A multi-dimensional active damping method and system based on visual prediction

Abstract

The application provides a multi-dimensional active damping method and system based on visual prediction, and the method comprises the following steps: acquiring road surface information in front of a vehicle in real time through a visual sensor; processing the road surface information to obtain a time domain signal curve of the road surface in front; obtaining predicted vibration information of a vehicle chassis according to a prediction model and the time domain signal curve; and sending a control instruction in advance to perform active damping according to the predicted vibration information. Compared with a traditional damping device, the processing scheme provided by the application can predict road surface obstacles and undulation information in front of the vehicle in advance through a visual sensor and computer vision technology, provide sufficient control time for a damping device in the vehicle, and consider the influence of mechanical conduction on delay, so that the damping effect is better.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of vibration reduction technology, in particular to a multi-dimensional active vibration reduction method and system based on visual prediction. BACKGROUND

[0002] With the development of automatic driving and auxiliary driving technology, visual sensors (multi-view cameras, depth cameras, laser radars, millimeter wave radars, etc.) are increasingly used in vehicles, and related visual perception algorithms are increasingly mature. Using visual sensors to detect road obstacles and undulations in front of the vehicle in advance and dynamically adjusting the vehicle suspension system has become a new development trend, such as the predictive suspension system of Audi A8. However, current vehicle suspensions are basically passive and semi-active suspensions, such as controllable electrorheological, controllable magnetorheological, adjustable air springs, and oil-gas suspensions, and active suspension technology is still not mature, and cannot achieve perfect vibration reduction effect. SUMMARY

[0003] To solve the above technical problems, the present application provides a multi-dimensional active vibration reduction method and system based on visual prediction. The multi-dimensional active vibration reduction device based on visual prediction has multi-dimensional active vibration reduction functions in vertical, pitch, roll and other directions. The device eliminates the delay effect through visual prediction, realizes early detection, early control and real-time vibration reduction, effectively reduces the vibration in the vehicle, and can be used in the transportation field of personnel and precision equipment.

[0004] The technical solutions provided by the present application are as follows:

[0005] In some embodiments, the present application provides a multi-dimensional active vibration reduction method based on visual prediction, which comprises the following steps:

[0006] Real-time acquisition of road surface information in front of the vehicle through a visual sensor;

[0007] Processing the road surface information to obtain a time-domain signal curve of the road surface in front of the vehicle;

[0008] Obtaining predicted vibration information of the vehicle chassis according to a prediction model and the time-domain signal curve;

[0009] According to the predicted vibration information, sending a control instruction in advance for active vibration reduction.

[0010] The advantages of this processing are that, compared with traditional passive vibration reduction devices, the road obstacles and undulations in front of the vehicle are predicted in advance through visual sensors and computer vision technology, sufficient control time is provided for the in-vehicle vibration reduction device, the delay of the system is eliminated through visual prediction, and better vibration reduction effect is achieved.

[0011] In some embodiments, the road surface information in front of the vehicle is acquired in real time by the visual sensor, specifically including:

[0012] According to the installation position of the visual sensor and the advancing direction of the vehicle, the maximum field of view distance and the minimum field of view distance of the visual sensor are acquired;

[0013] According to the maximum field of view distance and the minimum field of view distance, the field of view width of the visual sensor is obtained;

[0014] The road surface information within the field of view width is acquired in real time.

[0015] In some embodiments, the road surface information is processed to obtain a time-domain signal curve of the road surface in front, specifically including:

[0016] The road surface information is processed by computer vision technology to identify obstacle information and undulation information of the road surface in front;

[0017] According to the vehicle speed and the maximum field of view distance, a longest prediction time is acquired, which is greater than the delay time of the multi-dimensional active damping device;

[0018] According to the vehicle speed, the longest prediction time, and the obstacle information and undulation information of the road surface, a time-domain signal curve of the road surface in front is acquired.

[0019] In some embodiments, the predicted vibration information of the vehicle chassis is acquired according to a prediction model and the time-domain signal curve, specifically including:

[0020] According to a vehicle dynamics model, the time-domain signal curve is combined to obtain the predicted vibration information within the maximum prediction time, the predicted vibration information including vertical displacement, pitch angle, roll angle, and corresponding speed and acceleration information of the vehicle.

[0021] In some embodiments, the control instruction is sent in advance for active damping according to the predicted vibration information, specifically including:

[0022] The delay time of the active damping of the multi-dimensional active damping device is acquired;

[0023] According to the delay time, the control instruction is sent in advance for active damping, the delay time including software delay and mechanical delay.

[0024] In some embodiments, after the control instruction is sent in advance for active damping according to the predicted vibration information, it further includes:

[0025] The actual vibration information of the vehicle chassis is read in real time according to an inertial navigation unit installed on the vehicle chassis,

[0026] According to the actual vibration information and the predicted vibration information, the prediction model is adjusted.

[0027] The advantage of the processing is that the vehicle position information predicted by visual recognition and vehicle dynamics model is fused with the vehicle position information measured by the in-vehicle inertial navigation unit, the prediction model of the former is corrected according to the latter information, and better control effect is achieved.

[0028] In some embodiments, the application further provides a multi-dimensional active damping system based on visual prediction, comprising:

[0029] An information acquisition module is configured to acquire road surface information in front of the vehicle in real time through a visual sensor;

[0030] An information processing module is configured to process the road surface information to obtain a time-domain signal curve of the road surface in front of the vehicle;

[0031] A prediction module is configured to obtain predicted vibration information of a chassis of the vehicle according to a prediction model and the time-domain signal curve;

[0032] An active damping module is configured to send a control instruction in advance to perform active damping according to the predicted vibration information.

[0033] In some embodiments, the information processing module specifically comprises:

[0034] A computer vision sub-module is configured to identify the road surface information through computer vision technology;

[0035] A road surface information processing sub-module is configured to process the road surface information according to an identification result of the computer vision sub-module.

[0036] In some embodiments, the active damping module specifically comprises:

[0037] A software delay sub-module is configured to determine a time required for software delay;

[0038] A mechanical delay sub-module is configured to determine a time required for mechanical delay;

[0039] An active damping sub-module is configured to perform active damping according to the time required for the software delay and the mechanical delay.

[0040] In some embodiments, the multi-dimensional active damping system based on visual prediction further comprises:

[0041] A correction module is configured to obtain actual vibration information of the chassis of the vehicle and adjust the prediction model.

[0042] The technical solution provided by the application has at least the following beneficial effects:

[0043] 1. The application provides a multi-dimensional active damping method based on visual prediction, which adopts computer vision recognition technology, compared with traditional damping devices, the road surface obstacles and undulation information in front of the vehicle are predicted in advance through visual sensors and computer vision technology, enough control time is provided for the damping device in the vehicle, and the delay caused by mechanical conduction is considered, so that the active damping effect is better.

[0044] 2. The application provides a multi-dimensional active damping method based on visual prediction, through visual prediction of road surface excitation, the vehicle chassis pose information at future time is estimated through vehicle dynamics model analysis, including vertical, pitch and roll direction displacement, speed and acceleration; so as to realize multi-dimensional active damping.

[0045] 3. The application provides a multi-dimensional active damping method based on visual prediction, the vehicle pose information predicted through visual recognition and vehicle dynamics model and the vehicle pose information measured by the inertial navigation unit in the vehicle are fused, the prediction model of the former is corrected according to the latter information, and the prediction model is continuously optimized. BRIEF DESCRIPTION OF DRAWINGS

[0046] The above characteristics, technical features, advantages and implementation modes of the multi-dimensional active damping method and system based on visual prediction will be further described in the following in a clear and easy-to-understand manner combined with the preferred embodiments and the accompanying drawings.

[0047] Figure 1 is a flowchart of an embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0048] Figure 2 is a vehicle model reference diagram of an embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0049] Figure 3 is a vehicle 7-degree-of-freedom dynamics model in an embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0050] Figure 4 is another embodiment of the multi-dimensional active damping method based on visual prediction of the application, which can be used in a robot configuration;

[0051] Figure 5 is a time-domain curve of a section of road surface of an embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0052] Figure 6 is a vertical displacement time-domain curve of an embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0053] Figure 7 is a pitch angle time domain curve of one embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0054] Figure 8 is a roll angle time domain curve of one embodiment of the multi-dimensional active damping method based on visual prediction of the application;

[0055] Figure 9 is a module block diagram of one embodiment of the multi-dimensional active damping system based on visual prediction of the application. DETAILED DESCRIPTION

[0056] In the following description, for the purposes of explanation and not limitation, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the embodiments of the application. However, it will be apparent to those skilled in the art that the application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the application with unnecessary detail.

[0057] It should be understood that the term "comprising" as used in the specification and in the claims indicates the presence of the recited features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0058] It should be further understood that the term "and / or" as used in the specification and in the claims indicates any combination of one or more of the associated listed items and all possible combinations of the items.

[0059] In addition, in the description of the application, the terms "first", "second", etc. are only used to distinguish descriptions, and cannot be understood as indicating or implying relative importance.

[0060] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, specific implementation manners of the application will be described below with reference to the drawings. Obviously, the drawings in the following description are only some embodiments of the application, and for those skilled in the art, other drawings can be obtained from these drawings without creative labor, and other embodiments can also be obtained.

[0061] In one embodiment, the multi-dimensional active damping method based on visual prediction proposed by the application, with reference to Figure 1 , includes the steps of:

[0062] S100, acquiring road surface information in front of the vehicle in real time through a visual sensor;

[0063] Specifically, the visual sensor herein includes but is not limited to a multi-lens camera, a depth camera, a laser radar, a millimeter wave radar, etc. The visual sensor is used to capture the road surface information in front of the vehicle to make a visual prediction, such as the obstacle information and the road surface fluctuation information of the road in front.

[0064] S200, processing the road surface information to obtain a time-domain signal curve of the road in front;

[0065] Specifically, the range of the detected road surface, i.e. the field of view width, is determined according to the picture transmitted by the visual sensor, and then the time-domain signal curve of the road in front is determined according to the vehicle speed.

[0066] S300, obtaining the predicted vibration information of the vehicle chassis according to the prediction model and the time-domain signal curve;

[0067] Specifically, the predicted vibration information of the vehicle chassis at a future time, such as the vertical displacement, the pitch angle, the roll angle, and the corresponding speed and acceleration information of the vehicle, is obtained according to the prediction model and the vehicle dynamics model in combination with the time-domain signal curve of the road surface.

[0068] S400, sending a control instruction in advance to actively reduce vibration according to the predicted vibration information.

[0069] Specifically, the control instruction is sent to actively reduce vibration according to the obtained predicted vibration information in combination with the mechanical delay and the software delay of the multi-dimensional active vibration reduction device. The advantage of such a design is that the visual sensor is installed in front of the vehicle, the road surface obstacle and fluctuation information in front are detected and identified based on computer vision technology, the road surface obstacle and shape are identified using computer vision technology, the time-domain curve of the road surface is obtained, and the vibration in the vertical, pitch, and roll directions of the vehicle is reduced. The multi-dimensional active vibration reduction is realized without delay through visual prediction.

[0070] In the above embodiment, step S100 specifically includes:

[0071] According to the installation position of the visual sensor and the forward direction of the vehicle, the maximum field of view distance and the minimum field of view distance of the visual sensor are obtained;

[0072] The field of view width of the visual sensor is obtained according to the maximum field of view distance and the minimum field of view distance;

[0073] The road surface information within the field of view width is obtained in real time.

[0074] Specifically, the installation position of the visual sensor and the multi-dimensional active vibration reduction device on the vehicle model can refer to Figure 2The multi-dimensional active damping device body in the embodiment comprises an inertial navigation unit, an encoder, a multi-axis servo actuator, a control unit and other components. The multi-dimensional damping device body is installed in the vehicle cabin, and a visual sensor is installed in front of the vehicle for visual prediction. The visual sensor detects the road obstacle and undulation information in front, and according to the installation position and active direction of the sensor on the vehicle body, it is determined that the road range that can be detected in front is L min ~L max , wherein L max -L min is the field of view width.

[0075] In the above embodiment, step S200 specifically comprises:

[0076] The road information is processed by computer vision technology to identify the obstacle information and undulation information of the road in front;

[0077] According to the vehicle speed and the maximum field of view distance, the longest prediction time is obtained, and the longest prediction time is greater than the delay time of the multi-dimensional active damping device;

[0078] According to the vehicle speed, the longest prediction time, and the obstacle information and undulation information of the road, the time domain signal curve of the road in front is obtained.

[0079] Specifically, the visual sensor can continuously detect the road information at 0~T est time in the future, wherein T est =L max / v car is the longest prediction time, wherein v car is the vehicle speed. The computer vision technology is used for road obstacle and shape recognition to obtain the time domain curve of the road.

[0080] In one embodiment, on the basis of the above embodiment, step S300 specifically comprises:

[0081] According to the vehicle dynamics model, the time domain signal curve is obtained to obtain the predicted vibration information within the maximum prediction time, and the predicted vibration information includes the vertical displacement, pitch angle, roll angle and corresponding speed and acceleration information of the vehicle.

[0082] Specifically, different models can be selected for the vehicle dynamics model, and the selected model is a 7-degree-of-freedom dynamics model, which is described in reference Figure 3 , assuming that the car is left-right symmetrical, that is, the vehicle distance, suspension parameters and tire parameters on the left and right sides are the same. The parameters of the vehicle are shown in Table 1,

[0083] Vehicle parameters Symbol Value m c ]]> Vehicle body mass m sa ]]> 1380 kg m sb ]]> Front wheel mass (one side) I ox ]] 40.5 kg I oy ]] Rear wheel mass (one side) L x ]]> 45.4 kg [[ L y ]]> Rolling moment of inertia of the vehicle body about the center axis k sa ]]> 235 kg.m2 k sb ]]> Pitching moment of inertia of the vehicle body about the center axis k ta ]]> 1222 kg.m2 k tb ]]> Front-rear wheel base c sa ]]> Left-right wheel base c sb ]]>

[0084] Table 1

[0085] The rest of the kinematic parameters are described as follows:

[0086] (1) r la , r ra , r lb , r rb are respectively the road excitation of the left front, right front, left rear and right rear tire contact points;

[0087] (2) h la , h ra , h lb , h rb are respectively the absolute displacement of the left front, right front, left rear and right rear wheel centers;

[0088] (3) z la , z ra , z lb , z rb are respectively the absolute displacement of the left front, right front, left rear and right rear suspension connection points;

[0089] (4) The vertical absolute displacement of the center point of the vehicle floor is z A , the angle of rotation around the x-axis is a A , and the angle of rotation around the y-axis is b A .

[0090] Based on this, the dynamics equation of the vehicle is:

[0091]

[0092] The input quantity of the above dynamics equation is the road excitation r la , r ra , r lb , r rb of the four tire contact points, which can be obtained by scanning the road information through the aforementioned visual recognition system. Therefore, according to the vehicle dynamics model, the vibration information of the vehicle chassis at future 0~T est time can be predicted, including the vertical displacement, pitch angle, roll angle and corresponding speed, acceleration information of the vehicle.

[0093] On the basis of the above embodiment, step S400 specifically comprises:

[0094] Obtaining the delay time of active damping of the multi-dimensional active damping device;

[0095] According to the delay time, the control instruction is sent in advance for active damping, and the delay time includes software delay and mechanical delay.

[0096] Specifically, the multi-dimensional active damping device body of the present scheme can be selected from parallel robots, serial robots or series-parallel hybrid robot configurations, for reference Front suspension spring rate coefficientBut at least should have vertical, pitch, roll three degrees of freedom. Above robot only represents the robot configuration adopted in the embodiment, but is not limited to only using such robot configuration.

[0097] The control rate is designed in the robot task space (operation space) to ensure the universality of the control method under different robot configurations.

[0098] The vehicle chassis vibration input vector predicted based on the visual recognition system and the vehicle dynamics equation is

[0099] X A =[α A β A z A ] T ,

[0100] The coordinates of the multi-dimensional active vibration reduction device task space are

[0101] Delta X = [Delta alpha Delta beta Delta z] T ,

[0102] Wherein, Delta alpha, Delta beta, Delta z are the relative roll angle, the relative pitch angle and the relative vertical displacement of the upper platform relative to the base respectively;

[0103] The vibration output vector of the upper platform is wherein alpha, beta and z are the roll angle, the pitch angle and the vertical displacement of the upper platform respectively. The design goal of the task space control rate is to make the vibration output as small as possible, and various control theories such as PID, ADRC, optimal control, robust control and adaptive control can be used for design. Here, an optimal scheme is proposed, and the control rate of the task space is set as

[0104]

[0105] Spring damping + sky damping control is adopted in the vertical direction, wherein the stiffness coefficient is k z , the damping coefficient is c z , and the sky damping coefficient is s z . PID control is adopted in the pitch and roll directions, k α , k β are proportional parameters, and c α , c β are differential parameters.

[0106] According to the kinematics relationship of the robot, the control instruction of the task space is calculated to the joint space, and is sent to the servo driver of each axis, so that the actuator executes according to the expected position, speed and torque, and the actual execution is fed back to the control system through the encoder to realize closed-loop control.

[0107] The use effect of the application will be described below:

[0108] The dynamics parameters of the vehicle are shown in Table 2,

[0109] 17 kN / m Rear suspension spring rate coefficient 22 kN / m Front wheel radial stiffness coefficient m c ]]> 192 kN / m Rear wheel radial stiffness coefficient m sa ]]> 192 kN / m Front suspension damper damping coefficient m sb ]]> 1500 N.s / m Rear suspension damper damping coefficient I ox ]] 1500 N.s / m Figure 5 I oy ]] Figures 6 to 8 Figure 6 [[ L x ]]> 2m Figure 7 L y ]]> 1.3m Figure 8 k sa ]]> Figure 9 ​ k sb ]]> ​ ​ k ta ]]> ​ ​ k tb ]]> ​ ​ c sa ]]> ​ ​ c sb ]]> ​

[0110] Table 2

[0111] The laser radar is installed in the front of the vehicle as a visual sensor, and the maximum detection distance is 150 m. When the vehicle travels at a constant speed of 50 km / h on a D-level road, the maximum prediction time of the visual sensing system is T est = L max / v car = 10.8 s, and the time-domain curve of the road surface detected by the visual system is shown in ​ .

[0112] In this example, the multi-dimensional active damping device has vertical, pitch and roll degrees of freedom. The average mechanical delay time of the device is T delay = 0.08 s. Because T delay < T est , the visual sensing system provides sufficient prediction time for the device to perform early control.

[0113] The simulation results are shown in ​ . ​ The solid line is the vertical displacement curve of the vehicle chassis, the dashed line is the displacement curve of the damping device without visual prediction, and the dotted line is the displacement curve of the damping device based on visual prediction. As can be seen from the figure, the damping device without visual prediction has a mechanical delay of 80 ms, which greatly reduces the damping effect. The damping device based on visual prediction compensates for the delay of the system through early perception and early control, thereby achieving better damping effect and a very flat displacement curve, and the corresponding vertical acceleration is also very small.

[0114] ​ The upper half of the figure of ​ is the pitch angle curve and the roll angle curve of the vehicle chassis, and the solid line in the lower half of the figure is the pitch angle and roll angle curve of the damping device without visual prediction, and the dotted line is the pitch angle and roll angle curve of the damping device based on visual prediction. As can be seen from the figure, the damping effect in the pitch and roll directions is significantly improved after visual prediction.

[0115] In one embodiment, on the basis of the above-mentioned embodiment, after step S400, step S500 is further included, comprising:

[0116] The actual vibration information of the vehicle chassis is read in real time according to the inertial navigation unit installed on the vehicle chassis,

[0117] According to the actual vibration information and the predicted vibration information, the prediction model is adjusted.

[0118] Specifically, the robot mechanism includes n execution units, n is the degree of freedom of the robot, and the execution units can be servo electric cylinders, servo hydraulic cylinders, reduction machine connecting rods, etc. In addition, the multi-dimensional active damping device further includes:

[0119] The inertial navigation unit is installed on the base and includes a 3-axis accelerometer, a 3-axis gyroscope, a 3-axis magnetometer, a GPS, etc. The vibration information of the vehicle chassis can be sensed, and the vertical displacement, the pitch angle, the roll angle, and the speed and acceleration information of the corresponding dimensions of the chassis can be obtained through the sensing fusion algorithm processing. The obtained vibration information is compared with the predicted vibration information, and the prediction model is continuously optimized.

[0120] The control unit, the driver, and the encoder: the control unit executes the control algorithm, forms the driving instruction, and sends it to each execution unit through the driver for execution, and the actual execution situation is fed back to the control unit through the encoder.

[0121] In one embodiment, based on the same technical concept, the present application also provides a multi-dimensional active damping system based on visual prediction, which refers to ​ , comprising:

[0122] The information acquisition module 10 is used to acquire the road surface information in front of the vehicle in real time through the visual sensor;

[0123] The information processing module 20 is used to process the road surface information to obtain the time-domain signal curve of the front road surface;

[0124] The prediction module 30 is used to obtain the predicted vibration information of the vehicle chassis according to the prediction model and the time-domain signal curve;

[0125] The active damping module 40 is used to send the control instruction in advance for active damping according to the predicted vibration information.

[0126] Specifically, the information acquisition module 10 is mainly used to acquire the road surface information in front of the vehicle, and then the information processing module 20 is used to process the road surface information to obtain the time-domain signal curve of the front road surface in combination with the vehicle speed and other necessary parameters. The vibration information of the vehicle chassis at the future time is predicted through the prediction module 30, and finally the active damping module 40 is used for active damping.

[0127] In one embodiment, the information processing module 20 specifically includes:

[0128] The computer vision sub-module is used to identify the road surface information through computer vision technology;

[0129] The road surface information processing submodule is configured to process the road surface information according to the recognition result of the computer vision submodule.

[0130] Specifically, the computer vision submodule identifies the road surface information captured by the vision sensor through computer vision technology, and then processes the road surface information in combination with the vehicle speed and other parameters through the road surface information processing submodule to obtain the time-domain signal curve of the road surface in front of the vehicle.

[0131] In one embodiment, the active damping module 40 specifically includes:

[0132] The software delay submodule is configured to determine the time required for software delay.

[0133] The mechanical delay submodule is configured to determine the time required for mechanical delay.

[0134] The damping submodule is configured to actively damp according to the time required for software delay and mechanical delay.

[0135] Specifically, there are two delays in the active damping process, one is the delay of software control, which is determined by the software delay submodule, and the other is the delay time of mechanical transmission, which is determined by the mechanical delay submodule. After determining the total delay time according to the two submodules, the damping submodule is used to actively damp by sending control instructions in advance.

[0136] In one embodiment, the multi-dimensional active damping system based on visual prediction provided by the present application further includes:

[0137] The correction module is configured to obtain the actual vibration information of the vehicle chassis and adjust the prediction model.

[0138] Specifically, the multi-dimensional active damping device includes an inertial navigation unit that can sense the vibration information of the vehicle chassis. After processing by a sensing fusion algorithm, the actual vertical displacement, pitch angle, roll angle, and corresponding dimensional speed and acceleration information of the chassis can be obtained. The actual vibration information and the predicted vibration information are compared, and the prediction model is continuously optimized and adjusted to achieve better use effect.

[0139] They can be implemented by program codes executable by a computing device, so that they can be stored in a storage device for execution by the computing device, or they can be respectively manufactured into individual integrated circuit modules, or a plurality of modules or steps in them can be manufactured into a single integrated circuit module. Thus, the present application is not limited to any specific combination of hardware and software.

[0140] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the related description of other embodiments.

[0141] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0142] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e. they can be located in one place or distributed on multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiment according to actual needs.

[0143] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0144] It should be noted that the above embodiments can be freely combined as needed. The above is only the preferred embodiment of the present application. It should be noted that for those skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which should be considered within the scope of the present application.

Claims

1. A multi-dimensional active damping method based on visual prediction, characterized in that, The method comprises the following steps: real-time acquisition of road surface information in front of the vehicle through a visual sensor; processing of the road surface information to obtain a time-domain signal curve of the road surface in front of the vehicle; construction of a vehicle dynamics equation according to a vehicle dynamics model and in combination with the time-domain signal curve, and acquisition of predicted vibration information within a maximum prediction time based on the vehicle dynamics equation, the predicted vibration information comprising vertical displacement, pitch angle, roll angle, and corresponding speed and acceleration information of the vehicle, and input parameters of the vehicle dynamics equation being road excitation of four tire contact points of the vehicle; advance sending of a control instruction for active vibration reduction according to the predicted vibration information; wherein the vehicle dynamics equation is: r la is the road excitation for the left front tire contact point, r ra is the road excitation for the right front tire contact point, r lb is the road excitation for the left rear tire contact point, r rb is the road excitation for the right rear tire contact point; h la is the absolute displacement of the center of the left front wheel, h ra is the absolute displacement of the center of the right front wheel, h lb is the absolute displacement of the center of the left rear wheel, h rb is the absolute displacement of the center of the right rear wheel; z la is the absolute displacement of the left front suspension connection point, z ra is the absolute displacement of the right front suspension connection point, z lb is the absolute displacement of the left rear suspension connection point, z rb is the absolute displacement of the right rear suspension connection point; the vertical absolute displacement of the center point of the vehicle floor is z A , the angle of rotation about the x-axis is a A , the angle of rotation about the y-axis is β A ; m c is the total mass of the vehicle body, m sa is the mass of the single front wheel, m sb is the mass of the single rear wheel, I ox is the roll moment of inertia of the vehicle body about the center axis, I oy is the pitch moment of inertia of the vehicle body about the center axis, L x is the distance between the front and rear axles, L y is the distance between the left and right wheels, k sa is the front suspension spring stiffness coefficient, k sb is the rear suspension spring stiffness coefficient, k ta is the front wheel radial stiffness coefficient, k tb is the rear wheel radial stiffness coefficient, c sa is the front suspension damper damping coefficient, c sb is the rear suspension damper damping coefficient.

2. The multi-dimensional active damping method based on visual prediction according to claim 1, characterized in that, the real-time acquisition of road surface information in front of the vehicle through a visual sensor specifically comprises the following steps: acquisition of a maximum field of view distance and a minimum field of view distance of the visual sensor according to an installation position of the visual sensor and a forward direction of the vehicle; acquisition of a field of view width of the visual sensor according to the maximum field of view distance and the minimum field of view distance; real-time acquisition of the road surface information within the field of view width.

3. The multi-dimensional active damping method based on visual prediction according to claim 2, characterized in that, the processing of the road surface information to obtain a time-domain signal curve of the road surface in front of the vehicle specifically comprises the following steps: processing of the road surface information through computer vision technology to identify obstacle information and undulation information of the road surface in front of the vehicle; acquisition of a longest prediction time according to a vehicle speed of the vehicle and the maximum field of view distance, the longest prediction time being greater than a delay time of a multi-dimensional active vibration reduction device; acquisition of a time-domain signal curve of the road surface in front of the vehicle according to the vehicle speed of the vehicle, the longest prediction time, and the obstacle information and undulation information of the road surface.

4. The multi-dimensional active damping method based on visual prediction according to claim 1, characterized in that, the advance sending of a control instruction for active vibration reduction according to the predicted vibration information specifically comprises the following steps: acquisition of a delay time of active vibration reduction of the multi-dimensional active vibration reduction device; advance sending of a control instruction for active vibration reduction according to the delay time, the delay time comprising software delay and mechanical delay.

5. The multi-dimensional active damping method based on visual prediction according to any one of claims 1-4, characterized in that, after the advance sending of a control instruction for active vibration reduction according to the predicted vibration information, the method further comprises the following steps: real-time reading of actual vibration information of the vehicle chassis according to an inertial navigation unit installed on the vehicle chassis, comparison of the actual vibration information and the predicted vibration information to adjust the prediction model.

6. A multi-dimensional active damping system based on visual prediction, characterized in that, The method comprises the following steps: an information acquisition module for real-time acquisition of road surface information in front of the vehicle through a visual sensor; an information processing module for processing of the road surface information to obtain a time-domain signal curve of the road surface in front of the vehicle; a prediction module for acquisition of predicted vibration information of a vehicle chassis according to a prediction model and the time-domain signal curve; an active vibration reduction module for advance sending of a control instruction for active vibration reduction according to the predicted vibration information; wherein the prediction module constructs a vehicle dynamics equation according to a vehicle dynamics model and in combination with the time-domain signal curve, and acquires the predicted vibration information within a maximum prediction time based on the vehicle dynamics equation, the predicted vibration information comprising vertical displacement, pitch angle, roll angle, and corresponding speed and acceleration information of the vehicle. And wherein the input parameters of the vehicle dynamics equation are the road excitations of the four tire contact points of the vehicle; Wherein the vehicle dynamics equation is: r la is the road excitation for the left front tire contact point, r ra is the road excitation for the right front tire contact point, r lb is the road excitation for the left rear tire contact point, r rb is the road excitation for the right rear tire contact point; h la h is the absolute displacement of the center of the left front wheel, ra h is the absolute displacement of the center of the right front wheel, lb h is the absolute displacement of the center of the left rear wheel, rb h is the absolute displacement of the center of the right rear wheel; z la is the absolute displacement of the left front suspension connection point, z ra is the absolute displacement of the right front suspension connection point, z lb is the absolute displacement of the left rear suspension connection point, z rb is the absolute displacement of the right rear suspension connection point; the absolute vertical displacement of the center point of the vehicle floor is z A is the angle of rotation about the x-axis, a A is the angle of rotation about the y-axis, β A ; m c is the total mass of the vehicle body, m sa is the mass of the single front wheel, m sb is the mass of the single rear wheel, I ox is the roll moment of inertia of the vehicle body about the center axis, I oy is the pitch moment of inertia of the vehicle body about the center axis, L x is the distance between the front and rear axles, L y is the distance between the left and right wheels, k sa is the front suspension spring stiffness coefficient, k sb is the rear suspension spring stiffness coefficient, k ta is the front wheel radial stiffness coefficient, k tb is the rear wheel radial stiffness coefficient, c sa is the front suspension damper damping coefficient, c sb is the rear suspension damper damping coefficient.

7. The multi-dimensional active damping system based on visual prediction of claim 6, wherein, The information processing module specifically includes: A computer vision submodule for identifying the road surface information through computer vision technology; A road surface information processing submodule for processing the road surface information according to the identification result of the computer vision submodule.

8. The multi-dimensional active damping system based on visual pre-judgment according to claim 6, characterized in that, The active damping module specifically includes: A software delay submodule for determining the time required for software delay; A mechanical delay submodule for determining the time required for mechanical delay; A damping submodule for actively damping according to the time required for software delay and mechanical delay.

9. A multi-dimensional active damping system based on visual prediction according to any one of claims 6-8, characterized in that, Also includes: A correction module for obtaining the actual vibration information of the vehicle chassis and adjusting the prediction model.

Images