" and "second" in the first and second estimation results are only used to distinguish whether the estimation results come from the local control unit or from other control units.

[0181] For example, the second estimation result is transmitted in a multicast format. Here, multicast can be either broadcast or multicast.

[0182] As an example, when the second estimation result is transmitted via multicast, the second estimation result is carried in a multicast message. In addition to carrying the second estimation result, this multicast message may also carry several of the following information:

[0183] Multicast IP address: The multicast IP address is used to identify the multicast group receiving the second estimation result;

[0184] Parameter encoding information: Parameter encoding information is used to indicate that the calculated vehicle parameters are those of the target vehicle; and,

[0185] Intermediate process variables: Intermediate process variables are used to indicate intermediate parameters involved in the calculation of the second estimation result.

[0186] In some schemes, where vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data for the first period, and the multicast message carrying the second estimation result also carries the following information:

[0187] At least one historical calculation result of the target vehicle parameters, wherein the at least one historical calculation result comes from at least one other control unit, each historical result is associated with second operating data, the second operating data being vehicle operating data for a second cycle, the second cycle being the cycle preceding the first cycle; and,

[0188] Verification information is used to determine whether the second estimation result calculated in the current period is valid.

[0189] Please refer to the description of the relevant content in the multicast message mentioned above, which carries the first estimation result, for the content of each item in the multicast message mentioned above. It will not be repeated here.

[0190] The data exchange process between different control components is described in detail below with reference to Figure 5. Referring to Figure 5, Figure 5 is a schematic diagram of the transmission of estimation results of different control components in a parameter estimation method provided by an embodiment of this application.

[0191] In Figure 5, multiple control components include a first control component, a second control component, and a third control component. Each control component acquires first operating data through the vehicle bus and / or signal lines. Subsequently, the first control component uses a first calculation model to perform calculations on the first operating data to obtain an estimation result 1. The second control component uses a second calculation model to perform calculations on the first operating data to obtain an estimation result 2. The third control component uses a third calculation model to perform calculations on the first operating data to obtain an estimation result 3.

[0192] For the first control unit, the first control unit transmits the estimation result 1 to the multicast IP address 1. The multicast group identified by the multicast IP address 1 includes two members: the second control unit and the third control unit. That is, the first control unit sends the estimation result 1 to the second control unit and the third control unit in a multicast manner.

[0193] For the second control unit, the second control unit transmits the estimation result 2 to the multicast IP address 2. The multicast group identified by the multicast IP address 2 includes two members: the first control unit and the third control unit. That is, the second control unit sends the estimation result 2 to the first control unit and the third control unit in a multicast manner.

[0194] For the third control unit, the third control unit transmits the estimation result 3 to the multicast IP address 3. The multicast group identified by the multicast IP address 3 includes two members: the first control unit and the second control unit. That is, the third control unit sends the estimation result 3 to the first control unit and the second control unit in a multicast manner.

[0195] It is understood that in the above scheme, the members of the multicast group can include not only other control components among multiple control components, but also other processors that the upper-layer application needs to obtain the estimation results.

[0196] Referring to S301-S303 above, the first estimation result and the second estimation result are explained below with reference to Figure 5.

[0197] When the local control unit is the first control unit, the other control units among the above-mentioned multiple control units include the second control unit and the third control unit. Estimation result 1 is the first estimation result mentioned above, estimation result 2 is the second estimation result from the second control unit, and estimation result 3 is the second estimation result from the third control unit.

[0198] When the local control unit is the second control unit, the other control units among the above-mentioned multiple control units include the first control unit and the third control unit. Estimation result 2 is the first estimation result mentioned above, estimation result 1 is the second estimation result from the first control unit, and estimation result 3 is the second estimation result from the third control unit.

[0199] When the local control unit is the third control unit, the other control units among the above-mentioned multiple control units include the first control unit and the second control unit. The estimation result 3 is the first estimation result mentioned above, the estimation result 1 is the second estimation result from the first control unit, and the estimation result 2 is the second estimation result from the second control unit.

[0200] In Figure 5, the order in which different control components execute steps is not limited. For example, each control component may simultaneously calculate the corresponding estimation result using its locally deployed computational model on the first running data. For a single control component, the estimation result is multicast after it is obtained. In some schemes, where complex computational models are distributed across multiple control components, it is also possible that the second control component executes the calculation to obtain estimation result 2 only after the first control component multicasts estimation result 1, and the third control component executes the calculation to obtain estimation result 3 only after the second control component multicasts estimation result 2.

[0201] In Figure 5, in some schemes, the data relied upon for calculating the estimation result includes not only the first operating data but also the estimation results of other control components. For example, in the case where a complex computational model is distributed across multiple control components, the data relied upon for calculating estimation result 2 includes the first operating data and estimation result 1, and the data relied upon for calculating estimation result 3 includes the first operating data, estimation result 1, and estimation result 2. This implementation can be referred to in the description of the embodiment in Figure 6 below.

[0202] Here, Figure 5 is merely an example of the transmission of estimation results between different control units. In some schemes, it is not required that each control unit receive a second estimation result from other control units, nor is it required that each control unit perform local arbitration to determine the calculation result of the target vehicle parameters. For example, in the case where a complex calculation model is distributed across multiple control units, assuming that the calculation of estimation result 2 depends on estimation result 1, and the calculation of estimation result 3 depends on estimation result 1 and estimation result 2, then the first control unit used to calculate estimation result 2 may not be a member of the multicast group used to receive estimation result 2, nor may it be a member of the multicast group used to receive estimation result 3. That is, the first control unit does not need to receive estimation result 2 from other control units (e.g., the second control unit), nor does it need to receive estimation result 3 from other control units (e.g., the third control unit). In this case, the calculation result of the target vehicle parameters after local arbitration can be sent by the second and third control units, or by the third control unit alone.

[0203] S304: At least one control unit obtains the calculation results of the target vehicle parameters based on the first estimation result and the received second estimation result.

[0204] Taking the local control unit as the first control unit as an example, the first control unit obtains the calculation result of the target vehicle parameters based on the first estimation result and the received second estimation result. This means that the first control unit obtains the calculation result of the target vehicle parameters based on the first estimation result, the second estimation result from the second control unit, and the second estimation result from the third control unit.

[0205] For example, referring to Figure 5, when the local control unit is the first control unit, the first control unit obtains the calculated result 1 of the target vehicle parameters based on estimation result 1, estimation result 2, and estimation result 3; when the local control unit is the second control unit, the second control unit obtains the calculated result 2 of the target vehicle parameters based on estimation result 2, estimation result 1, and estimation result 3; when the local control unit is the third control unit, the third control unit obtains the calculated result 3 of the target vehicle parameters based on estimation result 3, estimation result 1, and estimation result 2. Here, calculation result 1, calculation result 2, and calculation result 3 are the same.

[0206] Because the computational models on multiple control components are deployed in different ways, the preset rules on which the control components arbitrate to obtain the calculation results of the target vehicle parameters may be different. For example, the calculation results of the target vehicle parameters may be a new calculation result calculated based on the first estimation result and the received second estimation result, or the calculation of the target vehicle parameters may be the first estimation result or a certain second estimation result.

[0207] The preset rules configured on different control components are the same. Taking the first control component as an example, the first control component obtains the calculation result of the target vehicle parameters based on the first estimation result and the received second estimation result. This includes: the first control component arbitrates the first estimation result and the received second estimation result based on preset rules to obtain the calculation result of the target vehicle parameters. The following describes different cases, please refer to the descriptions of cases 1-3 below.

[0208] Scenario 1: The same computing model is deployed on multiple control components.

[0209] In Case 1, as an example, the preset rule is that when all estimation results are the same, the calculated result of the target vehicle parameter is determined to be any one of the estimation results. Referring to Figure 5, the first control unit arbitrates estimation result 1, estimation result 2, and estimation result 3 according to the preset rule to obtain the calculated result 1 of the target vehicle parameter, including: when estimation result 1, estimation result 2, and estimation result 3 are all equal, the first control unit determines the calculated result 1 of the target vehicle parameter as estimation result 1.

[0210] As another example, when the control unit needs to arbitrate more than three estimation results, the preset rule can also be that the number of estimation results with the same value among the three or more estimation results reaches a preset proportion. For example, the first control unit arbitrates estimation result 1, estimation result 2, and estimation result 3 according to a preset rule. The preset rule is that when at least two estimation results are the same, the calculated result of the target vehicle parameter is determined to be any of the same estimation results. Assuming that estimation result 1 is the same as estimation result 2, but estimation result 2 is different from estimation result 3, the preset rule is satisfied, and the first control unit determines that the calculated result of the target vehicle parameter 1 is estimation result 1 after arbitration.

[0211] In some schemes, the preset rules may also be to average multiple estimation results or take the estimation result corresponding to the smallest error.

[0212] In some solutions, truth tables can also be used as preset rules. Taking multiple control components, including control component 1 and control component 2, as an example, control component 1 and control component 2 are deployed with the same calculation model. When control component 1 determines that estimation result 1 is consistent with estimation result 2 from control component 2, it returns True; when control component 2 determines that estimation result 2 is consistent with estimation result 1 from control component 1, it returns True. For control component 1, it only transmits estimation result 1 to at least one of the upper-layer application, sensor system, and execution system when it obtains two True values. Similarly, for control component 2, it only transmits estimation result 2 to at least one of the upper-layer application, sensor system, and execution system when it obtains two True values. It can be understood that when control component 1 or control component 2 only obtains one True value, a secondary calculation or alarm will be performed. In this way, the reliability and security of parameter estimation are improved.

[0213] For example, assuming that both control unit 1 and control unit 2 perform parameter estimation with a first duration as the period, in order to improve data density, an interpolation algorithm can be set on control unit 1 and / or control unit 2. Control unit 1 or control unit 2 can interpolate the estimation result 1 and the estimation result 2 of control unit 1 to obtain a set of estimation results calculated with a second duration as the period, wherein the second duration is shorter than the first duration. In this way, the sampling rate of the signal is significantly improved.

[0214] In this way, multiple control components execute the same calculation model and arbitrate the estimation results of multiple control components, making full use of the computing resources of the control components to achieve redundant calculation.

[0215] Scenario 2: Different computing models are deployed on multiple control components.

[0216] In scenario 2, as an example, when different calculation models are used for the same working condition, the preset rule can be to use the target estimation result as the calculation result of the target vehicle parameters, where the accuracy of the target estimation result is greater than the accuracy of other estimation results. For example, referring to Figure 5, the first control unit arbitrates estimation result 1, estimation result 2, and estimation result 3 according to the preset rule. Assuming that the accuracy of estimation result 1 is greater than the accuracy of estimation result 2, and the accuracy of estimation result 2 is greater than the accuracy of estimation result 3, then the first control unit determines through arbitration that the calculation result 1 of the target vehicle parameters is estimation result 1, that is, estimation result 1 is the aforementioned target estimation result.

[0217] In some schemes, when different control components are configured with different calculation models, for safety reasons, the preset rule can also be set to take the smaller value. For the estimation results obtained by each control component, the smaller estimation result is output as the calculation result of the target vehicle parameters.

[0218] As another example, when different calculation models are used for different operating conditions, the preset rule can be to use the estimation result under the operating condition at the time of the estimation as the calculation result of the target vehicle parameters. For example, referring to Figure 5, assuming that the first calculation model is used to calculate the road adhesion coefficient in a dry road environment, the second calculation model is used to calculate the road adhesion coefficient in a wet road environment, and the third calculation model is used to calculate the road adhesion coefficient in an icy and snowy road environment, and the vehicle is on a dry road when the road adhesion coefficient is estimated, and the estimation result corresponding to the first calculation model is estimation result 1, then the first control unit determines through arbitration that the calculation result 1 of the target vehicle parameters is estimation result 1.

[0219] For example, referring to Figure 5, taking the calculation of road surface adhesion coefficient as an example, assuming that the first calculation model is configured based on rapid acceleration, steering or braking conditions, the second calculation model is configured based on constant speed conditions, and the third calculation model is configured based on slow acceleration and deceleration conditions, after each control component obtains the road surface adhesion coefficient under the corresponding conditions, the synthesis algorithm deployed on one of the control components can process the road surface adhesion coefficient under each condition to obtain the calculation results of the target vehicle parameters, thus realizing the accurate calculation of the road surface adhesion coefficient under all conditions, which can meet the application requirements of upper-level applications and execution systems.

[0220] In this way, multiple control components execute different calculation models. When different calculation models are applied to the same working condition, they can provide estimation results with different levels of accuracy, thus achieving accurate estimation of vehicle parameters. When different calculation models are applied to different working conditions, the estimation of target vehicle parameters is differentiated by scenario, thus achieving refined estimation of vehicle parameters, which helps to improve the accuracy of the estimated vehicle parameters.

[0221] Scenario 3: Different modules of the same computing model are deployed on multiple control components.

[0222] In scenario 3, the preset rule can be to use the output result of the target vehicle parameters as the calculation result of the target vehicle parameters. For example, the tire force calculation model is distributed across multiple control units. The tire force calculation model is ultimately used to obtain the tire lateral force (i.e., the target vehicle parameters). Referring to Figure 5, assuming that the first calculation model is used to calculate the tire vertical force, the second calculation model is used to calculate the tire longitudinal force, and the third calculation model is used to calculate the tire lateral force, then estimation result 1 is the tire vertical force, estimation result 2 is the tire longitudinal force, and estimation result 3 is the tire lateral force. Among these, the calculation of the tire longitudinal force depends on the tire vertical force, and the calculation of the tire lateral force depends on both the tire vertical force and the tire longitudinal force. In this case, each control unit obtains the calculation result of the target vehicle parameters through local arbitration based on the above preset rule. It can be seen that the output result of the target vehicle parameters is estimation result 3. Therefore, the calculation result 1 obtained by the arbitration of the first control unit is estimation result 3, the calculation result 2 obtained by the arbitration of the second control unit is estimation result 3, and the calculation result 3 obtained by the arbitration of the third control unit is estimation result 3.

[0223] In this way, a complex computational model is distributed across multiple control units, with each control unit participating in a portion of the computation of the complex model. This fully utilizes the computing resources of each control unit, improving not only the efficiency of parameter estimation but also ensuring the accuracy and precision of the estimated target vehicle parameters. Furthermore, multiple control units can quickly exchange estimation results via a Type I bus (e.g., automotive Ethernet), enabling result sharing and ensuring that each control unit can obtain the calculated target vehicle parameters.

[0224] Optionally, in some possible embodiments, the following S305 may also be performed.

[0225] S305: At least one control unit sends the calculated results of the target vehicle parameters to the sensor system and / or the execution system.

[0226] In one implementation, taking the first control unit as an example, the first control unit sends the calculation results of the target vehicle parameters to the execution system, including: the first control unit sends a control command to the execution system, wherein the control command includes the calculation results of the target vehicle parameters.

[0227] For example, control commands can be transmitted via unicast, multicast, or broadcast.

[0228] Referring to Figure 5, the first control unit sends the calculation result 1 of the target vehicle parameters to the connected sensor system and / or execution system through the aforementioned second type of bus and / or signal line; the second control unit sends the calculation result 2 of the target vehicle parameters to the connected sensor system and / or execution system through the aforementioned second type of bus and / or signal line; and the third control unit sends the calculation result 3 of the target vehicle parameters to the connected sensor system and / or execution system through the aforementioned second type of bus and / or signal line.

[0229] It is understandable that different control components have different computing capabilities, so the time it takes for different control components to obtain the calculation results of the target vehicle parameters may be different. Setting up a mechanism whereby each control component sends the calculation results of the target vehicle parameters after obtaining them through arbitration can enable the sensor system and / or execution system to obtain the calculation results of the target vehicle parameters more quickly.

[0230] In the embodiment shown in Figure 3, multiple control units mounted on the same first-type bus and the same second-type bus form an on-board computing architecture. Utilizing the broadcast characteristics of the second-type bus in this architecture, the multiple control units can acquire vehicle operation data, achieving data alignment. By deploying computational models on multiple control units, these units can jointly participate in estimating the same target vehicle parameters, fully utilizing the computing resources of the control units and providing more complex and accurate parameter estimation services. Furthermore, these multiple control units can exchange estimation results via the first-type bus, achieving result sharing. This allows at least one control unit to locally arbitrate and obtain the calculation results of the target vehicle parameters, improving both the efficiency and accuracy of the estimated target vehicle parameters.

[0231] The application of the above method is illustrated below with a specific scenario. Referring to Figure 6, which is a schematic diagram of an application scenario provided by an embodiment of this application, the complex tire force calculation model is broken down into multiple parts (e.g., tire longitudinal force calculation, tire vertical force calculation, and tire lateral force calculation) and distributed across multiple control components. These control components communicate with each other via an in-vehicle Ethernet connection. These multiple control components include VDC-MCU-CoreX, VIU1-MCU-CoreX, and VIU2-MCU-CoreX. The VDC-MCU-CoreX is equipped with a tire longitudinal force calculation to calculate the tire longitudinal force, the VIU1-MCU-CoreX is equipped with a tire vertical force calculation to calculate the tire vertical force, and the VIU2-MCU-CoreX is equipped with a tire lateral force calculation to calculate the tire lateral force. In other words, the algorithm deployed on each control component is used to calculate the force in one direction.

[0232] For example, in Figure 6, each control unit acquires vehicle operating data in real time through the chassis bus. The vehicle operating data includes information such as longitudinal acceleration, yaw rate, longitudinal vehicle speed, lateral acceleration, steering wheel angle, tire parameters, and vertical load. The calculations performed locally by each control unit are as follows.

[0233] For VIU1-MCU-CoreX: VIU1-MCU-CoreX uses a locally deployed tire vertical force calculation formula to calculate the tire vertical force based on the acquired longitudinal acceleration, yaw rate, and longitudinal vehicle speed. VIU1-MCU-CoreX then multicasts this tire vertical force via the vehicle Ethernet. The members of this multicast group include at least VDC-MCU-CoreX and VIU2-MCU-CoreX, so that VDC-MCU-CoreX and VIU2-MCU-CoreX can obtain the tire vertical force from VIU1-MCU-CoreX.

[0234] For VDC-MCU-CoreX: VDC-MCU-CoreX uses a locally deployed tire longitudinal force calculation formula to calculate the tire longitudinal force based on the acquired tire parameters, vertical load, and tire vertical force. VDC-MCU-CoreX then multicasts this tire longitudinal force via the vehicle Ethernet. The members of this multicast group include at least VIU1-MCU-CoreX and VIU2-MCU-CoreX, so that VIU1-MCU-CoreX and VIU2-MCU-CoreX can obtain the tire longitudinal force from VDC-MCU-CoreX.

[0235] For VIU2-MCU-CoreX: VIU2-MCU-CoreX uses a locally deployed tire lateral force calculation formula to calculate the acquired longitudinal acceleration, lateral acceleration, steering wheel angle, tire vertical force, and tire longitudinal force to obtain the tire lateral force. VIU2-MCU-CoreX multicasts this tire lateral force through the vehicle Ethernet. The members of this multicast group include at least VDC-MCU-CoreX and VIU1-MCU-CoreX, so that VDC-MCU-CoreX and VIU1-MCU-CoreX obtain the tire lateral force from VIU2-MCU-CoreX.

[0236] Finally, VDC-MCU-CoreX, VIU1-MCU-CoreX, and VIU2-MCU-CoreX each arbitrate the tire vertical force, tire longitudinal force, and tire lateral force, and determine that the tire lateral force will be transmitted to the connected sensor system and / or execution system via the chassis bus.

[0237] In some schemes, after VDC-MCU-CoreX calculates the longitudinal force of the tire, it may not need to transmit the longitudinal force of the tire to VIU1-MCU-CoreX; after VIU2-MCU-CoreX calculates the lateral force of the tire, it may not need to transmit the lateral force of the tire to VIU1-MCU-CoreX. In this way, VIU1-MCU-CoreX does not need to perform subsequent arbitration operations locally, and at least one of VDC-MCU-CoreX and VIU2-MCU-CoreX can perform the arbitration operation according to the above description.

[0238] In this way, multiple control components mounted on the same chassis bus and the same vehicle Ethernet form an in-vehicle computing architecture. Using this in-vehicle computing architecture, complex computing models are decomposed and distributed across multiple control components, enabling multiple control components to work together to complete parameter estimation. This fully utilizes the computing resources of the control components, making the operation of complex computing models feasible, improving estimation accuracy, and also improving the execution efficiency of complex computing models.

[0239] In some possible embodiments, where vehicle operation data is acquired periodically, a single control unit may rely on historical calculation results from the previous cycle when calculating target vehicle parameters (e.g., displacement) locally using a computational model. In this case, the process executed by the single control unit can be referred to the embodiment in Figure 7 below. Referring to Figure 7, which is a schematic flowchart of parameter estimation performed by a single control unit according to an embodiment of this application, in Figure 7, taking the first type of bus as vehicle Ethernet and the second type of bus as an in-vehicle bus as an example, the execution process is described in steps 1-8 below.

[0240] Step 1: Calculate task timing to synchronize the execution time of each control component for the calculation task.

[0241] Step 2: The control unit acquires vehicle operation data for the current cycle (e.g., the first cycle mentioned above) via the in-vehicle bus and / or signal lines.

[0242] Step 3: The control unit uses a local calculation model to perform calculations on the vehicle data of the current cycle and at least one historical calculation result from the previous cycle (e.g., the second cycle mentioned above) to obtain the first estimation result mentioned above.

[0243] In this process, at least one historical calculation result from the previous cycle is obtained by the control unit from at least one multicast message transmitted from the vehicle Ethernet. The at least one multicast message comes from at least one other control unit, and each multicast message carries a historical calculation result from the previous cycle.

[0244] For example, when the same computational model is deployed on multiple control components or when different computational models are deployed on multiple control components, the historical computational results can be the output results of the target vehicle parameters; when different parts of the same complex algorithm are deployed on multiple control components (i.e., complex algorithms are deployed in a distributed manner on multiple control components), the historical computational results can also be an intermediate result on which the calculation of the target vehicle parameters depends.

[0245] Step 4: The control unit transmits the first estimation result to other control units via the vehicle Ethernet.

[0246] Step 5: The control unit obtains at least one second estimation result for the current period via the vehicle Ethernet. This second estimation result comes from at least one other control unit, which calculates the result using locally deployed computational models on vehicle operation data and other data for the current period. In some schemes, the calculation of the second estimation result may also involve at least one historical calculation result.

[0247] Step 6: The control unit performs local arbitration on the first estimation result and at least one of the above-mentioned second estimation results to obtain the calculation result of the target vehicle parameters. Here, please refer to the relevant description of the arbitration process in the embodiment S304 of Figure 3 above, and it will not be repeated here.

[0248] Step 7: The control unit transmits the calculation results of the target vehicle parameters to other control units via the vehicle Ethernet.

[0249] Step 8: The control unit sends control commands to the execution system via the vehicle bus. These commands carry the calculated results of the target vehicle parameters. Accordingly, after the execution system performs the corresponding operation based on the control commands, it transmits the execution results back to the vehicle bus and / or signal lines. This feedback indicates the execution status of the control commands for the current cycle.

[0250] Figure 7 is only an example of a process for estimating vehicle parameters by a single control unit. In Figure 7, steps 5-8 can be optional. In some schemes, when complex algorithms are distributed across multiple control units, some control units, when calculating the first estimation result, rely not only on the vehicle operation data of the current period and at least one historical calculation result from the previous period, but also on at least one second estimation result of the current period.

[0251] As can be seen, by periodically acquiring vehicle operation data and rapidly exchanging estimation results between multiple control components on the vehicle Ethernet, the control components can continuously estimate the target vehicle parameters, which not only improves the accuracy of the estimated vehicle parameters but also helps to improve the estimation efficiency of vehicle parameters.

[0252] In some solutions, based on the architecture shown in Figure 1A or Figure 1B, when multiple control components are equipped with computational models, if one control component fails to issue control commands (carrying the aforementioned computational results) normally, the system can use control commands (carrying the aforementioned computational results) issued normally by other control components, without affecting the normal operation of the system. This achieves functional backup.

[0253] For example, the reasons why the control unit cannot issue control commands normally include one or more of the following: a failure in the execution of the computational model within the control unit; a failure in the data link between the control unit and the sensor system; a failure in the communication link between the control unit and the execution system; and a failure in the communication link between the control unit and the upper-layer application.

[0254] For example, when the computing models deployed on multiple control components are related to the implementation of intelligent driving system functions, it is assumed that one control component acts as the master control component, responsible for upstream and downstream data transmission and computation, while the other control components act as slave control components, performing synchronous data computation. When the master control component fails, the slave control components can take over from the master control component to perform upstream and downstream data transmission and computation.

[0255] For example, multiple control components simultaneously perform upstream and downstream data transmission and calculation, with the upper-layer application or actuator arbitrating the calculation results. When one control component fails, the upper-layer application or actuator can adopt the calculation results of the other control components for processing.

[0256] The above method is applicable to functional modules such as DC-DC control, TMS thermal management control, BMS control, and gear control. By deploying computational models on multiple control components, backup control of functional modules can be achieved, increasing the system's fault tolerance.

[0257] Referring to Figure 8, which is a schematic diagram of a parameter estimation device provided in an embodiment of this application, the parameter estimation device 30 includes an acquisition unit 310, a processing unit 312, and a transmission unit 314. The parameter estimation device 30 can be implemented in hardware, software, or a combination of both. In some solutions, the acquisition unit 310 and the transmission unit 314 can also be collectively referred to as a communication unit.

[0258] For example, the parameter estimation device 30 may be any of the control components in the above embodiments or may be included in a control component. This control component is connected to other control components via a first type of bus, and is connected to the sensor system and / or execution system via a second type of bus. In some embodiments, the control component may also be connected to the sensor system via a signal line.

[0259] The acquisition unit 310 is used to acquire first operating data via signal lines and / or a second type of bus; the processing unit 312 is used to perform calculations on the first operating data using a local calculation model to obtain a first estimation result; the sending unit 314 is used to transmit the first estimation result to other control components via a first type of bus; the acquisition unit 310 is also used to receive a second estimation result from other control components, the second estimation result being associated with the first operating data; the processing unit 312 is also used to obtain the calculation result of the target vehicle parameters based on the first estimation result and the received second estimation result. In some embodiments, the sending unit 314 is also used to send the calculation result of the target vehicle parameters to the sensor system via a second type of bus and / or signal lines.

[0260] The parameter estimation device 30 can be used to implement the control component-side method described in the embodiment of FIG3. In the embodiment of FIG3, the acquisition unit 310 can be used to execute S301 and S303, the processing unit 312 can be used to execute S302 and S304, and the sending unit 314 can be used to execute S305. The parameter estimation device 30 can also be used to implement any of the control component-side methods described in the embodiment of FIG5, and can also be used to implement the method described in the embodiment of FIG7. For the sake of brevity, it will not be described again here.

[0261] It should be understood that the division of the units in the parameter estimation device 30 above is only a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, the units in the device can be implemented by a processor calling software; for example, the device includes a processor connected to a memory containing instructions. The processor calls the instructions stored in the memory to implement any of the above methods or to implement the functions of each unit in the device. The processor can be, for example, a general-purpose processor, such as a central processing unit (CPU) or a microprocessor, and the memory can be internal or external to the device. Alternatively, the units in the device can be implemented as hardware circuits. The functionality of some or all units can be achieved through the design of these hardware circuits, which can be understood as one or more processors. For example, in one implementation, the hardware circuit is an application-specific integrated circuit (ASIC). The functionality of some or all of the above units is achieved through the design of the logical relationships between the components within the circuit. In another implementation, the hardware circuit can be implemented using a programmable logic device (PLD). Taking a field-programmable gate array (FPGA) as an example, it can include a large number of logic gates. The connection relationships between the logic gates are configured through a configuration file, thereby achieving the functionality of some or all of the above units. All units of the above device can be implemented entirely through processor-invoked software, entirely through hardware circuits, or partially through processor-invoked software with the remaining parts implemented through hardware circuits.

[0262] In this application embodiment, a processor is a circuit with signal processing capabilities. In one implementation, the processor can be a circuit with instruction reading and execution capabilities, such as a central processing unit (CPU), a microprocessor, a graphics processing unit (GPU) (which can be understood as a type of microprocessor), or a digital signal processor (DSP). In another implementation, the processor can implement certain functions through the logical relationships of hardware circuits. These logical relationships of hardware circuits are fixed or reconfigurable. For example, the processor is a hardware circuit implemented using an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document and configuring the hardware circuit can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units. Furthermore, it can also be a hardware circuit designed for artificial intelligence, which can be understood as a type of ASIC, such as a neural network processing unit (NPU), a tensor processing unit (TPU), a deep learning processing unit (DPU), etc.

[0263] As can be seen, each unit in the above device can be one or more processors (or processing circuits) configured to implement the above methods, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor forms.

[0264] Furthermore, the units in the above devices can be integrated in whole or in part, or they can be implemented independently. In one implementation, these units are integrated together as a system-on-a-chip (SOC). The SOC may include at least one processor for implementing any of the above methods or implementing the functions of the units in the device. The at least one processor may be of different types, such as CPU and FPGA, CPU and artificial intelligence processor, CPU and GPU, etc.

[0265] Referring to Figure 9, which is a schematic diagram of a computing device according to an embodiment of this application, the computing device 40 includes a processor 401, a communication interface 402, a memory 403, and a bus 404. The processor 401, the memory 403, and the communication interface 402 communicate with each other via the bus 404. It should be understood that this application does not limit the number of processors and memories in the computing device 40.

[0266] In one implementation, the computing device 40 may be the control component described above. Here, the control component is described in the preceding embodiments and will not be repeated here.

[0267] Bus 404 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, only one line is used in Figure 9, but this does not imply that there is only one bus or one type of bus. Bus 404 can include pathways for transmitting information between various components of computing device 40 (e.g., memory 403, processor 401, communication interface 402).

[0268] The processor 401 can be referred to the relevant description of the processor in the above embodiments, and will not be repeated here.

[0269] Memory 403 provides storage space, which can store data such as the operating system and computer programs. Memory 403 can be one or a combination of several of the following: random access memory (RAM), erasable programmable read-only memory (EPROM), read-only memory (ROM), or compact disc read memory (CD-ROM). Memory 403 can exist alone or be integrated into processor 401.

[0270] The communication interface 402 can be used to provide information input or output to the processor 401. Alternatively, the communication interface 402 can be used to receive and / or send data to externally transmitted data, and can be a wired link interface including an Ethernet cable, or a wireless link interface (such as Wi-Fi, Bluetooth, general wireless transmission, etc.). Alternatively, the communication interface 402 may also include a transmitter (such as an RF transmitter, antenna, etc.) or a receiver coupled to the interface.

[0271] The processor 401 in the computing device 40 is used to read the computer program stored in the memory 403 to execute the aforementioned method, such as the method on the control component side described in FIG3 or FIG5, or the method described in FIG7.

[0272] In one possible design, computing device 40 may be one or more modules in an execution entity (e.g., a control unit) that performs the method shown in FIG3. The processor 401 may be used to read one or more computer programs stored in memory for performing the following operations:

[0273] First operating data is acquired from the second type of bus and / or signal lines by the acquisition unit 310;

[0274] The first estimation result is obtained by performing calculations on the first running data based on the locally deployed computing model;

[0275] The first estimation result is transmitted to other control components via the sending unit 314 based on the first type of bus;

[0276] The acquisition unit 310 receives a second estimation result from other control components on the first type of bus. The other control components are equipped with a calculation model for calculating the target vehicle parameters. The second estimation result is associated with the first operating data.

[0277] Based on the first estimation result and the received second estimation result, the calculation results of the target vehicle parameters are obtained.

[0278] Furthermore, the calculation results of the target vehicle parameters are transmitted to the sensor system and / or execution system via the transmitting unit 314 based on the second type of bus and / or signal line.

[0279] In the embodiments described above, each embodiment has its own emphasis. For parts not described in detail in a particular embodiment, please refer to the relevant descriptions in other embodiments. Furthermore, in the embodiments of this application, unless otherwise specified or logically conflicting, the terminology and / or descriptions between the embodiments are consistent and can be mutually referenced. Technical features from different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0280] It should be noted that those skilled in the art will recognize that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. This program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0281] The technical solution of this application, in essence, or the part that makes the contribution, or all or part of the technical solution, can be embodied in the form of a software product. The computer program product is stored in a storage medium and includes several instructions to cause a device (which may be a personal computer, server, network device, robot, microcontroller, chip, robot, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

Claims

1. A parameter estimation system, characterized in that, The parameter estimation system includes multiple control components, which are connected to each other via a first type of bus and connected to a sensor system and / or an execution system via a second type of bus. Each control component has a computational model deployed on it, which is used to calculate the parameters of the target vehicle. At least one of the plurality of control components is configured to perform the following operation: The first operating data is obtained through the second type of bus; The first estimation result is transmitted to other control components among the plurality of control components. The first estimation result is obtained by the control component using a local calculation model to perform calculations on the first running data. Receive a second estimation result from the other control components, the second estimation result being associated with the first operating data; Based on the first estimation result and the received second estimation result, the calculation result of the target vehicle parameters is obtained.

2. The system according to claim 1, characterized in that, The at least one control component is also used for: The calculation results of the target vehicle parameters are transmitted to the sensor system and / or the execution system.

3. The system according to claim 1 or 2, characterized in that, The plurality of control components include a first control component and a second control component, wherein the computational model deployed on the first control component is the same as the computational model deployed on the second control component.

4. The system according to claim 1 or 2, characterized in that, The plurality of control components include a first control component and a second control component, wherein the computing model deployed on the first control component is a first computing model, and the computing model deployed on the second control component is a second computing model, and the first computing model and the second computing model are different.

5. The system according to claim 4, characterized in that, The operating conditions applicable to the first calculation model are different from those applicable to the second calculation model.

6. The system according to claim 4, characterized in that, The first calculation model is applicable to the same working conditions as the second calculation model, and the at least one control component is specifically used for: Arbitrate the first estimation result and the received second estimation result based on preset rules to obtain the calculation result of the target vehicle parameters.

7. The system according to claim 1 or 2, characterized in that, The plurality of control components include a first control component and a second control component, wherein the computational model deployed on the first control component is a part of the target computational model, and the computational model deployed on the second control component is another part of the target computational model.

8. The system according to any one of claims 1-7, characterized in that, When vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data for the first cycle, and the at least one control component is further configured to: The first estimation result is obtained by performing calculations on the first running data and at least one historical calculation result of the target vehicle parameters using a local calculation model. The at least one historical calculation result comes from at least one of the other control components, and the historical calculation result is associated with second operating data, which is vehicle operating data for a second cycle, and the second cycle is the previous cycle of the first cycle.

9. The system according to any one of claims 1-8, characterized in that, The first estimation result is carried in a multicast message, which also carries several of the following information: A multicast IP address, wherein the multicast IP address is used to identify the multicast group receiving the first estimation result, and the multicast group includes the other control components; Parameter encoding information, wherein the parameter encoding information is used to indicate that the calculated vehicle parameters are the target vehicle parameters; and, Intermediate process variables, which are used to indicate intermediate parameters involved in the calculation of the first estimation result.

10. The system according to claim 9, characterized in that, When the transmission of vehicle operation data is periodic, the first operation data is the vehicle operation data for the first period, and the multicast message also carries the following information: At least one historical calculation result of the target vehicle parameters, wherein the at least one historical calculation result is involved in the calculation of the first estimation result, the at least one historical calculation result comes from at least one of the other controllers, and the historical calculation result is associated with second operating data, the second operating data being vehicle operating data for a second period, the second period being the period preceding the first period; and, Verification information, which is used to determine whether the first estimation result calculated in the current period is valid.

11. The system according to any one of claims 1-10, characterized in that, The first estimation result or the second estimation result is an intermediate result obtained in the process of calculating the target vehicle parameters or the output result of the target vehicle parameters.

12. The system according to any one of claims 1-11, characterized in that, The first type of bus is an in-vehicle Ethernet bus, and the second type of bus is an in-vehicle bus.

13. The system according to any one of claims 1-12, characterized in that, The plurality of control components includes a first control component, which is directly connected to the sensor system via a signal line; the at least one control component is specifically used for: The first operating data is obtained through the second type of bus and the signal line.

14. A parameter estimation method, characterized in that, The method, applied to a first control unit having a computational model deployed thereon, the computational model being used to calculate target vehicle parameters, includes: First operating data is acquired via signal lines and / or a second type of bus; The first estimation result is transmitted to other control components via a first type of bus. The first estimation result is obtained by the first control component using the calculation model to perform calculations on the first running data. The second estimation result is received from the other control components via the first type of bus. The other control components are equipped with a calculation model for calculating the parameters of the target vehicle. The second estimation result is associated with the first operating data. Based on the first estimation result and the received second estimation result, the calculation result of the target vehicle parameters is obtained.

15. The method according to claim 14, characterized in that, The method further includes: The calculated results of the target vehicle parameters are sent to the sensor system and / or the execution system via the second type of bus and / or the signal line.

16. The method according to claim 14 or 15, characterized in that, The computational model deployed on the first control unit and the computational models deployed on the other control units satisfy any of the following conditions: same; Different; Partially the same; or, Deploy different modules of the same computing model separately.

17. The method according to any one of claims 14-16, characterized in that, When vehicle operation data is transmitted periodically, the first operation data is the vehicle operation data for the first period. The step of obtaining the calculation result of the target vehicle parameters based on the first estimation result and the received second estimation result includes: Calculations are performed based on the first estimation result, the received second estimation result, and the historical calculation results of at least one of the target vehicle parameters to obtain the first estimation result; The at least one historical calculation result comes from at least one of the other control components, and the historical calculation result is associated with second operating data, which is vehicle operating data for a second cycle, and the second cycle is the previous cycle of the first cycle.

18. The method according to any one of claims 14-17, characterized in that, The first estimation result is carried in a multicast message, which also carries several of the following information: A multicast IP address, wherein the multicast IP address is used to identify the multicast group receiving the first estimation result, and the multicast group includes the other control components; Parameter encoding information, wherein the parameter encoding information is used to indicate that the calculated vehicle parameters are the target vehicle parameters; and, Intermediate process variables, which are used to indicate intermediate parameters involved in the calculation of the first estimation result.

19. The method according to claim 18, characterized in that, When the transmission of vehicle operation data is periodic, the first operation data is the vehicle operation data for the first period, and the multicast message also carries the following information: At least one historical calculation result of the target vehicle parameters, wherein the at least one historical calculation result is involved in the calculation of the first estimation result, the at least one historical calculation result comes from at least one of the other controllers, and the historical calculation result is associated with second operating data, the second operating data being vehicle operating data for a second period, the second period being the period preceding the first period; and, Verification information, which is used to determine whether the first estimation result calculated in the current period is valid.

20. The method according to any one of claims 14-19, characterized in that, The first estimation result or the second estimation result is an intermediate result obtained in the process of calculating the target vehicle parameters or the output result of the target vehicle parameters.

21. An apparatus for parameter estimation, characterized in that, The device includes a communication unit and a processing unit, and is used to implement the method according to any one of claims 14-20.

22. A chip for parameter estimation, characterized in that, The chip includes a memory and a processor, the memory storing computer program instructions, and the processor executing the computer program instructions to cause the chip to perform the method as described in any one of claims 14-20.

23. A vehicle, characterized in that, The vehicle includes a parameter estimation system as described in any one of claims 1-13, or an apparatus as described in any one of claims 21, or a chip as described in claim 22.

24. A computer-readable storage medium containing computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the method as described in any one of claims 14-20.

25. A computer program product containing instructions, characterized in that, When the instruction is executed by the parameter estimation device, the computing device performs the method as described in any one of claims 14-20.

Images