Vehicle anti-roll cooperative control method and device, electronic equipment and storage medium

Abstract

The invention discloses a vehicle anti-roll cooperative control method and device based on multi-source signal fusion, electronic equipment and a storage medium. The method comprises the steps that multi-source signals including vehicle three-axis acceleration, three-axis angular velocity, front road curvature information, vehicle speed and front wheel steering angles are obtained; after fusing the multi-source signals, estimating through an extended Kalman filter to obtain a vehicle roll state; based on the vehicle roll state and a multi-source signal, outputting a vehicle control quantity through a sliding mode controller of an enhanced sliding mode surface; a QP optimization algorithm is adopted to decompose the vehicle control quantity into suspension roll moment and active front wheel steering correction; and the suspension rigidity and damping are adjusted according to the suspension roll moment, the front wheel steering angle is controlled according to the active front wheel steering correction, and cooperative control is completed. According to the invention, based on multi-source data fusion driving cooperative control, an information island problem of a traditional independent closed-loop control architecture is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle roll control, in particular to a vehicle roll control cooperative control method and device, an electronic device and a storage medium. BACKGROUND

[0002] The current mainstream vehicle roll control active control system mainly includes: active steering system (AFS), active suspension system and active stabilizer bar system.

[0003] The traditional suspension system (such as CDC continuous damping control) and AFS (active front wheel steering) adopt independent closed-loop control architecture. The suspension control target is to suppress the roll angle by increasing the damping force; the AFS control target is to correct the yaw rate by active steering. The prior art regards the suspension and steering as two independent subsystems, the suspension system does not perceive the steering intention, and the AFS system does not obtain the roll moment information. When the vehicle enters a curve: the AFS increases the steering angle to improve the yaw tracking, which will generate additional roll moment; when the suspension increases the damping force to resist the roll, it will weaken the tire lateral force, and the two systems will offset each other, reducing the control efficiency and increasing the energy consumption.

[0004] The existing feedback control relies on real-time vehicle state feedback signals such as IMU / wheel speed sensors. For sudden road disturbances (such as sudden curvature change, road unevenness impact), the system needs to wait for the vehicle attitude to change significantly before responding, forming a passive chasing type control, which may cause uncontrollable roll angle peak, control quantity overshoot (leading to excessive compensation of the actuator), and simultaneous deterioration of ride comfort and safety. SUMMARY

[0005] Therefore, it is necessary to provide a vehicle roll control cooperative control method, device, electronic device and storage medium which can solve the problems of multi-system control conflict, response lag, poor parameter adaptability and actuator chattering in the prior art.

[0006] To solve the above technical problems, the technical solutions of the present application are as follows: The first aspect of the present application provides a vehicle roll control cooperative control method based on multi-source signal fusion, comprising: obtaining multi-source signals including vehicle three-axis acceleration, three-axis angular velocity, front road curvature information, vehicle speed and front wheel steering angle; After fusing the multi-source signals, the vehicle roll state is estimated by an extended Kalman filter; Based on the vehicle roll state and multi-source signals, a vehicle control quantity is output by a sliding mode controller with enhanced sliding surface; The vehicle control quantity is decomposed into suspension roll moment and active front wheel steering correction quantity by using QP optimization algorithm; According to the suspension roll moment adjustment suspension stiffness and damping, according to the active front wheel steering correction amount control front wheel steering angle, complete the coordinated control.

[0007] Further, the vehicle roll state is estimated by an extended Kalman filter, including: In the prediction stage of the extended Kalman filter, a discretized vehicle dynamics model and the state at the previous time , combined with the control input , the state prediction value and the covariance matrix are calculated; In the update stage of the extended Kalman filter, the measurement vector is constructed by fusing the multi-source signals, and the vehicle roll state is obtained by correcting the state estimation through the Kalman gain.

[0008] Further, the discretized vehicle dynamics model is obtained by discretizing the vehicle dynamics model, and the vehicle dynamics model is:

[0009] In the formula, represents the vehicle dynamics model; represents the state variable, represents the mass side slip angle, represents the yaw rate, represents the roll angle, represents the roll rate, represents the longitudinal vehicle speed; represents the control input, represents the front wheel steering angle, represents the active suspension roll moment; it is assumed that the longitudinal vehicle speed remains constant, that is , and it is assumed that the roll angle is small, that is , ; , represents the front wheel lateral force, the rear wheel lateral force, , , , represents the front, rear wheel tire cornering stiffness, 、 represents the front, rear wheel tire cornering angle, , ; represents the total vehicle mass; , represents the roll moment of inertia of the vehicle around the longitudinal axis, the roll moment of inertia around the vertical axis; , respectively denote the longitudinal distance from the front wheel to the vehicle center of mass, the longitudinal distance from the rear wheel to the vehicle center of mass; denotes the equivalent roll stiffness of the suspension system; denotes the equivalent roll damping of the suspension system; denotes the gravitational acceleration; denotes the height of the vehicle center of mass; denotes the unmodeled disturbance term in roll direction, which is assumed to be bounded: .

[0010] Further, the enhanced sliding mode surface comprises a basic sliding mode surface and a preview energy term :

[0011]

[0012]

[0013] wherein, denotes the enhanced sliding mode surface, denotes the preview weight, and ; , denote weight coefficients, and , , is a time constant, which is an indicator of the desired response speed of the control system, the smaller, the faster the system response is required (e.g. in emergency obstacle avoidance scenario), the larger, the slower the system response is allowed (e.g. in cruising working condition); = 0 denotes the ideal roll angle; denotes the forward road curvature information along which the vehicle is driving; denotes the future time window, , is the preview distance; , is the future roll state extrapolated based on the dynamics model and the road curvature , which is calculated using the quasi-static model method.

[0014] Further, the preview distance is adaptively adjusted according to the current road curvature rate of change :

[0015] wherein, is a basic weight coefficient, is a curvature change sensitive adjustment factor, and , are all positive numbers.

[0016] Further, the vehicle control amount is decomposed into a suspension roll moment and an active front wheel steering correction amount by using a QP optimization algorithm, comprising: designing a target function as:

[0017] wherein, denotes a discrete time step number, is a prediction horizon length. , , is a center of mass side slip angle error weight, is a yaw rate error weight, denotes a roll angle error weight, denotes a roll angle velocity error weight, denotes a front wheel steering angle weight, denotes a suspension moment weight; is a terminal state weight matrix; the constraint condition is:

[0018] wherein, denotes a maximum value, denotes a maximum value, denotes a maximum value, denotes a maximum value; by the target function and the constraint condition, the distributor state vector is redesigned, the control vector is , the suspension roll moment and the active front wheel steering correction amount are obtained.

[0019] Further, the suspension stiffness and damping are adjusted according to the suspension roll moment, and the front wheel steering angle is controlled according to the active front wheel steering correction amount, comprising: the suspension roll moment is sent to the active suspension through a CAN bus, and the active suspension adjusts the suspension stiffness and damping according to the suspension roll moment; the active front wheel steering correction amount is sent to the active front wheel steering actuator through a CAN bus, and the active front wheel steering actuator controls the front wheel steering angle according to the active front wheel steering correction amount.

[0020] The second aspect of the present application provides a vehicle anti-roll cooperative control device based on multi-source signal fusion, comprising: An acquisition module acquires multi-source signals including vehicle three-axis acceleration, three-axis angular velocity, front road curvature information, vehicle speed and front wheel steering angle; A first calculation module estimates the vehicle roll state through an extended Kalman filter after fusing the multi-source signals; A second calculation module outputs a vehicle control quantity through a sliding mode controller based on the vehicle roll state and multi-source signals; A third calculation module decomposes the vehicle control quantity into suspension roll torque and active front wheel steering correction quantity by using a QP optimization algorithm; A control module adjusts suspension stiffness and damping according to the suspension roll torque, and controls the front wheel steering angle according to the active front wheel steering correction quantity, to complete the cooperative control.

[0021] The third aspect of the present application provides an electronic device, comprising a processor, a memory and a program stored on the memory and executable on the processor, wherein the program is executed by the processor to implement the steps of the vehicle anti-roll cooperative control method based on multi-source signal fusion.

[0022] The fourth aspect of the present application is a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the vehicle anti-roll cooperative control method based on multi-source signal fusion.

[0023] Compared with the prior art, the beneficial effects of the technical scheme of the present application are: 1) The present application is based on multi-source data fusion driven cooperative control, which breaks the information island problem of traditional independent closed-loop control architecture (such as the mutual unawareness of the suspension and AFS systems in the background technology), realizes full-dimensional perception of the vehicle state, road environment and driving behavior by constructing a cross-dimensional information acquisition network, and provides bottom-layer data support for multi-system cooperative control; 2) The pre-look information preprocessing of the present application addresses the defects of traditional feedback control relying on lagging attitude feedback (such as the uncontrollable peak value of roll angle caused by "passive pursuit type control" in the background technology), realizes the advanced analysis of road features by acquiring front road curvature pre-look information (such as curvature mutation, continuous curves, etc.) and through a weighted fusion algorithm, the weight being dynamically determined by the sensor confidence), and combines the pre-look time parameter to convert the environmental information into control pre-input, so that the system starts the pre-control strategy before the vehicle enters the complex working condition, and shortens the response delay. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 This is a flowchart illustrating a vehicle anti-roll cooperative control method provided in an embodiment of the present invention; Figure 2 This is a system operation principle diagram of a vehicle anti-roll cooperative control method provided in an embodiment of the present invention; Figure 3 This is a schematic diagram showing the comparison of roll suppression effects provided in an embodiment of the present invention; Figure 4 A diagram illustrating the roll angle suppression rate provided in an embodiment of the present invention; Figure 5 This is a structural block diagram of a vehicle anti-roll cooperative control device provided in an embodiment of the present invention; Detailed Implementation To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0025] This invention provides a vehicle anti-roll cooperative control method based on multi-source signal fusion, such as... Figure 1 As shown, it includes: Acquire multi-source signals including vehicle three-axis acceleration, three-axis angular velocity, road curvature information ahead, vehicle speed, and front wheel steering angle; In a specific embodiment, a vehicle anti-roll cooperative control method is implemented by constructing a vehicle anti-roll cooperative control system, such as... Figure 2 As shown, the system includes a signal input layer, a signal processing layer, a signal control layer, a signal execution layer, and a vehicle dynamics system. The signal input layer specifically includes an IMU (Inertial Measurement Unit), a forward-facing camera, a V2X (Vehicle-to-Everything) communication module, a vehicle speed sensor, and a front wheel steering angle sensor. The IMU measures the vehicle's three-axis acceleration and angular velocity, while the forward-facing camera and V2X communication module simultaneously acquire information about the curvature of the road ahead. The vehicle speed sensor is used to measure vehicle speed signals. After fusing the multi-source signals, the vehicle roll state is estimated using an extended Kalman filter. In a specific embodiment, the fusion of the multi-source signals and the estimation of the vehicle roll state through the extended Kalman filter can be achieved through signal preprocessing of the signal processing layer of the vehicle anti-roll cooperative control system and the EKF state observer. The signal preprocessing mainly includes temperature compensation of the IMU and fusion calculation of the predicted road curvature to improve the reliability of the data. Specifically: IMU has drift phenomenon in temperature change, so the system needs to preprocess the output signal. The system adopts linear compensation method:

[0026] In the formula, respectively, axis acceleration signal value and its preprocessed value; , respectively, angular velocity around the axis and its preprocessed value; respectively represent three axes; , respectively, accelerometer temperature drift coefficient and gyroscope temperature drift coefficient matrix. The difference between the current actual temperature and the reference temperature, the reference temperature is 25°. Specifically ， . , The parameters in the matrix represent the corresponding axis coefficients, and their actual values are calibrated by temperature gradient experiments in the laboratory; The front road curvature information collected by the front-view camera and the V2X communication module is respectively . The actual curvature of the road can be . The size of the road curvature is determined according to the confidence of the front-view camera and the V2X communication module. Specifically, , , represents the noise variance of the front-view camera data, represents the noise variance of the V2X communication module data.

[0027] The EKF state observer mainly fuses different types of sensor data, and provides reliable estimates of the roll state and for the sliding mode controller in the presence of external disturbances and measurement noise. At the same time, the influence of the disturbance is reflected indirectly through state estimation, which assists the controller design, thereby significantly improving the overall performance of the system.

[0028] Based on the vehicle roll state and multi-source signals, a sliding mode controller with enhanced sliding surface is used to output the vehicle control quantity; In a specific embodiment, the vehicle control quantity is output by a sliding mode controller based on preview information in the system control layer; The QP optimization algorithm is used to decompose the vehicle control quantity into suspension roll torque and active front wheel steering correction; In a specific embodiment, the vehicle control quantity is decomposed into suspension roll moment and active front wheel steering correction quantity by the tire force distributor in the system control layer; The suspension stiffness and damping are adjusted according to the suspension roll moment, and the front wheel steering angle is controlled according to the active front wheel steering correction amount to complete the coordinated control. In a specific embodiment, coordinated control is achieved through the active suspension actuator and the AFS actuator in the actuator layer.

[0029] In a further embodiment, the vehicle roll state is estimated using an extended Kalman filter, including: In the prediction phase of the extended Kalman filter, a discretized vehicle dynamics model and the previous time-state are utilized. Combined with control input Calculate the state prediction value Covariance Matrix ,Right now:

[0030]

[0031] In the formula, Sampling time, for Discretized form, This is process noise. The process noise covariance matrix can be calibrated offline using the maximum likelihood estimation method based on real vehicle data. The state transition Jacobian matrix is ​​also relevant. The expression is:

[0032] During the update phase of the extended Kalman filter, the multi-source signals are fused to construct a measurement vector, and the vehicle roll state is obtained by correcting the state estimation through Kalman gain. In a specific embodiment, IMU data is fused ( , , Vehicle speed data from vehicle speed sensor Construct measurement vectors .exist When smaller, The measurement equation is:

[0033] The measurement Jacobian matrix is = :

[0034] The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The state estimation is corrected by the Kalman gain The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as

[0035] The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as

[0036] The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as

[0037] The mapping relationship between the state and the measurement is described as R The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as

[0038] In further embodiments, the discretized vehicle dynamics model is obtained by discretizing a vehicle dynamics model, and in consideration of system disturbances, a coupling nonlinear vehicle dynamics model of AFS steering angle and active anti-suspension roll torque is established as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as

[0039] The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as 、 The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as The mapping relationship between the state and the measurement is described as Iyand Iy denote the roll moment of inertia of the vehicle about the longitudinal axis and the roll moment of inertia about the vertical axis, respectively; , Lff and Lrr denote the longitudinal distance from the front wheels to the center of mass of the vehicle and the longitudinal distance from the rear wheels to the center of mass of the vehicle, respectively; Keff denotes the equivalent roll stiffness of the suspension system; Ceff denotes the equivalent roll damping of the suspension system; g denotes the gravitational acceleration; h denotes the height of the center of mass of the vehicle; denotes the unmodeled disturbance in the roll direction, which is assumed to be bounded: .

[0040] In further embodiments, a sliding mode controller, incorporating preview information, is implemented to achieve anticipatory control of the roll motion of the vehicle, reducing the hysteresis of traditional feedback control and improving the roll suppression capability at high speed cornering. To this end, an enhanced sliding surface is designed, which includes both a basic sliding surface and a preview energy term. The enhanced sliding surface includes a basic sliding surface and a preview energy term :

[0041]

[0042]

[0043] wherein denotes the enhanced sliding surface, denotes the preview weight, and ; , denote weight coefficients, respectively; = 0 denotes the ideal roll angle, the goal of which is to suppress roll, such that the actual roll angle tracks the zero reference; denotes the forward road curvature information traveling theoretical yaw rate; denotes the future time window, , is the preview distance; , is the future roll state extrapolated based on the dynamics model and the road curvature extrapolation; is the roll angular velocity, which is added to enhance the sensitivity of the controller to the roll angular velocity; in order to better utilize the preview information and make the control respond to the road curvature changes in advance, reducing the hysteresis, the value of estimated by EKF is used to design the preview energy term embedded in the sliding surface, Typically, the fourth-order Runge-Kutta method, the first-order lag model method, or the quasi-static model method can be used for calculation. This embodiment uses the quasi-static model method, that is:

[0044] In the formula, , For the sprung mass; In practical engineering, the fourth-order Runge-Kutta method can also be used to... Perform discretization prediction, such as:

[0045] 10ms is acceptable. . The aiming time parameter represents a future time window used to predict the vehicle's roll state during that period, enabling the controller to respond in advance to changes in road curvature and solving the lag problem of traditional feedback control. Because , The aiming distance represents the range of road information the system needs to detect in advance. The higher the vehicle speed, the longer the aiming distance. Based on Table 1, this embodiment is designed. The possible values ​​are as follows:

[0046] Table 1

[0047] In a further embodiment, dynamic adjustment To avoid overshooting, such as based on the current rate of change of road curvature. Adaptive adjustment:

[0048] In the formula, Based on the weighting coefficient, It is a curvature-sensitive adjustment factor, and , All are positive numbers, and their specific values ​​can be calibrated experimentally. This invention calibrates them through high-speed dual-track switching simulation. , This can reduce the roll overshoot by 15%. When the curvature changes drastically ( (increase) Dynamic attenuation to suppress control oscillations caused by excessively long aiming distances; on straight roads ( Maintain sufficient aiming capability when ≈0).

[0049] In a further embodiment, the Lyapunov function is defined as follows:

[0050] Taking derivative of the sliding surface, we have , and substituting the derivative of the sliding surface, we have

[0051] Combining the roll motion equation , we have

[0052] To ensure the stability of the controller, we need to satisfy ; is the weight coefficient of roll angle, which adjusts the convergence speed of the roll angle error . Increasing can accelerate the convergence speed of the roll angle, but may exacerbate the load of the suspension actuator. In the roll motion equation, is the damping term, which hinders the change of roll angle. In order to make the sliding mode control effectively overcome the influence of the damping term, combined with the expression of , we need to be able to offset the influence of , that is . We can know that

[0053] is the weight coefficient of yaw rate. It adjusts the response speed of the yaw rate error . Increasing can enhance the tracking ability of the road curvature change, but may introduce the risk of oversteering. The value of can be determined according to the time constant of yaw dynamics . reflects the inherent characteristics of the vehicle yaw dynamics system, which represents the fast and slow degree of the change of yaw rate. The value of the yaw response time constant s( is determined according to the actual vehicle performance requirements and driving safety factors). Thus, the value range of can be determined:

[0054] In the formula, is the time constant, which is the expected response speed index of the control system, the smaller, the faster the system response is required (such as emergency obstacle avoidance scene), the larger, the slower the system response is allowed (such as cruising working condition); design index approach rate ​The control law of the sliding mode controller can be obtained as follows:

[0055] In the formula, For the gain of the switching item, The thickness of the boundary layer of the sliding surface. and The value of is determined to ensure that the sliding surface converges to . At the same time, minimize the jitter amplitude of the controller.

[0056] Substituting the control law of the aforementioned sliding mode controller into... We can obtain:

[0057] because Only ≥ ,ensure The system is asymptotically stable.

[0058] In a further embodiment, the QP optimization algorithm is used to decompose the vehicle control quantity into suspension roll moment and active front wheel steering correction, so that the two work together to meet the control requirements of vehicle yaw, roll and lateral motion, ensuring the stability and handling of the vehicle under complex conditions. The QP objective function and constraints are set to ensure reasonable and feasible allocation, and the distributor state vector is redesigned. The control vector is ,include: Design the objective function for:

[0059] In the formula, , , The weight of the centroid sideslip angle error. The range is generally , The weight of the yaw rate error is generally within the range of... , This represents the weight of the roll angle error, since roll suppression is the core objective, and its range is taken as... , The weight of the roll rate error is taken as the range of... , This indicates the weight of the front wheel steering angle, representing steering energy consumption and wear, with a value range of [value missing]. Avoid over-steering. This indicates the suspension torque weight, representing the suspension's kinetic energy consumption and lifespan, and its range is within... ; a terminal state weight matrix, , denotes a discrete time step number, is a prediction time domain length; In order to protect the actuator mechanical structure and avoid overload, the constraint condition is:

[0060] In the formula, denotes the maximum value of , denotes the maximum value of , denotes the maximum value of , denotes the maximum value of , takes , takes , takes , takes ; By the objective function and the constraint condition, the distributor state vector is redesigned, the control vector is , the suspension roll moment and the active front wheel steering correction amount are obtained.

[0061] In a further embodiment, the suspension stiffness and damping are adjusted according to the suspension roll moment, and the front wheel steering angle is controlled according to the active front wheel steering correction amount, comprising: The suspension roll moment is sent to the active suspension through the CAN bus, and the active suspension adjusts the suspension stiffness and damping according to the suspension roll moment; The active front wheel steering correction amount is sent to the active front wheel steering actuator through the CAN bus, and the active front wheel steering actuator controls the front wheel steering angle according to the active front wheel steering correction amount.

[0062] The actual simulation results of the system are shown in Figure 3 and Figure 4 It can be seen that after using the vehicle anti-roll collaborative control method of the embodiment, the roll suppression effect is obviously improved.

[0063] The second embodiment of the application also provides a vehicle anti-roll collaborative control device based on multi-source signal fusion, as shown in Figure 5 , comprising: An acquisition module, which acquires multi-source signals including vehicle three-axis acceleration, three-axis angular velocity, front road curvature information, vehicle speed and front wheel steering angle; a first calculation module, which estimates a vehicle roll state through an extended Kalman filter after fusing the multi-source signals; a second calculation module, which outputs a vehicle control variable through a sliding mode controller based on the vehicle roll state and the multi-source signals; a third calculation module, which decomposes the vehicle control variable into a suspension roll torque and an active front wheel steering correction amount through a QP optimization algorithm; a control module, which adjusts suspension stiffness and damping according to the suspension roll torque and controls a front wheel steering angle according to the active front wheel steering correction amount, thereby completing the cooperative control.

[0064] The third embodiment of the present application also provides an electronic device, which comprises a processor, a memory, and a program stored in the memory and executable on the processor, and the program, when executed by the processor, implements the steps of the vehicle anti-roll cooperative control method based on multi-source signal fusion.

[0065] The fourth embodiment of the present application also provides a computer readable storage medium, which stores a computer program, and the computer program, when executed by a processor, implements the steps of the vehicle anti-roll cooperative control method based on multi-source signal fusion.

[0066] Obviously, the above embodiments of the present application are merely examples for clearly illustrating the present application, and are not intended to limit the implementation modes of the present application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, it is unnecessary and impossible to enumerate all the implementation modes. Any modification, equivalent replacement, and improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the claims of the present application.

Claims

1. A vehicle anti-roll cooperative control method based on multi-source signal fusion, characterized in that, The method comprises the following steps: Obtaining multi-source signals including three-axis acceleration, three-axis angular velocity, front road curvature information, vehicle speed and front wheel steering angle of the vehicle; After fusing the multi-source signals, a vehicle roll state is estimated by an extended Kalman filter; Based on the vehicle roll state and the multi-source signals, a vehicle control amount is output by a sliding mode controller with an enhanced sliding surface; The vehicle control amount is decomposed into a suspension roll torque and an active front wheel steering correction amount by using a QP optimization algorithm; The suspension stiffness and damping are adjusted according to the suspension roll torque, and the front wheel steering angle is controlled according to the active front wheel steering correction amount, so as to complete the cooperative control.

2. The vehicle anti-roll cooperative control method based on multi-source signal fusion according to claim 1, characterized in that, The vehicle roll state is estimated by the extended Kalman filter, which comprises the following steps: In the prediction phase of the extended Kalman filter, the discretized vehicle dynamics model and the state of the previous time instant are used in combination with the control input to compute the state prediction and the covariance matrix ; In the update stage of the extended Kalman filter, a measurement vector is constructed by fusing the multi-source signals, and the vehicle roll state is obtained by correcting the state estimation by using the Kalman gain.

3. The vehicle anti-roll collaborative control method based on multi-source signal fusion according to claim 2, characterized in that, The discrete vehicle dynamics model is obtained by discretizing a vehicle dynamics model, and the vehicle dynamics model is as follows: wherein denotes the vehicle dynamics model; denotes the state variable, denotes the side slip angle, denotes the yaw rate, denotes the roll angle, denotes the roll rate, denotes the longitudinal vehicle speed; denotes the control input, denotes the front wheel steering angle, denotes the active anti-roll moment of the suspension; it is assumed that the longitudinal vehicle speed is constant, i.e. , while the roll angle is assumed to be small, i.e. , ; , denote the front and rear wheel side forces, , , , denote the front and rear wheel tire cornering stiffness, 、 denote the front and rear tire side slip angle, , ; denotes the total vehicle mass; , denote the roll moment of inertia of the vehicle about the longitudinal axis and the roll moment of inertia about the vertical axis, respectively; , denote the longitudinal distance of the front and rear wheels to the vehicle center of mass; denotes the equivalent roll stiffness of the suspension system; denotes the equivalent roll damping of the suspension system; denotes the gravitational acceleration; denotes the height of the vehicle center of mass; denotes the unmodeled disturbance in roll direction, which is assumed to be bounded: .

4. The vehicle anti-roll cooperative control method based on multi-source signal fusion according to claim 3, characterized in that, The enhanced slip surface includes a base slip surface and a preview energy term : wherein represents the enhanced sliding surface, represents the preview weight, and ; , represent the weight coefficients, respectively, and , , represents the desired response speed index; = 0 represents the ideal roll angle; represents the theoretical yaw rate along the front road curvature information for the vehicle to travel; represents the future time window, , is the preview distance; , is the future roll state extrapolated based on the dynamics model and the road curvature extrapolation.

5. The vehicle anti-roll collaborative control method based on multi-source signal fusion according to claim 4, characterized in that, Also in accordance with the current rate of change of road curvature To Adaptively adjust: wherein is a base weight coefficient, is a curvature change sensitivity adjustment factor, and , are all positive numbers.

6. The vehicle anti-roll collaborative control method based on multi-source signal fusion according to claim 5, characterized in that, The vehicle control amount is decomposed into a suspension roll torque and an active front wheel steering correction amount by using a QP optimization algorithm, which comprises the following steps: Design objective function is: wherein , , is a center of mass side slip angle error weight, is a yaw rate error weight, denotes a roll angle error weight, denotes a roll rate error weight, denotes a front wheel steering angle weight, denotes a suspension force moment weight; is a terminal state weight matrix, denotes a discretized time step number, is a prediction horizon length; The constraint condition is as follows: wherein denotes the maximum value of denotes the maximum value of denotes the maximum value of denotes the maximum value of denotes the maximum value of denotes the maximum value of denotes the maximum value of denotes the maximum value of By the objective function and the constraints, the redesign allocator state vector is obtained, the control vector is , the suspension roll moment and the active front wheel steering correction are obtained.

7. The vehicle anti-roll collaborative control method based on multi-source signal fusion according to claim 6, characterized in that, The suspension stiffness and damping are adjusted according to the suspension roll torque, and the front wheel steering angle is controlled according to the active front wheel steering correction amount, which comprises the following steps: The suspension roll torque is sent to the active suspension through a CAN bus, and the active suspension adjusts the suspension stiffness and damping according to the suspension roll torque; The active front wheel steering correction amount is sent to the active front wheel steering actuator through the CAN bus, and the active front wheel steering actuator controls the front wheel steering angle according to the active front wheel steering correction amount.

8. A vehicle anti-roll cooperative control device based on multi-source signal fusion, characterized in that, The method comprises the following steps: An acquisition module is configured to acquire multi-source signals including three-axis acceleration, three-axis angular velocity, front road curvature information, vehicle speed and front wheel steering angle of the vehicle; A first calculation module is configured to fuse the multi-source signals, and estimate a vehicle roll state by using an extended Kalman filter; A second calculation module is configured to output a vehicle control amount by using a sliding mode controller with an enhanced sliding surface based on the vehicle roll state and the multi-source signals; A third calculation module is configured to decompose the vehicle control amount into a suspension roll torque and an active front wheel steering correction amount by using a QP optimization algorithm; A control module is configured to adjust the suspension stiffness and damping according to the suspension roll torque, and control the front wheel steering angle according to the active front wheel steering correction amount, so as to complete the cooperative control.

9. An electronic device, comprising: The method comprises the following steps: A processor, a memory and a program stored in the memory and executable on the processor are provided, and the program is executed by the processor to implement the steps of the vehicle anti-roll cooperative control method based on multi-source signal fusion according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, A computer program is stored in the computer readable storage medium, and the computer program is executed by the processor to implement the steps of the vehicle anti-roll cooperative control method based on multi-source signal fusion according to any one of claims 1 to 7.

Images