Method for determining motion information, electronic device, and vehicle

Abstract

This application provides a method for determining motion information, an electronic device, and a vehicle. The method involves acquiring first motion information of the vehicle, which is calculated based on first raw data determined by the vehicle's inertial navigation unit. The first motion information includes first position information and first attitude information. The method further involves acquiring reference navigation data and determining reference attitude information corresponding to the first position information based on the reference navigation data. It then determines attitude deviation information between the first attitude information and the reference attitude information. Based on the attitude deviation information, it determines error parameters for the inertial navigation unit. Finally, the method acquires second motion information of the vehicle, which is calculated based on corrected second raw data. The corrected second raw data is obtained by correcting the second raw data determined by the inertial navigation unit using the error parameters. The second motion information includes second position information and second attitude information. This approach allows for more accurate determination of the vehicle's motion information.

Description

Technical Field

[0001] This application relates to the field of navigation technology, and in particular to a method for determining motion information, an electronic device, and a vehicle. Background Technology

[0002] With the development of navigation technology, the requirements for navigation accuracy are becoming increasingly stringent. In related technologies, the vehicle's motion information can be determined using its inertial measurement unit (IMU), and navigation can then be performed based on this motion information.

[0003] However, the inventors' long-term research revealed that the determination of vehicle motion information through inertial navigation units can be inaccurate. Summary of the Invention

[0004] This application provides a method for determining motion information, an electronic device, and a vehicle, aiming to improve the technical problem that the motion information of a vehicle determined by an inertial navigation unit is not accurate enough, thereby improving the accuracy of determining the motion information of a vehicle by an inertial navigation unit.

[0005] In a first aspect, embodiments of this application propose a method for determining motion information, comprising: acquiring first motion information of a vehicle, the first motion information being calculated based on first raw data determined by the vehicle's inertial navigation unit, the first motion information including first position information and first attitude information; acquiring reference navigation data, and determining reference attitude information corresponding to the first position information based on the reference navigation data; determining attitude deviation information between the first attitude information and the reference attitude information; determining error parameters of the inertial navigation unit based on the attitude deviation information; and acquiring second motion information of the vehicle, the second motion information being calculated based on corrected second raw data, the corrected second raw data being obtained by correcting the second raw data determined by the inertial navigation unit based on the error parameters, the second raw data being data determined after the first raw data, and the second motion information being information determined after the first motion information, the second motion information including second position information and second attitude information.

[0006] In one possible implementation, the first attitude information includes at least one of a first heading angle, a first pitch angle, and a first roll angle; the reference attitude information includes at least one of a reference heading angle, a reference pitch angle, and a reference roll angle; the attitude deviation information includes at least one of a heading angle deviation, a pitch angle deviation, and a roll angle deviation; and determining the attitude deviation information between the first attitude information and the reference attitude information includes at least one of the following: determining the heading angle deviation between the first heading angle and the reference heading angle; or, determining the pitch angle deviation between the first pitch angle and the reference pitch angle; or, determining the roll angle deviation between the first roll angle and the reference roll angle.

[0007] In one possible implementation, the error parameters include at least one of gyroscope zero bias and accelerometer zero bias. Determining the error parameters of the inertial navigation unit based on attitude deviation information includes: determining at least one of gyroscope zero bias and accelerometer zero bias of the inertial navigation unit based on attitude deviation information.

[0008] In one possible implementation, determining the error parameters of the inertial navigation unit based on attitude deviation information includes: determining a state quantity matrix based on attitude deviation information; acquiring observation information, including position deviation, attitude deviation information, and error parameters, wherein the position deviation is determined based on the difference between first position information and reference position information, and the reference position information is determined based on reference navigation data; determining an observation matrix based on the observation information; acquiring a mapping matrix, which represents the mapping between the state quantity matrix and the observation matrix; and determining the error parameters of the inertial navigation unit based on the state quantity matrix, the observation matrix, and the mapping matrix.

[0009] In one possible implementation, the error parameters of the inertial navigation unit are determined based on the state matrix, the observation matrix, and the mapping matrix, including: obtaining a matrix function, which represents the sum of the matrix product of the observation matrix and the target observation noise, the matrix product including the product of the mapping matrix and the state matrix; and substituting the state matrix, the observation matrix, and the mapping matrix into the matrix function to solve for the error parameters of the inertial navigation unit.

[0010] In one possible implementation, the method further includes: obtaining the current temperature and a mapping relationship, the mapping relationship being used to represent the relationship between temperature and observation noise; and determining the target observation noise corresponding to the current temperature based on the mapping relationship.

[0011] In one possible implementation, the first motion information further includes first velocity information, and the observation information further includes velocity deviation, which is determined based on the difference between the first velocity information and the reference velocity information. Determining the observation matrix based on the observation information includes: determining the observation row vector based on the observation information, wherein the observation row vector includes a vector of position deviation, a vector of velocity deviation, a vector of attitude deviation information, and an error parameter vector; and transposing the observation row vector to obtain the observation matrix.

[0012] In one possible implementation, the reference navigation data includes digital map data and vehicle-road cooperative data sent from the vehicle-road cooperative unit. Determining the reference attitude information corresponding to the first position information based on the reference navigation data includes: determining the reference attitude information corresponding to the first position information based on at least one of the digital map data and the vehicle-road cooperative data.

[0013] Secondly, embodiments of this application propose a motion information determination device, comprising: an acquisition module for acquiring first motion information of a vehicle, the first motion information being calculated based on first raw data determined by the vehicle's inertial navigation unit, the first motion information including first position information and first attitude information; acquiring reference navigation data and determining reference attitude information corresponding to the first position information based on the reference navigation data; a deviation determination module for determining attitude deviation information between the first attitude information and the reference attitude information; a calibration module for determining error parameters of the inertial navigation unit based on the attitude deviation information; and a motion information determination module for acquiring second motion information of the vehicle, the second motion information being calculated based on corrected second raw data, the corrected second raw data being obtained by correcting the second raw data determined by the inertial navigation unit based on the error parameters, the second raw data being data determined after the first raw data, and the second motion information being information determined after the first motion information, the second motion information including second position information and second attitude information.

[0014] Thirdly, embodiments of this application propose an electronic device, including a processor and a memory, wherein: the memory is used to store computer programs; and the processor is used to execute the programs stored in the memory to implement the method of the first aspect.

[0015] Fourthly, embodiments of this application propose a vehicle that includes the electronic equipment of the third aspect.

[0016] Fifthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method of the first aspect. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for determining motion information according to an embodiment of this application.

[0018] Figure 2 This is a flowchart illustrating a method for determining motion information according to another embodiment of this application.

[0019] Figure 3 This is a flowchart illustrating a method for determining motion information according to another embodiment of this application.

[0020] Figure 4 This is a structural block diagram of a motion information determination device according to an embodiment of this application.

[0021] Figure 5 This is a structural diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0022] To make the technical problems, technical solutions, and beneficial effects solved by this application clearer, the following detailed description is provided in conjunction with embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0023] In some scenarios, such as long tunnels, Global Navigation Satellite System (GNSS) signals may be lost. Motion information can be determined using a combination of Pure Inertial Navigation System (INS) and odometry (OD). This method relies entirely on the Inertial Measurement Unit (IMU) for dead reckoning. However, its drawback is that IMU errors (especially zero bias) accumulate over time, causing significant drift in positioning results and failing to meet the accuracy requirements of long-distance tunnels. Although IMU error parameters are calibrated at the factory, these parameters may change with use, leading to inaccuracies in the vehicle's motion estimation by the IMU.

[0024] Another possible implementation is to determine motion information using Simultaneous Localization and Mapping (SLAM) based on vision and LiDAR. This method utilizes environmental features for localization and does not rely on GNSS. However, its drawbacks include sparse and repetitive environmental features within tunnels, as well as drastic changes in lighting, which can easily lead to feature matching failures, tracking loss, and insufficient reliability. This can result in inaccurate determination of vehicle motion information.

[0025] Another possible implementation is a matching localization method based on prior maps: this method matches real-time sensor data with a high-precision map to correct the position. When the accumulated error of the IMU is large, the calculated vehicle pose differs too much from the actual pose, which can lead to mismatches in the matching algorithm (e.g., matching the oncoming lane or the wrong slope). Once a mismatch occurs, the system will receive an incorrect correction value, causing the localization to fail completely and be difficult to recover from.

[0026] In view of this, embodiments of this application provide a method for determining motion information, an electronic device, and a vehicle. The method involves acquiring first motion information of the vehicle, which is calculated based on first raw data determined by the vehicle's inertial navigation unit. The first motion information includes first position information and first attitude information. The method further involves acquiring reference navigation data and determining reference attitude information corresponding to the first position information based on the reference navigation data; determining attitude deviation information between the first attitude information and the reference attitude information; determining error parameters of the inertial navigation unit based on the attitude deviation information; and acquiring second motion information of the vehicle, which is calculated based on corrected second raw data. The second original data, determined by the inertial navigation unit, is obtained by correcting the error parameters. The second original data is the data determined after the first original data, and the second motion information is the information determined after the first motion information. The second motion information includes second position information and second attitude information. In this way, the attitude deviation information between the estimated first attitude information and the reference attitude information determined by the reference navigation data can be used to determine the error parameters of the inertial navigation unit. Then, the error parameters are used to correct the original data output by the inertial navigation unit, which makes the output original data more accurate. Therefore, when the motion information is estimated using the corrected original data, the determined motion information of the vehicle can be more accurate.

[0027] The following section provides a detailed explanation of the methods for determining motion information.

[0028] Please see Figure 1 , Figure 1 This is a flowchart illustrating a method for determining motion information according to an embodiment of this application. Figure 1 The method shown can be executed by an electronic device, which may include a terminal or a server. The terminal can be a smartphone, tablet, laptop, desktop computer, smart TV, smart home device, in-vehicle terminal, etc., without specific limitations. The server can be a standalone physical server, a server cluster consisting of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. Figure 1 The methods shown may include: S110. Obtain the first motion information of the vehicle. The first motion information is calculated based on the first raw data determined by the vehicle's inertial navigation unit. The first motion information includes first position information and first attitude information.

[0029] The inertial navigation unit (INS) is a sensor component that integrates an accelerometer, a gyroscope, and in some models, a magnetometer. It measures the linear acceleration, angular rate (angular velocity), and (optionally) magnetic field strength of the vehicle in real time, and outputs raw data on the vehicle's motion state. It is a hardware unit of the inertial navigation system (INS). The INS does not rely on external signals (such as GPS or vision) and can independently provide short-term motion state information of the vehicle. In this embodiment, the first motion information of the vehicle can be estimated (or calculated) based on the first raw data output by the vehicle's INS.

[0030] In this embodiment, the inertial navigation unit can be activated to determine the vehicle's first motion information in a preset scenario, such as entering a long tunnel or a scenario with severe interference, which would cause GNSS signal loss or other navigation capabilities such as lidar failure. The first motion information includes first position information and first attitude information, where the position information represents the vehicle's position and the attitude information represents the vehicle's attitude or pose.

[0031] S120. Acquire reference navigation data and determine the reference attitude information corresponding to the first position information based on the reference navigation data.

[0032] In this embodiment, the reference navigation data can be from a more reliable data source, such as digital map data or vehicle-to-infrastructure (V2I) data. The digital map data can be high-precision map data. The V2I data can be data sent by the V2I unit. In this embodiment, the reference attitude information can be more accurate attitude information than the first attitude information. The reference attitude information in this embodiment can be considered as the true attitude information.

[0033] S130. Determine the attitude deviation information between the first attitude information and the reference attitude information.

[0034] In this embodiment, the attitude deviation information is used to represent the deviation between the first attitude information and the reference attitude information.

[0035] S140. Based on attitude deviation information, determine the error parameters of the inertial navigation unit.

[0036] In this embodiment, the error parameter can refer to the parameter that causes inaccuracy in the original data output by the inertial navigation unit. In this embodiment, the error parameter can represent the deterministic error of the first original data. In this embodiment, since the attitude deviation information is the deviation between the first attitude information and the reference attitude information, and the reference attitude information can be considered the true attitude information, the error parameter of the inertial navigation unit can be determined through the attitude deviation information. IMU error parameters mainly include deterministic errors, which directly cause the integration results to diverge. These can be mainly divided into: accelerometer bias, gyroscope bias, accelerometer scale factor error, gyroscope scale factor error, and installation misalignment angle / non-orthogonality error.

[0037] S150. Obtain the second motion information of the vehicle. The second motion information is calculated based on the corrected second original data. The corrected second original data is obtained by correcting the second original data determined by the inertial navigation unit based on the error parameters. The second original data is the data determined after the first original data. The second motion information is the information determined after the first motion information. The second motion information includes second position information and second attitude information.

[0038] In this embodiment, when determining motion information after the first motion information, the second original data determined by the inertial navigation unit can be corrected based on the error parameter, and then the second motion information can be determined using the corrected second original data.

[0039] In this embodiment, the first motion information may be determined based on the first raw data output by the inertial navigation unit at a first moment, and the second motion information may be determined based on the second raw data output by the inertial navigation unit after correction at a second moment, wherein the first moment is later than the second moment.

[0040] In this embodiment, by acquiring the vehicle's first motion information, which is calculated based on first raw data determined by the vehicle's inertial navigation unit and includes first position information and first attitude information; acquiring reference navigation data and determining reference attitude information corresponding to the first position information based on the reference navigation data; determining attitude deviation information between the first attitude information and the reference attitude information; determining the error parameter of the inertial navigation unit based on the attitude deviation information; and acquiring the vehicle's second motion information, which is calculated based on corrected second raw data. The corrected second raw data is obtained by correcting the second raw data determined by the inertial navigation unit based on the error parameter. The second raw data is data determined after the first raw data, and the second motion information is information determined after the first motion information. The second motion information includes second position information and second attitude information. In this way, the attitude deviation information between the estimated first attitude information and the reference attitude information determined by the reference navigation data can be used to determine the error parameter of the inertial navigation unit, and then the error parameter can be used to correct the raw data output by the inertial navigation unit. This makes the output raw data more accurate, and when the motion information is estimated using the corrected raw data, the determined vehicle motion information can be more accurate.

[0041] In one possible implementation, the first motion information may further include first speed information, and the second motion information may further include second speed information, whereby the speed information is used to represent the vehicle's speed. This enriches the motion information and thereby improves the accuracy of navigation.

[0042] In one possible implementation, the reference navigation data includes digital map data and vehicle-to-infrastructure (V2I) data transmitted from the V2I unit. Based on the reference navigation data, reference attitude information corresponding to the first position information is determined, including: The reference attitude information corresponding to the first location information is determined based on at least one of digital map data and vehicle-road cooperative data.

[0043] For example, taking digital map data, based on the first location information, the map features corresponding to the first location information can be determined from the digital map data. Then, the reference attitude information can be determined based on the map features. For example, if the first location information is uphill and a left turn, the reference attitude information of the vehicle can be determined accordingly. Taking vehicle-road cooperative data as an example, after the vehicle-road cooperative unit captures the image of the vehicle, it determines the reference attitude information of the vehicle based on the image and then sends it to the electronic device.

[0044] In this embodiment, the reference attitude information corresponding to the first position information can be determined based on at least one of digital map data and vehicle-road cooperative data. In this way, even when GNSS signals cannot be received, the reference attitude information can be determined using highly reliable reference navigation data, and then the error parameters can be determined. This can improve the applicable scenarios for determining error parameters, and thus improve the applicable scenarios for determining motion information.

[0045] In one possible implementation, the first attitude information includes at least one of a first heading angle, a first pitch angle, and a first roll angle; the reference attitude information includes at least one of a reference heading angle, a reference pitch angle, and a reference roll angle; the attitude deviation information includes at least one of a heading angle deviation, a pitch angle deviation, and a roll angle deviation; and the attitude deviation information between the first attitude information and the reference attitude information includes at least one of the following: The method involves determining the yaw angle deviation between the first yaw angle and the reference yaw angle; or, determining the pitch angle deviation between the first pitch angle and the reference pitch angle; or, determining the roll angle deviation between the first roll angle and the reference roll angle. The yaw angle represents the angle between the direction of travel and the reference direction (usually due north), describing the direction of travel. The pitch angle represents the forward and backward roll angle, describing pitching up or down, such as a vehicle climbing or descending a slope. The roll angle represents the left and right roll angle, describing roll, such as a vehicle cornering. Optionally, the second attitude information in this embodiment may include at least one of the second yaw angle, the second pitch angle, and the second roll angle.

[0046] In this embodiment, at least one of the heading angle deviation, pitch angle deviation, and roll angle deviation can be used to determine the error parameters of the inertial navigation unit.

[0047] In this embodiment, by determining the heading angle deviation between the first heading angle and the reference heading angle; or, determining the pitch angle deviation between the first pitch angle and the reference pitch angle; or, determining the roll angle deviation between the first roll angle and the reference roll angle, the deviations of attitude angles in different dimensions can be obtained to determine the error parameters of the inertial measurement unit. This can further improve the accuracy of the correction of the original data determined by the inertial navigation unit, thereby further improving the accuracy of determining the vehicle's motion information.

[0048] In one possible implementation, the error parameters include at least one of gyroscope zero bias and accelerometer zero bias. Based on attitude deviation information, the error parameters of the inertial navigation unit are determined, including: Based on attitude deviation information, determine at least one of the gyroscope zero bias and accelerometer zero bias of the inertial navigation unit.

[0049] Among these, the gyroscope measures the angular rate around the three-dimensional axes, which is the core input for attitude calculation. Gyroscope bias, a systematic error of the gyroscope, is a constant output when there is no motion, leading to long-term attitude angle drift. Specific force (including gravity and acceleration) is measured and is the input for velocity and position calculation. Accelerometer bias, a systematic error of the accelerometer, is a constant output when there is no motion, leading to the accumulation of velocity and position integration errors.

[0050] In this embodiment, at least one of the gyroscope zero bias and accelerometer zero bias of the inertial navigation unit can be determined based on attitude deviation information. Determining the gyroscope zero bias improves the accuracy of vehicle attitude determination, and the gyroscope zero bias can be used for correction of the second raw data, resulting in more accurate second attitude information of the determined second motion information. Similarly, determining the accelerometer zero bias improves vehicle position determination, and the accelerometer zero bias can be used for correction of the second raw data, resulting in more accurate second position information of the determined second motion information. When the second motion information includes second velocity information, the accelerometer zero bias is used for correction, further improving the accuracy of the determined second velocity information.

[0051] The following section explains how to determine the error parameters of the inertial navigation unit.

[0052] In one possible implementation, the error parameters of the inertial navigation unit are determined based on attitude deviation information, including: The state quantity matrix is ​​determined based on attitude deviation information; observation information is acquired, including position deviation, attitude deviation information, and error parameters. The position deviation is determined based on the difference between the first position information and the reference position information, and the reference position information is determined based on the reference navigation data; the observation matrix is ​​determined based on the observation information; a mapping matrix is ​​acquired, which is used to represent the mapping between the state quantity matrix and the observation matrix; and the error parameters of the inertial navigation unit are determined based on the state quantity matrix, the observation matrix, and the mapping matrix.

[0053] In this embodiment, determining the state quantity matrix based on attitude deviation information can be achieved by determining state quantity row vectors based on attitude deviation information and then transposing these row vectors to obtain the state quantity matrix. The state quantity row vectors in this embodiment can include vectors representing attitude deviation information. Similarly, determining the observation matrix based on observation information can be achieved by determining observation row vectors based on observation information and then transposing these row vectors to obtain the observation matrix. The observation row vectors in this embodiment can include vectors representing position deviations, vectors representing attitude deviation information, and vectors representing error parameters. The vector of error parameters in the observation matrix can be a vector to be solved.

[0054] The determination of the reference location information is explained below.

[0055] In this embodiment, the reference position information for the vehicle at the first moment can be determined based on the reference navigation data, since the vehicle may be located at the position indicated by the first position information at the first moment. For example, using digital map data, the vehicle's position on the digital map at the first moment can be used as the reference position information; using vehicle-to-infrastructure (V2I) data, the reference position information for the vehicle at the first moment can be the V2I information indicated in the V2I data sent by the V2I unit. The mapping matrix in this embodiment can be predetermined, for example, determined experimentally or trained using an artificial intelligence (AI) model. The position deviation in this embodiment can be the difference between the first position information and the reference position information.

[0056] For example, taking the mapping matrix obtained from training an artificial intelligence model as an example, sample data can be obtained to train the AI ​​model. This sample data may include a state sample matrix, an observation sample matrix, and error sample parameters. The mapping matrix is ​​the matrix to be trained. Then, the state sample matrix and the observation sample matrix are used as inputs to the model. The model makes predictions based on the state sample matrix, the observation sample matrix, and the mapping matrix to obtain error prediction parameters. Then, the loss value is calculated using the error prediction parameters and the error sample parameters. The mapping matrix is ​​then updated based on the loss value until the loss value meets the training termination condition. This training termination condition may be, for example, less than a loss value threshold and tending to stabilize.

[0057] In this embodiment, the state quantity matrix is ​​determined based on attitude deviation information; observation information is acquired, including position deviation, attitude deviation information, and error parameters. The position deviation is determined based on the difference between first position information and reference position information, and the reference position information is determined based on reference navigation data; an observation matrix is ​​determined based on the observation information; a mapping matrix is ​​acquired, which represents the mapping between the state quantity matrix and the observation matrix; and the error parameters of the inertial navigation unit are determined based on the state quantity matrix, the observation matrix, and the mapping matrix. In this way, the error parameters can be determined mathematically, thereby improving the interpretability of the determined error parameters and thus improving the interpretability of the determined motion information.

[0058] In another possible implementation, attitude deviation information can be input into a large language model to obtain the error parameters of the inertial navigation unit that generates the output of the large language model.

[0059] In one possible implementation, the error parameters of the inertial navigation unit are determined based on the state matrix, the observation matrix, and the mapping matrix, including: Obtain the matrix function, which represents the sum of the result of the matrix product of the observation matrix and the target observation noise. The result of the matrix product includes the product between the mapping matrix and the state matrix. Substitute the state matrix, the observation matrix, and the mapping matrix into the matrix function to solve for the error parameters of the inertial navigation unit.

[0060] In this embodiment, the target observation noise can be a predetermined observation noise. Observation noise can characterize the superposition of random factors not covered by the system model.

[0061] In this embodiment, a matrix function is obtained, which represents the sum of the matrix product of the observation matrix and the target observation noise. The matrix product includes the product of the mapping matrix and the state matrix. The state matrix, observation matrix, and mapping matrix are substituted into the matrix function for solution to obtain the error parameters of the inertial navigation unit. In this way, the observation noise is taken into account when determining the error parameters, which makes the determined error parameters more accurate and helps to improve the accuracy of the determined motion information.

[0062] In another possible implementation, the matrix function can also be set as the product of the mapping matrix and the state matrix. This can reduce the computing resources required to determine the error parameters and thus reduce the computing resources required to determine the motion information.

[0063] In one possible implementation, the method also includes: Obtain the current temperature and the mapping relationship, which represents the relationship between temperature and observation noise; based on the mapping relationship, determine the target observation noise corresponding to the current temperature.

[0064] In this embodiment, the inventors have discovered through long-term research that observation noise changes with temperature. Therefore, this embodiment can determine the target observation noise corresponding to the current temperature based on the mapping relationship between temperature and observation noise.

[0065] In this embodiment, by obtaining the current temperature and the mapping relationship, the mapping relationship is used to represent the relationship between temperature and observation noise. Based on the mapping relationship, the target observation noise corresponding to the current temperature is determined. This makes the determined target observation noise more accurate, which is beneficial to the more accurate determination of error parameters and the improvement of the accuracy of the determined motion information.

[0066] In another possible implementation, the target observation noise can also be a fixed value, that is, independent of temperature. This can improve the efficiency of determining error parameters, and thus improve the efficiency of determining motion information.

[0067] In one possible implementation, the first motion information further includes first velocity information, and the observation information further includes velocity deviation, which is determined based on the difference between the first velocity information and the reference velocity information. The observation matrix is ​​determined based on the observation information, including: The observation row vector is determined based on the observation information. The observation row vector includes the vector of position deviation, the vector of velocity deviation, the vector of attitude deviation information, and the vector of error parameters. The observation row vector is transposed to obtain the observation matrix.

[0068] In this embodiment, the reference speed information can be based on the speed information detected by the vehicle's wheel speed sensors, or it can be the reference speed information of the vehicle determined by the vehicle-to-infrastructure (V2I) unit and sent to the electronic device. The speed deviation can be the difference between the first speed information and the reference speed information. Optionally, the reference speed information in this embodiment can be the reference speed information at a first moment.

[0069] In this embodiment, the first motion information also includes first velocity information, and the observation information also includes velocity deviation, which is determined based on the difference between the first velocity information and the reference velocity information. An observation row vector is determined based on the observation information, which includes a vector of position deviation, a vector of velocity deviation, a vector of attitude deviation information, and an error parameter vector. The observation row vector is transposed to obtain the observation matrix. In other words, the observation matrix is ​​also determined in conjunction with the velocity deviation, which helps to improve the accuracy of the determined error parameters, and thus helps to improve the accuracy of the determined motion information.

[0070] In another possible implementation, the observation matrix can be determined without velocity deviation, which can reduce the required computing resources.

[0071] To facilitate understanding, another embodiment will be described below in conjunction with the above embodiments.

[0072] Please see Figure 2 , Figure 2 This is a flowchart illustrating a method for determining motion information according to another embodiment of this application. Figure 2 The method shown can be performed by an electronic device, such as Figure 2 The methods shown may include: S210. Obtain the vehicle's first motion information, which is calculated based on the first raw data determined by the vehicle's inertial navigation unit. The first motion information includes first position information, first speed information, first heading angle, first pitch angle, and first roll angle.

[0073] S220. Determine the reference attitude information corresponding to the first position information based on the reference navigation data. The reference attitude information includes the reference heading angle, reference pitch angle and reference roll angle.

[0074] S230. Determine the heading angle deviation between the first heading angle and the reference heading angle, determine the pitch angle deviation between the first pitch angle and the reference pitch angle, and determine the roll angle deviation between the first roll angle and the reference roll angle.

[0075] S240. Determine the state variable row vector based on the heading angle deviation, pitch angle deviation, and roll angle deviation, and transpose the state variable row vector to obtain the state variable matrix.

[0076] S250, determine the position deviation based on the first position information and the reference position information, and determine the speed deviation based on the first speed information and the reference speed information, wherein the reference position information is determined based on reference navigation data, and the speed deviation is determined based on the difference between the first speed information and the reference speed information.

[0077] S260. Based on position deviation, velocity deviation, heading angle deviation, pitch angle deviation, roll angle deviation, gyroscope zero bias, and accelerometer zero bias, determine the observation row vector, and transpose the observation row vector to obtain the observation matrix.

[0078] S270. Obtain the matrix function. The matrix function is used to represent the sum of the result of the matrix product of the observation matrix and the target observation noise. The matrix product result includes the product result between the mapping matrix and the state variable matrix. The target observation noise is determined based on the current temperature and the mapping relationship. The mapping relationship is used to represent the relationship between temperature and observation noise.

[0079] S280. Substitute the state matrix, observation matrix, and mapping matrix into the matrix function to solve for the zero bias of the gyroscope and the zero bias of the accelerometer of the inertial navigation unit.

[0080] S290. Obtain the second motion information of the vehicle. The second motion information is calculated based on the corrected second original data. The corrected second original data is obtained by correcting the second original data determined by the inertial navigation unit based on the gyroscope zero bias and accelerometer zero bias. The second original data is the data determined after the first original data. The second motion information is the information determined after the first motion information. The second motion information includes second position information, second speed information, second heading angle, second pitch angle and second roll angle.

[0081] This embodiment can be referred to the description of the above embodiment, and will not be repeated here.

[0082] The following describes an embodiment of this application that uses high-precision map data (a type of digital map data) as the reference navigation data in the case of GNSS signal loss in a long tunnel scenario.

[0083] Please see Figure 3 , Figure 3 This is a flowchart illustrating a method for determining motion information according to another embodiment of this application. Figure 3 The methods shown may include: S310 and IMU data preprocessing.

[0084] The IMU in this embodiment can be described with reference to the original data description in the above embodiment, and will not be repeated here. S320, inertial navigation solution.

[0085] In this embodiment, an open-loop calculation method can be used for the solution. This step can output a navigation state prediction value (including error). The navigation state prediction value can be referred to the description of motion information in the above embodiment, and will not be repeated here.

[0086] S330, Map Matching and Difference Calculation.

[0087] In this embodiment, theoretical geometric features can be queried from a priori high-precision map database to determine the pose difference. This step can output the pose difference, which can be referred to the description of the pose deviation information in the above embodiment, and will not be repeated here.

[0088] S340, IMU error parameter back estimation.

[0089] In this embodiment, the gyroscope zero bias and accelerometer zero bias of the IMU can be output.

[0090] S350, navigation filter update and feedback.

[0091] This step can output the optimal navigation state. Furthermore, the IMU error parameters can be fed back to the inertial navigation calculation stage, which is beneficial to the accuracy of subsequent navigation calculations.

[0092] S360, outputs the final accurate pose.

[0093] In this embodiment, the precise pose can be referred to the description of the second motion information, which will not be repeated here.

[0094] In general, this embodiment can be divided into several stages: Phase 1: Matching – generating observational differences.

[0095] Phase Two: Estimate – Diagnosing the root causes of errors.

[0096] Phase 3: Feedback – Correcting the system status.

[0097] The following sections will explain each of the several stages.

[0098] Phase 1: Matching – Generating Observational Differences The goal of this stage is to compare the calculation results of the inertial navigation system with "absolute truth" (such as a priori high-precision maps) to generate an observation signal that characterizes the current error magnitude.

[0099] Step 1.1: Inertial Navigation Calculation (Open-Loop Estimation): Input: Raw IMU data (acceleration a_raw, angular velocity ω_raw), odometry pulse velocity v_odom, and IMU error parameters fed back from the previous moment (accel_bias_prev, gyro_bias_prev).

[0100] Process: Data correction: Real-time compensation of the raw IMU data is performed using the feedback error parameters.

[0101] a_corrected=a_raw-accel_bias_prev.

[0102] ω_corrected=ω_raw-gyro_bias_prev.

[0103] Where a_corrected and ω_corrected are the corrected original data, a_raw and ω_raw are the original data, and accel_bias_prev and gyro_bias_prev represent error parameters.

[0104] Mechanical arrangement: Using the compensated "clean" data, the current navigation state prediction value is calculated by performing integral calculations through the inertial navigation mechanical arrangement equations.

[0105] [p_ins, v_ins, ψ_ins, θ_ins, φ_ins]=INS_Mechanization(a_corrected, ω_corrected, v_odom) Output: Navigation status prediction values ​​at the current moment, including position p_ins, velocity v_ins, heading angle ψ_ins, pitch angle θ_ins, and roll angle φ_ins.

[0106] The position p_ins can be found in the description of the first position information, the velocity v_ins can be found in the description of the first velocity information, the heading angle ψ_ins can be found in the description of the first heading angle, the pitch angle θ_ins can be found in the description of the first pitch angle, and the roll angle φ_ins can be found in the description of the first roll angle.

[0107] This step corresponds to the inertial navigation solution in the diagram.

[0108] Step 1.2: Map Query and Difference Calculation Input: The predicted position p_ins calculated by inertial navigation.

[0109] Process: Map query: Using p_ins as the index, query the prior high-precision map database to obtain the theoretical geometric features of the tunnel corresponding to the point.

[0110] [ψ_map, θ_map, φ_map]=Query_HD_Map(p_ins).

[0111] ψ_map: The design heading angle for this point (calculated from the direction of the road centerline).

[0112] θ_map: This point is used to design the longitudinal slope angle.

[0113] φ_map: The design superelevation angle of the cross slope at this point.

[0114] The ψ_map can be found in the description of the reference heading angle, the θ_map in the description of the reference pitch angle, and the φ_map in the description of the reference roll angle.

[0115] Difference calculation: The difference between the pose calculated by INS and the theoretical pose provided by the map is used to generate observations.

[0116] Δψ = ψ_ins - ψ_map.

[0117] Δθ = θ_ins - θ_map.

[0118] Δφ = φ_ins - φ_map.

[0119] Among them, Δψ can be referred to the explanation of heading angle deviation, Δθ can be referred to the explanation of pitch angle deviation, and Δφ can be referred to the explanation of roll angle deviation.

[0120] Output: Attitude difference observation matrix (or simply observation matrix) Z = [Δψ, Δθ, Δφ]^T. Where [Δψ, Δθ, Δφ] represents the observation row vector, and "^T" denotes the transpose.

[0121] This step corresponds to the map query and difference calculation in the diagram.

[0122] Phase Two: Estimate – Diagnosing the Root Causes of Error The goal of this stage is to analyze the observational discrepancies generated in the previous step and diagnose the root causes of these discrepancies—that is, which error parameters of the IMU ontology are responsible for them.

[0123] Step 2.1: Inverse estimation of IMU error parameters Input: Attitude difference observations Z = [Δψ, Δθ, Δφ]^T.

[0124] Procedure: This step is performed during the update step of the Extended Kalman Filter (EKF).

[0125] State definition: The filter state variable X not only includes navigation error, but more importantly, it includes IMU error parameters.

[0126] X=[δp, δv, δψ, δθ, δφ, accel_bias, gyro_bias]^T.

[0127] Among them, δp can refer to the explanation of position deviation, δv can refer to the explanation of velocity deviation, δψ can refer to the explanation of yaw angle deviation, δθ can refer to the explanation of pitch angle, δφ can refer to the explanation of roll angle, accel_bias (also represented as b_a) can refer to the explanation of accelerometer bias, and gyro_bias (also represented as b_g) can refer to the explanation of gyroscope bias.

[0128] Observation model: Establish the mathematical relationship between the observed quantity Z and the state variable X (observation matrix H). The design of the H matrix ensures that the attitude difference Z is mainly strongly correlated with the attitude error [δψ, δθ, δφ] and the IMU error parameters [accel_bias, gyro_bias].

[0129] Z = H*X + v (v is the observation noise).

[0130] Where H represents the mapping matrix.

[0131] In the Kalman filter update step X_corrected=X_prior+K*Z, the calculation of the gain matrix K is highly dependent on the observation matrix H.

[0132] Here, H defines the "association rules" (which states are affected by which observations); K calculates the "correction weights" based on the "association rules" and "uncertainties" (P and R); and finally, state optimization is achieved through X_corrected = X_prior + K·Z.

[0133] For example, in IMU bias estimation, H needs to clearly define the "correlation between attitude observation Δψ and gyroscope bias b_gz" (modeled through integration time Δt) so that K can calculate a reasonable bias correction weight and effectively suppress IMU drift. If H does not consider the correlation between b_gz and Δψ (the corresponding element of H is set to 0), then K will not correct b_gz, and the accumulation of bias will lead to a continuous increase in attitude error.

[0134] Since the H matrix explicitly contains strong correlation terms between Z and b_g, b_a, the calculated gain matrix K will assign very high weights to these correlation terms.

[0135] When there is a significant attitude difference Z, the filter update process will prioritize using a large portion of the information in the observation matrix Z to adjust the state variables b_g and b_a, rather than just adjusting the current δp, δv, and δθ.

[0136] Results: The system outputs not only more accurate position and attitude, but more importantly, the optimal estimate of the error of the IMU sensor itself [accel_bias_est, gyro_bias_est].

[0137] The construction of the observation matrix H here is the mathematical foundation for realizing "error root cause diagnosis".

[0138] The observation matrix H is essentially the Jacobian matrix of the nonlinear observation equation at the state estimation point. It describes how a small change in the state variable X causes a small change in the observation Z. That is, H = Z / X.

[0139] The construction of this matrix will be broken down in detail below, especially the part related to the IMU error parameters.

[0140] a. Definitions of state variables and observables: State quantity X: [δp, δv, δψ, δθ, δφ, b_a, b_g]^T Where δψ, δθ, and δφ are attitude error angles (small angle assumption).

[0141] The observation Z: [Δψ, Δθ, Δφ]^T is the difference between the pose obtained by map matching and the pose inferred by INS.

[0142] b. Linearization of the observation equation: The most critical observation is the heading angle error Δψ. We will use it as an example to derive the terms in the H matrix related to the gyroscope's zero bias b_g.

[0143] Dynamic relationship: The heading angle ψ_ins calculated by INS is obtained by integrating the angular velocity ω_z measured by the gyroscope.

[0144] ψ_ins(t)=∫[ω_z_measured(τ)-b_g]dτ (integrate from 0 to t).

[0145] Error relationship: The true heading angle ψ_true is obtained by integrating the true angular velocity.

[0146] ψ_true(t)=∫ω_z_true(τ)dτ.

[0147] Therefore, the heading angle error δψ is approximately: δψ(t)≈ψ_ins(t)-ψ_true(t)=∫[(ω_z_measured-b_g)-ω_z_true]dτ=∫[dω_z-b_g]dτ.

[0148] Where dω_z is the white noise measured by the gyroscope. In Kalman filtering, b_g is modeled as a state variable.

[0149] The derivative is obtained by taking the elements of the H matrix: We take the partial derivative of the observation Δψ with respect to the state variable b_g.

[0150] .

[0151] This derivative result is crucial: H_Δψ, b_g = -Δt.

[0152] Physical meaning: The sensitivity of the observed heading angle deviation to the gyroscope's zero bias is proportional to the integration time Δt. This means that the longer the map matching interval, the greater the contribution of the gyroscope's zero bias to the observed heading error. This single element of the H matrix precisely quantifies the numerical relationship between the "symptom (Δψ)" and the "root cause (b_g)".

[0153] c. Complete H-matrix structure (taking horizontal orientation as an example): Based on a similar derivation, we can obtain the approximate structure of the H matrix. It clearly demonstrates the "strong correlation," as shown in Table 1: Table 1 Observation Z State variable X (partial) H matrix element values ​​(example) Physical meaning interpretation Δψ (heading difference) δψ (heading error) 1 Directly relevant: The observed heading differences directly reflect the heading errors in the navigation solution. This is also included in traditional models. b_gz (Z-axis gyroscope zero bias) -Δt Core innovation: Quantifying the drift in heading caused by the cumulative effect of gyroscope bias over time. This is key to correlating attitude observations with the IMU's intrinsic error parameters, achieving a leap from "attitude correction" to "bias estimation". Δθ (pitch difference) δθ (Pitch error) 1 Directly related: Pitch difference reflects pitch error. b_ax (X-axis accelerometer zero bias) <![CDATA[-k*Δt 2 ]]> Core Innovation: In hill-climbing scenarios, pitch angle is related to acceleration. Accelerometer bias can lead to incorrect speed estimation, which in turn affects pitch angle estimation through the vehicle kinematics model. The coefficient k is related to vehicle speed and slope curvature. This demonstrates a deep integration with the vehicle motion model. Kalman Update: Performs the standard EKF update steps, using the observation Z to make an optimal estimate of the state variable X. The core output of this process is the latest estimate of the IMU error parameters, accel_bias_est and gyro_bias_est.

[0154] Specifically, the state variable X is optimally estimated using the observation Z, and the difference between external observations (map matching) and internal estimation (IMU) is used to simultaneously accomplish two tasks through a carefully designed model (observation matrix H): Temporary fix: Correcting the navigation error (position, velocity, attitude) at the current moment.

[0155] The root cause solution is to estimate and correct the source of these errors (IMU sensor error parameters).

[0156] By using a strongly correlated observation model, a leap from "treating the symptoms" to "treating the root cause" has been achieved, thus obtaining long-term accuracy and stability that traditional methods cannot match.

[0157] Output: Optimal estimates of the IMU error parameters [accel_bias_est, gyro_bias_est].

[0158] Estimation of error parameters corresponding to the flowchart.

[0159] Phase 3: Feedback – Correcting System Status This stage feeds back the diagnosed "cause" (error estimate) to the front end of the system, and performs real-time correction on the "lesion" (IMU raw data and solution model), thereby achieving closed-loop control.

[0160] Step 3.1: Navigation Filter Update and Feedback: Input: The optimal estimates of the IMU error parameters [accel_bias_est, gyro_bias_est].

[0161] Process: State feedback: The estimated error parameter value is directly assigned to the corresponding term in the state variable of the EKF filter and used as the basis for prediction in the next filtering cycle.

[0162] X(accel_bias) = ​​accel_bias_est.

[0163] X(gyro_bias) = ​​gyro_bias_est.

[0164] Model calibration feedback: Simultaneously, these two values ​​are also fed back to the "data calibration" step in step 1.1 for real-time compensation of the raw IMU data input at the next time step. This forms a closed loop from the backend to the frontend.

[0165] Output: Feedback signal: Corrected IMU parameters, sent to step 1.1.

[0166] Final output: The final optimal navigation state (position, velocity, attitude) after overall optimization by EKF.

[0167] The navigation filter update and feedback corresponding to the flowchart will output the final result, and there is a clear feedback path pointing to the inertial navigation solution.

[0168] It is understandable that this embodiment does not simply link "map matching" and "filtering algorithms" in a linear fashion. Instead, it redefines the role of map information in the navigation system and designs a closed-loop system architecture to address the fundamental problem of IMU error accumulation. Existing technologies lack such a complete solution that combines external absolute references with online calibration of internal sensor parameters to form a closed loop. This is reflected in the following three closely linked and progressively layered technical points: 1. Technical point one: Attitude difference observation based on geometric features of high-precision maps. This step corresponds to the "Map Query and Difference Calculation" step in the workflow.

[0169] Improvements: Unlike existing technologies that only use ground matching results to correct pose or rely solely on vehicle motion constraints (such as zero speed), this embodiment transforms the precise 3D tunnel geometry information (heading angle, pitch angle, roll angle) contained in the high-precision map into a powerful and continuous "absolute attitude reference source." By calculating the difference (Δψ, Δθ, Δφ) between the inertial navigation-derived attitude and the theoretical attitude on the map, a direct, high-precision observation vector strongly correlated with IMU error is generated. This provides the possibility for subsequent accurate diagnosis of IMU errors.

[0170] 2. Technical point two: Using attitude difference as input to back-estimate IMU ontological error parameters. The relevant step is "Inverse estimation of IMU error parameters" in the corresponding process.

[0171] Improvements: Unlike existing technologies that only use map information to correct navigation status (position, attitude), this invention breaks with traditional thinking. Instead of using map matching results as a "correction factor" for pose, it uses them as a "diagnostic signal" for IMU sensor errors. Within the filter framework, attitude difference observations are strongly correlated with IMU error parameters (zero bias) in the observation model, thereby achieving reverse tracing and optimal estimation from "pose error" to the "root cause of sensor error." This is a crucial leap from "treating the symptoms" to "addressing the root cause."

[0172] 3. Technical point three: "Matching-Estimation-Feedback" closed-loop correction circuit. Related steps: This corresponds to the entire process, especially the navigation filter update and feedback steps.

[0173] Improvements: This system breaks away from the traditional open-loop or semi-open-loop navigation process, establishing a complete negative feedback closed-loop system. This system continuously and in real-time feeds back the estimated IMU error parameters diagnosed in "Technical Point Two" to the front end of the system for: (a) Real-time compensation and correction of raw IMU measurement data to improve quality from the source of data.

[0174] (b) Update the error state model inside the navigation filter to make the prediction more accurate.

[0175] This closed-loop mechanism, which uses backend diagnostic results for real-time frontend correction, enables the system to have self-calibration, self-learning, and self-adaptation capabilities. It can effectively cope with the time-varying characteristics of IMU parameters and fundamentally guarantee long-term navigation accuracy.

[0176] The technical effects of this embodiment include, but are not limited to, the following: It greatly improves the robustness and reliability of map matching and effectively avoids mismatches. The corresponding technical points are: Technical Point 1 (attitude difference observation based on geometric features of high-precision maps) and Technical Point 2.

[0177] Technical solution: This invention utilizes the geometric features of high-precision maps to generate observation values, which are then used to improve the accuracy of inertial navigation.

[0178] Mechanism of Action: The direct cause of map matching mismatches is the excessive error in the initial pose prediction value provided by inertial navigation. Therefore, it is first ensured that the initial pose prediction value provided for map matching is of high accuracy.

[0179] The derivation shows that the high-precision initial pose significantly reduces the solution range that the matching algorithm needs to search, enabling it to quickly and accurately find the unique and correct matching position, thereby greatly reducing the probability of mismatches. Even if the matching algorithm itself remains unchanged, its reliability and robustness experience a qualitative leap due to the improvement in input quality.

[0180] It endows the system with self-learning and self-adaptation capabilities to adapt to complex working conditions.

[0181] The corresponding technical point: Technical point three ("matching-estimation-feedback" closed-loop correction loop).

[0182] Technical solution: The estimated error parameters are continuously fed back and used for correction at the next time step.

[0183] Mechanism of operation: This is a continuous process of "learning" and "adaptation". During operation, the system continuously learns about the "operating status" of the IMU device (such as the specific value of its zero bias under the current temperature and vibration environment) and adjusts the compensation amount in real time.

[0184] The resulting effect is that the system is no longer based on fixed parameters, but rather on adaptive parameters. It can cope with IMU parameter drift caused by vehicle startup, temperature changes in the tunnel, and long-term operation, exhibiting the intelligent characteristic of "becoming more accurate over time," a capability that traditional open-loop systems completely lack.

[0185] In summary, this embodiment can solve the problem of cumulative positioning accuracy decay caused by time-varying IMU error parameters in long tunnel navigation, as well as the problem of mismatch in prior map matching caused by excessive cumulative IMU error. This embodiment constructs a closed-loop system of "matching-estimation-feedback", which uses the results of high-precision map matching to estimate and compensate for the IMU error parameters online in reverse, thereby suppressing error accumulation at its source.

[0186] This embodiment relates to a tightly coupled navigation method for online calibration of inertial measurement unit (IMU) errors based on high-precision maps in long tunnel environments without Global Navigation Satellite System (GNSS) signals. This method primarily addresses positioning drift and map mismatch issues caused by IMU error accumulation in long tunnels. The technical solution is as follows: First, autonomous navigation calculations are performed using the IMU and odometry. Second, the calculated trajectory is matched with the geometric features of the high-precision tunnel map, and the pose difference is calculated. Then, the pose difference is used to construct observation equations to back-estimate IMU error parameters (such as gyroscope bias and accelerometer bias). Finally, the estimated error parameters are fed back to the navigation filter to update the IMU parameters in real time, improving subsequent navigation accuracy. By constructing a closed-loop system of "matching-estimation-feedback," online self-calibration of IMU errors is achieved, effectively suppressing the accumulation of positioning errors, significantly improving the accuracy and reliability of navigation in long tunnels, and reducing dependence on high-cost IMUs.

[0187] Please see Figure 4 , Figure 4 This is a structural block diagram of a motion information determination device according to an embodiment of this application. Figure 4 The device can be applied to electronic devices, such as Figure 4 The device may include an acquisition module 410, a deviation determination module 420, a calibration module 430, and a motion information determination module 440, wherein: The acquisition module 410 is used to acquire first motion information of the vehicle, which is calculated based on first raw data determined by the vehicle's inertial navigation unit. The first motion information includes first position information and first attitude information. It also acquires reference navigation data and determines reference attitude information corresponding to the first position information based on the reference navigation data. The deviation determination module 420 is used to determine the attitude deviation information between the first attitude information and the reference attitude information. The calibration module 430 is used to determine the error parameters of the inertial navigation unit based on the attitude deviation information. The motion information determination module 440 is used to acquire second motion information of the vehicle, which is calculated based on corrected second raw data. The corrected second raw data is obtained by correcting the second raw data determined by the inertial navigation unit based on the error parameters. The second raw data is data determined after the first raw data, and the second motion information is information determined after the first motion information. The second motion information includes second position information and second attitude information.

[0188] In one possible implementation, the first attitude information includes at least one of a first heading angle, a first pitch angle, and a first roll angle; the reference attitude information includes at least one of a reference heading angle, a reference pitch angle, and a reference roll angle; the attitude deviation information includes at least one of a heading angle deviation, a pitch angle deviation, and a roll angle deviation; and when the deviation determination module 420 determines the attitude deviation information between the first attitude information and the reference attitude information, it is used for at least one of the following: determining the heading angle deviation between the first heading angle and the reference heading angle; or determining the pitch angle deviation between the first pitch angle and the reference pitch angle; or determining the roll angle deviation between the first roll angle and the reference roll angle.

[0189] In one possible implementation, the error parameters include at least one of gyroscope zero bias and accelerometer zero bias. When the calibration module 430 determines the error parameters of the inertial navigation unit based on the attitude deviation information, it is used to: determine at least one of the gyroscope zero bias and accelerometer zero bias of the inertial navigation unit based on the attitude deviation information.

[0190] In one possible implementation, when the calibration module 430 determines the error parameters of the inertial navigation unit based on the attitude deviation information, it is used to: determine the state quantity matrix based on the attitude deviation information; acquire observation information, including position deviation, attitude deviation information, and error parameters, wherein the position deviation is determined based on the difference between the first position information and the reference position information, and the reference position information is determined based on the reference navigation data; determine the observation matrix based on the observation information; acquire a mapping matrix, which is used to represent the mapping between the state quantity matrix and the observation matrix; and determine the error parameters of the inertial navigation unit based on the state quantity matrix, the observation matrix, and the mapping matrix.

[0191] In one possible implementation, when the calibration module 430 determines the error parameters of the inertial navigation unit based on the state matrix, the observation matrix, and the mapping matrix, it is used to: obtain a matrix function, which represents the sum of the matrix product of the observation matrix and the target observation noise, and the matrix product includes the product between the mapping matrix and the state matrix; and substitute the state matrix, the observation matrix, and the mapping matrix into the matrix function to solve for the error parameters of the inertial navigation unit.

[0192] In one possible implementation, the calibration module 430 is further configured to: obtain the current temperature and a mapping relationship, wherein the mapping relationship represents the relationship between temperature and observation noise; and determine the target observation noise corresponding to the current temperature based on the mapping relationship.

[0193] In one possible implementation, the first motion information also includes first velocity information, and the observation information also includes velocity deviation, which is determined based on the difference between the first velocity information and the reference velocity information. When the calibration module 430 determines the observation matrix based on the observation information, it is used to: determine the observation row vector based on the observation information, which includes a vector of position deviation, a vector of velocity deviation, a vector of attitude deviation information, and an error parameter vector; and transpose the observation row vector to obtain the observation matrix.

[0194] In one possible implementation, the reference navigation data includes digital map data and vehicle-road cooperative data sent from the vehicle-road cooperative unit. When the acquisition module 410 determines the reference attitude information corresponding to the first position information based on the reference navigation data, it is used to: determine the reference attitude information corresponding to the first position information based on at least one of the digital map data and the vehicle-road cooperative data.

[0195] The apparatus in this embodiment can be described with reference to the above method, and will not be repeated here.

[0196] This application also provides an electronic device, please refer to... Figure 5 , Figure 5 The electronic device 500 shown includes a processor 510 and a memory 520, wherein the memory 510 is used to store computer programs; and the processor 520 is used to execute the programs stored in the memory 510 to implement the methods described in any embodiment of this application.

[0197] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in any embodiment of this application.

[0198] In this application, "multiple" refers to two or more.

[0199] In this application, unless otherwise expressly defined, 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 between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0200] The terms “first,” “second,” “third,” “fourth,” etc., in this application (if present) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0201] In this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, in this application, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0202] Unless otherwise specified, all steps in this application may be performed sequentially or randomly. For example, if a method includes steps A and B, it means that the method may include steps A and B performed sequentially, or it may include steps B and A performed sequentially. For example, if a method may also include step C, it means that step C may be added to the method in any order. For example, the method may include steps A, B, and C, or it may include steps A, C, and B, or it may include steps C, A, and B, etc.

[0203] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for determining motion information, characterized in that, include: The vehicle's first motion information is obtained, which is calculated based on first raw data determined by the vehicle's inertial navigation unit. The first motion information includes first position information and first attitude information. Acquire reference navigation data, and determine reference attitude information corresponding to the first position information based on the reference navigation data; Determine the attitude deviation information between the first attitude information and the reference attitude information; Based on the attitude deviation information, the error parameters of the inertial navigation unit are determined; The second motion information of the vehicle is obtained. The second motion information is calculated based on the corrected second original data. The corrected second original data is obtained by correcting the second original data determined by the inertial navigation unit based on the error parameters. The second original data is the data determined after the first original data. The second motion information is the information determined after the first motion information. The second motion information includes second position information and second attitude information.

2. The method according to claim 1, characterized in that, The first attitude information includes at least one of a first heading angle, a first pitch angle, and a first roll angle; the reference attitude information includes at least one of a reference heading angle, a reference pitch angle, and a reference roll angle; the attitude deviation information includes at least one of a heading angle deviation, a pitch angle deviation, and a roll angle deviation; and determining the attitude deviation information between the first attitude information and the reference attitude information includes at least one of the following: Determine the heading angle deviation between the first heading angle and the reference heading angle; Determine the pitch angle deviation between the first pitch angle and the reference pitch angle; Determine the roll angle deviation between the first roll angle and the reference roll angle.

3. The method according to claim 1, characterized in that, The error parameters include at least one of gyroscope zero bias and accelerometer zero bias. Determining the error parameters of the inertial navigation unit based on the attitude deviation information includes: Based on the attitude deviation information, at least one of the gyroscope zero bias and accelerometer zero bias of the inertial navigation unit is determined.

4. The method according to any one of claims 1-3, characterized in that, Determining the error parameters of the inertial navigation unit based on the attitude deviation information includes: The state quantity matrix is ​​determined based on the attitude deviation information; Acquire observation information, which includes position deviation, attitude deviation information, and error parameters. The position deviation is determined based on the difference between the first position information and the reference position information, and the reference position information is determined based on the reference navigation data. Determine the observation matrix based on the observation information; Obtain a mapping matrix, which represents the mapping between the state quantity matrix and the observation matrix; The error parameters of the inertial navigation unit are determined based on the state quantity matrix, the observation matrix, and the mapping matrix.

5. The method according to claim 4, characterized in that, The step of determining the error parameters of the inertial navigation unit based on the state quantity matrix, the observation matrix, and the mapping matrix includes: Obtain a matrix function, which represents the observation matrix as the sum of the matrix product result and the target observation noise, wherein the matrix product result includes the product result between the mapping matrix and the state variable matrix; The error parameters of the inertial navigation unit are obtained by substituting the state quantity matrix, the observation matrix, and the mapping matrix into the matrix function and solving for the error parameters.

6. The method according to claim 5, characterized in that, The method further includes: Obtain the current temperature and the mapping relationship, which represents the relationship between temperature and observation noise; Based on the mapping relationship, the target observation noise corresponding to the current temperature is determined.

7. The method according to claim 4, characterized in that, The first motion information further includes first velocity information, and the observation information further includes velocity deviation, which is determined based on the difference between the first velocity information and the reference velocity information. Determining the observation matrix based on the observation information includes: The observation vector is determined based on the observation information. The observation vector includes a vector of position deviation, a vector of velocity deviation, a vector of attitude deviation information, and a vector of error parameters. The observation row vector is transposed to obtain the observation matrix.

8. The method according to any one of claims 1-7, characterized in that, The reference navigation data includes digital map data and vehicle-to-infrastructure (V2I) data sent from the V2I unit. Determining the reference attitude information corresponding to the first location information based on the reference navigation data includes: The reference attitude information corresponding to the first location information is determined based on at least one of the digital map data and the vehicle-road cooperative data.

9. An electronic device, characterized in that, Includes processor and memory, of which: Memory, used to store computer programs; A processor for executing a program stored in memory to implement the method described in any one of claims 1-8.

10. A vehicle, characterized in that, Including the electronic device as described in claim 9.

Images