Vehicle driving leveling method based on suspension active control

Abstract

The application discloses a vehicle driving leveling method based on suspension active control and belongs to the technical field of suspension control. The method comprises only utilizing physical quantities such as suspension dynamic stroke, pitch angle and roll angle which can be directly measured by a vehicle-mounted sensor to construct a dynamic reference and reference error. The dynamic reference can reflect the general trend of the overall fluctuation of a road surface and ensure that the reference error to be adjusted does not exceed the effective stroke of an actuator. When the vehicle is stationary on a horizontal road surface, the dynamic reference returns to zero. The method solves the practical problems of the dependence and limitation of the existing method on the vertical height of the vehicle body. The application can make the design idea of the driving leveling system more clear, the control method more simple, break through the technical bottleneck of the dependence on the vertical height of the vehicle body and open up a new idea for vehicle driving leveling.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of suspension control, and particularly relates to a vehicle driving leveling method based on suspension active control. BACKGROUND

[0002] A vehicle needs to have the ability to level the vehicle body posture during driving to provide stable and reliable support conditions for the normal operation of special devices carried by the vehicle. For example, an anti-air laser weapon combat vehicle needs to keep the vehicle body horizontal when chasing unmanned aerial vehicles and other targets, so as to facilitate the high-energy laser cannon to lock the laser on a certain point of the target for continuous irradiation to achieve energy attack. When a high-rise fire-fighting robot follows the fire to implement mobile high-rise water spraying operation, the chassis needs to be kept horizontal at all times to prevent the robot with a high-rise arm from falling over. The driving leveling technology of the vehicle is a common demand of important fields related to the national economy and people's livelihood, such as national defense and military affairs, space launch, emergency rescue, and agricultural production.

[0003] Most of the existing methods are directly based on the whole vehicle driving dynamics model, and a controller is designed to output the expected control force or the expected displacement, which is then executed by the actuator to realize driving leveling by adjusting the vehicle body pitch angle, roll angle, and body mass center vertical height (absolute vertical displacement of the vehicle body mass center in space) to converge to a constant reference position. However, this exposes two shortcomings of the existing design method.

[0004] Firstly, the whole vehicle model is suitable for representing the driving mechanical properties of the vehicle, but it is not a friendly form for controller design. On the one hand, the whole vehicle driving dynamics model has more system inputs than controlled states. Directly designing a controller based on this model will result in the need to solve the right inverse matrix of the gain matrix when solving the control algorithm, which is very troublesome. On the other hand, directly taking the vehicle body pitch angle, roll angle, and body mass center vertical height as the controlled states, the mixed control of displacement and attitude needs to be realized, and the control algorithm design and parameter tuning are relatively difficult. Therefore, establishing a model form convenient for controller design will be an important breakthrough for the optimization of control method design.

[0005] Secondly, the way of adjusting the vertical height of the vehicle body needs to be regulated to converge to a constant reference value. On the one hand, the spatial coordinates of the vehicle cannot be measured by external calibration in actual production, and the vertical height of the vehicle body cannot be accurately obtained, which makes it difficult to put the existing control method into practice. On the other hand, the idea of tracking the constant reference value of the vertical height of the vehicle body actually hinders the cooperation between the actuators. Because when the wheel encounters road excitation, the constant reference value only allows the actuator corresponding to the wheel to adjust, and other actuators must remain unchanged to ensure that the vertical height of the vehicle body remains at the reference position. Even when the amplitude of the road excitation exceeds the stroke of the actuator, other actuators cannot provide stroke compensation in time. This will cause the actuator to reach the upper limit of the stroke setting and cause a limit collision, resulting in a poor riding experience and safety hazards. Therefore, the dependence and limitation of the vertical height of the vehicle body have become a prominent technical bottleneck restricting the development of the ride leveling method. Therefore, a ride leveling method for a vehicle using a non-constant reference value is needed to solve the dependence on the vertical height of the vehicle body during the ride leveling process. SUMMARY

[0006] The technical problem to be solved by the present application is to provide a vehicle ride leveling method based on active suspension control, which balances the leveling difficulty and accuracy, and solves the dependence on the vertical height of the vehicle body during the ride leveling process.

[0007] To solve the above technical problems, the technical scheme adopted by the present application is as follows: a vehicle ride leveling method based on active suspension control, comprising the following steps:

[0008] Step 1: Grouping the suspension nodes in the vertical model of the whole vehicle;

[0009] Step 2: Building a dynamic reference based on the dynamic stroke of the suspension;

[0010] Step 3: Building a reference error representing the difference between the state of the spring-loaded part of the suspension node i and the dynamic reference;

[0011] Step 4: Controlling each suspension node i actuator to output adjustment according to the reference error, so that the vertical height of the vehicle body center of gravity converges to the dynamic reference to realize vehicle ride leveling.

[0012] The further improvement of the technical scheme of the present application is that the specific steps of step 1 are as follows:

[0013] Step 1.1: Decomposing the vertical model of the whole vehicle into multi-agent suspension nodes i driven by actuators with mutual coupling characteristics;

[0014] Step 1.2: Directly measuring the suspension dynamic stroke, pitch angle and roll angle physical quantities by the on-board sensor, and building geometric relationship formulas (1) and (2) based on these physical quantities

[0015]

[0016]

[0017] Among them, l a and l b Respectively represent the distance from the front axle and rear axle to the center of mass, l c and l d Respectively represent the vertical distance from the left and right sides of the axle to the center of mass, usually l c =l d = 1 / 2 axis length. θ is the vehicle body pitch angle, is the body roll angle.

[0018] Step 1.3: Group the suspension nodes i into Ω l Group and Ω k Specifically, the vertical height of the suspension node i is compared with the plumb height of the vehicle body mass center, and the geometric relationship (1) can be used to obtain z si -z s , with z si -z s The positive or negative value is used as the criterion. If the vertical height of the suspension node i is above the center of mass of the vehicle body, that is, z si -z s When >0, the corresponding actuator needs to be compressed, divided into Ω l If the vertical height of the suspension node i is below the center of mass of the vehicle body, that is, z si -z s When <0, the corresponding actuator needs to be extended, divided into Ω k Group.

[0019] The further improvement of the technical solution of the present invention is that the specific steps of step 2 are as follows:

[0020] Step 2.1: Design dynamic benchmark z s0,m , as shown in (3)

[0021]

[0022] Where (3) represents the equation l The average value of the motion state of all suspension nodes in the group is equal to Ω k The average value of the motion state of all suspension nodes in the group is summed up first and then divided by 2, h 0,m It is called the dynamic equilibrium quantity, as shown in (5), z tl,m Represents Ω l The state quantity of suspension node i in the group, z tk,m Represents Ω k The state quantity of the suspension node i in the group, the subscript m represents the derivative order, m = 0, 1; n ldenotes Ω l the number of suspension nodes i in the group, n k denotes Ω k the number of suspension nodes i in the group, n

[0023] where, when all suspension nodes i belong to the group Ω l n l =n, n k =0; when all suspension nodes i belong to the group Ω k n l =0, n k =n, where n denotes the total number of suspension nodes, n=n l +n k , the dynamic reference needs to be calculated by formula (4)

[0024]

[0025] Step 2.2: Construct the dynamic balance quantity, introduce the dynamic balance quantity h 0,m in formula (3) and formula (4), which is used to balance the effective stroke of the actuator of the group Ω p and the group Ω q , as shown in (5)

[0026]

[0027] where, Ω p denotes the set of suspension nodes i in the compression state, Ω q denotes the set of suspension nodes i in the extension state; n p denotes the number of suspension nodes i in the set Ω p , n q denotes the number of suspension nodes i in the set Ω q ; Δz stp denotes the dynamic stroke of the actuator in the set Ω q , Δz stq denotes the dynamic stroke of the actuator in the set Ω q ;

[0028] where, when all actuators are compressed, n p =n, n q =0; when all actuators are extended, n p =0, n q =n; in the formula, n is the total number of suspension nodes, n=n p +n q , at this time, the dynamic balance quantity h 0,m needs to be calculated by formula (6);

[0029]

[0030] A further improvement of the technical solution of the present invention is that the specific steps of step 3 are as follows:

[0031] Step 3.1: Obtain the dynamic baseline state information of all suspension nodes i;

[0032] Step 3.2: Directly measure the suspension dynamic travel, pitch angle, and roll angle physical quantities using the onboard sensors on the vehicle body to construct a benchmark error, the benchmark error e si,m , as shown in (7)

[0033]

[0034] A further improvement of the technical solution of the present invention is that the difference between the state of the sprung portion of the suspension node i and the dynamic reference is determined according to step 3.2, specifically as follows:

[0035] Substitute the dynamic reference formula (3) (4) into the reference error formula (7), and use z ti,m =z si,m -Δz sti,m By replacing the variables, the specific benchmark error can be obtained by calculation, as shown in (8)

[0036]

[0037] Where, represents z si.m and Ω l The average value of the differences in the kinematic states of all suspension nodes in the group is calculated; represents z si.m and Ω k The average value of the differences in the kinematic states of all suspension nodes in the group is calculated; Represents the average motion state of all suspension node actuators;

[0038] All actuators belong to Ω l Group time, n l =n,n k =0; all actuators belong to Ω k Group time, n l =0,n k =nAt this time, the reference error needs to be calculated using formula (9)

[0039]

[0040] Where,

[0041] in and Calculated based on geometric relationships (1) and (2); Obtained by actuator stroke sensor measurement value calculation.

[0042] By adopting the technical scheme, the technical progress achieved by the application is as follows: the suspension nodes are decomposed from the whole vehicle model, the suspension nodes are controlled, the pose mixed control problem based on the hyperstatic whole vehicle vertical dynamics model is converted into the simple displacement control problem based on the full-drive type suspension node dynamics model. The dynamic reference is constructed and is completely different from the constant reference. The dynamic reference is dynamically adjusted in real time according to the motion state of the unsprung part, can adapt to the case that the road surface fluctuation is large in a wide range of maneuvering, can improve the adaptability of the vehicle to complex terrain, and the dynamic reference can reflect the general trend of the road surface fluctuation. The dynamic equilibrium quantity is introduced, the effective stroke of the two groups of actuators can be evenly distributed, and the passability of the vehicle can be improved. Finally, by proposing and designing the dynamic reference and the reference error, the technical bottleneck of relying on and limiting the vehicle body vertical height is broken, the practical problem that the existing method relies on and limits the vehicle body vertical height is solved, and the passability of the vehicle is further improved while the ride leveling is realized, which will well support the design of the subsequent low-complexity ride leveling control method. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings;

[0044] Figure 1 is a flow chart of the method of the present application;

[0045] Figure 2 is a schematic diagram of the geometry of the suspension node; DETAILED DESCRIPTION

[0046] The present application will be further described in detail below in combination with embodiments:

[0047] As shown in Figure 1 , it is a flow chart of a vehicle ride leveling method based on suspension active control, which specifically includes the following steps.

[0048] Step 1: grouping the suspension nodes in the whole vehicle vertical model.

[0049] Step 1.1: decomposing the whole vehicle vertical model into a plurality of intelligent agent suspension nodes i driven by actuators with mutual coupling characteristics;

[0050] Step 1.2: as Figure 2As shown, by directly measuring the suspension dynamic stroke, pitch angle and roll angle physical quantities through the vehicle-mounted sensors, geometric relations (1) and (2) are constructed based on these physical quantities

[0051]

[0052]

[0053] wherein l a and l b represent the distance from the front axle and the rear axle to the center of mass, l c and l d represent the vertical distance from the left side and the right side of the axle to the center of mass, and l c = l d = 1 / 2 axle length. θ is the pitch angle of the vehicle body, is the roll angle of the vehicle body. Subsequently, the pose hybrid control problem based on the statically indeterminate whole-vehicle vertical dynamics model can be converted into a pure displacement control problem based on the all-wheel drive suspension node dynamics model.

[0054] Step 1.3: Grouping the suspension nodes i into Ω l group and Ω k group, specifically, comparing the plumb height of the suspension node i with the plumb height of the center of mass of the vehicle body, taking the positive and negative of z si -z s as the criterion, if the plumb height of the suspension node i is above the center of mass of the vehicle body, i.e., z si -z s > 0, then the corresponding actuator needs to be compressed and is divided into Ω l group, if the plumb height of the suspension node i is below the center of mass of the vehicle body, i.e., z si -z s < 0, then the corresponding actuator needs to be stretched and is divided into Ω k group. Wherein the plumb height of the suspension node i and the plumb height of the center of mass of the vehicle body do not need to be measured separately, after the difference is taken, the z si -z s required for grouping the suspension node model can be directly obtained according to the geometric relation of formula (1).

[0055] Step 2: Constructing a dynamic reference based on the suspension dynamic stroke;

[0056] Step 2.1: Designing a dynamic reference z s0,m as shown in (3)

[0057]

[0058] wherein formula (3) represents the average value of the motion state of all suspension nodes in Ω l group and Ω kThe average of all suspension node motion states within the group, h 0,m The dynamic equilibrium quantity, z tl,m denotes Ω l The state quantity of suspension node i within the group, z tk,m denotes Ω k The state quantity of suspension node i within the group, the subscript m denotes the derivative order, m = 0, 1; n l denotes Ω l The number of suspension nodes i within the group, n k denotes Ω k The number of suspension nodes i within the group;

[0059] Where, when all suspension nodes i belong to the Ω l group, n l = n, n k = 0; all suspension nodes i belong to the Ω k group, n l = 0, n k = n, where n denotes the total number of suspension nodes, n = n l + n k At this time, the dynamic reference needs to be calculated through formula (4)

[0060]

[0061] Step 2.2: Construct the dynamic equilibrium quantity in the dynamic reference, introduce the dynamic equilibrium quantity h 0,m in formula (3) and formula (4), which is used to evenly distribute the effective stroke of the actuator of the Ω p group and the Ω q group,

[0062]

[0063] Where, Ω p denotes the set of suspension nodes i in the compression state, Ω q denotes the set of suspension nodes i in the extension state; n p denotes the number of suspension nodes i in the set Ω p , n q denotes the number of suspension nodes i in the set Ω q ; Δz stp denotes the dynamic stroke of the actuator in the set Ω q , Δz stq denotes the dynamic stroke of the actuator in the set Ω q ;

[0064] Where, when all actuators are compressed, n p = n, n q= 0; all actuators are extended, n p = 0, n q = n; in the formula, n is the total number of suspension nodes, n = n p + n q At this time, the dynamic balance amount h 0,m needs to be calculated by formula (6);

[0065]

[0066] It can be seen that the dynamic reference constructed by formula (3) and formula (4) is very different from the setting method of the constant reference. The dynamic reference will follow the real-time dynamic adjustment of the motion state of the unsprung part, can adapt to the case that the road surface fluctuation changes greatly in a wide range of maneuvering process, and can improve the adaptability of the vehicle to complex terrain. One very important setting of the present application is that if the parameters in formula (3) and formula (4) are calculated according to the absolute space vertical motion state of the unsprung part, it is difficult to realize in practice, and the low complexity control method proposed in the present application does not need to directly calculate the dynamic reference, but the difference between the motion state of the sprung part of each suspension node and the dynamic reference is really needed to be calculated.

[0067] Step 3: Construct a reference error for representing the difference between the motion state of the sprung part of each suspension node i and the dynamic reference;

[0068] Step 3.1: Obtain the dynamic reference state information of all suspension nodes i;

[0069] Step 3.2: Directly measure the suspension dynamic stroke, pitch angle and roll angle physical quantities on the vehicle body by the vehicle-mounted sensor, based on which, the reference error is constructed and solved, and the reference error e si,m As shown in (7)

[0070]

[0071] According to step 3.2, the difference between the state of the sprung part of the suspension node i and the dynamic reference is determined, which is as follows:

[0072] The dynamic reference formula (3) (4) is brought into the reference error formula (7), and z ti,m = z si,m - Δz sti,m is substituted, and the specific reference error can be obtained by calculation, as shown in (8)

[0073]

[0074] In the formula, represents z si.m and Ω l The average value of the motion states of all suspension nodes in the group is respectively subtracted. denotes z si.m with Ω k the average of the difference of all suspension node motion states in the group respectively; denotes the average of all suspension node actuator motion states;

[0075] When all actuators belong to the group Ω l n l = n, n k = 0; When all actuators belong to the group Ω k n l = 0, n k = n, the reference error needs to be calculated by equation (9)

[0076]

[0077] In the equation,

[0078] It should be noted that, and are calculated according to geometric relations (1) and (2); can be calculated by the actuator stroke sensor measurement.

[0079] Step 4: Control each suspension node i actuator to output adjustment according to the reference error, so that the vehicle body mass center vertical height converges to the dynamic reference to realize vehicle driving leveling.

[0080] Characteristic analysis

[0081] Through the construction of dynamic reference and reference error, the four characteristics of vehicle driving leveling can be ensured.

[0082] First, the reference error can be calculated based on the measurement value of the on-board sensor. Thanks to the ingenious design of dynamic reference (3) and (4), the difference between the state of the spring-loaded part of the suspension node i and the dynamic reference, i.e. dynamic error e si,m can be calculated by equation (8) and equation (9).

[0083] Second, the dynamic reference z s0,m can reflect the overall trend of the road surface fluctuation, and can ensure that the reference error e si,m that needs to be adjusted will not exceed the effective stroke of the actuator. On the one hand, equations (3) and (4) show that the essence of the dynamic reference is Ω l and Ω kThe average of the vertical heights of the unsprung sections of the two groups can reflect the overall undulation trend of the road surface. On the other hand, the upper and lower bounds of the dynamic reference are defined by Equation (4), which are the upper and lower limits of the suspension travel setting. When the actuators are fully compressed, the upper limit of the dynamic reference is obtained, which is the upper limit of the suspension travel setting; when the actuators are fully extended, the lower limit of the dynamic reference is obtained, which is the lower limit of the suspension travel setting. In other cases, the dynamic reference will be between the upper and lower limits of the suspension travel setting.

[0084] Third, dynamic benchmark z s0,m Based on the plumb bob height design of the unsprung part, this ensures that when the vehicle is stationary and on a level surface, the dynamic reference z s0,m Return to zero. When the road is absolutely flat, ti =0, i=1,2,3,4, at this time, the dynamic reference z can be obtained according to formula (3) or formula (4): s0,m =0.

[0085] Fourth, dynamic equilibrium h 0,m The introduction of Ω p and Ω q The effective travel of the two actuator groups is evenly distributed. This, in turn, improves the vehicle's maneuverability. Ultimately, the proposal and design of a dynamic benchmark and benchmark error overcome the technical bottleneck of relying on and limiting the vertical height of the vehicle body, resolving the practical issues of existing methods that rely on and limit the vertical height of the vehicle body. This further improves the vehicle's maneuverability while achieving driving leveling, and will support the design of subsequent low-complexity driving leveling control methods.

[0086] The embodiments described above are merely descriptions of preferred implementations of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A vehicle leveling method based on active suspension control, characterized in that: Here are the steps: Step 1: Group the suspension nodes in the vertical model of the vehicle; Step 2: Construct a dynamic benchmark based on suspension travel. The specific steps of Step 2 are as follows: Step 2.1: Design dynamic benchmark z s0,m , as shown in the following formula (3) Wherein formula (3) represents the l The average value of the motion state of all suspension nodes in the group is equal to Ω k The average value of the motion state of all suspension nodes in the group is summed up first and then divided by 2, h 0,m It is called dynamic balance, as shown in the following formula (5), z tl,m Represents Ω l The state quantity of suspension node i in the group, z tk,m Represents Ω k The state quantity of the suspension node i in the group, the subscript m represents the derivative order, m = 0, 1; n l Represents Ω l The number of suspension nodes i in the group, n k Represents Ω k The number of suspension nodes i in the group; Among them, when all suspension nodes i belong to Ω l When group, n l =n,n k = 0; all suspension nodes i belong to Ω k When group, n l =0,n k =n, where n represents the total number of suspension nodes, n=n l +n k , at this time the dynamic benchmark needs to be calculated by formula (4) Step 2.2: Construct dynamic balancing quantity. In equations (3) and (4), introduce dynamic balancing quantity h 0,m , as shown in (5), which is used to evenly distribute Ω p Group and Ω q The effective stroke of the actuator, Among them, Ω p represents the set of suspension nodes i in compression state, Ω q represents the set of suspension nodes i in the extended state; n p Represents the set Ω p The number of suspension nodes i, n q Represents the set Ω q The number of suspension nodes i; Δz stp Represents the set Ω p Dynamic stroke of the actuator, Δz stq Represents the set Ω q Dynamic stroke of the actuator; Among them, when all actuators are compressed, n p =n,n q =0; when all actuators are extended, n p =0,n q =n; where n is the total number of suspension nodes, n=n p +n q , at this time, the dynamic equilibrium h 0,m It needs to be calculated by formula (6); Step 3: Construct a reference error representing the difference between the state of the sprung part of the suspension node i and the dynamic reference; Step 4: Control the actuators of each suspension node i to adjust the output according to the reference error, so that the vertical height of the vehicle body's center of mass converges to the dynamic reference to achieve vehicle leveling during driving.

2. The vehicle driving leveling method based on active suspension control according to claim 1, characterized in that: Step 1 The specific steps are as follows: Step 1.1: Decompose the vertical model of the vehicle into a multi-agent suspension node i driven by actuators with mutually coupled characteristics; Step 1.2: Directly measure the suspension dynamic travel, pitch angle, and roll angle physical quantities through on-board sensors, and construct geometric relationships (1) and (2) based on these physical quantities. Among them, l a and l b Respectively represent the distance from the front axle and rear axle to the center of mass, l c and l d Respectively represent the vertical distances from the left and right sides of the axle to the center of mass, l c =l d =1 / 2 axis length; θ is the vehicle body pitch angle, is the body roll angle; Step 1.3: Group the suspension nodes i into Ω l Group and Ω k Specifically, the vertical height of the suspension node i is compared with the plumb height of the vehicle body mass center, and the geometric relationship (1) can be used to obtain z si -z s , with z si -z s The positive or negative value is used as the criterion. If the vertical height of the suspension node i is above the center of mass of the vehicle body, that is, z si -z s When >0, the corresponding actuator needs to be compressed, divided into Ω l If the vertical height of the suspension node i is below the center of mass of the vehicle body, that is, z si -z s When <0, the corresponding actuator needs to be extended, divided into Ω k Group.

3. The vehicle driving leveling method based on active suspension control according to claim 1, characterized in that: Step 3: Step 3.1: Obtain the dynamic baseline state information of all suspension nodes i; Step 3.2: Directly measure the suspension dynamic travel, pitch angle, and roll angle physical quantities using the onboard sensors on the vehicle body to construct a benchmark error, the benchmark error e si,m , as shown in the following formula (7) 4. The vehicle driving leveling method based on active suspension control according to claim 3, characterized in that: Determine the difference between the state of the sprung portion of suspension node i and the dynamic reference according to step 3.2, as follows: Substitute the dynamic reference formula (3) (4) into the reference error formula (7), and use z ti,m =z si,m -Δz sti,m By replacing the variables, the specific benchmark error can be obtained by calculation, as shown in the following formula (8): Where, represents z si.m and Ω l The average value of the differences in the kinematic states of all suspension nodes in the group is calculated; represents z si.m and Ω k The average value of the differences in the kinematic states of all suspension nodes in the group is calculated; Represents the average motion state of all suspension node actuators; All actuators belong to Ω l When group, n l =n,n k =0; all actuators belong to Ω k Group time, n l =0,n k =nAt this time, the reference error needs to be calculated using formula (9) Where,

Images