Vehicle airbag control method and device, medium, vehicle and program product

Abstract

The invention relates to a vehicle airbag control method and device, a medium, a vehicle and a program product. The method comprises the steps that optimal estimation inclination information of a first sampling moment in a first direction is determined; according to the sampling frequency, the optimal estimated inclination information of the vehicle in the first direction at the first sampling moment and the angular velocity of the vehicle in the first direction at the second sampling moment, the rolling angles of the vehicle with different attributes in the first direction at the second sampling moment are determined, the second sampling moment is a sampling moment after the first sampling moment; and according to the rolling angles with different attributes in the first direction at the second sampling moment, performing ignition control on the vehicle airbag. In this way, the risks of vehicle rolling misjudgment and air bag mistaken explosion can be effectively reduced.

Description

Technical Field

[0001] The present disclosure relates to the field of automobile safety technology, and in particular, to a vehicle airbag control method, device, medium, vehicle and program product. Background Art

[0002] The rollover protection function is widely used in active and passive vehicle safety technologies. By identifying the risk of a vehicle rolling over, the airbag is deployed to protect the occupants. The rollover detection or identification method directly affects the airbag detonation time and the occupant protection effect. At the same time, the rollover detection or identification method must avoid misjudging non-rollover situations as rollovers, causing the airbag to be accidentally deployed, causing harm to the occupants. Summary of the invention

[0003] The purpose of the present disclosure is to provide a vehicle airbag control method, device, medium, vehicle and program product.

[0004] In a first aspect of the present disclosure, a vehicle airbag control method is provided, comprising: determining optimal estimated tilt information in a first direction at a first sampling moment; determining a roll angle of different attributes of the vehicle in the first direction at the second sampling moment according to a sampling frequency, the optimal estimated tilt information in the first direction at the first sampling moment, and an angular velocity in the first direction at a second sampling moment, wherein the second sampling moment is a sampling moment after the first sampling moment; and performing detonation control on the vehicle airbag according to the roll angle of different attributes in the first direction at the second sampling moment.

[0005] According to a second aspect of the present disclosure, an electronic device is provided, comprising: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the executable instructions stored in the memory to implement any one of the methods according to the first aspect.

[0006] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the methods described in the first aspect are implemented.

[0007] According to a fourth aspect of the present disclosure, a vehicle is provided, comprising the electronic device as described in the second aspect or the computer-readable storage medium as described in the third aspect.

[0008] In a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the steps of any one of the methods in the first aspect.

[0009] The above technical solution has at least the following beneficial effects: determining the optimal estimated tilt information in the first direction at the first sampling moment, and then determining the roll angle of different attributes of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment. Combining the optimal estimated tilt information in the first direction and the angular velocity in the first direction, it is possible to identify the roll angle of the vehicle at different time points in the future, and then control the detonation of the vehicle airbag according to the roll angle of different attributes in the first direction at the second sampling moment. In this way, not only can the vehicle rollover condition be identified more accurately, the problem of misjudgment of vehicle rollover and misdetonation of airbags in certain scenes with initial roll angles be solved, unnecessary detonation of airbags can be avoided, and resource waste and mis-triggering can be reduced, but also the detonation timing and degree of the airbag can be accurately controlled, and timely and effective protection can be provided when the vehicle rolls over, the risk of passenger injury can be reduced, and the overall efficiency and reliability of the system can be improved.

[0010] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings: Figure 1 The figure is a flow chart of a vehicle airbag control method according to an exemplary embodiment.

[0012] Figure 2 According to an exemplary embodiment, an implementation Figure 1 Flow chart of step S3 in FIG.

[0013] Figure 3 The figure is an example graph showing a relative angle of a vehicle rollover according to an exemplary embodiment.

[0014] Figure 4 The figure is an example graph of a curve of the absolute angle of a vehicle rollover according to an exemplary embodiment.

[0015] Figure 5 According to an exemplary embodiment, an implementation Figure 1 Flowchart of step S2 in FIG.

[0016] Figure 6 According to an exemplary embodiment, an implementation Figure 1 Flow chart of step S1 in FIG.

[0017] Figure 7 According to an exemplary embodiment, an implementation Figure 2Flowchart of step S14 in FIG.

[0018] Figure 8 According to an exemplary embodiment, an implementation Figure 3 Flowchart of step S141 in FIG.

[0019] Fig. 9 According to an exemplary embodiment, an implementation Figure 3 Flowchart of step S142 in FIG.

[0020] Fig.10 According to an exemplary embodiment, an implementation Figure 2 Flowchart of step S12 in FIG.

[0021] Fig.11 According to an exemplary embodiment, an implementation Figure 6 Flowchart of step S122 in FIG.

[0022] Fig.12 According to an exemplary embodiment, an implementation Figure 2 Flowchart of step S13 in FIG.

[0023] Fig.13 According to an exemplary embodiment, an implementation Figure 2 Flowchart of step S11 in FIG.

[0024] Fig.14 The present invention is a flow chart showing a method of calculating a vehicle roll angle using a Kalman filter algorithm according to an exemplary embodiment.

[0025] Fig.15 The flowchart of another vehicle airbag control method is shown according to an exemplary embodiment.

[0026] Fig.16 The present invention is a block diagram of a vehicle airbag control device according to an exemplary embodiment.

[0027] Fig.17 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0028] The specific implementation of the present disclosure is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the present disclosure, and is not used to limit the present disclosure.

[0029] Before introducing a vehicle airbag control method, device, medium, vehicle and program product provided by the embodiments of the present disclosure, the contents in the relevant scenarios are first introduced. In the relevant scenarios, the perceived angular velocity at the current moment is obtained through the sensor, and the predicted angular velocity is obtained through the angular momentum conservation formula of the vehicle rolling over around a single wheel under the lateral acceleration ay. The integration is performed with a certain moment as the starting point to obtain the perceived angle and predicted angle at the current moment. The sum of the predicted angular velocity and the perceived angular velocity at the current moment is taken as the vehicle angular velocity at the current moment, and the sum of the predicted angle and the perceived angle at the current moment is taken as the vehicle angle at the current moment, and the vehicle rollover is predicted.

[0030] In another related scenario, the angular velocity ωx, acceleration ay, and acceleration az are collected and processed simultaneously, and the angle around the X-axis, the acceleration around the Y-axis, and the Z-axis are obtained through integral calculation. The three direction signals are compared with the set threshold value to accurately identify the rollover state of the vehicle. While protecting the safety of the occupants in the rollover state, the phenomenon of occupant damage caused by airbag deployment due to misjudgment of the vehicle state is avoided.

[0031] However, the above methods mainly collect the vehicle's X-axis angular velocity, Y-axis acceleration and Z-axis acceleration, integrate the X-axis angular velocity starting from a certain moment to obtain an angle as the vehicle's rollover angle, compare the angular velocity, acceleration or related characteristics at this time with a preset threshold, and judge that the vehicle is in a rollover state when all threshold conditions are met, thereby outputting a signal to control airbags and other occupant protection devices. This method uses the relative angle obtained by integrating the angular velocity over a period of time as one of the core conditions for judging vehicle rollover, but rarely involves the vehicle's initial roll angle or the absolute angle relative to the horizontal ground. In some cases, the change in the relative angle cannot reflect the actual change in the body angle, so that a vehicle in a non-rollover state is misjudged as a rollover state, resulting in a high risk of false airbag explosion.

[0032] In view of this, the present disclosure provides a vehicle airbag control method, equipment, medium, vehicle and program product, which calculates the relative angle and absolute angle of the vehicle rollover, and then makes a judgment based on the relative angle and the absolute angle to obtain a signal of whether to trigger the airbag to pop out. This ensures that the airbag can pop out accurately and timely in a rolling state to protect the safety of the occupants, while reducing the risk of misjudgment of vehicle non-rollover and accidental explosion of the airbag.

[0033] The present disclosure provides a vehicle airbag control method. Figure 1 As shown, the following steps may be included.

[0034] In step S1, determining optimal estimated tilt information in a first direction at a first sampling moment; In the disclosed embodiment, when the vehicle is in a driving state, the optimal estimated tilt information in the first direction at any first sampling time is determined. In determining the optimal estimated tilt information of the vehicle, the tilt angle and angular velocity of the vehicle can be used as state variables, and the Kalman filter algorithm can be used for optimal estimation. Among them, for estimating the tilt angle, the angular velocity measured by the gyroscope can be integrated to predict the current tilt angle.

[0035] In step S2, the rolling angles of different attributes of the vehicle in the first direction at the second sampling moment are determined according to the sampling frequency, the optimal estimated tilt information in the first direction at the first sampling moment, and the angular velocity in the first direction at the second sampling moment.

[0036] The second sampling time is a sampling time after the sampling time corresponding to the optimal estimated tilt information is obtained.

[0037] The sampling frequency is the number of times the vehicle collects angular velocity and acceleration per unit time. The roll angle can be the angle at which the vehicle rotates around its longitudinal axis (or other axis), which can be used to describe the roll state of the vehicle.

[0038] In the disclosed embodiment, based on the known sampling frequency, the optimal estimated tilt information obtained at the first sampling moment is used as the initial condition, and the angular velocity in the first direction (such as the longitudinal rolling direction) at the subsequent second sampling moment is combined to calculate the roll angle of the vehicle at different second sampling moments by integration or other mathematical methods. This process takes into account the influence of the sampling frequency on the result, ensuring that the calculated roll angle can accurately reflect the actual roll state of the vehicle and its changes.

[0039] For example, assume that the sampling frequency is 100 Hz, that is, the angular velocity and acceleration are collected 100 times per second. First, use the best estimated tilt information as the starting point, and then use the angular velocity in the longitudinal roll direction at each subsequent moment to calculate the roll angle of each subsequent sampling point by integration. For example, starting from the initial 0°, as the angular velocity changes, the roll angle gradually increases to 5°, 10°, etc., until a certain stable state is reached or a new change occurs.

[0040] In step S3, the vehicle airbag is controlled to be detonated according to the roll angles of different attributes in the first direction at the second sampling moment.

[0041] Among them, Deployment Control refers to activating the airbag and popping out the airbag when a detonation signal is received.

[0042] In the disclosed embodiment, according to the calculated rollover angle in the first direction at the second sampling moment, combined with the preset safety threshold and logic judgment rules, it is determined whether the vehicle's airbag system needs to be activated. If the rollover angle exceeds a certain safety threshold and meets other detonation conditions (such as vehicle speed, collision sensor signal, etc.), a command is issued to detonate the airbag at an appropriate time to reduce the damage to passengers caused by the collision.

[0043] For example, assume that the preset rollover angle safety threshold is 20°. At a certain moment, it is detected that the rollover angle has reached or exceeded the threshold, and the vehicle speed is within the range of the safety system activation, and the collision sensor also detects a possible impact signal. Based on these conditions, it is determined that the airbag needs to be activated, and a detonation command is immediately sent, causing the airbag to inflate quickly within milliseconds to provide protection for the passengers.

[0044] The above technical solution determines the optimal estimated tilt information in the first direction at the first sampling moment, and then determines the roll angle of different attributes of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment. Combining the optimal estimated tilt information in the first direction and the angular velocity in the first direction, it is possible to identify the roll angle of the vehicle at different time points in the future, and then control the detonation of the vehicle airbag according to the roll angle of different attributes in the first direction at the second sampling moment. In this way, not only can the vehicle rollover condition be identified more accurately, the problem of misjudgment of vehicle rollover and misdetonation of airbags in certain scenarios with initial roll angles be solved, unnecessary detonation of airbags can be avoided, resource waste and mis-triggering can be reduced, and the detonation timing and degree of the airbags can be accurately controlled, timely and effective protection can be provided when the vehicle rolls over, the risk of passenger injury can be reduced, and the overall efficiency and reliability of the system can be improved.

[0045] Further, in step S1, determining the optimal estimated tilt information in the first direction at the first sampling moment includes: Based on the angular velocity of the vehicle in the first direction and the acceleration in multiple second directions at the first sampling moment, and the estimated tilt information at the previous sampling moment of the first sampling moment, the optimal estimated tilt information in the first direction at the first sampling moment is determined, wherein the first direction is different from the second direction.

[0046] Among them, the first direction is different from the second direction, which can be understood as the first direction corresponding to the angular velocity is different from the multiple second directions corresponding to the acceleration. For example, in the vehicle rollover scenario illustrated in this article, the first direction is the X-axis direction along the front and rear wheels of the vehicle, and the acceleration is the Y-axis direction along the left and right wheels of the vehicle and the Z-axis direction in the vertical direction.

[0047] Among them, Angular Velocity is a physical quantity used to describe the speed of a vehicle's rotation around a certain point (such as the vehicle's gyroscope) or axis (such as any axis of a gyroscope). It can be represented by a vector, with a direction perpendicular to the rotation plane and a magnitude equal to the angle that the rotating object rotates per unit time. Acceleration is used to describe the speed of a vehicle's speed change, that is, the derivative of speed with respect to time. It is a vector, including magnitude and direction. Optimal Estimated Tilt Information is the optimal estimate of the vehicle's tilt state in a certain direction using algorithms such as Kalman filtering, combined with current and historical data.

[0048] In the embodiment of the present disclosure, the X-axis angular velocity ω of the vehicle is collected after the vehicle is started. X 、Y-axis angular velocity ω Y , Y-axis acceleration a Y and the Z-axis acceleration signal a Z , low-pass filtering is performed on the collected signal, and the low-pass filtered signal is used to determine the optimal estimated tilt information in the first direction at the first sampling moment.

[0049] In the embodiment of the present disclosure, after the vehicle is started, the X-axis angular velocity ω of the vehicle is continuously collected through a signal acquisition module, such as an angular velocity sensor, an acceleration sensor or a multi-axis gyroscope. X , Y-axis acceleration a Y and Z-axis acceleration a Z The vehicle's X-axis angular velocity, Y-axis acceleration, and Z-axis acceleration signals collected by the signal acquisition module are low-pass filtered to obtain filtered signals. When the vehicle is in the starting state, the Kalman filter algorithm is used to continuously calculate the vehicle's optimal estimated tilt information, such as the roll angle, based on the signal obtained after low-pass filtering.

[0050] In the disclosed embodiment, the angular velocity of the vehicle in a first direction (such as the rolling direction) and the accelerations in multiple directions (such as the longitudinal, lateral, and vertical directions) can be obtained, and these angular velocities and accelerations can reflect the current dynamic state of the vehicle. Subsequently, the estimated tilt information (including the estimated state and the covariance matrix) at the last sampling moment is used, combined with the observed data (angular velocity and acceleration) at the current moment, and the optimal estimated tilt information in the first direction at the current moment is updated. In this way, the two steps of prediction and update can be continuously iterated to minimize the estimation error and obtain the most accurate tilt state estimation.

[0051] For example, suppose that when the vehicle is turning at high speed, the angular velocity in the rolling direction is detected to increase, and the longitudinal and lateral accelerations exceed the threshold. Get the X-axis angular velocity ω after low-pass filtering X, Y-axis acceleration a Y and Z-axis acceleration a Z The signal is input into the Kalman filter algorithm. By combining these data with the Kalman filter results of the previous moment, the optimal estimated tilt angle of the vehicle in the rolling direction at the current moment is calculated. For example, the calculated tilt angle is 15°, which is more accurate than relying solely on angular velocity or acceleration.

[0052] In this way, the optimal estimated tilt information in the first direction at the first sampling moment is determined according to the angular velocity of the vehicle in the first direction and the acceleration in multiple second directions at the first sampling moment, and the estimated tilt information at the last sampling moment before the first sampling moment. Combining the vehicle state at the current moment and the estimated tilt information at the last moment, the tilt state of the vehicle at the current moment can be estimated more accurately, reducing the influence of noise and error.

[0053] In one possible implementation, see Figure 2 As shown, in step S3, the vehicle airbag is detonated and controlled according to the rollover angles of different attributes in the first direction at the second sampling moment, including: In step S31, according to the roll angles of different attributes in the first direction at the second sampling moment, angular velocity thresholds corresponding to the attributes are calculated respectively; In the disclosed embodiment, the roll angles of different attributes include the roll relative angle and the roll absolute angle. The roll relative angle in the first direction at the second sampling moment is input into the roll judgment algorithm, and the first angular velocity threshold corresponding to the roll relative angle is determined by, for example, a linear table lookup, wherein the linear table in this step is a linear relationship between the roll relative angle and the angular velocity threshold.

[0054] In an embodiment of the present disclosure, the absolute roll angle in the first direction at the second sampling moment is input into the roll algorithm, and the second angular velocity threshold corresponding to the absolute roll angle is determined by, for example, a linear table lookup, wherein the linear table in this step is a linear relationship between the absolute roll angle and the angular velocity threshold.

[0055] In step S32, an ignition signal corresponding to each of the attributes is determined according to the angular velocity in the first direction at each of the second sampling moments and an angular velocity threshold corresponding to each of the attributes.

[0056] In the disclosed embodiment, the first angular velocity threshold is compared with the angular velocity in the first direction at the second sampling moment to obtain the first ignition signal corresponding to the second sampling moment. The input conditions of the rollover judgment algorithm include but are not limited to the vehicle angle, the angular velocity and acceleration of the vehicle in the lateral, longitudinal and vertical directions, and the derived characteristic conditions calculated based on these conditions.

[0057] Among them, if the angular velocity in the first direction at the second sampling moment is greater than the first angular velocity threshold, the first ignition signal corresponding to the second sampling moment is determined to be the vehicle rolling over, and the airbag is controlled to be deployed; if the angular velocity in the first direction at the second sampling moment is less than or equal to the first angular velocity threshold, the first ignition signal corresponding to the second sampling moment is determined to be the vehicle not rolling over, and the airbag is not controlled to be deployed.

[0058] The first ignition signal can be represented by a logic signal, that is, a logic signal of "0" indicates that the judgment result is "the vehicle has not rolled over, and the airbag is not controlled to be deployed", and a logic signal of "1" indicates that the judgment result is "the vehicle has rolled over, and the airbag is controlled to be deployed". Similarly, the second angular velocity threshold is compared with the angular velocity in the first direction at the second sampling moment to obtain a second ignition signal corresponding to the second sampling moment.

[0059] Among them, if the angular velocity in the first direction at the second sampling moment is greater than the second angular velocity threshold, the second ignition signal corresponding to the second sampling moment is determined to be the vehicle rolling over, and the airbag is controlled to be deployed; if the angular velocity in the first direction at the second sampling moment is less than or equal to the second angular velocity threshold, the second ignition signal corresponding to the second sampling moment is determined to be the vehicle not rolling over, and the airbag is not controlled to be deployed.

[0060] Among them, the second ignition signal can also be represented by a logic signal, that is, the logic signal is "0" indicates that the judgment result is "the vehicle has not rolled over, and the airbag is not controlled to be deployed", and the logic signal is "1" indicates that the judgment result is "the vehicle has rolled over, and the airbag is controlled to be deployed".

[0061] In the disclosed embodiment, an ignition signal for controlling the detonation of the vehicle airbag can be determined according to any one of the first ignition signal and the second ignition signal corresponding to the first direction at the second sampling time. When the ignition signal indicates that the vehicle has rolled over, the airbag is detonated to protect the passengers; when the ignition signal indicates that the vehicle has not rolled over, the airbag is kept in a non-detonated state.

[0062] In a possible implementation, in step S33, controlling the detonation of the vehicle airbag according to the plurality of ignition signals includes: When the ignition signal corresponding to any one of the attributes at the second sampling time indicates that the vehicle has not rolled over, maintaining the vehicle airbag in an unfired state; When the ignition signals corresponding to each of the attributes at the second sampling moment indicate that the vehicle has rolled over, the vehicle airbag is detonated.

[0063] The method flow of the present invention is explained by taking an actual scenario as an example: in the case of a vehicle rollover error, there is a scenario where one side of the vehicle is lifted and falls to the ground. One side of the vehicle lying on the horizontal ground is lifted so that there is a certain angle between the side and the ground (generally not more than 45°), and then the restraint force is removed to allow the lifted side to fall freely to the ground. After the vehicle is powered on and started normally, the signal acquisition module continuously collects vehicle motion status information, including the X-axis angular velocity ω X , Y-axis acceleration a Y and Z-axis acceleration a Z , the collected signal is low-pass filtered to obtain the filtered signal, and the Kalman filter algorithm continuously calculates the roll angle of the vehicle.

[0064] Since the vehicle is lifted on one side, the Kalman filter algorithm will calculate that the vehicle has a certain roll angle. When the vehicle loses its restraint and starts to fall freely, the vehicle motion state at a certain sampling time satisfies the rollover tendency condition, and the vehicle is judged to have a rollover tendency. This sampling time is the T0 sampling time. The angular velocity is integrated with the T0 sampling time as the starting point to obtain the relative rollover angle of the vehicle. The integral operation is not started before the T0 sampling time, so the relative angle is zero. The relative rollover angle curve of the vehicle is as follows: Figure 3 shown.

[0065] At sampling time T0, the initial roll angle calculated by the Kalman filter algorithm is latched and recorded, and added to the calculated relative rollover angle of the vehicle to obtain the absolute rollover angle of the vehicle. An example of a vehicle rollover absolute angle curve is shown in Figure 3 As shown, before the sampling time T0, the roll angle is obtained by continuous calculation, and after the sampling time T0, the roll absolute angle is obtained by adding the latched initial roll angle and the rollover relative angle. Figure 3 and Figure 4 The positive and negative values ​​of the middle angle only indicate the rollover direction of the vehicle. During the free fall of a vehicle on one side, the physical manifestation is that the angle between the vehicle and the ground decreases and tends to zero. The X-axis angular velocity ω of the vehicle before the wheel touches the ground X Continue to grow. Figure 3 The relative angle change trend of the vehicle rollover is gradually increasing from zero. This trend is obviously inconsistent with the actual physical performance of the vehicle. The increase in the body angle is a necessary condition for judging that the vehicle is in a rollover state, which has a greater risk of causing misjudgment of rollover and accidental explosion of airbags. Figure 4 The absolute angle of the vehicle rollover gradually decreases from a certain value and approaches zero, which conforms to the movement law of the vehicle in this scenario. It does not meet the necessary conditions for judging that the vehicle is in a rollover state and will not be misjudged as a rollover.

[0066] Therefore, the rollover ignition signal output based on the absolute angle of the vehicle rollover is "0", regardless of whether the rollover ignition signal based on the relative angle of the vehicle rollover is "0" or "1", after the two ignition signals are logically ANDed, the output of the final ignition signal obtained by comprehensive judgment must be "0", and the airbag will not be controlled to be deployed. Therefore, the risk of misjudgment of rollover and misdeployment of airbags in similar scenarios can be effectively reduced.

[0067] In one possible implementation, see Figure 5 As shown, in step S2, determining the roll angle of different attributes of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information in the first direction at the first sampling moment, and the angular velocity in the first direction at the second sampling moment includes: In step S21, a relative rolling angle of the vehicle in the first direction at the second sampling moment is determined according to the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment; In the disclosed embodiment, the calculation of the relative roll angle is based on the time integration of the angular velocity. The angular velocity is the rate of change of the angular position over time. Therefore, by integrating the angular velocity, the change in the angular position, that is, the relative roll angle, can be obtained.

[0068] In step S22, the absolute rolling angle of the vehicle in the first direction at the second sampling moment is determined according to the optimal estimated tilt information corresponding to the first sampling moment and the relative rolling angle corresponding to the second sampling moment.

[0069] In the disclosed embodiment, the sampling frequency is used to determine the time step of integration, and the optimal estimated tilt information at the first sampling moment and the angular velocity at the second sampling moment are used to predict the relative roll angle.

[0070] In a possible implementation manner, the optimal estimated tilt information includes an optimal estimated tilt angle. In step S21, determining the relative roll angle of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment includes: Taking the optimal estimated tilt angle corresponding to the first sampling moment as the initial integral angular velocity, integrating the angular velocity in the first direction at the second sampling moment to obtain the integrated angular velocity corresponding to the second sampling moment; The relative rolling angle of the vehicle in the first direction at the second sampling moment is determined according to the integrated angular velocity corresponding to the second sampling moment and the sampling frequency.

[0071] In the embodiment of the present disclosure, the angular velocity ω of the vehicle in the first direction (X axis) is calculated by taking the sampling time T0 as the starting point of the integration. X Integrate to get the relative roll angle, as shown in the following formula:

[0072] in, is the relative angle of vehicle rollover at the first sampling moment, is the sampling frequency.

[0073] In a possible implementation, in step S22, determining the absolute roll angle of the vehicle in the first direction at the second sampling moment according to the optimal estimated tilt information corresponding to the first sampling moment and the relative roll angle corresponding to the second sampling moment includes: Calculating a sum of the optimal estimated tilt angle corresponding to the first sampling moment and the relative roll angle corresponding to the second sampling moment; The sum is determined as the absolute rolling angle of the vehicle in the first direction at the second sampling moment.

[0074] In the embodiment of the present disclosure, the calculated optimal estimated tilt angle is used as the initial roll angle and added to the calculated relative roll angle to obtain the absolute roll angle at the first sampling moment. Therefore, in the embodiment of the present disclosure, the absolute roll angle can be calculated as follows:

[0075] in, is the absolute angle of vehicle rollover at the first sampling moment, is the relative rolling angle at the first sampling moment, is the optimal estimated tilt angle as the initial roll angle.

[0076] In a possible implementation manner, the estimated tilt information includes: optimal estimated tilt information and an updated covariance matrix, see Figure 6 As shown, determining the optimal estimated tilt information in the first direction at the first sampling moment according to the angular velocity of the vehicle in the first direction at the first sampling moment and the accelerations in multiple second directions, and the estimated tilt information at the previous sampling moment of the first sampling moment, includes: In step S11, the predicted tilt information at the first sampling moment is determined according to a preset state transfer matrix, a preset state control matrix, an angular velocity in the first direction at the first sampling moment, and the optimal estimated tilt information at a previous sampling moment of the first sampling moment; In Kalman filtering, the state transition matrix is ​​used to describe the linear mapping of the vehicle from one state to the next. It represents the information about the change of the vehicle state over time in the absence of external control input. The state control matrix is ​​used to describe the effect of control input on the vehicle state. In the context of vehicle tilt estimation, if acceleration or angular velocity is used as control input to adjust the system state, this matrix will reflect the effect of these inputs on the tilt angle estimate.

[0077] In the disclosed embodiment, the change of the vehicle's tilt angle (or tilt rate) over time is linear and is affected by the vehicle's angular velocity. Therefore, a state transfer matrix can be constructed to describe this linear relationship. At the same time, since acceleration (especially the gravity acceleration component) may also contribute to the tilt estimation (through integration, etc.), a state control matrix can be used to incorporate these external influences.

[0078] In the embodiment of the present disclosure, the time interval between two adjacent samplings can be determined according to the sampling rate, and then the time interval is used as the time step, and the state transfer matrix and the state control matrix at the first sampling moment are constructed according to the time step. Therefore, the state transfer matrix and the state control matrix at the first sampling moment can be respectively expressed as:

[0079]

[0080] Among them, A is the state transfer matrix, B is the state control matrix, and dt is the time step, that is, the time difference between the first sampling moment and the previous sampling moment.

[0081] The predicted tilt information is an estimated value of the system state at the current moment calculated in the prediction step of the Kalman filter based on the optimal estimate at the previous moment and the current system dynamics (such as angular velocity). In the disclosed embodiment, the predicted tilt information at the first sampling moment is calculated using the constructed state transfer matrix and state control matrix, combined with the angular velocity input at the first sampling moment and the optimal estimated tilt information at the previous sampling moment.

[0082] In step S12, a prediction covariance matrix corresponding to the first sampling moment is determined according to the state transfer matrix, the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, and the prediction tilt information at the first sampling moment; The update covariance matrix represents the uncertainty (or covariance of error) of the state estimate at the previous moment in the Kalman filter. It is a measure of the quality of the state estimate. The prediction covariance matrix is ​​the uncertainty of the state prediction at the current moment calculated after the prediction step based on the update covariance matrix at the previous moment, the current state transfer matrix, and the system dynamics.

[0083] For example, the updated covariance matrix at the previous sampling moment is used in combination with the state transfer matrix at the first sampling moment to calculate the predicted covariance matrix at the first sampling moment. This is because the uncertainty (i.e., error) of the state will propagate over time and with changes in the system dynamics. Through the state transfer matrix, the uncertainty at the previous moment can be mapped to the current moment.

[0084] In step S13, the observed tilt information at the first sampling moment is determined according to the preset observation matrix and the accelerations in the plurality of second directions at the first sampling moment; The preset observation matrix is ​​used to map the system state to the observation space in the Kalman filter. It describes the linear relationship between the system state and the observation value. The observation tilt information is the observation value calculated based on the observation data (such as acceleration) at the first sampling moment and the observation matrix.

[0085] For example, the observed tilt information is calculated using a preset observation matrix H and the accelerations in multiple second directions at the first sampling moment. The preset observation matrix H defines how to convert the system state (such as the tilt angle) into an observable quantity (such as the acceleration component). Then, the observed tilt information is calculated using these acceleration data and H.

[0086] In step S14, the optimal estimated tilt information of the vehicle at the first sampling moment is determined based on the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix.

[0087] In the disclosed embodiment, in the Kalman filter, the real state of the system is estimated by combining the predicted value (forward prediction based on the model) and the observed value (actual measurement from the sensor). In this specific scenario, the state of the system is the tilt information of the vehicle, and the observed value is the reading of the vehicle tilt provided by the sensor.

[0088] Among them, the prediction stage: at each sampling moment, the current state (i.e., predicted tilt information) is predicted based on the state estimate (or initial state) of the previous moment and the dynamic model of the system (which may include information such as the vehicle's acceleration and angular velocity). At the same time, the uncertainty of the state is also predicted, which is represented by the prediction covariance matrix. The prediction covariance matrix measures the possible difference between the predicted value and the actual value.

[0089] Observation phase: At each sampling moment, observations (i.e., observation tilt information) are obtained from the sensor. These observations also have uncertainties, which are usually quantified by the observation noise covariance matrix. The observation matrix (H) is used to map the state space of the system to the observation space. In simple cases, it may be an identity matrix or a diagonal matrix, indicating that each state component directly corresponds to an observation value. But in more complex cases, the observation matrix may contain more complex linear transformations.

[0090] Update phase: In order to obtain the best estimated tilt information at the first sampling moment, the predicted value and the observed value are combined. This is achieved by calculating the Kalman gain.

[0091] The Kalman gain is a weighting factor that balances the contribution of the predicted and observed values ​​to the state update based on their uncertainty. Specifically, it is calculated using the prediction covariance matrix and the observation noise covariance matrix, and takes into account the effect of the observation matrix. Once the Kalman gain is calculated, it can be used to update the state estimate. The update process is to add the predicted value to a weighted observation error (i.e., the observed value minus the predicted observation value, which is weighted by the Kalman gain). This process is repeated at each sampling time to continuously update the vehicle's tilt information estimate.

[0092] In one possible implementation, see Figure 7 As shown, in step S14, determining the optimal estimated tilt information of the vehicle at the first sampling moment according to the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix includes: In step S141, the Kalman gain at the first sampling moment is determined according to the preset observation matrix, the observation tilt information at the first sampling moment, and the prediction covariance matrix; The Kalman gain is a weighting factor used in the Kalman filter to weigh the predicted value against the observed value. It determines how much trust should be given to the observed value when updating the state estimate.

[0093] In the disclosed embodiment, the calculation of the Kalman gain is based on the prediction covariance matrix, the preset observation matrix and the observation tilt information. The prediction covariance matrix represents the uncertainty of the prediction value, and the observation tilt information represents the uncertainty of the observation value. With this information, the Kalman gain can be calculated, which reflects the credibility of the observation value relative to the prediction value.

[0094] In step S142, the optimal estimated tilt information of the vehicle at the first sampling moment is determined based on the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix.

[0095] In the disclosed embodiment, after the Kalman gain is obtained, it is used to update the state estimate. Specifically, the predicted value is added to a weighted observation error (i.e., the observed value minus the predicted observed value, which is weighted by the Kalman gain) to obtain the optimal estimate. This weighting process ensures that when the observed value is more reliable than the predicted value (i.e., when the observation error is small and the Kalman gain is large), the observed value has a greater impact on the final estimate.

[0096] In one possible implementation, see Figure 8 As shown, in step S141, determining the Kalman gain at the first sampling moment according to the preset observation matrix and the observation tilt information and the prediction covariance matrix at the first sampling moment includes: In step S1411, the observed tilt angle variance at the first sampling moment is determined according to the observed tilt information at the first sampling moment; The observed tilt angle variance (or observed noise variance) is a measure of the uncertainty of the observed data. In the disclosed embodiment, the observed tilt angles at multiple moments including the first sampling moment and before are averaged, and then the observed tilt angle variance at the first sampling moment is determined based on the average and the observed tilt angle at the first sampling moment.

[0097] In step S1412, the Kalman gain at the first sampling moment is determined according to the preset observation matrix, the observed tilt angle variance at the first sampling moment, and the predicted covariance matrix.

[0098] In the disclosed embodiment, the Kalman gain is calculated using the preset observation matrix H, the observed tilt angle variance, and the predicted covariance matrix at the first sampling moment. The Kalman gain is a key intermediate variable that determines how much trust should be given to the observed value when updating the state estimate.

[0099] Specifically, the calculation of the Kalman gain takes into account the uncertainty of the predicted value and the uncertainty of the observed value. If the predicted value is very uncertain, the Kalman gain will tend to give more weight to the observed value; conversely, if the observed value is very uncertain, the Kalman gain will tend to reduce the influence of the observed value on the state update.

[0100] In a possible implementation, in step S1412, determining the Kalman gain at the first sampling moment according to the preset observation matrix and the observed tilt angle variance and the predicted covariance matrix at the first sampling moment includes: Determine a first Kalman parameter at a first sampling moment according to the preset observation matrix and the predicted covariance matrix and the observed tilt angle variance at a first sampling moment; In the disclosed embodiment, the product of the preset observation matrix and the prediction covariance matrix at the first sampling moment is calculated, and then the transpose of the preset observation matrix is ​​multiplied by the product to express the uncertainty of the prediction value in the observation space. Then, the product is summed with the variance of the observed tilt angle to obtain the first Kalman parameter at the first sampling moment.

[0101] Determine a second Kalman parameter at the first sampling moment according to the preset observation matrix and the predicted covariance matrix at the first sampling moment; In the disclosed embodiment, the product of the transposed matrix of the preset observation matrix and the predicted covariance matrix at the first sampling moment is calculated to obtain the second Kalman parameter at the first sampling moment.

[0102] A Kalman gain at a first sampling moment is determined according to the first Kalman parameter and the second Kalman parameter at a first sampling moment.

[0103] In the embodiment of the present disclosure, the ratio of the second Kalman parameter to the first Kalman parameter at the first sampling moment is calculated to determine the Kalman gain at the first sampling moment.

[0104] The Kalman gain may include a roll angle Kalman gain and an angular velocity zero bias value Kalman gain. Therefore, in the embodiment of the present disclosure, the Kalman gain may be calculated in the following manner:

[0105] in, is the roll angle Kalman gain calculated at the first sampling moment, is the angular velocity zero bias Kalman gain calculated at the first sampling moment, HT is the transposed matrix of the observation matrix, is the variance of the observed tilt angle.

[0106] In a possible implementation, the method further includes: According to the preset two-dimensional unit matrix, the preset observation matrix and the Kalman gain at the first sampling moment, the predicted covariance matrix at the first sampling moment is updated to obtain an updated covariance matrix at the first sampling moment.

[0107] In the disclosed embodiment, the product of the preset observation matrix and the Kalman gain at the first sampling moment can be calculated, and then the difference between the preset two-dimensional unit matrix and the product can be calculated. The product of the difference and the predicted covariance matrix at the first sampling moment is calculated to obtain the updated covariance matrix at the first sampling moment.

[0108] In the embodiment of the present disclosure, the updated covariance matrix can be calculated as follows:

[0109]

[0110] in, is the updated covariance matrix at the first sampling moment, I is the preset two-dimensional unit matrix, is the prediction covariance matrix at the first sampling moment.

[0111] In one possible implementation, see Fig. 9 As shown, in step S142, determining the optimal estimated tilt information of the vehicle at the first sampling moment according to the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix includes: In step S1421, first estimated tilt information at the first sampling moment is determined according to the preset observation matrix and the predicted tilt information at the first sampling moment; In the disclosed embodiment, the product of the preset observation matrix and the predicted tilt matrix at the first sampling moment is calculated to obtain a new matrix which is the first estimated tilt information at the first sampling moment. The predicted tilt matrix is ​​a two-dimensional matrix composed of predicted tilt angles and predicted angular velocity zero bias values.

[0112] In step S1422, second estimated tilt information at the first sampling moment is determined according to the observed tilt information at the first sampling moment and the first estimated tilt information; In the disclosed embodiment, the difference between the observed tilt matrix and the new matrix corresponding to the product of the preset observation matrix and the predicted tilt matrix at the first sampling moment is calculated, and the matrix corresponding to the difference is used as the second estimated tilt information at the first sampling moment.

[0113] In step S1423, third estimated tilt information at the first sampling moment is determined according to the second estimated tilt information at the first sampling moment and the prediction covariance matrix; In the disclosed embodiment, the matrix corresponding to the difference calculated in the previous step is multiplied by the predicted covariance matrix, and the newly obtained matrix is ​​determined as the third estimated tilt information at the first sampling moment.

[0114] In step S1424, the optimal estimated tilt information of the vehicle at the first sampling moment is determined based on the predicted tilt information at the first sampling moment and the third estimated tilt information.

[0115] In the disclosed embodiment, the matrix corresponding to the third estimated tilt information is added to the predicted tilt matrix to obtain the optimal estimated tilt information including the optimal estimated roll angle and the optimal estimated angular velocity zero bias value. That is, the optimal estimated tilt information includes the optimal estimated roll angle and the optimal estimated angular velocity zero bias value.

[0116] Therefore, in the embodiment of the present disclosure, the optimal estimated tilt information at the first sampling moment can be calculated in the following manner:

[0117] in, is the optimal estimated roll angle at the first sampling moment, is the optimal estimated angular velocity zero bias value at the first sampling moment. is the predicted tilt angle calculated at the first sampling moment, It is the predicted angular velocity zero bias value calculated at the first sampling moment. is the observed tilt angle calculated at the first sampling moment.

[0118] In one possible implementation, see Fig.10 As shown, in step S12, determining the prediction covariance matrix corresponding to the first sampling moment according to the state transfer matrix, the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, and the prediction tilt information at the first sampling moment includes: In step S121, the state covariance at the first sampling moment is determined according to the predicted tilt information at the first sampling moment; In the embodiment of the present disclosure, the variance of the predicted roll angle is calculated according to the predicted roll angle in the predicted tilt information. , the variance of the predicted angular velocity zero bias value is calculated based on the predicted angular velocity zero bias value in the predicted tilt information Then, according to the variance of the predicted roll angle and the variance of the predicted angular velocity zero bias Construct a matrix to obtain the state covariance at the first sampling moment.

[0119] In step S122, a prediction covariance matrix corresponding to the first sampling moment is determined according to the state transfer matrix and the state covariance, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment.

[0120] In the disclosed embodiment, the prediction covariance matrix is ​​the covariance matrix calculated in the time update step during the Kalman filter process, and is used to predict the uncertainty of the state estimate for the next time step based on the current state estimate and the state transfer matrix, as well as the state covariance.

[0121] In one possible implementation, see Fig.11As shown, in step S122, determining the prediction covariance matrix corresponding to the first sampling moment according to the state transfer matrix and the state covariance, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, includes: In step S1221, the covariance transfer parameter of the first sampling moment is determined according to the state transfer matrix and the corresponding transposed matrix, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment; In the disclosed embodiment, the matrix product of the state transfer matrix, the corresponding transposed matrix, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment is calculated, and the matrix product is used as the covariance transfer parameter of the first sampling moment.

[0122] In step S1222, a prediction covariance matrix corresponding to the first sampling moment is determined according to the covariance transfer parameter and the state covariance at the first sampling moment.

[0123] In the disclosed embodiment, the matrix corresponding to the covariance transfer parameter at the first sampling moment and the matrix corresponding to the state covariance are summed to obtain the predicted covariance matrix corresponding to the first sampling moment.

[0124] Among them, the prediction covariance matrix corresponding to the first sampling time can be calculated as follows:

[0125]

[0126] in, is the prediction covariance matrix at the first sampling moment, is the covariance matrix at the previous sampling moment, AT is the transposed matrix of the state transfer matrix, Q is the covariance of the predicted roll angle and the predicted angular velocity zero bias value, is the variance of the predicted roll angle, is the variance of the predicted angular velocity zero bias value.

[0127] In one possible implementation, see Fig.12 As shown, in step S13, determining the observed tilt information at the first sampling moment according to the preset observation matrix and the acceleration in multiple second directions at the first sampling moment includes: In step S131, according to the acceleration in the plurality of second directions at the first sampling moment, the angle between the vehicle body and the vertical direction at the first sampling moment is determined; In the disclosed embodiment, the acceleration data of the vehicle in multiple second directions is used to estimate the angle between the vehicle body and the vertical direction, and then the vehicle angle matrix is ​​constructed, and finally the observed tilt information at the first sampling moment is determined in combination with the preset observation matrix. This process is a common sensor data fusion step in vehicle state estimation, especially when it comes to tilt angle estimation based on inertial sensors.

[0128] Among them, the accelerometer is usually able to measure the acceleration of the vehicle in all directions, including the gravity acceleration component. When the vehicle is in a non-horizontal state, the gravity acceleration component will produce different readings on different axes of the sensor. By analyzing these readings, the angle between the vehicle body and the vertical direction (i.e. the direction of gravity) can be calculated. This usually involves converting the readings of the accelerometer from the sensor coordinate system to the vehicle coordinate system and applying trigonometric functions to calculate the angle.

[0129] In step S132, a vehicle angle matrix at the first sampling moment is constructed according to the included angle; In the disclosed embodiment, a vehicle angle matrix (also called a rotation matrix or a direction cosine matrix) can be constructed based on the angle between the vehicle body and the vertical direction (which may include a pitch angle and a roll angle, which correspond to the front-back and left-right tilt of the vehicle, respectively). This matrix is ​​used to describe the rotation state of the vehicle body relative to a reference coordinate system (such as the earth coordinate system or the initial vehicle coordinate system). Through this matrix, a vector in the vehicle coordinate system can be converted to the reference coordinate system, or vice versa.

[0130] In step S133, the observation tilt information at the first sampling moment is determined according to the vehicle angle matrix and the preset observation matrix.

[0131] In the disclosed embodiment, after the vehicle angle matrix is ​​obtained, it can be combined with the preset observation matrix to determine the observed tilt information at the first sampling moment. For example, the vehicle angle matrix (or the tilt angle extracted therefrom) is regarded as part of the observation data and used together with other parts of the Kalman filter (such as the predicted state, the predicted covariance matrix, etc.) to update the state estimate.

[0132] In a possible implementation, the accelerations in the multiple second directions include a first acceleration in a direction perpendicular to the vehicle's travel and a second acceleration in a gravity direction. In step S131, determining the angle between the vehicle body and the vertical direction at the first sampling moment according to the accelerations in the multiple second directions at the first sampling moment includes: calculating a ratio of the first acceleration to the second acceleration; Among them, the acceleration sensor can measure the acceleration of the vehicle in various directions, including the gravity acceleration component: the lateral acceleration perpendicular to the vehicle's driving direction (the first acceleration) and the longitudinal acceleration along the direction of gravity (the second acceleration).

[0133] When the vehicle is tilted, the gravity acceleration component will produce different readings on different axes of the sensor. The longitudinal acceleration (i.e., the acceleration in the direction of gravity) will no longer be equal to the free fall acceleration (about 9.81 m / s²), and the lateral acceleration will have a non-zero value. By calculating the ratio of the lateral acceleration to the longitudinal acceleration, a quantity related to the vehicle's tilt angle can be obtained.

[0134] According to the ratio, the angle between the vehicle body and the vertical direction at the first sampling moment is determined.

[0135] In the embodiment of the present disclosure, the tilt angle can be calculated using trigonometric functions (such as inverse tangent function). Therefore, the observed tilt angle can be calculated as follows:

[0136]

[0137]

[0138] in, is the observed tilt angle calculated at the first sampling moment, H is the observation matrix, is the vehicle Y-axis acceleration collected and filtered at the first sampling moment, is the vehicle Z-axis acceleration collected and filtered at the first sampling moment, The first sampling moment is based on and The calculated angle of the vehicle.

[0139] In one possible implementation, see Fig.13 As shown, in step S11, the predicted tilt information at the first sampling moment is determined according to the preset state transfer matrix, the preset state control matrix, the angular velocity in the first direction at the first sampling moment, and the optimal estimated tilt information at the previous sampling moment of the first sampling moment, including: In step S111, a state transfer parameter at the first sampling moment is determined according to the state transfer matrix and the optimal estimated tilt information corresponding to the previous sampling moment of the first sampling moment; In the disclosed embodiment, the state transfer matrix (also called the dynamic model matrix) describes how the system state is transferred from the previous moment to the current moment in the absence of external control input. This matrix usually reflects the natural change of the vehicle tilt angle (or related state variables) between adjacent sampling moments. The state transfer parameter is obtained by multiplying the optimal estimated tilt information (as an estimate of the system state) at the previous sampling moment with the state transfer matrix. It represents the expected value of the system state at the first sampling moment without considering the current control input and external interference.

[0140] In the embodiment of the present disclosure, by multiplying the state control matrix and the matrix composed of the optimal estimated roll angle and the optimal estimated angular velocity zero bias value corresponding to the optimal estimated tilt information corresponding to the previous sampling moment of the first sampling moment, a prediction of the system state change caused by this control input, that is, the state transfer parameter, can be obtained.

[0141] In step S112, a state control parameter at the first sampling moment is determined according to the state control matrix and the angular velocity in the first direction at the first sampling moment; In the disclosed embodiment, the state control matrix and the angular velocity at the first sampling moment are used together to calculate the state control parameter. In the context of vehicle tilt estimation, the angular velocity (especially the angular velocity around a certain axis, such as the vertical axis) is one of the main factors affecting the change in the vehicle tilt angle. By multiplying the angular velocity by the state control matrix, a prediction of the system state change caused by this control input can be obtained.

[0142] In step S113, the predicted tilt information corresponding to the first sampling moment is determined according to the state transfer parameter and the state control parameter at the first sampling moment.

[0143] In the disclosed embodiment, the state transfer parameter and the state control parameter are combined, for example, the sum of the state transfer parameter and the state control parameter is calculated to obtain a matrix corresponding to the predicted tilt information to determine the predicted tilt information at the first sampling moment. This usually involves adding the two together (or performing a more complex combination based on the system model). The predicted tilt information represents the best prediction of the system state at the first sampling moment without considering the current observation data.

[0144] In a possible implementation manner, the predicted tilt information includes a predicted tilt angle and a predicted angular velocity zero bias value, and the optimal estimated tilt information includes an optimal estimated tilt angle and an optimal estimated angular velocity zero bias value.

[0145] In the embodiment of the present disclosure, the predicted tilt angle and the predicted angular velocity zero bias value can be determined by the following formula:

[0146] Among them, k=0, 1, 2, 3...N are discrete sampling points, is the predicted tilt angle calculated at the first sampling moment, It is the predicted angular velocity zero bias value calculated at the first sampling moment, where the angular velocity zero bias value is the average output offset value of the angular velocity sensor when there is no rotation input. It is an inherent property of instruments such as gyroscopes. The sensor reading minus the zero bias value is the true angular velocity detection value. The zero bias value can be optimally estimated through Kalman filtering. is the optimal estimated tilt angle calculated at the previous sampling moment, is the optimal estimated angular velocity zero bias value calculated at the previous sampling moment, A is the state transfer matrix, B is the state control matrix, It is the vehicle X-axis angular velocity collected and filtered at the first sampling moment.

[0147] The following is an example of calculating the vehicle roll angle using the Kalman filter algorithm. Fig.14 As shown, in step 201, the acquired angular velocity and acceleration are low-pass filtered to obtain the filtered angular velocity w X , acceleration a along the Y axis Y and the acceleration a along the Z axis Z In step 202, based on the angular velocity w X Calculate the predicted roll angle. In step 203, the predicted covariance matrix corresponding to the first sampling moment is determined based on the state transfer matrix, the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, and the predicted roll angle at the first sampling moment. In step 204, the predicted covariance matrix is ​​calculated based on the predicted observation matrix, the acceleration a along the Y axis direction, and the predicted roll angle at the first sampling moment. Y and the acceleration a along the Z axis Z , determine the observed roll angle at the first sampling moment. In step 205, determine the Kalman gain at the first sampling moment according to the preset observation matrix, the observed roll angle at the first sampling moment and the predicted covariance matrix. In step 206, update the predicted covariance matrix at the first sampling moment according to the preset two-dimensional unit matrix, the preset observation matrix and the Kalman gain at the first sampling moment to obtain the updated covariance matrix at the first sampling moment. In step 207, determine the optimal estimated tilt information of the vehicle at the first sampling moment according to the preset observation matrix, the predicted roll angle at the first sampling moment, the Kalman gain and the observed roll angle.

[0148] In a possible implementation manner, before determining the optimal estimated tilt information at the first sampling moment in the first direction, the method includes: Determining whether the vehicle satisfies a roll condition at the first sampling moment according to the angular velocity in the first direction and the accelerations in multiple second directions at the first sampling moment; In the embodiment of the present disclosure, the X-axis angular velocity ω X , Y-axis acceleration a Y The change in the Z-axis acceleration a Z The changes in the values ​​are compared with the preset thresholds to identify whether the vehicle has a tendency to roll over. If all three conditions are greater than the corresponding thresholds at a certain moment, the vehicle is determined to have a tendency to roll over, that is, the vehicle meets the rollover condition at the first sampling moment, and this moment is defined as the T0 sampling moment; otherwise, the angular velocity and acceleration in multiple directions are re-acquired.

[0149] When the vehicle meets the roll condition at the first sampling moment, the optimal estimated tilt information at the first sampling moment is latched.

[0150] In the embodiment of the present disclosure, the optimal estimated roll angle calculated by the Kalman filter algorithm at the sampling time T0 is latched and recorded as the initial roll angle of the vehicle rollover, as shown in the following formula:

[0151] in, is the initial roll angle of the vehicle, It is the optimal estimated roll angle calculated by the Kalman filter algorithm at sampling time T0.

[0152] In a possible implementation manner, determining whether the vehicle satisfies the roll condition at the first sampling moment according to the angular velocity in the first direction and the accelerations in multiple second directions at the first sampling moment includes: Compare the angular velocity along the first direction at the first sampling moment with a preset angular velocity threshold to determine a corresponding first comparison result; In the disclosed embodiment, in the vehicle dynamic monitoring, the angular velocity on the roll axis (i.e., the axis perpendicular to the vehicle's travel direction) is particularly important for determining the roll. By comparing the angular velocity along the first direction (especially the roll axis direction) at the first sampling moment with a preset angular velocity threshold, it can be preliminarily determined whether the vehicle is experiencing abnormal rotational motion.

[0153] For example, assume that at the first sampling moment, the angular velocity on the roll axis of the vehicle is ω_roll, and the preset angular velocity threshold is ω_threshold. Compare ω_roll with ω_threshold: If |ω_roll|>ω_threshold, it indicates that the rotation speed on the roll axis of the vehicle exceeds the normal range, which may be a sign of roll, and the corresponding first comparison result is "abnormal". If |ω_roll| ≤ ω_threshold, it indicates that the rotation speed of the vehicle in this direction is within the normal range, and the corresponding first comparison result is "normal". Similarly, the angular velocities on the pitch axis and the yaw axis can also be compared.

[0154] Compare the change in acceleration along each direction at the first sampling moment with a preset change threshold to determine a corresponding second comparison result; In the embodiment of the present disclosure, in determining whether the vehicle is rolling, the acceleration in various directions (especially the acceleration perpendicular to the driving direction) can provide direct information about the change in the vehicle's posture. By comparing the acceleration in various directions at the first sampling moment with a preset acceleration threshold, it can be further confirmed whether the vehicle is in a rolling state.

[0155] For example, assume that at the first sampling moment, the acceleration of the vehicle in the vertical direction (i.e., the rolling direction) is a_roll, the change in acceleration is △a_roll, and the preset change threshold is a_threshold. Compare △a_roll with a_threshold: If △|a_roll|>a_threshold, it indicates that the vehicle is subjected to a large force in this direction, which may be caused by roll, and the corresponding second comparison result is "abnormal". If △|a_roll| ≤ a_threshold, it indicates that the acceleration of the vehicle in this direction is within the normal range, and the corresponding second comparison result is "normal".

[0156] In the disclosed embodiment, it is determined simultaneously whether the changes in acceleration in multiple directions exceed corresponding preset change thresholds respectively, so that multiple second comparison results can be obtained.

[0157] According to the first comparison result and the second comparison result, it is determined whether the vehicle satisfies the roll condition at the first sampling moment.

[0158] In the disclosed embodiment, a first comparison result and one or more second comparison results may be subjected to a logical AND judgment, that is, if the first comparison result and one or more second comparison results both indicate that the vehicle has a tendency to roll over, then it is determined that the vehicle satisfies the roll condition at the first sampling moment; if any one of the first comparison result and one or more second comparison results indicates that the vehicle does not have a tendency to roll over, then it is determined that the vehicle does not satisfy the roll condition at the first sampling moment.

[0159] The method disclosed in the present invention is described in detail below through specific embodiments: Fig.15 As shown, in step 101, the angular velocity w is collected X , acceleration a along the Y axis Y and the acceleration a along the Z axis Z In step 102, the angular velocity w X , acceleration a Y and acceleration a Z Perform low-pass filtering. In step 103, the Kalman filter algorithm calculates the optimal estimated tilt angle. In step 104, determine whether the vehicle meets the roll condition. If the vehicle does not meet the roll condition, return to step 101; if the vehicle meets the roll condition, execute step 105, record the optimal estimated tilt angle calculated in step 103 as the initial roll angle, and execute step 106 to calculate the angular velocity w X Begin integration to obtain the relative angle of rollover. In step 107, the absolute angle of rollover is calculated based on the initial roll angle and the relative angle of rollover. In step 108, the relative angle of rollover and the absolute angle of rollover are respectively input into the rollover judgment algorithm. In step 109, a rollover ignition signal is obtained based on the measured absolute angle of rollover. In step 110, a rollover ignition signal is obtained based on the relative angle of vehicle rollover. In step 111, a logical AND operation is performed on the rollover ignition signal obtained in step 109 based on the measured absolute angle of rollover and the rollover ignition signal obtained in step 110 based on the relative angle of vehicle rollover. In step 112, a final ignition signal is obtained. If the ignition signal indicates that the vehicle has rolled over, the airbag detonation control is executed; if the ignition signal indicates that the vehicle has not rolled over, the airbag is kept in the non-detonated state.

[0160] In the disclosed embodiments, only the first direction is used to represent the angular velocity of the vehicle in any direction (for example, the left and right wheels of the vehicle are used as the first direction to determine whether the vehicle has rolled over, or the front and rear of the vehicle are used as the first direction to determine whether the vehicle has pitched forward and backward, or the left front and right rear of the vehicle are used as the first direction to determine whether the vehicle has rolled over at an angle). In actual scene applications, it is possible to simultaneously determine whether the vehicle has rolled over in the first direction based on the angular velocities in multiple first directions, rather than being limited to judging in only one direction.

[0161] The above technical solution uses the collected vehicle angular velocity and acceleration motion signals to continuously calculate the vehicle's roll angle using the Kalman filter algorithm. The sampling time when the vehicle has a rollover tendency is the T0 sampling time. The initial roll angle calculated by the Kalman filter algorithm at the T0 sampling time is latched and recorded. The X-axis angular velocity ω is calculated starting from the T0 sampling time. X The relative angle of vehicle rollover is obtained by integration, and the absolute angle of vehicle rollover is obtained by adding the initial roll angle recorded in the latch to the relative angle of vehicle rollover obtained by integration. The rollover ignition signal based on the absolute angle of rollover and the rollover ignition signal based on the relative angle of rollover are logically ANDed, and the final ignition signal is outputted through comprehensive judgment to control the airbag, which can effectively reduce the risk of misjudgment of vehicle rollover and accidental explosion of airbag.

[0162] The present disclosure also provides a vehicle airbag control device, see Fig.16 As shown, including: A first determination module 160 is configured to determine optimal estimated tilt information in a first direction at a first sampling moment; The second determination module 162 is configured to determine the roll angle of different attributes of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information in the first direction at the first sampling moment, and the angular velocity in the first direction at the second sampling moment, wherein the second sampling moment is a sampling moment after the sampling moment corresponding to the optimal estimated tilt information is obtained; The control module 164 is configured to control the deployment of the vehicle airbag according to the roll angles of different attributes in the first direction at the second sampling moment.

[0163] In one possible implementation, the first determination module 160 is configured to determine the optimal estimated tilt information in the first direction at the first sampling moment based on the angular velocity of the vehicle in the first direction and the acceleration in multiple second directions at the first sampling moment, and the estimated tilt information at the previous sampling moment of the first sampling moment, wherein the first direction is different from the second direction.

[0164] In a possible implementation, the control module 164 is configured to: According to the rolling angles of different attributes in the first direction at the second sampling moment, the angular velocity thresholds corresponding to the attributes are calculated respectively; according to the angular velocity in the first direction at each of the second sampling moments and the angular velocity thresholds corresponding to each of the attributes, the ignition signal corresponding to each of the attributes is determined; and according to the multiple ignition signals, the vehicle airbag is detonated and controlled.

[0165] In a possible implementation, the control module 164 is configured to: When the ignition signal corresponding to any of the attributes at the second sampling moment indicates that the vehicle has not rolled over, the vehicle airbag is kept in an undetonated state; when the ignition signal corresponding to each of the attributes at the second sampling moment indicates that the vehicle has rolled over, the vehicle airbag is detonated.

[0166] In a possible implementation manner, the second determining module 162 is configured to: The relative rolling angle of the vehicle in the first direction at the second sampling moment is determined based on the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment; the absolute rolling angle of the vehicle in the first direction at the second sampling moment is determined based on the optimal estimated tilt information corresponding to the first sampling moment and the relative rolling angle corresponding to the second sampling moment.

[0167] In a possible implementation manner, the optimal estimated tilt information includes an optimal estimated tilt angle, and the second determining module 162 is configured to: Taking the optimal estimated tilt angle corresponding to the first sampling moment as the initial angular velocity of integration, the angular velocity in the first direction at the second sampling moment is integrated to obtain the integrated angular velocity corresponding to the second sampling moment; according to the integrated angular velocity corresponding to the second sampling moment and the sampling frequency, the relative rolling angle of the vehicle in the first direction at the second sampling moment is determined.

[0168] In a possible implementation manner, the second determining module 162 is configured to: Calculate the sum of the optimal estimated tilt angle corresponding to the first sampling moment and the relative rolling angle corresponding to the second sampling moment; and determine the sum as the absolute rolling angle of the vehicle in the first direction at the second sampling moment.

[0169] In a possible implementation manner, the estimated tilt information includes: optimal estimated tilt information and an updated covariance matrix, and the first determining module 160 includes: The first determination submodule is configured to determine the predicted tilt information at the first sampling moment based on a preset state transfer matrix, a preset state control matrix, the angular velocity in the first direction at the first sampling moment, and the optimal estimated tilt information at the previous sampling moment of the first sampling moment; the second determination submodule is configured to determine the predicted covariance matrix corresponding to the first sampling moment based on the state transfer matrix, the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, and the predicted tilt information at the first sampling moment; the third determination submodule is configured to determine the observed tilt information at the first sampling moment based on a preset observation matrix and the acceleration in multiple second directions at the first sampling moment; the fourth determination submodule is configured to determine the optimal estimated tilt information of the vehicle at the first sampling moment based on the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix.

[0170] In a possible implementation manner, the fourth determining submodule includes: The first determination unit is configured to determine the Kalman gain at the first sampling moment based on the preset observation matrix, the observed tilt information at the first sampling moment, and the predicted covariance matrix; the second determination unit is configured to determine the optimal estimated tilt information of the vehicle at the first sampling moment based on the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information, and the predicted covariance matrix.

[0171] In a possible implementation manner, the first determining unit includes: A first determining subunit is configured to determine the observed tilt angle variance at the first sampling moment according to the observed tilt information at the first sampling moment; The second determination subunit is configured to determine the Kalman gain at the first sampling moment according to the preset observation matrix, the observed tilt angle variance at the first sampling moment, and the predicted covariance matrix.

[0172] In a possible implementation manner, the second determining subunit is configured to: Determine the first Kalman parameter at the first sampling moment according to the preset observation matrix and the predicted covariance matrix and the observed tilt angle variance at the first sampling moment; determine the second Kalman parameter at the first sampling moment according to the preset observation matrix and the predicted covariance matrix at the first sampling moment; determine the Kalman gain at the first sampling moment according to the first Kalman parameter and the second Kalman parameter at the first sampling moment.

[0173] In a possible implementation, the first determination module 160 further includes: a fifth determination submodule, configured to update the predicted covariance matrix at the first sampling moment according to a preset two-dimensional unit matrix, the preset observation matrix and the Kalman gain at the first sampling moment to obtain an updated covariance matrix at the first sampling moment.

[0174] In a possible implementation manner, the fifth determining submodule is configured as follows: According to the preset observation matrix and the Kalman gain at the first sampling moment, the gain observation matrix at the first sampling moment is determined; according to the preset two-dimensional unit matrix and the gain observation matrix at the first sampling moment, the unit observation matrix at the first sampling moment is determined; according to the predicted covariance matrix at the first sampling moment and the unit observation matrix, the updated covariance matrix at the first sampling moment is determined.

[0175] In a possible implementation manner, the second determining unit includes: The third determination subunit is configured to determine the first estimated tilt information at the first sampling moment based on the preset observation matrix and the predicted tilt information at the first sampling moment; the fourth determination subunit is configured to determine the second estimated tilt information at the first sampling moment based on the observed tilt information at the first sampling moment and the first estimated tilt information; the fifth determination subunit is configured to determine the third estimated tilt information at the first sampling moment based on the second estimated tilt information at the first sampling moment and the predicted covariance matrix; the sixth determination subunit is configured to determine the optimal estimated tilt information of the vehicle at the first sampling moment based on the predicted tilt information at the first sampling moment and the third estimated tilt information.

[0176] In a possible implementation manner, the second determining submodule includes: The third determination unit is configured to determine the state covariance at the first sampling moment based on the predicted tilt information at the first sampling moment; the fourth determination unit is configured to determine the predicted covariance matrix corresponding to the first sampling moment based on the state transfer matrix and the state covariance, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment.

[0177] In a possible implementation manner, the fourth determining unit is configured to: Determine the transposed matrix of the state transfer matrix; determine the covariance transfer parameters of the first sampling moment according to the state transfer matrix and the corresponding transposed matrix, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment; determine the predicted covariance matrix corresponding to the first sampling moment according to the covariance transfer parameters of the first sampling moment and the state covariance.

[0178] In a possible implementation manner, the third determining submodule is configured as follows: According to the acceleration in multiple second directions at the first sampling moment, the angle between the vehicle body and the vertical direction at the first sampling moment is determined; according to the angle, the vehicle angle matrix at the first sampling moment is constructed; according to the vehicle angle matrix and the preset observation matrix, the observed tilt information at the first sampling moment is determined.

[0179] In a possible implementation manner, the accelerations in the multiple second directions include a first acceleration in a direction perpendicular to the vehicle's travel direction and a second acceleration in a gravity direction, and the third determination submodule is configured as follows: The ratio of the first acceleration to the second acceleration is calculated; and according to the ratio, the angle between the vehicle body and the vertical direction at the first sampling moment is determined.

[0180] In a possible implementation manner, the first determining submodule is configured to: The state transfer parameter at the first sampling moment is determined according to the state transfer matrix and the optimal estimated tilt information corresponding to the previous sampling moment of the first sampling moment; the state control parameter at the first sampling moment is determined according to the state control matrix and the angular velocity in the first direction at the first sampling moment; the predicted tilt information corresponding to the first sampling moment is determined according to the state transfer parameter and the state control parameter at the first sampling moment.

[0181] In a possible implementation manner, the predicted tilt information includes a predicted tilt angle and a predicted angular velocity zero bias value, and the optimal estimated tilt information includes an optimal estimated tilt angle and an optimal estimated angular velocity zero bias value.

[0182] In a possible implementation manner, the second determining module 162 is configured to: The relative rolling angle of the vehicle in the first direction at the second sampling moment is determined based on the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment; the absolute rolling angle of the vehicle in the first direction at the second sampling moment is determined based on the optimal estimated tilt information corresponding to the first sampling moment and the relative rolling angle corresponding to the second sampling moment.

[0183] In a possible implementation, the second determination module 162 is configured to determine whether the vehicle satisfies the roll condition at the first sampling moment based on the angular velocity in the first direction at the first sampling moment and the acceleration in multiple second directions before determining the optimal estimated tilt information in the first direction at the first sampling moment, and latch the optimal estimated tilt information at the first sampling moment if the vehicle satisfies the roll condition at the first sampling moment.

[0184] In a possible implementation manner, the second determining module 162 is configured to: The angular velocity along the first direction at the first sampling moment is compared with a preset angular velocity threshold to determine a corresponding first comparison result; the change in acceleration along each direction at the first sampling moment is compared with a preset change threshold to determine a corresponding second comparison result; based on the first comparison result and the second comparison result, it is determined whether the vehicle meets the roll condition at the first sampling moment.

[0185] The present disclosure also provides an electronic device, see Fig.17 As shown, including: A processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute the executable instructions stored in the memory to implement the method of any one of the aforementioned embodiments.

[0186] The embodiments of the present disclosure further provide a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method described in any one of the aforementioned embodiments are implemented.

[0187] The present disclosure also provides a vehicle, including the electronic device described in the above embodiment or the computer-readable storage medium described in the above embodiment. The present disclosure also provides a computer program product, including a computer program, which implements the steps of any one of the methods in the aforementioned embodiments when executed by a processor.

[0188] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings; however, the present disclosure is not limited to the specific details in the above embodiments. Within the technical concept of the present disclosure, a variety of simple modifications can be made to the technical solution of the present disclosure, and these simple modifications all fall within the protection scope of the present disclosure.

[0189] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0190] In addition, various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

Claims

1. A vehicle airbag control method, characterized in that: include: Determine optimal estimated tilt information in a first direction at a first sampling time; Determine the roll angle of different attributes of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information in the first direction at the first sampling moment, and the angular velocity in the first direction at the second sampling moment, wherein the second sampling moment is a sampling moment after the first sampling moment; The vehicle airbag is controlled to be detonated according to the rolling angles of different attributes in the first direction at the second sampling moment.

2. The method according to claim 1, characterized in that The determining of the optimal estimated tilt information in the first direction at the first sampling moment includes: Based on the angular velocity of the vehicle in the first direction and the acceleration in multiple second directions at the first sampling moment, and the estimated tilt information at the previous sampling moment of the first sampling moment, the optimal estimated tilt information in the first direction at the first sampling moment is determined, wherein the first direction is different from the second direction.

3. The method according to claim 1, characterized in that: The method of controlling the vehicle airbag to explode according to the rolling angles of different attributes in the first direction at the second sampling moment includes: Calculate the angular velocity thresholds corresponding to the attributes respectively according to the rolling angles of different attributes in the first direction at the second sampling moment; determining an ignition signal corresponding to each of the attributes according to the angular velocity in the first direction at each of the second sampling moments and an angular velocity threshold corresponding to each of the attributes; The vehicle airbag is ignited and controlled according to the plurality of ignition signals.

4. The method according to claim 3, characterized in that The method of controlling the detonation of the vehicle airbag according to the plurality of ignition signals comprises: When the ignition signal corresponding to any one of the attributes at the second sampling time indicates that the vehicle has not rolled over, maintaining the vehicle airbag in an unfired state; When the ignition signals corresponding to each of the attributes at the second sampling moment indicate that the vehicle has rolled over, the vehicle airbag is detonated.

5. The method according to claim 1, characterized in that The roll angles of different attributes include a relative roll angle and an absolute roll angle, and determining the roll angles of different attributes of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information in the first direction at the first sampling moment, and the angular velocity in the first direction at the second sampling moment includes: Determine a relative rolling angle of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment; The absolute rolling angle of the vehicle in the first direction at the second sampling moment is determined according to the optimal estimated tilt information corresponding to the first sampling moment and the relative rolling angle corresponding to the second sampling moment.

6. The method according to claim 5, characterized in that The optimal estimated tilt information includes an optimal estimated tilt angle, and determining the relative rolling angle of the vehicle in the first direction at the second sampling moment according to the sampling frequency, the optimal estimated tilt information corresponding to the first sampling moment, and the angular velocity in the first direction at the second sampling moment includes: Taking the optimal estimated tilt angle corresponding to the first sampling moment as the initial integral angular velocity, integrating the angular velocity in the first direction at the second sampling moment to obtain the integrated angular velocity corresponding to the second sampling moment; The relative rolling angle of the vehicle in the first direction at the second sampling moment is determined according to the integrated angular velocity corresponding to the second sampling moment and the sampling frequency.

7. The method according to claim 6, characterized in that The determining, according to the optimal estimated tilt information corresponding to the first sampling moment and the relative rolling angle corresponding to the second sampling moment, the absolute rolling angle of the vehicle in the first direction at the second sampling moment includes: Calculating a sum of the optimal estimated tilt angle corresponding to the first sampling moment and the relative roll angle corresponding to the second sampling moment; The sum is determined as the absolute rolling angle of the vehicle in the first direction at the second sampling moment.

8. The method according to claim 2, characterized in that: The estimated tilt information includes: optimal estimated tilt information and an updated covariance matrix, wherein the optimal estimated tilt information in the first direction at the first sampling moment is determined according to the angular velocity of the vehicle in the first direction and the accelerations in multiple second directions at the first sampling moment, and the estimated tilt information at the previous sampling moment of the first sampling moment, including: Determine the predicted tilt information at the first sampling moment according to a preset state transfer matrix, a preset state control matrix, the angular velocity in the first direction at the first sampling moment, and the optimal estimated tilt information at a previous sampling moment of the first sampling moment; Determine a prediction covariance matrix corresponding to the first sampling moment according to the state transfer matrix, the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, and the prediction tilt information at the first sampling moment; Determine the observed tilt information at the first sampling moment according to the preset observation matrix and the acceleration in the plurality of second directions at the first sampling moment; The optimal estimated tilt information of the vehicle at the first sampling moment is determined according to the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix.

9. The method according to claim 8, characterized in that The determining the optimal estimated tilt information of the vehicle at the first sampling moment according to the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix comprises: Determine the Kalman gain at the first sampling moment according to the preset observation matrix, the observation tilt information at the first sampling moment, and the prediction covariance matrix; The optimal estimated tilt information of the vehicle at the first sampling moment is determined according to the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix.

10. The method according to claim 9, characterized in that The determining of the Kalman gain at the first sampling moment according to the preset observation matrix, the observation tilt information at the first sampling moment, and the prediction covariance matrix includes: Determining the observed tilt angle variance at the first sampling moment according to the observed tilt information at the first sampling moment; The Kalman gain at the first sampling moment is determined according to the preset observation matrix, the observed tilt angle variance at the first sampling moment, and the predicted covariance matrix.

11. The method according to claim 9, characterized in that The determining of the Kalman gain at the first sampling moment according to the preset observation matrix and the observed tilt angle variance and the predicted covariance matrix at the first sampling moment comprises: Determine a first Kalman parameter at a first sampling moment according to the preset observation matrix and the predicted covariance matrix and the observed tilt angle variance at a first sampling moment; Determine a second Kalman parameter at the first sampling moment according to the preset observation matrix and the predicted covariance matrix at the first sampling moment; A Kalman gain at a first sampling moment is determined according to the first Kalman parameter and the second Kalman parameter at a first sampling moment.

12. The method according to claim 9, characterized in that The method further comprises: According to the preset two-dimensional unit matrix, the preset observation matrix and the Kalman gain at the first sampling moment, the predicted covariance matrix at the first sampling moment is updated to obtain an updated covariance matrix at the first sampling moment.

13. The method according to claim 8, characterized in that The determining the optimal estimated tilt information of the vehicle at the first sampling moment according to the preset observation matrix, the predicted tilt information at the first sampling moment, the observed tilt information and the predicted covariance matrix comprises: Determining first estimated tilt information at a first sampling moment according to the preset observation matrix and the predicted tilt information at a first sampling moment; Determining second estimated tilt information at the first sampling moment according to the observed tilt information at the first sampling moment and the first estimated tilt information; Determining third estimated tilt information at the first sampling moment according to the second estimated tilt information at the first sampling moment and the prediction covariance matrix; The optimal estimated tilt information of the vehicle at the first sampling moment is determined based on the predicted tilt information at the first sampling moment and the third estimated tilt information.

14. The method according to claim 8, characterized in that The step of determining the prediction covariance matrix corresponding to the first sampling moment according to the state transfer matrix, the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, and the prediction tilt information at the first sampling moment comprises: Determining the state covariance at the first sampling moment according to the predicted tilt information at the first sampling moment; A prediction covariance matrix corresponding to the first sampling moment is determined according to the state transfer matrix and the state covariance, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment.

15. The method according to claim 14, characterized in that Determining the prediction covariance matrix corresponding to the first sampling moment according to the state transfer matrix and the state covariance, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment, comprises: Determine the covariance transfer parameter at the first sampling moment according to the state transfer matrix and the corresponding transposed matrix, and the updated covariance matrix corresponding to the previous sampling moment of the first sampling moment; A prediction covariance matrix corresponding to the first sampling moment is determined according to the covariance transfer parameter and the state covariance at the first sampling moment.

16. The method according to claim 8, characterized in that The step of determining the observed tilt information at the first sampling moment according to the preset observation matrix and the accelerations in the plurality of second directions at the first sampling moment comprises: Determine the angle between the vehicle body and the vertical direction at the first sampling moment according to the acceleration in the plurality of second directions at the first sampling moment; According to the angle, construct a vehicle angle matrix at a first sampling moment; The observation tilt information at a first sampling moment is determined according to the vehicle angle matrix and the preset observation matrix.

17. The method according to claim 16, characterized in that The accelerations in the multiple second directions include a first acceleration in a direction perpendicular to the vehicle's travel and a second acceleration in a gravity direction, and determining the angle between the vehicle body and the vertical direction at the first sampling moment according to the accelerations in the multiple second directions at the first sampling moment includes: calculating a ratio of the first acceleration to the second acceleration; According to the ratio, the angle between the vehicle body and the vertical direction at the first sampling moment is determined.

18. The method according to claim 8, characterized in that The step of determining the predicted tilt information at the first sampling moment according to a preset state transfer matrix, a preset state control matrix, an angular velocity in the first direction at the first sampling moment, and the optimal estimated tilt information at a previous sampling moment of the first sampling moment comprises: Determine the state transfer parameter at the first sampling moment according to the state transfer matrix and the optimal estimated tilt information corresponding to the previous sampling moment of the first sampling moment; Determine a state control parameter at the first sampling moment according to the state control matrix and the angular velocity in the first direction at the first sampling moment; The predicted tilt information corresponding to the first sampling moment is determined according to the state transfer parameter and the state control parameter at the first sampling moment.

19. The method according to claim 18, characterized in that The predicted tilt information includes a predicted tilt angle and a predicted angular velocity zero bias value, and the optimal estimated tilt information includes an optimal estimated tilt angle and an optimal estimated angular velocity zero bias value.

20. The method according to any one of claims 1 to 19, characterized in that Before determining the optimal estimated tilt information in the first direction at the first sampling moment, the method includes: Determining whether the vehicle satisfies a roll condition at the first sampling moment according to the angular velocity in the first direction and the accelerations in multiple second directions at the first sampling moment; When the vehicle meets the roll condition at the first sampling moment, the optimal estimated tilt information at the first sampling moment is latched.

21. The method according to claim 20, characterized in that The determining whether the vehicle satisfies the roll condition at the first sampling moment according to the angular velocity in the first direction and the accelerations in the plurality of second directions at the first sampling moment comprises: Compare the angular velocity along the first direction at the first sampling moment with a preset angular velocity threshold to determine a corresponding first comparison result; Compare the change in acceleration along each direction at the first sampling moment with a preset change threshold to determine a corresponding second comparison result; According to the first comparison result and the second comparison result, it is determined whether the vehicle satisfies the roll condition at the first sampling moment.

22. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to execute the executable instructions stored in the memory to implement the method according to any one of claims 1 to 21.

23. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 21 are implemented.

24. A vehicle comprising the electronic device of claim 22 or the computer-readable storage medium of claim 23.

25. A computer program product, characterized in that The invention comprises a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 21.

Images