Vehicle fusion positioning method and device and vehicle

Abstract

The application provides a vehicle fusion positioning method and device and a vehicle, and relates to the technical field of automobile electronics and intelligent control.The method comprises the following steps: identifying a current driving scene of the vehicle; adjusting a weight parameter of each sensor according to the driving scene; fusing positioning data collected by each sensor based on the weight parameter; and outputting current positioning information of the vehicle.The vehicle fusion positioning method and device and the vehicle provided by the application can fuse the positioning data collected by each sensor based on the weight parameter and output the current positioning information of the vehicle.By adjusting the weight parameter of each sensor, the sensor can be more suitable for the current application scene, and by fusing the positioning data of multiple sensors, the precision and accuracy of the positioning data can be effectively improved.

Description

Technical Field

[0001] This invention relates to the field of automotive electronics and intelligent control technology, and in particular to a vehicle fusion positioning method, device, and vehicle. Background Technology In the rapid development of vehicle electrification and intelligence, vehicle positioning and navigation capabilities have become the core supporting technologies for realizing advanced driver assistance systems (ADAS) and autonomous driving. Their accuracy directly affects driving safety, autonomous driving path planning, obstacle avoidance control, and human-machine interaction experience.

[0002] Current mainstream vehicle positioning methods are often relatively simple and have significant shortcomings in complex scenarios, leading to inaccurate vehicle positioning. This not only affects driving safety but also reduces the driver's driving experience. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a vehicle fusion positioning method, device and vehicle to alleviate the technical problem that the vehicle positioning and navigation capabilities in related technologies are difficult to meet the positioning and navigation needs in complex scenarios.

[0004] In a first aspect, embodiments of the present invention provide a vehicle fusion positioning method applied to a vehicle microcontroller, wherein the vehicle is equipped with a sensor layer, the sensor layer including multiple sensors for vehicle positioning; the method includes: identifying the current driving scenario of the vehicle; adjusting the weight parameters of each sensor according to the driving scenario; wherein the weight parameters are used to characterize the contribution of each sensor in different driving scenarios; fusing the positioning data collected by each sensor based on the weight parameters, and outputting the current positioning information of the vehicle.

[0005] In conjunction with the first aspect, the present invention provides a first possible implementation of the first aspect, wherein the method further includes: during the vehicle's operation, acquiring positioning data collected by each of the sensors, and timestamp information corresponding to the positioning data; performing time synchronization on the positioning data collected by each of the sensors based on the timestamp information to obtain synchronized positioning data corresponding to each of the sensors; and performing feature extraction on each of the synchronized positioning data to obtain a feature vector corresponding to each of the sensors.

[0006] In conjunction with the first possible implementation of the first aspect, this embodiment of the invention provides a second possible implementation of the first aspect, wherein the microcontroller integrates a lightweight AI positioning model; the step of fusing the positioning data collected by each of the sensors based on the weight parameters and outputting the current positioning information of the vehicle includes: inputting the weight parameters of each of the sensors and the feature vector corresponding to each of the sensors into the AI ​​positioning model, and performing weighted fusion of each feature vector based on the weight parameters by the AI ​​positioning model to obtain the current positioning information of the vehicle; wherein the positioning information includes the current real-time position and vehicle attitude of the vehicle.

[0007] In conjunction with the first aspect, this embodiment of the invention provides a third possible implementation of the first aspect, wherein the method further includes: in response to any one of the sensors malfunctioning, reducing the weight parameter of the malfunctioning sensor according to a pre-configured malfunction handling strategy.

[0008] Secondly, embodiments of the present invention also provide a vehicle fusion positioning device applied to a vehicle microcontroller, wherein the vehicle is equipped with a sensor layer, the sensor layer including a plurality of sensors for vehicle positioning; the device includes: an identification module for identifying the current driving scene of the vehicle; an adjustment module for adjusting the weight parameters of each sensor according to the driving scene; wherein the weight parameters are used to characterize the contribution of each sensor under different driving scenes; and a positioning module for fusing the positioning data collected by each sensor based on the weight parameters and outputting the current positioning information of the vehicle.

[0009] Thirdly, embodiments of the present invention also provide a vehicle, the vehicle being configured with a sensor layer and a microcontroller; wherein the microcontroller is configured with the vehicle fusion positioning device described in the second aspect above; the sensor layer includes a plurality of sensors for vehicle positioning; the microcontroller is used to identify the current driving scene of the vehicle; adjust the weight parameters of each sensor according to the driving scene; fuse the positioning data collected by each sensor based on the weight parameters, and output the current positioning information of the vehicle.

[0010] In conjunction with the third aspect, the present invention provides a first possible implementation of the third aspect, wherein the sensors in the above-mentioned sensor layer include at least: a GNSS positioning module, an IMU positioning module, an odometer positioning module, a visual sensor positioning module, and a vehicle network positioning module.

[0011] In conjunction with the third aspect, this embodiment of the invention provides a second possible implementation of the third aspect, wherein the microcontroller integrates a lightweight AI positioning model; the microcontroller is further configured to input the weight parameters and positioning data collected by each of the sensors into the AI ​​positioning model, so that the AI ​​positioning model fuses the positioning data collected by each of the sensors based on the weight parameters and outputs the current positioning information of the vehicle, wherein the positioning information includes the current real-time position and vehicle attitude of the vehicle; wherein the AI ​​positioning model is a lightweight AI model; the AI ​​positioning model includes an input layer, a feature extraction layer, a multimodal fusion layer, a weight layer, and an output layer; wherein the input layer is configured to acquire the positioning data collected by each of the sensors; the feature extraction layer is configured to extract feature vectors of each of the positioning data, the feature vectors including numerical feature vectors and image feature vectors; the weight layer is configured to adjust the weight parameters of each of the sensors according to the driving scenario; the multimodal fusion layer is configured to fuse the numerical feature vectors and image feature vectors included in the feature vectors based on the weight parameters to obtain a fused vector; the output layer is configured to output the current positioning information of the vehicle based on the fused vector.

[0012] In conjunction with the second possible implementation of the third aspect, this embodiment of the invention provides a third possible implementation of the third aspect, wherein the vehicle is further configured with a programmable chip that works in conjunction with the microcontroller; the microcontroller is used to acquire positioning data from each of the sensors; and the programmable chip is responsible for the calculation process of the AI ​​positioning model.

[0013] In conjunction with the second possible implementation of the third aspect, this embodiment of the invention provides a fourth possible implementation of the third aspect, wherein the above-mentioned AI localization model is a lightweight AI model employing a convolutional neural network, a long short-term memory network, and an attention mechanism.

[0014] The embodiments of the present invention bring the following beneficial effects: This invention provides a vehicle fusion positioning method, device, and vehicle, which can identify the current driving scenario of the vehicle; adjust the weight parameters of each sensor according to the driving scenario; fuse the positioning data collected by each sensor based on the weight parameters, and output the current positioning information of the vehicle. By adjusting the weight parameters of each sensor, the weight parameters of sensors that are not applicable to the current driving scenario can be reduced, while the weight parameters of sensors that are applicable to the current driving scenario can be increased, avoiding the influence of sensors that are not applicable to the current driving scenario on the positioning information, thereby making the positioning process more applicable to the current driving scenario. Furthermore, by fusing the positioning data of multiple sensors, the accuracy and precision of the positioning data can be effectively improved.

[0015] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A flowchart of a vehicle fusion positioning method provided in an embodiment of the present invention; Figure 2 A system architecture diagram for vehicle fusion positioning provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of a vehicle fusion positioning device provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Currently, in the field of vehicle positioning, commonly used positioning methods include GNSS (Global Navigation Satellite System) positioning, IMU (Inertial Measurement Unit Positioning) positioning, odometry positioning, and V-SLAM (Visual Simultaneous Localization and Interference) positioning. Mapping (vision-based simultaneous localization and mapping) involves visual positioning and vehicle-to-everything (V2X) positioning. GNSS positioning is susceptible to signal blockage and multipath effects in areas like urban canyons, tunnels, and underground parking garages, leading to positioning drift or even complete failure. IMU positioning suffers from cumulative error, i.e., positioning drift, which increases rapidly over time; without external correction, it will deviate significantly from the true location over extended periods. Odometer positioning is easily affected by wheel slippage and wear, and error accumulation is significant. V-SLAM visual positioning is computationally intensive, highly dependent on lighting and weather conditions, and suffers performance degradation at night or in inclement weather. V2X positioning requires external infrastructure support and has poor universality. Therefore, none of the above-mentioned positioning technologies alone can meet the requirements for high accuracy and robustness, resulting in inaccurate vehicle positioning, which not only affects driving safety but also reduces the driver's experience.

[0021] Based on this, the vehicle fusion positioning method, device and vehicle provided in the embodiments of the present invention can alleviate the technical problem that the vehicle positioning and navigation capabilities in related technologies are difficult to meet the positioning and navigation needs in complex scenarios.

[0022] To facilitate understanding of this embodiment, a detailed description of the vehicle fusion positioning method disclosed in this invention will be provided first. Specifically, the vehicle fusion positioning method provided in this invention is applied to a vehicle's microcontroller. Specifically, this microcontroller is typically integrated into the vehicle's ECU (Electronic Control Unit). Furthermore, the vehicle in this invention is equipped with a sensor layer, which includes multiple sensors for vehicle positioning; specifically, such as... Figure 1 The flowchart shown illustrates a vehicle fusion localization method, which includes the following steps: Step S102: Identify the current driving scenario of the vehicle; Step S104: Adjust the weight parameters of each sensor according to the driving scenario; In this embodiment of the invention, the weighting parameter is used to characterize the contribution of each sensor in different driving scenarios; In practical use, a mapping table between scenarios and weights can be pre-configured in the microcontroller. For example, various driving scenarios and the weight parameters of each sensor in each driving scenario can be pre-configured. When the vehicle is driving, the current driving scenario can be determined based on electronic maps, meteorological environmental data, and parameters such as temperature and humidity. For example, urban canyons, tunnels, underground parking garages, or cloudy, sunny, rainy, snowy, or night driving. Then, the weight parameters of each sensor can be found in the mapping table between scenarios and weights, and then the following step S106 is executed.

[0023] Step S106: Based on the weight parameters, the positioning data collected by each sensor is fused, and the current positioning information of the vehicle is output.

[0024] In this embodiment of the invention, the positioning information includes the vehicle's current real-time location and vehicle attitude.

[0025] In practical use, each of the above sensors can be activated and collect positioning data during vehicle operation. In order to fuse the positioning data collected by each sensor in step S106, in this embodiment of the invention, during vehicle operation, the positioning data collected by each sensor and the timestamp information corresponding to the positioning data can also be obtained; the positioning data collected by each sensor is time-synchronized based on the timestamp information to obtain the synchronized positioning data corresponding to each sensor; and features are extracted from each synchronized positioning data to obtain the feature vector corresponding to each sensor.

[0026] Furthermore, the microcontroller in this embodiment of the invention integrates a lightweight AI positioning model; in step S106 above, when fusing the positioning data collected by each sensor based on the weight parameters, the weight parameters of each sensor and the feature vector corresponding to each sensor can be input into the AI ​​positioning model, and the AI ​​positioning model can perform weighted fusion of each feature vector based on the weight parameters to obtain the current positioning information of the vehicle.

[0027] In practical use, the sensors in the aforementioned sensor layer include at least: a GNSS positioning module, an IMU positioning module, an odometer positioning module, a visual sensor positioning module, and a vehicle-to-everything (V2X) positioning module. In this embodiment of the invention, in response to an anomaly in any sensor, the weight parameters of the malfunctioning sensor can be reduced according to a pre-configured anomaly handling strategy.

[0028] In practical implementation, the vehicle's microcontroller typically has corresponding sensor interfaces to collect multi-source data from each of the aforementioned sensors. Specifically, among these sensors, the GNSS positioning module usually provides absolute position data; the IMU positioning module provides acceleration and angular velocity; the odometer positioning module provides wheel speed and cumulative mileage; the visual sensor positioning module can provide image sequences or feature points; and the vehicle network positioning module provides base station positioning and roadside unit auxiliary information. In practical implementation, the data from each of these positioning modules can be preprocessed, including synchronized positioning data obtained through the aforementioned time synchronization method, as well as filtering and feature extraction, etc., to obtain the feature vector of each sensor.

[0029] Furthermore, the AI ​​positioning model used in this embodiment of the invention is a lightweight AI model; specifically, the AI ​​positioning model includes an input layer, a feature extraction layer, a multimodal fusion layer, a weighting layer, and an output layer; wherein, the input layer is used to acquire positioning data collected by each sensor; the feature extraction layer is used to extract feature vectors for each positioning data, the feature vectors including numerical feature vectors and image feature vectors; the weighting layer is used to adjust the weight parameters of each sensor according to the driving scenario; the multimodal fusion layer is used to fuse the numerical feature vectors and image feature vectors included in the feature vectors based on the weight parameters to obtain a fused vector; and the output layer is used to output the current positioning information of the vehicle based on the fused vector.

[0030] Furthermore, in this embodiment of the invention, the vehicle is also equipped with a programmable chip that works in conjunction with a microcontroller; the microcontroller acquires positioning data from each sensor; and the programmable chip is responsible for the calculation process of the AI ​​positioning model. For example, a microcontroller (MCU) and a programmable gate array (FPGA) can work together, with the MCU responsible for sensor data acquisition and preprocessing, and the FPGA responsible for AI model inference and high-parallel computation. The advantage of this collaborative work is that it can reduce latency and increase robustness. Another example is the collaboration between the MCU and a DSP (Digital Signal Processor), where the DSP processes high-frequency signals, such as positioning data from the IMU positioning module and the odometer positioning module; the MCU is used for fusion and AI inference. This collaborative approach is suitable for autonomous vehicles with extremely high real-time requirements. The specific collaborative method can be set according to actual usage, and this embodiment of the invention does not impose any limitations on it.

[0031] Specifically, for ease of understanding, Figure 2 A system architecture diagram for vehicle fusion localization is shown, including a microcontroller (MCU) and an AI localization module 200 integrated within the microcontroller. Figure 2The document also shows the host 201 and various sensors in the sensor layer, including the aforementioned GNSS positioning module 202, IMU positioning module 203, odometer positioning module 204, vision sensor positioning module 205, and vehicle-to-everything (V2X) positioning module 206, where the V2X positioning module is also referred to as a V2X module. Further, the aforementioned... Figure 2 The AI ​​positioning model in this system is a lightweight AI model employing convolutional neural networks, long short-term memory networks, and attention mechanisms. This host system can be located in a vehicle, such as in a smart cockpit, and can send the positioning information to the interactive screen in the driver's cab for display after acquiring it.

[0032] In practical use, the microcontroller in this embodiment of the invention ensures the alignment of multi-source data through a timestamp synchronization mechanism. Furthermore, the data collected by each sensor can be transmitted via CAN bus, Ethernet, or other means, ensuring that the data delay is less than 1ms. To avoid noise interference, the microcontroller can preprocess the raw data using low-pass filtering and Kalman filtering.

[0033] Furthermore, the aforementioned GNSS positioning module is typically mounted as an antenna in the center of the vehicle roof to ensure maximum satellite signal reception; the IMU positioning module is installed near the vehicle's center of gravity, such as in the center of the chassis, to reduce vibration interference; the odometer sensor of the odometer positioning module is usually integrated into the drive wheel motor or ABS wheel speed sensor; the vision sensor of the vision sensor positioning module is installed on the windshield or roof for forward environmental perception and positioning; and the vehicle-to-everything (V2X) positioning module communicates with the roadside unit via an antenna to obtain base station positioning information.

[0034] Based on the sensor layer composed of the aforementioned positioning modules, comprehensive data acquisition can be achieved. For example, the GNSS positioning module, when providing absolute position data, can provide a global absolute positioning reference, ensuring that the vehicle's global position reference does not drift; the IMU positioning module, when providing acceleration and angular velocity, can leverage its high-frequency sampling advantage to output the vehicle's acceleration and angular velocity data in real time, continuously calculating short-term attitude and relative displacement, unaffected by occlusion environments; the odometer positioning module, through wheel rotation speed signals, accurately calculates the vehicle's mileage and real-time speed, providing a reference for the vehicle's motion; the visual sensor positioning module's visual camera can acquire real-world road images, extract environmental feature points and road semantic information, and construct visual positioning constraints; the vehicle network positioning module enables vehicle-to-infrastructure and vehicle-to-vehicle information interaction, sharing surrounding positioning data and road condition information, filling in blind spots in single-vehicle perception. The data from these various sensors are standardized by the microprocessor (MCU) integrating the AI ​​positioning model, generating a unified format of multi-dimensional feature vectors, which are then incorporated into the AI ​​positioning model.

[0035] Furthermore, the AI ​​positioning model in this embodiment of the invention is a lightweight AI model employing a convolutional neural network (CNN), a long short-term memory (LSTM) network, and an attention mechanism. The CNN model focuses on spatial feature extraction, efficiently identifying key image features and spatial dimension features such as sensor signal waveforms. The LSTM network concentrates on mining temporal features, establishing a temporal correlation model of vehicle motion states, and reconstructing continuous vehicle trajectory and attitude change patterns. The attention mechanism incorporates a dynamic weight adjustment mechanism for AI positioning, intelligently allocating sensor weights according to the actual driving scenario. For example, in areas with weak GNSS signals, such as tunnels, high-rise buildings, or canyons, it automatically lowers the weight parameters of the GNSS positioning module and increases the weight parameters of the IMU positioning module, odometer positioning module, and visual sensor positioning module. This reduces the impact of inaccurate GNSS positioning and makes the positioning information more closely match the characteristics of the scene.

[0036] Furthermore, in this embodiment of the invention, when outputting positioning information, the output layer can rely on a dedicated microprocessor (MCU) for fusion positioning to complete the centralized calculation and integration of all perceived data and AI features, and finally output accurate and stable fusion estimates of the vehicle's real-time position and attitude.

[0037] Furthermore, in order to implement the vehicle fusion positioning method in this embodiment of the invention, the following hardware implementation can be adopted in the vehicle: (1) Microprocessor selection: Typically, in this embodiment of the invention, an automotive-grade microprocessor (MCU) is used as the control core, such as the NXP i.MXRT1170 (dual-core Arm Cortex-M7+M4, 1GHz); the STM32H7 series (Arm Cortex-M7, 480MHz); the TI TMS320F2838x series (DSP + MCU architecture), etc. Its advantages include: low power consumption (<1W), suitable for new energy vehicles; rich peripheral interfaces, such as CAN, LIN, UART, SPI, I2C, etc., facilitating the acquisition of positioning data from multiple sensors; and support for the TinyML inference framework, enabling the running of lightweight AI models.

[0038] (2) Sensor configuration: The GNSS positioning module supports GPS / BeiDou / GLONASS / Galileo positioning technologies and outputs position and speed information; the IMU positioning module typically uses a 6-axis (3-axis accelerometer + 3-axis gyroscope) sensor with a sampling rate of 200Hz; the odometer positioning module is integrated into the wheel speed sensor with a sampling frequency of 100Hz; the visual sensor positioning module uses monocular and binocular cameras with a resolution of 640×480 and a frame rate of 30fps; and the vehicle-to-everything (V2X) positioning module supports 5G NR-V2X or DSRC to obtain roadside positioning assistance information.

[0039] Furthermore, based on the aforementioned hardware implementation, a lightweight design strategy is adopted when deploying the AI ​​positioning model. The input layer includes: latitude and longitude coordinates and velocity from the GNSS positioning module, acceleration and angular velocity from the IMU positioning module, wheel speed from the odometer positioning module, keypoint coordinates and sparse features from the visual sensor positioning module, and base station coordinate correction data from the vehicle-to-everything (V2X) positioning module, etc. In the feature extraction layer, a Convolutional Neural Network (CNN) is used to extract visual features; a Long Short-Term Memory (LSTM) network is used to model temporal dependencies; the multimodal fusion layer further concatenates the numerical features from the input layer with the visual features; the weighting layer uses an attention mechanism to dynamically allocate the weight parameters of each sensor, for example, reducing the weight of the GNSS positioning module and increasing the weight parameters of the IMU positioning module in tunnels. The output layer can output the real-time vehicle position (x, y, z) and vehicle attitude (roll, pitch, yaw).

[0040] Furthermore, when designing a lightweight AI positioning model, the model can first be quantized, such as compressing a 32-bit floating-point number into an 8-bit fixed-point number to reduce computation; then network pruning can be performed, that is, removing redundant neurons to reduce the number of parameters; knowledge distillation can then be performed, with a large model in the cloud guiding the training of a lightweight AI positioning model on the microcontroller MCU; finally, the trained AI positioning model can be deployed on the microcontroller MCU.

[0041] Furthermore, when performing anomaly detection on the above-mentioned multiple sensors, self-encoder residual detection is usually adopted. That is, if the data reconstruction error of a certain sensor is too large, it is considered abnormal. This is usually achieved using statistical methods. For example, if the GNSS positioning module suddenly deviates by more than 3σ, the GNSS positioning module is determined to be abnormal.

[0042] Furthermore, to verify the effectiveness of the vehicle fusion positioning method provided in this embodiment of the invention, simulation experiments can be conducted using Python under laboratory conditions. These simulation experiments typically include the following steps: (1) Constructing simulation scenarios: Simulating vehicle driving on urban roads, GNSS signals are blocked and jump; IMU data has noise and drift; odometer has slippage error and other driving scenarios. (2) Setting up comparative experimental methods: Scheme A: positioning using a single GNSS positioning module; Scheme B: positioning using an IMU positioning module independently; Scheme C: in this embodiment of the invention, the AI ​​positioning model performs a fusion positioning method on positioning data from multiple sensors.

[0043] Typically, to verify the effectiveness of the vehicle fusion positioning method of this invention in practical applications, different road scenarios can be selected as driving scenarios for real-vehicle testing. Furthermore, any vehicle type can be used, and the aforementioned GNSS positioning module, IMU positioning module, odometer positioning module, visual sensor positioning module, and vehicle-to-everything (V2X) positioning module can be configured. Specifically, the driving scenario settings are as follows: (1) Highway (100 km / h): The GNSS signal is stable, verifying the real-time performance at high speed; (2) Urban canyons (densely populated areas with high-rise buildings): severe multipath interference with GNSS signals; (3) Tunnel scenario (2 km in total): GNSS signal is completely ineffective, mainly relying on IMU positioning module, odometer positioning module and visual sensor positioning module; (4) Underground parking lot (low speed 10 km / h): GNSS signal is completely lost, and the signal of the visual sensor positioning module is limited; (5) Rural roads (tree obstruction): GNSS signal is intermittently lost.

[0044] For each of the above driving scenarios, the three comparative experimental schemes were used for comparison, and the results are as follows: (1) Scheme A: Positioning is performed using a single GNSS positioning module; in highway scenarios: error <0.5m; in urban canyons: error increases to 5–10 m, and some locations are lost; in tunnel scenarios and underground parking lots, it is completely ineffective.

[0045] (2) Scheme B: Independent positioning by IMU positioning module and Kalman filtering algorithm; on highways: error 0.3–0.5m; in urban canyons: error 1–3m; in tunnel scenarios, positioning can be maintained for a short time, but the error accumulates rapidly (about 5 m / minute).

[0046] (3) Scheme C: In this embodiment of the invention, the AI ​​positioning model performs a fusion positioning method on positioning data from multiple sensors. In the highway scenario: error <0.2 m; in the urban canyon: error <0.5 m, which is significantly better than Kalman filtering; in the tunnel scenario: error <0.3 m over a 2 km route, with significantly improved stability; in the underground parking lot: error <0.4 m; in the rural road scenario: error <0.3 m.

[0047] Based on the experimental data above, it can be seen that in complex environments where GNSS signals are limited, the vehicle fusion positioning method provided by this embodiment of the invention can maintain good accuracy and can be applied to complex scenarios.

[0048] In summary, the vehicle fusion positioning method provided by this invention can identify the current driving scenario of the vehicle; adjust the weight parameters of each sensor according to the driving scenario; input the weight parameters and the positioning data collected by each sensor into the AI ​​positioning model, so that the AI ​​positioning model fuses the positioning data collected by each sensor based on the weight parameters and outputs the current positioning information of the vehicle. By adjusting the weight parameters of each sensor, the sensor can be made more suitable for the current application scenario, and by fusing the positioning data of multiple sensors, the accuracy and precision of the positioning data can be effectively improved.

[0049] Furthermore, based on the above embodiments, this invention also provides a vehicle fusion positioning device applied to a vehicle's microcontroller, wherein the vehicle is equipped with a sensor layer, the sensor layer including a plurality of sensors for vehicle positioning; such as Figure 3 The diagram shows a structural schematic of a vehicle fusion positioning device, which includes: The identification module 30 is used to identify the current driving scenario of the vehicle; The adjustment module 32 is used to adjust the weight parameters of each sensor according to the driving scenario; wherein the weight parameters are used to characterize the contribution of each sensor under different driving scenarios; The positioning module 34 is used to fuse the positioning data collected by each of the sensors based on the weight parameters and output the current positioning information of the vehicle.

[0050] Furthermore, this embodiment of the invention also provides a vehicle equipped with a sensor layer and a microcontroller; wherein the microcontroller is equipped with the aforementioned vehicle fusion positioning device; wherein, the vehicle in this embodiment of the invention can refer to Figure 2The system architecture diagram shown illustrates that the vehicle's sensor layer includes multiple sensors for vehicle positioning; a microcontroller identifies the vehicle's current driving scenario; the weight parameters of each sensor are adjusted according to the driving scenario; the positioning data collected by each sensor are fused based on the weight parameters, and the vehicle's current positioning information is output.

[0051] Furthermore, in this embodiment of the invention, the sensors in the vehicle's sensor layer include at least: a GNSS positioning module, an IMU positioning module, an odometer positioning module, a visual sensor positioning module, and a vehicle network positioning module.

[0052] Furthermore, the microcontroller in this embodiment of the invention integrates a lightweight AI positioning model; the microcontroller is also used to input the weight parameters and the positioning data collected by each of the sensors into the AI ​​positioning model, so that the AI ​​positioning model fuses the positioning data collected by each of the sensors based on the weight parameters and outputs the current positioning information of the vehicle, wherein the positioning information includes the current real-time position and vehicle attitude of the vehicle; wherein the AI ​​positioning model includes an input layer, a feature extraction layer, a multimodal fusion layer, a weight layer and an output layer; wherein the input layer is used to acquire the positioning data collected by each of the sensors; the feature extraction layer is used to extract the feature vector of each positioning data, the feature vector including numerical feature vector and image feature vector; the weight layer is used to adjust the weight parameters of each sensor according to the driving scenario; the multimodal fusion layer is used to fuse the numerical feature vector and the image feature vector included in the feature vector based on the weight parameters to obtain a fused vector; the output layer is used to output the current positioning information of the vehicle based on the fused vector.

[0053] Furthermore, the vehicle in this embodiment of the invention is also equipped with a programmable chip that works in conjunction with the microcontroller; the microcontroller is used to acquire the positioning data of each of the sensors; and the programmable chip is responsible for the calculation process of the AI ​​positioning model.

[0054] Furthermore, the AI ​​localization model in this embodiment of the invention is a lightweight AI model employing a convolutional neural network, a long short-term memory network, and an attention mechanism.

[0055] The device and vehicle provided in the embodiments of the present invention have the same technical features as the vehicle fusion positioning method provided in the above embodiments, so they can also solve the same technical problems and achieve the same technical effects.

[0056] Furthermore, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above method.

[0057] This invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the above-described method.

[0058] Furthermore, embodiments of the present invention also provide a schematic diagram of the structure of an electronic device, such as... Figure 4 The diagram shows the structure of the electronic device, which includes a processor 41 and a memory 40. The memory 40 stores computer-executable instructions that can be executed by the processor 41, and the processor 41 executes the computer-executable instructions to implement the above-described method.

[0059] exist Figure 4 In the illustrated embodiment, the electronic device further includes a bus 42 and a communication interface 43, wherein the processor 41, the communication interface 43, and the memory 40 are connected via the bus 42.

[0060] The memory 40 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 43 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 42 may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 42 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0061] Processor 41 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 41 or by software instructions. Processor 41 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in the memory, and the processor 41 reads the information in the memory and uses its hardware to complete the aforementioned method.

[0062] The vehicle fusion positioning method, device, and vehicle computer program product provided in this embodiment of the invention include a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0063] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the device and vehicle described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0064] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0065] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0066] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0067] Finally, it should be noted that the above embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A vehicle fusion positioning method, characterized in that, A microcontroller for use in a vehicle, the vehicle being equipped with a sensor layer including a plurality of sensors for vehicle positioning; the method includes: Identify the current driving scenario of the vehicle; The weight parameters of each sensor are adjusted according to the driving scenario; wherein the weight parameters are used to characterize the contribution of each sensor under different driving scenarios. The positioning data collected by each sensor is fused based on the weight parameters, and the current positioning information of the vehicle is output.

2. The method according to claim 1, characterized in that, The method further includes: During the vehicle's operation, the positioning data collected by each sensor and the timestamp information corresponding to the positioning data are acquired. Based on the timestamp information, the positioning data collected by each sensor is synchronized in time to obtain the synchronized positioning data corresponding to each sensor; Feature extraction is performed on each of the synchronous positioning data to obtain the feature vector corresponding to each sensor.

3. The method according to claim 2, characterized in that, The microcontroller integrates a lightweight AI positioning model; The step of fusing the positioning data collected by each sensor based on the weight parameters and outputting the current positioning information of the vehicle includes: The weight parameters of each sensor and the feature vector corresponding to each sensor are input into the AI ​​positioning model. The AI ​​positioning model performs weighted fusion on each feature vector based on the weight parameters to obtain the current positioning information of the vehicle. The positioning information includes the current real-time position and vehicle attitude of the vehicle.

4. The method according to claim 1, characterized in that, The method further includes: In response to any of the sensors malfunctioning, the weight parameters of the malfunctioning sensor are reduced according to a pre-configured malfunction handling strategy.

5. A vehicle fusion positioning device, characterized in that, A microcontroller for use in a vehicle, the vehicle being equipped with a sensor layer including a plurality of sensors for vehicle positioning; the device includes: The identification module is used to identify the current driving scenario of the vehicle; An adjustment module is used to adjust the weight parameters of each sensor according to the driving scenario; wherein the weight parameters are used to characterize the contribution of each sensor under different driving scenarios; The positioning module is used to fuse the positioning data collected by each of the sensors based on the weight parameters and output the current positioning information of the vehicle.

6. A vehicle, characterized in that, The vehicle is equipped with a sensor layer and a microcontroller; wherein the microcontroller is equipped with the vehicle fusion positioning device as described in claim 5; The sensor layer includes multiple sensors for vehicle positioning; The microcontroller is used to identify the current driving scenario of the vehicle; adjust the weight parameters of each sensor according to the driving scenario; fuse the positioning data collected by each sensor based on the weight parameters, and output the current positioning information of the vehicle.

7. The vehicle according to claim 6, characterized in that, The sensors in the sensor layer include at least: a GNSS positioning module, an IMU positioning module, an odometer positioning module, a visual sensor positioning module, and a vehicle network positioning module.

8. The vehicle according to claim 6, characterized in that, The microcontroller integrates a lightweight AI positioning model; The microcontroller is further configured to input the weight parameters and the positioning data collected by each of the sensors into the AI ​​positioning model, so that the AI ​​positioning model fuses the positioning data collected by each of the sensors based on the weight parameters and outputs the current positioning information of the vehicle, wherein the positioning information includes the current real-time position and vehicle attitude of the vehicle. The AI ​​localization model includes an input layer, a feature extraction layer, a multimodal fusion layer, a weight layer, and an output layer. The input layer is used to acquire the positioning data collected by each of the sensors; The feature extraction layer is used to extract feature vectors for each of the positioning data, and the feature vectors include numerical feature vectors and image feature vectors; The weighting layer is used to adjust the weight parameters of each sensor according to the driving scenario; The multimodal fusion layer is used to fuse the numerical feature vector and the image feature vector included in the feature vector based on the weight parameters to obtain a fused vector; The output layer is used to output the current location information of the vehicle based on the fusion vector.

9. The vehicle according to claim 8, characterized in that, The vehicle is also equipped with a programmable chip that works in conjunction with the microcontroller; The microcontroller is used to acquire the positioning data of each of the sensors; The programmable chip is responsible for the calculation process of the AI ​​positioning model.

10. The vehicle according to claim 8, characterized in that, The AI ​​localization model is a lightweight AI model that employs convolutional neural networks, long short-term memory networks, and attention mechanisms.

Images