Suspension systems with multi-dimensional pose adjustment function, their applications, and control methods

By employing a suspension system with 12 actuators and its control method in a split-type flying car, multi-dimensional attitude adjustment was achieved, solving the problems of non-achieving control intentions and mechanical interference in existing suspension systems in split-type flying cars, and improving the system's flexibility and safety.

CN122770673BActive Publication Date: 2026-11-10HEFEI UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611282465.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-24
Publication Date
2026-11-10
Estimated Expiration
2046-08-24

AI Technical Summary

Technical Problem

Existing suspension systems are difficult to achieve multi-dimensional attitude adjustment in split-type flying cars, resulting in the failure to achieve control intentions or mechanical interference. They cannot effectively counteract high-frequency airflow disturbances, and traditional suspensions have slow response speeds and insufficient flexibility.

Method used

The suspension system employs 12 mutually influencing actuators, including 4 hub motors, 4 servo motors, and 4 linear direct drive motors. Through a control method with multi-dimensional posture adjustment function, the actuator parameters are simulated and tire slip compensation and motion interference suppression are performed to achieve omnidirectional planar alignment and spatial posture compensation.

Benefits of technology

It improves the flexibility and safety of the suspension system, enabling it to maintain the consistency of attitude between the mobile chassis and the flight module during dynamic handover, optimize the center of gravity and space utilization, reduce slippage, and enhance resistance to lateral shear.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122770673B_ABST
    Figure CN122770673B_ABST
Patent Text Reader

Abstract

This invention relates to the field of suspension design technology, and discloses a suspension system with multi-dimensional attitude adjustment function, its application, and control method. The suspension system of this invention provides 12 mutually influencing actuators, enabling it to simultaneously and independently handle omnidirectional translation in a plane and three-degree-of-freedom attitude adjustment in vertical space. This invention uses this suspension system as the mobile chassis of a split-type flying car and proposes a targeted control method. First, it simulates the execution of parameter control reference quantities for the 12 actuators. Then, based on tire slippage, it performs hub motor target speed compensation and / or servo motor target rotation angle optimization as needed to ensure the wheels maintain rolling as much as possible and reduce slippage. Next, based on motion interference, it performs redundant coordination allocation as needed to achieve mechanism interference suppression, obtaining safe-to-execute parameter control quantities to achieve attitude tracking of the flight module by the mobile chassis while ensuring safety.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of suspension design technology, and particularly to: 1. a suspension system with multi-dimensional attitude adjustment function; 2. the application of this suspension system with multi-dimensional attitude adjustment function in a split-type flying car; 3. a control method for applying this suspension system with multi-dimensional attitude adjustment function to a split-type flying car. Background Technology

[0002] The suspension system generally refers to the elastic device that connects the wheels and the vehicle body. It is used to buffer road impacts, improve ride comfort, and ensure vehicle handling stability. It is a very important part of the vehicle structure.

[0003] The existing suspension system design lacks sufficient degrees of freedom, and its posture adjustment capability cannot meet the requirements of certain scenarios. For example, in the field of industrial operations, the chassis suspension of transport vehicles (such as AGVs) is generally a simple fixed frame design, and its movement direction is mostly linear, unable to achieve posture adjustment, resulting in low flexibility. In the field of split-type flying cars, air-to-ground docking technology mainly relies on the active landing of the flight module. The mobile chassis on the ground is usually in a stationary or constant-speed linear motion state. However, existing mobile chassis mostly use traditional passive suspension (such as coil springs and shock absorbers) or semi-active suspension (such as magnetorheological shock absorbers), which have slow response speed and insufficient adjustment freedom—they cannot actively adjust the chassis's height (Z-axis) and horizontal tilt angle (roll and pitch angles). This makes it difficult to counteract the high-frequency airflow disturbances generated by the split-type aircraft in the ground effect zone, and it is impossible to maintain the attitude consistency between the mobile chassis and the flight module during dynamic handover. Furthermore, existing mobile chassis mostly use non-omnidirectional movement structures such as Ackerman steering, whose steering angle is limited by mechanical linkages (usually less than 45°). This means that they cannot achieve lateral or diagonal movement in any direction during close-range precision alignment, and must complete attitude correction through multiple reciprocating changes, prolonging the docking time. Although researchers have tried to improve the control accuracy of the flight module, due to the uncertainty of the physical characteristics of near-ground airflow, it is difficult to completely eliminate deviations by relying solely on the control logic of the flight end.

[0004] Although there are some suspension systems that can achieve multi-degree-of-freedom design, their application in split-type flying cars can result in poor control performance. For example, the actual control may fail to achieve the intended control, or the target pose may be achieved but the actual mechanism may not be able to execute safely, and there may be significant interference in the movement of components. Summary of the Invention

[0005] To address the problems that existing suspension systems often fail to achieve the intended control when applied to split-type flying cars, or that the actual mechanism cannot execute safely and there is significant interference in component movement even though the target posture can be achieved, this invention provides a suspension system with multi-dimensional posture adjustment function, its application, and control method.

[0006] This invention is achieved using the following technical solution: In a first aspect, the present invention discloses a control method for a suspension system with multi-dimensional pose adjustment function, which is used to control the suspension system as a mobile chassis of a split flying car to achieve pose tracking of the flight module of the split flying car.

[0007] The mobile chassis has 12 interconnected actuators, including: 4 hub motors that drive the rotation of the 4 wheels, 4 servo motors that change the rolling direction of the 4 wheels, and 4 linear direct drive motors that adjust the spatial orientation of the mobile chassis.

[0008] Control methods for suspension systems with multi-dimensional pose adjustment capabilities include: I. Based on the relative pose error of the moving chassis and flight module at the current moment, the parameter control reference values ​​of 12 actuators are solved, and the execution is simulated. II. The motion trend of each wheel contact point is solved by vector superposition of the wheel-end active velocity, steering additional velocity, and wheel-end passive velocity; where, if the slip risk ratio of the i-th wheel is ρ i = |v perp,i | / (|v parallel,i | + |v perp,i If | + ε) < ρth, then the parameter control reference values ​​of the 12 actuators remain unchanged; otherwise, compensate for the target speed of the hub motor corresponding to the wheel, and / or optimize the target rotation angle of the servo motor corresponding to the wheel, to ensure that the wheel maintains rolling state as much as possible and reduces slippage; where i = 1, 2, 3, 4; v parallel,i v is the velocity component of the contact point velocity of the i-th wheel along the rolling direction; perp,i Let be the velocity component of the i-th wheel's contact point velocity perpendicular to the wheel's rolling direction; ε is a positive number to prevent the denominator from being zero; ρ th This indicates a preset safety threshold; III. If it is predicted that the moving chassis will interfere with each other, then the result of II will be used to suppress the interference and redundancy coordination allocation to obtain the actual action combination to be executed at the current moment; otherwise, the result of II will be used as the actual action combination to be executed at the current moment.

[0009] The control method for this suspension system with multi-dimensional pose adjustment function is implemented according to the method or process of the embodiments of this disclosure.

[0010] Secondly, the present invention discloses a suspension system with multi-dimensional posture adjustment function, which adopts the control method of the suspension system with multi-dimensional posture adjustment function disclosed in the first aspect.

[0011] The implementation of this suspension system with multi-dimensional pose adjustment function is based on the method or process of embodiments of this disclosure.

[0012] Thirdly, the present invention discloses the application of a suspension system with multi-dimensional posture adjustment function as disclosed in the second aspect in a split-type flying car: the suspension system with multi-dimensional posture adjustment function serves as the mobile chassis of the split-type flying car.

[0013] The present invention has the following beneficial effects: 1. This invention provides a suspension system with a multi-degree-of-freedom hardware architecture design, which provides 12 mutually influential actuators. By integrating 4 independent drive units, 4 independent steering units, and 4 active suspension units, the suspension system can simultaneously and independently handle omnidirectional translation in the plane and three-degree-of-freedom attitude adjustment in vertical space, resulting in high flexibility.

[0014] 2. This invention changes the traditional thinking of vertically installed suspension and spring shock absorption. It adopts a controllable telescopic component arranged at an angle as an active suspension. Combined with the parallelogram mechanism formed by the reinforcing rod, the axial telescopic movement of the controllable telescopic component can be converted into the constrained deflection movement of the wheel end actuator relative to the chassis frame. This not only optimizes the chassis center of gravity and internal space, but also enhances the resistance to lateral shear through structural mechanical properties.

[0015] 3. This invention designs a corresponding control method for the proposed suspension system with multi-dimensional attitude adjustment function as the mobile chassis of a split-type flying car. First, it simulates the execution of parameter control reference quantities of 12 actuators. Then, based on the tire slippage situation, it performs target speed compensation of the hub motor and / or optimization of the target rotation angle of the servo motor as needed to ensure that the wheels maintain a rolling state as much as possible and reduce the slippage. Then, based on the motion interference situation, it performs redundancy coordination allocation as needed to achieve mechanism interference suppression, thereby obtaining a safe-to-execute parameter control quantity to achieve attitude tracking of the flight module by the mobile chassis while ensuring safety.

[0016] 4. This invention uses a planar omnidirectional alignment algorithm to process the planar error between the mobile chassis and the flight module. Through rolling optimization prediction, multiple sets of candidate planar motion commands are preset in the current control cycle. After being substituted into the prediction model for evaluation, the optimal command is selected, thereby obtaining the target rotation speed of the four hub motors and the target rotation angle of the four servo motors to correctly represent the target motion intention of the mobile chassis on the plane.

[0017] 5. This invention employs an active suspension pose compensation algorithm to process the spatial attitude error between the mobile chassis and the flight module. First, the extension and retraction of the four active suspension units at the current moment are obtained. The impact of the extension and retraction of each suspension unit on the attitude of the mobile chassis is calculated within a small range. Then, rolling optimization prediction is performed. Multiple sets of candidate spatial motion commands are preset in the current control cycle. They are substituted into the prediction model and evaluated in combination with the local control relationship between each active suspension unit and the chassis frame attitude. The optimal command is then selected, thereby obtaining the target speed of the four linear direct drive motors to correctly represent the target motion intention of the mobile chassis in space. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a three-dimensional view of a suspension system with multi-dimensional pose adjustment function provided in Embodiment 1 of the present invention. Figure 2 This is a three-dimensional view of another form of the suspension system with multi-dimensional pose adjustment function provided in Embodiment 1 of the present invention; Figure 3 The flowchart shows the control method of the suspension system with multi-dimensional pose adjustment function provided in Embodiment 2 of the present invention.

[0020] The symbols in the attached diagram have the following meanings: 1. Chassis frame; 101. Triangle frame; 102. Support rod; 103. Tube sleeve; 2. Independent drive unit; 201. Wheel; 202. Hub motor; 3. Independent steering unit; 301. Steering knuckle; 302. Servo motor; 4. Active suspension unit; 401. Linear direct drive motor; 402. Reinforcing bar. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0022] Example 1 This embodiment 1 provides a suspension system with multi-dimensional posture adjustment function, which mainly includes: a chassis frame as the main structure, and four independent execution modules distributed at the four corners of the chassis frame (i.e., the i-th independent execution module is located at the i-th corner of the four corners of the chassis frame, i=1,2,3,4).

[0023] See Figure 1 , Figure 2 It demonstrates the different operating states of a suspension system with multi-dimensional posture adjustment capabilities.

[0024] ① For weight reduction and stability considerations, the chassis frame is designed with two triangular frame structures spaced apart along its length, connected by three support rods at the corresponding three corners to form a stable three-dimensional frame. The triangular frame structure can be integrally molded or made of three short rods spliced ​​together (e.g., connected by tubular sleeves or directly welded). Similarly, the support rods and short rods can be spliced ​​together (e.g., connected by tubular sleeves or directly welded). Figure 1 The image shows a connection achieved using a sleeve. Of course, screws can be added to the sleeve for further tightening.

[0025] ② The i-th independent execution module includes: the i-th independent drive unit, the i-th independent steering unit, and the i-th active suspension unit.

[0026] 1. The i-th independent drive unit independently drives the wheel located at the i-th corner; the i-th independent drive unit is connected to the i-th independent steering unit and supports independent steering within a range of ±90°.

[0027] Specifically, such as Figure 1 As shown, the i-th independent drive unit can be designed to include: an i-th hub motor and a wheel at the i-th corner. The output of the i-th hub motor is connected to the wheel at the i-th corner to drive the wheel at the i-th corner to rotate independently. The housing of the i-th hub motor is connected to the i-th independent steering unit through a steering knuckle, thereby enabling the i-th independent steering unit to act on the i-th independent drive unit.

[0028] Of course, the i-th independent drive unit can also be replaced by other solutions that can achieve independent drive, such as using a Mecanum wheel structure, with its central axis connected to the i-th independent steering unit through a steering knuckle.

[0029] The i-th independent steering unit can be designed to include: an i-th servo motor, the output of which is connected to a steering knuckle to drive the i-th independent drive unit to achieve independent steering within a range of ±90°. Alternatively, the i-th independent steering unit can be replaced with other solutions capable of achieving independent steering within a range of ±90°, such as: using an i-th steering motor and an i-th worm gear transmission mechanism. The i-th steering motor is connected to a steering knuckle via the i-th worm gear transmission mechanism to drive the i-th independent drive unit to achieve independent steering within a range of ±90°, thereby allowing the four wheels to freely switch between lateral and longitudinal directions.

[0030] In this way, through the coordinated work of four independent drive units and four independent steering units, the chassis frame can achieve longitudinal movement, lateral movement, diagonal movement, and steering.

[0031] It should also be noted that there may be an offset between the steering axis of the servo motor and the center of the corresponding wheel, which will cause additional wheel-end steering motion when the wheel turns.

[0032] II. It should be noted that the key structure of this suspension system lies in the spatial arrangement of the four active suspension units and the kinematic constraint relationship formed between them and the chassis frame and independent steering unit.

[0033] The i-th active suspension unit includes: the i-th controllable telescopic component; it may also include: 4 reinforcing rods; one end of the i-th controllable telescopic component is hinged to the chassis frame, and the other end is hinged to the i-th independent steering unit; the 4 reinforcing rods are arranged in parallel, one end of which is hinged to the chassis frame, and the other end is hinged to the i-th independent steering unit, forming an acute angle with the i-th controllable telescopic component; the i-th active suspension unit is used to realize the relative deflection between the i-th independent actuator and the chassis frame in a direction perpendicular to the length of the chassis frame.

[0034] See Figure 1 , 2 In the i-th active suspension unit, four reinforcing rods are arranged in pairs on both sides of the i-th controllable telescopic component. Viewed along the length of the chassis frame, the reinforcing rods on both sides form two parallel sides of a parallelogram, with the controllable telescopic component located on the diagonal of the parallelogram. Therefore, the controllable telescopic component does not directly push the chassis frame as a vertical lifting component, but rather drives the parallelogram mechanism to undergo constrained deformation through its own length change. This causes the corresponding independent actuator module to deflect relative to the chassis frame around the corresponding hinge point (i.e., deflection in a plane perpendicular to the length of the chassis frame). In other words, as the i-th controllable telescopic component extends or retracts, it causes the parallelogram containing the reinforcing rods to deform, resulting in a relative deflection between the i-th independent actuator module and the chassis frame perpendicular to the length of the chassis frame. Therefore, if the i-th independent actuator module remains stationary, the i-th corner of the chassis frame can be raised or lowered in the opposite direction.

[0035] The i-th controllable telescopic component adopts a non-vertical, inclined arrangement: the two controllable telescopic components at the front of the chassis frame are arranged in an outward octagonal shape, and the two controllable telescopic components at the rear of the chassis frame are also arranged in an outward octagonal shape. This inclined outward octagonal arrangement, in conjunction with the aforementioned parallelogram mechanism, allows the axial telescopic motion of the controllable telescopic component to be converted into a constrained deflection motion of the wheel-end actuator relative to the chassis frame. This approach abandons the traditional concept of vertical suspension mounting and spring shock absorption, which not only facilitates the optimization of the chassis's center of gravity and internal space but also enhances the resistance to lateral shear at the moment of docking through structural mechanical properties.

[0036] When the support position of the corresponding wheel relative to the ground remains essentially unchanged, the relative deflection generated by the active suspension unit can reversely change the spatial height of the corresponding corner point of the chassis frame. Therefore, the four active suspension units not only have corner height adjustment capabilities individually, but can also achieve multi-dimensional posture adjustment of the chassis frame through different combinations.

[0037] When the four controllable telescopic components adjust synchronously in the same direction, the overall height of the chassis frame can be changed. When the front and rear active suspension units use different telescopic combinations, the pitch attitude of the chassis frame can be changed. When the left and right active suspension units use different telescopic combinations, the roll attitude of the chassis frame can be changed. Thus, the four active suspension units, through corner differential adjustment, form a combined adjustment capability for the chassis frame height, pitch angle, and roll angle.

[0038] Furthermore, because the controllable telescopic component is arranged at an angle, its extension and retraction not only changes the height of the corresponding corner point of the chassis frame, but also causes lateral positional changes in the independent steering unit and wheels relative to the chassis frame. Therefore, this mechanism forms a spatially coupled motion relationship distinct from vertically lifting active suspensions. That is, the single axial movement of the controllable telescopic component simultaneously corresponds to changes in the height of the chassis frame corner point and changes in the relative position of the wheel ends. In other words, the extension and retraction of the controllable telescopic component causes displacement of the corresponding wheel relative to the chassis frame, thus creating passive wheel end movement caused by suspension motion.

[0039] Therefore, the above structural relationship provides a mechanical basis for determining the spatial attitude of the chassis frame based on the current extension and retraction of the four active suspension units, as well as for coordinating and compensating for the additional motion at the wheel ends.

[0040] Specifically, such as Figure 1 As shown, a linear direct-drive motor is recommended for the i-th controllable telescopic component. Of course, if the load is large, a hydraulic actuator with a servo proportional valve can also be used for the i-th controllable telescopic component.

[0041] The suspension system provided in this embodiment 1 does not simply combine independent drive, independent steering, and active suspension. Instead, it uses four sets of tilting outward-pointing controllable telescopic components, parallel reinforcing rods, and the parallelogram constraint mechanism they form to convert the axial telescopic motion of the active suspension into the relative deflection motion of the wheel ends under constraint. Furthermore, it uses four-corner differential to form the height, pitch, and roll adjustment capabilities of the chassis frame, establishing a unified connection between the omnidirectional motion actuator and the multi-dimensional posture adjustment mechanism at the mechanical structure level.

[0042] The suspension system with multi-dimensional pose adjustment function based on the above structure supports omnidirectional movement and multi-dimensional pose adjustment by configuring four independent execution modules.

[0043] Example 2 This embodiment 2 applies the suspension system with multi-dimensional posture adjustment function proposed in embodiment 1 to a transport vehicle—the suspension system with multi-dimensional posture adjustment function serves as the chassis suspension of the transport vehicle. Referring to the description in embodiment 1, this transport vehicle supports omnidirectional movement and multi-dimensional posture adjustment, greatly improving its flexibility.

[0044] Example 3 This embodiment 3 applies the suspension system with multi-dimensional posture adjustment function proposed in embodiment 1 to a split-type flying car—the suspension system with multi-dimensional posture adjustment function serves as the mobile chassis of the split-type flying car.

[0045] This embodiment 3 also provides a control method for a suspension system with multi-dimensional pose adjustment function in the above application scenario, which aims to control the suspension system with multi-dimensional pose adjustment function to achieve pose tracking of the flight module of the split flying car while ensuring safety.

[0046] In this embodiment 3, the i-th independent drive unit adopts the design of the i-th hub motor and the i-th wheel disclosed in embodiment 1; the i-th independent steering unit adopts the design of the i-th servo motor disclosed in embodiment 1. The i-th active suspension unit adopts the linear direct drive motor design disclosed in embodiment 1.

[0047] Therefore, the mobile chassis includes 4 hub motors (corresponding to drive the rotation of the 4 wheels), 4 servo motors (corresponding to change the rolling direction of the 4 wheels), and 4 linear direct drive motors (adjusting the spatial posture of the mobile chassis) – a total of 12 actuators. Referring to the description of Embodiment 1, these 12 actuators are not simply isolated integrations, but rather interact with each other and form a coupled structure – they will generate compound wheel-end motions when performing planar motion, steering, and active suspension posture adjustment.

[0048] It is important to emphasize that the control method provided in this embodiment 3 does not directly send action commands to the 12 actuators. Instead, it first obtains the relative pose error between the mobile chassis and the flight module, and then divides the relative pose error into planar error and spatial attitude error. Subsequently, on the one hand, a planar omnidirectional alignment algorithm is used to calculate the planar motion commands of the mobile chassis in the forward, backward, left, right and yaw directions for the planar error. On the other hand, an active suspension pose compensation algorithm is used to calculate the spatial motion commands of the mobile chassis in the height, pitch and roll directions for the spatial attitude error. Finally, the above motion commands are processed by the tire slip and mechanism interference suppression algorithm to achieve verification, compensation and redistribution, thereby obtaining the parameter control quantities that the 12 actuators can safely execute at the current moment.

[0049] See Figure 3 Control methods for suspension systems with multi-dimensional pose adjustment capabilities include: Step 1: Obtain the six-degree-of-freedom pose data of the moving chassis and flight module at the current moment, and calculate the relative pose error.

[0050] The dimensions of the six-degree-of-freedom pose data include: X-axis coordinate, Y-axis coordinate, Z-axis coordinate, pitch angle, roll angle, and yaw angle. Therefore, there are six relative errors (collectively referred to as relative pose errors): forward / backward error (i.e., the difference between the X-axis coordinates of the moving chassis and the flight module), left / right error (i.e., the difference between the Y-axis coordinates of the moving chassis and the flight module), altitude error (i.e., the difference between the Z-axis coordinates of the moving chassis and the flight module), roll angle error (i.e., the difference between the roll angles of the moving chassis and the flight module), pitch angle error (i.e., the difference between the pitch angles of the moving chassis and the flight module), and yaw angle error (i.e., the difference between the yaw angles of the moving chassis and the flight module).

[0051] These six relative errors can be divided into two main categories: forward / backward error, left / right error, and yaw angle error are classified as plane errors; altitude error, roll angle error, and pitch angle error are classified as spatial attitude errors.

[0052] It should be noted that if high precision is not required, six-axis attitude sensors can be installed on the mobile chassis and flight module to obtain the corresponding six-degree-of-freedom pose data.

[0053] If higher accuracy is required, in addition to replacing the six-DOF pose data with higher precision data, six-DOF pose data can also be provided by deploying a multi-source sensing array and using a multi-source data fusion algorithm.

[0054] The multi-source sensing array is deployed on the mobile chassis and / or flight module to provide multi-source sensing information. In this embodiment 3, the multi-source sensing array may include: 1. a vision camera, which is installed on the mobile chassis to identify the position and attitude of the lower surface of the flight module; 2. a depth camera, which is installed on the mobile chassis to obtain the spatial distance between the lower surface of the flight module and the mobile chassis; 3. a positioning base station or radar, which is installed on the mobile chassis and flight module to detect the relative position within a certain range; 4. a six-axis attitude sensor, which is installed on the mobile chassis and flight module to measure the six-axis attitude of the mobile chassis and flight module; 5. a wheel speed encoder, which is installed on the mobile chassis to collect the vehicle's rotational speed to characterize the motion state of the mobile chassis; 6. a telescopic component displacement sensor, which is installed on the telescopic components to obtain the telescopic amount of the four telescopic components.

[0055] It is important to note that: First, the data refresh rates of multi-source sensing arrays differ. Direct use would result in some data being from the current moment and others from the previous moment, leading to inaccurate pose calculations. Second, the data from multi-source sensing arrays reside in different coordinate systems, making direct comparison impossible. Therefore, the multi-source data fusion algorithm not only adds time stamps to the data collected by the multi-source sensing arrays and unifies them to the same time for time synchronization, but also transforms all coordinate relationships between the multi-source sensing arrays to a unified coordinate system (generally set as a moving chassis body coordinate system: the positive X-axis direction is the front of the chassis frame, the positive Y-axis direction is the left side of the chassis frame, the positive Z-axis direction is vertically upward, the positive direction of the pitch angle θ is the rise of the front end of the chassis frame relative to the rear end, and the side roll angle...). The positive direction is the left side of the chassis frame rising relative to the right side, and the positive direction of the yaw angle ψ is the clockwise direction when viewed from the positive direction of the Z-axis downwards, thus obtaining multi-source acquisition data at the same time and in the same coordinate system.

[0056] Secondly, since multi-source sensing arrays may contain data with obvious errors such as acquisition noise and outliers, multi-source data fusion algorithms remove outliers from multi-source acquisition data acquired at the same time and in the same coordinate system (for example, using Kalman filtering for noise reduction and using threshold judgment to extract outliers), and perform data weighted fusion according to the reliability of different sensors (for example, using the measurement variance of the corresponding pose components of each sensor to calculate the weight, and then weighting it with the corresponding pose component measurement value to achieve fusion), thereby obtaining six-degree-of-freedom pose data—which is smoother, more reliable, and more accurate.

[0057] Of course, other known data fusion methods can also be used to achieve this, but it is necessary to ensure that the six-DOF pose data is smoother, more reliable, and more accurate.

[0058] Regardless of which method of acquiring six degrees of freedom pose data is used, once the six degrees of freedom pose data is acquired, the relative errors in the corresponding six directions can be calculated. This allows us to know whether the moving chassis is forward or backward, left or right, higher or lower, lower on the left and higher on the right or lower on the left and higher on the right, lower in the front and higher in the back or higher in the front and lower in the back, and whether the vehicle body is aligned.

[0059] Step two: On the one hand, the planar omnidirectional alignment algorithm is used to process the planar error in order to obtain the parameter control quantities of the four independent drive units and the four independent steering units, namely the target speed of the four hub motors (to directly control the wheel speed at the four corners) and the target rotation angle of the four servos (to directly control the rotation angle of the four independent steering units).

[0060] On the other hand, an active suspension pose compensation algorithm is used to process the spatial attitude error in order to obtain the parameter control quantities of the four active suspension units—that is, the target speeds of the four linear direct drive motors (in order to directly control the extension and retraction speeds of the four active suspension units).

[0061] Specifically: 1. The planar omnidirectional alignment algorithm controls the mobile chassis to move in a planar manner to align the mobile chassis with the center of the flight module—handling deviations in the forward / backward, left / right, and yaw directions. In other words, the planar omnidirectional alignment algorithm requires determining where the mobile chassis should translate or rotate.

[0062] The processing procedure of the planar omnidirectional alignment algorithm is as follows: S2011, based on a threshold-based judgment method, the required planar motion direction correction for the moving chassis is determined according to the planar error: S2011 uses a threshold-based method to determine the relative deviation between the mobile chassis and the flight module in the planar dimension, thereby determining the planar motion direction that the mobile chassis needs to correct.

[0063] For ease of explanation, let the forward and backward error be e_x, the left and right error be e_y, and the yaw angle error be e_ψ, and set the corresponding error allowable thresholds ε_x (i.e., forward and backward error allowable threshold), ε_y (i.e., left and right error allowable threshold), and ε_ψ (i.e., yaw angle error allowable threshold).

[0064] ① If e_x > ε_x, it means the flight module is tilted forward relative to the mobile chassis, so the mobile chassis needs to be moved forward; if e_x < -ε_x, it means the flight module is tilted backward relative to the mobile chassis, so the mobile chassis needs to be moved backward; if |e_x| ≤ ε_x, it means the flight module is aligned with the mobile chassis in the front-back direction, so the mobile chassis does not need to be moved forward or backward.

[0065] ② If e_y > ε_y, it means the flight module is tilted to the left relative to the mobile chassis, so the mobile chassis needs to be moved to the left; if e_y < -ε_y, it means the flight module is tilted to the right relative to the mobile chassis, so the mobile chassis needs to be moved to the right; if |e_y| ≤ ε_y, it means the flight module is aligned with the mobile chassis in the left and right directions, so the mobile chassis does not need to be moved left or right.

[0066] ③ If e_ψ > ε_ψ, it means that the flight module has a clockwise yaw angle relative to the mobile chassis, and the mobile chassis needs to be rotated clockwise; if e_ψ < -ε_ψ, it means that the flight module has a counterclockwise yaw angle relative to the mobile chassis, and the mobile chassis needs to be rotated counterclockwise; if |e_ψ| ≤ ε_ψ, it means that the flight module is aligned with the mobile chassis in the yaw direction, and the mobile chassis does not need to be rotated.

[0067] S2012 directly predicts the control based on the plane error and the plane motion direction to be corrected, and selects the plane motion command that can reduce the plane error and satisfy the plane stability constraint.

[0068] It should be noted that S2012 calculates the desired speed (i.e., the ideal forward / backward, left / right, and yaw angular velocities of the mobile chassis) based on the plane error and the chassis's motion capability (i.e., the forward / backward, left / right, and yaw angular velocities achievable by the mobile chassis under the action of the corresponding motors). This is not simply designed so that a larger plane error results in a larger desired speed. Instead, it considers the optimal movement of the mobile chassis within the prediction time domain (i.e., the next short period of time, comprising N_p control cycles; N_p is an integer greater than 1, adaptively set according to the mobile chassis's response speed, actuator bandwidth, and docking accuracy requirements). Because subsequent docking is required—which demands meeting plane stability constraints (smooth movement on the plane, no frequent changes in direction, no sudden increases in speed, and compliance with motor rated limits)—S2012 performs rolling optimization prediction: predicting multiple control cycles into the future within the current control cycle. This means that although control commands for multiple future control cycles are generated, only the first set of control commands affecting the next control cycle is executed; after re-acquiring the relative pose error in the next control cycle, the commands are repeated.

[0069] Let's take the current control cycle as an example to illustrate: At the beginning of the current control cycle, multiple sets of candidate planar motion commands are preset for the mobile chassis (which are generated by expanding outwards from the planar motion commands of the previous control cycle, including multiple sets of forward and backward speeds, left and right speeds, and yaw rate; the step is determined by the maximum acceleration, control cycle, and actuator limits). Then, each set of candidate planar motion commands is substituted into the planar motion prediction model of the mobile chassis to calculate the state changes of the mobile chassis in the prediction time domain (including changes in position, attitude, and planar error). Commands are then eliminated according to safety limits (for example, if a candidate planar motion command causes speed, rudder angle, or wheel speed to exceed the limit, then that candidate planar motion command is eliminated). Subsequently, among the remaining candidate planar motion commands, the set of candidate planar motion commands with the smallest planar evaluation value is selected as the planar motion command for the current control cycle.

[0070] Among them, the planar evaluation value (which includes at least the planar error variation term, speed magnitude term, speed variation term, and actuator constraint penalty term in the planar dimension) characterizes the unstable impact of the candidate planar motion command on the mobile chassis, and the smaller the value, the better.

[0071] After the above prediction and selection, the most suitable forward and backward speed, left and right speed and yaw rate at the current moment are obtained, which is the planar motion command that can reduce planar error and meet the planar stability constraint.

[0072] The goal of S2012 is not to achieve the fastest possible alignment between the mobile chassis and the flight module center, but rather to strike a balance between minimizing errors, maintaining stable speed, and ensuring actuators do not exceed limits. For example, if the flight module deviates a certain distance to the left relative to the mobile chassis, the mobile chassis will not immediately provide the maximum leftward movement speed. Instead, it will predict the changes in the mobile chassis state under different candidate left and right speeds and select a set of left and right speeds that minimizes left and right errors without causing sudden speed changes or large-angle servo swings as the output of the current control cycle. Other situations are similar and will not be elaborated here.

[0073] S2013 decomposes the planar motion command into four-wheel motion to obtain the target steering angle and target wheel speed of the four wheels.

[0074] The target wheel speed of the four wheels is the target rotational speed of the four hub motors.

[0075] Since the movement of the mobile chassis actually relies on the four hub motors driving the four wheels, based on the planar motion command, the position of each wheel on the mobile chassis is used to calculate which direction it needs to roll in order to achieve the movement of the entire mobile chassis and at what speed.

[0076] Specifically, Establish a vehicle coordinate system with the geometric center of the moving chassis as the origin, and let the position of the i-th wheel relative to the geometric center be (x_i, y_i); where x_i represents the position of the i-th wheel along the X-axis of the vehicle body, and y_i represents the position of the i-th wheel along the Y-axis of the vehicle body.

[0077] If the planar motion command of the current control cycle is forward / backward velocity v_x, left / right velocity v_y, and yaw rate ω_z, then the target velocity at the center of the i-th wheel can be divided into: the X-axis velocity at the center of the i-th wheel v_x - ω_z × y_i; and the Y-axis velocity at the center of the i-th wheel v_y + ω_z × x_i. Then, the target steering angle and target wheel speed of the i-th wheel can be calculated: the target steering angle of the i-th wheel is the direction angle formed by the Y-axis velocity and the X-axis velocity at the center of the i-th wheel; the target wheel speed of the i-th wheel is the target speed at the center of the i-th wheel divided by the wheel radius.

[0078] From a macroscopic perspective, if the movable chassis only needs to move horizontally to the left, then the directions of the four wheels will be aligned as much as possible with the lateral movement direction; if the movable chassis needs to rotate in place, then the directions and speeds of the four wheels will combine to form a rotation around the center of the vehicle; if the movable chassis needs to move horizontally and rotate simultaneously, then each wheel will receive a different direction and speed to satisfy this complex motion; other cases are as listed and will not be elaborated further.

[0079] In either case, the motion of the four wheels must be decomposed to obtain the target steering angle and target wheel speed of the four wheels, so that their combined action on the moving chassis conforms to the planar motion command.

[0080] S2014, based on the target steering angles of the four wheels, performs equivalent solution selection processing to obtain the target rotation angles of the four servos.

[0081] It should be noted that when a wheel rolls in a certain direction, there are two equivalent solutions: one is that the servo motor associated with that wheel turns in that direction, causing the wheel to rotate clockwise; the other is that the servo motor associated with that wheel turns in the opposite direction, causing the wheel to rotate counterclockwise. The two solutions may appear to have the same motion effect, but the target rotation angle of the servo motor will differ.

[0082] Therefore, S2014 performs equivalent solution selection, prioritizing the solution that is closer to the current servo angle, thereby avoiding large-angle rapid oscillations of the servo and achieving smoother and easier execution.

[0083] At this point, the target speeds of the four hub motors and the target rotation angles of the four servos obtained above cannot be directly executed—because these are only the commands required for ideal planar alignment, and do not yet take into account steering axis offset, suspension extension / retraction, and tire slippage. Therefore, the output of the planar omnidirectional alignment algorithm is the intended movement of the moving chassis on the plane, not the final motor commands.

[0084] II. For the active suspension attitude compensation algorithm, its function is to control the lifting, lowering, and tilting of the mobile chassis to ensure it aligns with the flight module—handling deviations in altitude, pitch, and roll. In other words, the active suspension attitude compensation algorithm aims to determine how the mobile chassis should lift, lower, and tilt.

[0085] The processing procedure of the active suspension pose compensation algorithm is as follows: S2021, based on a threshold-based judgment method, determines the spatial motion direction that the mobile chassis needs to compensate for according to the spatial attitude error: S2021 uses a threshold-based method to determine the relative deviation between the mobile chassis and the flight module in the spatial dimension, thereby determining the spatial motion direction that the mobile chassis needs to correct.

[0086] For ease of description, let the altitude error be e_z, the pitch angle error be e_θ, and the roll angle error be... And set the corresponding error allowable thresholds ε_z (i.e., altitude error allowable threshold) and ε_θ (i.e., pitch angle error allowable threshold). (i.e., the allowable threshold for roll angle error).

[0087] ① If e_z > ε_z, it means the flight module is higher than the mobile chassis, so the mobile chassis will rise; if e_z < -ε_z, it means the flight module is lower than the mobile chassis, so the mobile chassis will descend; if |e_z| ≤ ε_z, it means the flight module and the mobile chassis are level, so the mobile chassis does not need to be raised or lowered.

[0088] ② If e_θ > ε_θ, it means that the front end of the flight module is higher than the rear end, and the mobile chassis should perform corresponding pitch compensation; if e_θ < -ε_θ, it means that the front end of the flight module is lower than the rear end, and the mobile chassis should perform corresponding pitch compensation; if |e_θ| ≤ ε_θ, it means that the front and rear ends of the flight module are level, and the mobile chassis does not need to perform pitch compensation.

[0089] ③ If This indicates that the left side of the flight module is higher than the right side, so the moving chassis will perform corresponding tilt compensation; if This indicates that the left side of the flight module is lower than the right side, so the moving chassis will perform corresponding tilt compensation; if This indicates that the left and right sides of the flight module are level, so the mobile chassis does not need to be tilted.

[0090] S2022 generates altitude reference targets, pitch reference targets, and roll reference targets based on spatial attitude errors and the spatial motion directions that need to be compensated.

[0091] The altitude reference target is determined by the altitude of the lower surface of the flight module, the current altitude of the chassis frame, and the preset safety clearance, and is generated by a smooth ascent trajectory; the pitch angle reference target is determined by the current pitch angle of the flight module or its filtered value; and the roll angle reference target is determined by the current roll angle of the flight module or its filtered value.

[0092] These three reference targets are used to subsequently evaluate the effectiveness of each candidate spatial motion command in reducing attitude error.

[0093] S2023 performs control prediction based on spatial attitude error, the required spatial motion direction for compensation, and the local control relationship between each active suspension unit and the chassis frame attitude. It also selects spatial motion commands that can reduce attitude error and meet spatial stability constraints by combining altitude reference target, pitch angle reference target, and roll angle reference target.

[0094] It should be noted that the height of the chassis frame will not suddenly rise to the height reference target, but should gradually approach the lower surface of the flight module along a smooth trajectory—generally, the speed should be small at the beginning and small at the end to avoid excessive impact at the beginning of the ascent and to avoid colliding with the flight module when approaching the target.

[0095] When the flight module experiences slight pitch fluctuations, the chassis frame cannot simply remain level; it must adapt to these pitch changes. If the front of the flight module tilts slightly upward, the chassis frame must compensate for the pitch by differential extension and retraction of the front and rear linear direct drive motors. If the flight module tilts to the side, the chassis frame must compensate for the roll by differential extension and retraction of the left and right linear direct drive motors.

[0096] In other words, the altitude direction is for smooth ascent, while the pitch and roll directions follow the attitude fluctuations of the flight module in real time.

[0097] Because the four active suspension units are often in a tilted state, their extension and retraction do not have a simple one-to-one correspondence with the chassis frame attitude: the extension of one active suspension unit may simultaneously affect the height, pitch angle, and roll angle of the moving chassis, and may also cause the wheels to move relative to the chassis frame. Therefore, it cannot be simply assumed that the extension of a certain active suspension unit will raise the chassis frame connected to it by the same amount; rather, it is necessary to first calculate the impact of the active suspension unit's travel on the chassis frame attitude.

[0098] Specifically: S2023 first obtains the extension and retraction amounts of the four active suspension units at the current moment. Then, it performs small-scale trial calculations near the current position of the moving chassis: if one active suspension unit extends or retracts slightly, how do the height, pitch angle, and roll angle of the moving chassis change? If another active suspension unit extends or retracts slightly, how do the height, pitch angle, and roll angle of the moving chassis change? It iterates through the four active suspension units to obtain the influence relationship between the extension and retraction of the four active suspension units on the chassis frame attitude at the current moment—that is, the local control relationship between each active suspension unit and the chassis frame attitude; then, it performs rolling optimization prediction: Let's take the current control cycle as an example to illustrate: At the beginning of the current control cycle, multiple sets of candidate spatial motion commands are preset for the mobile chassis (which are generated by expanding the spatial motion commands of the previous control cycle, including multiple sets of active suspension unit extension and retraction speeds, and the step is determined by the maximum acceleration, control cycle, and actuator limits). Then, each set of candidate spatial motion commands is substituted into the active suspension nonlinear geometric model or local linearization model of the mobile chassis. Based on the local control relationship between each active suspension unit and the chassis frame attitude, the state changes of the mobile chassis in the prediction time domain are calculated (including: position, attitude, and spatial attitude error changes, where the spatial attitude error change is obtained by comparing the position and attitude of the mobile chassis after executing the candidate spatial motion commands with the height reference target, pitch angle reference target, and roll angle reference target). Commands are eliminated according to safety limits (for example, if a candidate spatial motion command causes the height, pitch angle, or roll angle to exceed the limit, then the candidate spatial motion command is eliminated). Then, among the remaining candidate spatial motion commands, the set of candidate spatial motion commands with the smallest spatial evaluation value is selected as the spatial motion command for the current control cycle.

[0099] Among them, the spatial evaluation value (which includes at least the spatial attitude error change term, velocity magnitude term, velocity change term, and actuator constraint penalty term in the spatial dimension) characterizes the unstable impact of candidate spatial motion commands on the mobile chassis, and the smaller the value, the better.

[0100] After the above prediction and selection, the most suitable extension and retraction speed of the active suspension unit at the current moment is obtained, which is the spatial motion command that can reduce attitude error and meet spatial stability constraints.

[0101] In short, S2023 first identifies the current state of the moving chassis, then identifies the specific control effect of each active suspension unit on the moving chassis, and then combines the three reference targets of S2022 to perform control prediction and selection to obtain appropriate spatial motion commands.

[0102] S2024 performs linear conversion based on spatial motion commands to obtain the target speeds of the four linear direct drive motors.

[0103] As described above, since there is a linear relationship between the extension and retraction speed of the active suspension unit and the displacement speed of the linear direct drive motor inside it, the spatial motion command is converted according to this linear relationship to obtain the target speed of the four linear direct drive motors.

[0104] Therefore, the target speeds of the four linear direct drive motors obtained above cannot be directly executed—because the rotation of the linear direct drive motors will cause the wheels to shift relative to the chassis frame, which may lead to passive slippage of the wheel contact point; at the same time, the differential action between different linear direct drive motors may also cause interference in the mechanism's motion. Therefore, the active suspension pose compensation algorithm outputs the target motion intention of the moving chassis in space, rather than the final motor command.

[0105] In other words, the target rotation speeds of the four hub motors, the target rotation angles of the four servo motors, and the target speeds of the four linear direct drive motors obtained in step two are only reference values ​​for parameter control of the 12 actuators.

[0106] Step 3: The tire slippage and mechanism interference suppression algorithm is used to verify and correct the parameter control reference quantities (i.e., the target speed of the four hub motors, the target rotation angle of the four servo motors, and the target speed of the four linear direct drive motors) of the four independent drive units, four independent steering units, and four active suspension units, so as to obtain the parameter control quantities that the 12 actuators can safely execute and execute.

[0107] As described above, the parameter control reference values ​​of the 12 actuators are only the ideal motion of the moving chassis at the current moment, and cannot be directly sent to the 12 actuators for execution.

[0108] This is because, for a mobile chassis: First, the steering axis of the servo motor is offset from the center of the wheel, so when the wheel turns, it will generate additional steering motion at the wheel end: instead of rotating in place around its own center, it will sweep across a small arc around the offset steering axis. Thus, even if the mobile chassis only intends to change the direction of the wheel, the wheel may drag the ground and slide laterally, causing the mobile chassis to generate additional displacement or yaw. Second, the linear direct drive motor is tilted. When it extends or retracts, it will not only change the height and attitude of the mobile chassis, but may also push the wheel away from the outside of the chassis frame or pull it closer to the inside of the chassis frame, thereby causing passive movement of the wheel contact point. Third, the mobile chassis has a total of 12 actuators. The large number and high degree of freedom make it easy for incoordination to occur, causing some actions to cancel each other out or interfere with each other. Therefore, the tire slippage and mechanism interference suppression algorithm performs safety checks and motion corrections before execution. It determines whether the target speeds of the four hub motors, the target rotation angles of the four servo motors, and the target speeds of the four linear direct drive motors can be smoothly executed by the moving chassis. If direct execution would cause abnormal operation (including tire slippage, vehicle body deviation, mutual interference of motion, etc.), the target speeds of the four hub motors, the target rotation angles of the four servo motors, and the target speeds of the four linear direct drive motors are redistributed to achieve correction, so that the final alignment effect remains unchanged, but the execution process is more stable.

[0109] Specifically: S301, obtain the current state of the moving chassis - including: the steering angle and wheel speed of the four wheels at the current moment, the stroke and displacement speed of the linear direct drive motor, and the attitude and planar speed of the chassis frame.

[0110] The current state of the moving chassis described above serves as the basis for determining slippage and interference.

[0111] S302 performs execution prediction on the mobile chassis based on the target speed of the four hub motors, the target rotation angle of the four servo motors, and the target speed of the four linear direct drive motors. When an operational abnormality occurs, it performs corresponding parameter control corrections to obtain the parameter control quantities that the 12 actuators can safely execute at the current moment.

[0112] The specific procedure for S302 is as follows: S3021 simulates the target rotational speed of four hub motors, the target rotational angle of four servo motors, and the target speed of four linear direct drive motors.

[0113] It should be noted that, at this point, the following has been accomplished: I. Based on the relative pose error of the moving chassis and flight module at the current moment, the parameter control reference quantities of the 12 actuators have been calculated and simulated.

[0114] S3022 predicts the changes in the contact points of the four wheels and performs wheel speed compensation and / or rudder angle optimization as needed to obtain the target rotation speed after compensation of the four hub motors and the target rotation angle after optimization of the four servos.

[0115] The wheel contact point is determined by three factors: first, the target rotational speed of the four hub motors—reflected in the active wheel-end speed resulting from the overall translation and rotation of the moving chassis; second, the target rotation angle of the four servo motors—reflected in the additional speed caused by the steering axis offset when the servo motors turn; and third, the target speed of the four linear direct drive motors—reflected in the passive wheel-end speed caused by the suspension movement when the linear direct drive motors extend and retract. The S3022 then vector-superimposes the active wheel-end speed, the additional steering speed, and the passive wheel-end speed to determine the motion trend of each wheel contact point.

[0116] Specifically, it determines whether the movement of a wheel's contact point conforms to the wheel's rolling direction. If yes, no processing is required; the target speeds of the four hub motors simulated by S3021 are used as the compensated target speeds of the four hub motors, and the target rotation angles of the four servos simulated by S3021 are used as the optimized target rotation angles of the four servos, while keeping the target speeds of the four linear direct drive motors unchanged. Otherwise, the target speed of the hub motor corresponding to the wheel is compensated to obtain the compensated target speed, and / or the target rotation angle of the servo corresponding to the wheel is optimized to obtain the optimized target rotation angle.

[0117] In other words, the S3022 focuses on optimizing hub motors and servos, and does not involve linear direct drive motors for the time being.

[0118] It should be noted that the velocity at the wheel's contact point can be decomposed into two components: one is the velocity component along the wheel's rolling direction (also known as the rollable part), and the other is the velocity component perpendicular to the wheel's rolling direction (also known as the slip risk part). The former is achieved through the rotation of the hub motor, while the latter is the potential slip.

[0119] If the velocity at the wheel's contact point is mainly along the wheel's rolling direction, it means that the wheel's contact point movement conforms to the wheel's rolling direction, i.e., the wheel is rolling normally. If the velocity at the wheel's contact point has a large velocity component perpendicular to the wheel's rolling direction, it means that the wheel's contact point movement does not conform to the wheel's rolling direction and there is a risk of slippage. That is, the wheel has an undesired velocity component perpendicular to the rolling direction, which can easily cause slippage, tire wear, and vehicle body disturbance. Therefore, wheel speed compensation and / or rudder angle optimization should be performed next.

[0120] To quantify this, we define the slip risk ratio ρ of the i-th wheel. i = |v perp,i| / (|v parallel,i | +|v perp,i | + ε); Among them, v parallel,i v is the velocity component of the contact point velocity of the i-th wheel along the rolling direction; perp,i Let ε be the velocity component of the contact point velocity of the i-th wheel perpendicular to the rolling direction of the wheel; ε is a very small positive number to prevent the denominator from being zero.

[0121] If ρ i Less than ρ th If the speed at the wheel's contact point is positive, it indicates that the speed is primarily along the rolling direction; otherwise, it indicates that the wheel does not conform to its rolling direction and there is a risk of slippage. Here, ρth represents a preset safety threshold, set based on tire adhesion conditions, ground material, and the permissible slippage of the mobile chassis.

[0122] 1. To facilitate understanding, an example of wheel speed compensation is provided: If the wheel's contact point is dragged along due to steering shaft offset when the servo is turning, the hub motor will rotate synchronously for a short distance to allow the wheel to release this motion through rolling rather than being dragged on the ground. If the linear direct drive motor extends and pushes the wheel end outward, the hub motor's compensation speed will be increased or decreased according to the passive motion direction of the wheel end to keep the wheel's contact point in a rolling state as much as possible. Other situations are illustrated in the example and will not be repeated here.

[0123] In summary, wheel speed compensation works by adjusting the target speed of the hub motor to ensure that the wheel keeps rolling as much as possible and reduces slippage.

[0124] 2. Rudder angle optimization is still based on the theoretical foundation that there are two equivalent schemes for a wheel rolling in a certain direction—both schemes can achieve the same wheel movement, but the servo rotation angle and slip risk are different. Therefore, the scheme that is closer to the current servo angle is selected first, so that the servo steering action is smaller and smoother.

[0125] In other words, rudder angle optimization is to adjust the target rotation angle of the servo motor to ensure that the wheels keep rolling as much as possible and reduce slippage.

[0126] It should be noted that this completes the following: II. Vector superposition of the wheel-end active speed, steering additional speed, and wheel-end passive speed to solve for the motion trend of each wheel contact point. Where the slip risk of the i-th wheel is greater than ρ i = |v perp,i | / (|v parallel,i | + |v perp,iIf | +ε)<ρth, then keep the parameter control reference values ​​of the 12 actuators unchanged; otherwise, compensate for the target speed of the hub motor corresponding to the wheel, and / or optimize the target rotation angle of the servo motor corresponding to the wheel, so as to ensure that the wheel keeps rolling as much as possible and reduce slippage.

[0127] S3023 predicts whether there is mutual interference in the movement of the mobile chassis, and performs redundancy coordination and allocation as needed to obtain the actual combination of actions to be executed at the current moment (i.e., including 12 actuator action commands) as the parameter control quantity that the 12 actuators can safely execute at the current moment.

[0128] As described above, the extension and retraction of the linear direct drive motors not only affect the attitude of the mobile chassis but also change the position of the wheels relative to the chassis frame—that is, if a linear direct drive motor extends, it may cause the wheel to move away from the chassis frame; if a linear direct drive motor shortens, it may cause the wheel to move closer to the chassis frame. Therefore, S3023 predicts the positional state of each component of the mobile chassis in the next control cycle based on the current stroke, displacement speed, and target speed of the four linear direct drive motors, and determines whether there is motion interference. Motion interference is determined to exist if any of the following conditions are met: First, the predicted stroke of any linear direct drive motor is less than the minimum allowable stroke or greater than the maximum allowable stroke; second, the predicted distance between any wheel and the chassis frame is less than the preset minimum safety distance; third, the predicted angle between any reinforcing rod and the chassis frame or independent steering unit exceeds the preset safety angle range; fourth, the target rotation speed after compensation for any hub motor, the target rotation angle after optimization for the servo motor, or the target speed of any linear direct drive motor exceeds the allowable range of the corresponding actuator; fifth, the predicted slip risk ratio of any wheel after processing by S3022 still exceeds the preset safety threshold. That is, the above five conditions are in an "or" relationship.

[0129] In other words, S3023 will continue to predict the movement of the mechanism based on the results of S3022, thus realizing the risk assessment of whether there is mutual interference between the movements of the mobile chassis.

[0130] Of course, if there is no mutual interference of motion, it means that the results of S3022—including the target speeds of the four hub motors after compensation, the target rotation angles of the four servos after optimization, and the target speeds of the four linear direct drive motors (without optimization / compensation)—can be safely executed and can be directly used as the actual combination of actions to be executed at the current moment. If there is mutual interference of motion, it will not simply stop the whole thing, but will enter a redundant coordination allocation to obtain the actual combination of actions to be executed at the current moment.

[0131] Redundancy coordination allocation utilizes the redundancy of the 12 actuators (the same chassis frame attitude change can often be achieved by more than one combination of linear direct drive motors; the same chassis frame planar motion can also be achieved by more than one combination of hub motors and servo motors) to redistribute them to suppress interference.

[0132] Specifically, the redundancy coordination allocation is based on the results of S3022 and is extended within the allowable adjustment range (e.g., reducing the speed of the linear direct drive motor with risk, increasing or decreasing the target speed of the hub motor of the corresponding wheel, selecting the equivalent solution of the rudder angle, reducing the participation of actuators that are close to the limit, and compensating for the remaining motion by other actuators that are not close to the limit, etc.) to obtain multiple candidate correction schemes.

[0133] For each set of candidate correction schemes, the plane error, spatial attitude error, slip risk ratio, and actuator stroke after simulation are recalculated. Then, the schemes are eliminated according to safety constraints (if a set of candidate correction schemes causes any of the following to occur: plane error exceeding the limit, spatial attitude error exceeding the limit, slip risk ratio exceeding the limit, or actuator stroke exceeding the limit, then the set of candidate correction schemes is eliminated). From the remaining candidate correction schemes, the set with the smallest comprehensive evaluation value is selected as the actual action combination to be executed at the current moment - which is a safer allocation scheme.

[0134] Among them, the comprehensive evaluation value (including at least: residual docking error, slip risk ratio, linear direct drive motor travel margin, servo angle change, hub motor speed change, and linear direct drive motor speed change) characterizes the impact of the candidate correction scheme on the state of the mobile chassis, and the smaller the value, the better.

[0135] This translates to the following for each component: For example, if extending a linear direct drive motor further would cause an excessive outward movement of a wheel, then the speed of that linear direct drive motor would be appropriately reduced, and other linear direct drive motors would take on more of the attitude compensation; if a wheel's steering would produce significant slippage, then other wheels would take on more planar motion compensation; if an actuator is approaching its speed or stroke limit, then its participation would be reduced, and the remaining actuators would be used for execution. Other situations are illustrated in the examples provided and will not be elaborated further.

[0136] It is important to note that the aforementioned redistribution does not mandate the complete elimination of interference. Minor interferences (which can be termed acceptable interferences) that do not affect the planar motion or attitude of the chassis frame are permissible. For example, even if the steering directions of some wheels are not entirely consistent, as long as the chassis frame can still move stably forward and backward or laterally, this is acceptable. In other words, what S3023 truly aims to suppress are motion interferences (which can be termed dangerous interferences) that would cause significant tire slippage, additional vehicle body sway, actuator over-limits, mechanism jamming, or amplified docking errors. This process may involve adjusting the parameter control values ​​of up to 12 actuators: priority is given to adjusting the actuators directly related to dangerous interferences, and other actuators are used for coordination and compensation when necessary; in extreme cases, all 12 actuators may be involved.

[0137] S303 is used to execute parameter control quantities that can be safely executed by 12 actuators at the current moment.

[0138] For the four hub motors, the action command they receive is the target wheel speed after the above processing; for the four servo motors, the action command they receive is the target rotation angle after the above processing; and for the four linear direct drive motors, the action command they receive is the target speed after the above processing.

[0139] It should be noted that this completes the following: if the predicted movement of the mobile chassis interferes with each other, the result of II is used to suppress the interference and redundancy coordination allocation to obtain the actual action combination to be executed at the current moment; otherwise, the result of II is used as the actual action combination to be executed at the current moment.

[0140] After execution is complete, return to step one, and then adjust the pose again as time progresses to achieve real-time looping.

[0141] Example 4 This embodiment 4 discloses a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the control method for the suspension system with multi-dimensional pose adjustment function disclosed in embodiment 3.

[0142] The computer equipment can be either a mobile terminal or a fixed terminal. Examples of the former include mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (such as in-vehicle navigation terminals); examples of the latter include digital TVs and desktop computers.

[0143] This embodiment 4 also discloses a readable storage medium storing computer program instructions. When the computer program instructions are read and executed by a processor, the steps of the control method for the suspension system with multi-dimensional pose adjustment function disclosed in embodiment 3 are performed.

[0144] The readable storage medium may include, but is not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the above.

[0145] This embodiment 4 also discloses a computer program product, including a computer program. When executed by a processor, the computer program implements the steps of the control method for the suspension system with multi-dimensional pose adjustment function disclosed in embodiment 3.

[0146] It should be noted that the computer program used to execute the above can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—as well as conventional procedural programming languages—such as C or similar languages. The computer program can be executed entirely on the user's computer, partially on the user's computer, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer through any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN).

[0147] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0148] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A control method for a suspension system with multi-dimensional pose adjustment function, characterized in that, It is used to control the suspension system as the mobile chassis of the split flying car to achieve attitude tracking of the flight module of the split flying car while ensuring safety. The mobile chassis has 12 mutually influential actuators, including: 4 hub motors that drive the rotation of 4 wheels, 4 servos that change the rolling direction of 4 wheels, and 4 linear direct drive motors that adjust the spatial attitude of the mobile chassis. The control method includes: I. Based on the relative pose error of the moving chassis and flight module at the current moment, the parameter control reference values ​​of 12 actuators are solved, and the execution is simulated. II. The motion trend of each wheel contact point is solved by vector superposition of the wheel-end active velocity, steering additional velocity, and wheel-end passive velocity; where, if the slip risk ratio of the i-th wheel is ρ i = |v perp,i | / (|v parallel,i | +|v perp,i If | + ε) < ρth, then the parameter control reference values ​​of the 12 actuators remain unchanged; otherwise, compensate for the target speed of the hub motor corresponding to the wheel, and / or optimize the target rotation angle of the servo motor corresponding to the wheel, to ensure that the wheel maintains rolling state as much as possible and reduces slippage; where i = 1, 2, 3, 4; v parallel,i v is the velocity component of the contact point velocity of the i-th wheel along the rolling direction; perp,i Let be the velocity component of the i-th wheel's contact point velocity perpendicular to the wheel's rolling direction; ε is a positive number to prevent the denominator from being zero; ρ th This indicates a preset safety threshold; III. If it is predicted that the moving chassis will interfere with each other, then the result of II will be used to suppress the interference and redundancy coordination allocation to obtain the actual action combination to be executed at the current moment; otherwise, the result of II will be used as the actual action combination to be executed at the current moment. Redundancy coordination and allocation methods include: Based on the results of II, several candidate modification schemes were obtained by expanding within the allowable adjustment range; For each set of candidate correction schemes, the plane error, spatial attitude error, slip risk ratio, and actuator stroke after simulation are recalculated. Schemes are then eliminated according to safety constraints, and the set with the smallest comprehensive evaluation value among the remaining candidate correction schemes is selected as the actual action combination to be executed at the current moment.

2. The control method for a suspension system with multi-dimensional pose adjustment function according to claim 1, characterized in that, In section I, the parameter control reference quantities for the 12 actuators include: the target speed of the 4 hub motors, the target rotation angle of the 4 servo motors, and the target speed of the 4 linear direct drive motors; The methods for solving the parameter control reference values ​​of the 12 actuators include: The system acquires the six-degree-of-freedom pose data of the moving chassis and flight module at the current moment and calculates the relative pose error. The relative pose error includes: forward / backward error, left / right error, yaw angle error, altitude error, roll angle error, and pitch angle error. Forward / backward error, left / right error, and yaw angle error are classified as planar errors. Altitude error, roll angle error, and pitch angle error are classified as spatial attitude errors. A planar omnidirectional alignment algorithm is used to process the planar error in order to obtain the target rotation speed of the four hub motors and the target rotation angle of the four servo motors. An active suspension pose compensation algorithm is used to process spatial attitude errors in order to obtain the target speeds of the four linear direct drive motors.

3. The control method for a suspension system with multi-dimensional pose adjustment function according to claim 2, characterized in that, Six-degree-of-freedom pose data is provided by deploying a multi-source sensing array and processing it using a multi-source data fusion algorithm. Or / and, the dimensions of six-DOF pose data include: X-axis coordinate, Y-axis coordinate, Z-axis coordinate, pitch angle, roll angle, and yaw angle.

4. The control method for a suspension system with multi-dimensional pose adjustment function according to claim 2, characterized in that, Methods for handling planar errors using planar omnidirectional alignment algorithms include: S2011, The planar motion direction that the mobile chassis needs to correct is determined based on the planar error using a threshold-based judgment method; S2012, based on the plane error and the plane motion direction to be corrected, directly performs control prediction and selects a plane motion command that can reduce the plane error and satisfy the plane stability constraint. S2012 performs rolling optimization prediction: At the beginning of the current control cycle, multiple sets of candidate planar motion commands are preset for the mobile chassis. Then, each set of candidate planar motion commands is substituted into the planar motion prediction model of the mobile chassis to calculate the state change of the mobile chassis in the prediction time domain. Commands are eliminated according to safety constraints. Then, among the remaining candidate planar motion commands, the set of candidate planar motion commands with the smallest planar evaluation value is selected as the planar motion command for the current control cycle. S2013 decomposes the planar motion command into four-wheel motion to obtain the target steering angle and target wheel speed of the four wheels; among them, the target wheel speed of the four wheels is the target rotational speed of the four hub motors. S2014, based on the target steering angles of the four wheels, the equivalent solution selection process is performed to obtain the target rotation angles of the four servos; Among them, the equivalent solution is selected from the schemes that are closer to the servo angle at the current moment.

5. The control method for a suspension system with multi-dimensional pose adjustment function according to claim 2, characterized in that, Methods for processing spatial attitude errors using active suspension pose compensation algorithms include: S2021, based on a threshold-based judgment method, determines the spatial motion direction that the mobile chassis needs to compensate for according to the spatial attitude error: S2022, Generate altitude reference target, pitch reference target, and roll reference target based on spatial attitude error and the spatial motion direction to be compensated; S2023, based on spatial attitude error, the required spatial motion direction for compensation and the local control relationship between each active suspension unit and the chassis frame attitude, performs control prediction, and selects spatial motion commands that can reduce attitude error and meet spatial stability constraints by combining height reference target, pitch angle reference target and roll angle reference target. S2023 first obtains the extension and retraction of the four active suspension units at the current moment, and then performs a small-scale trial calculation near the current position of the moving chassis to obtain the influence relationship between the extension and retraction of the four active suspension units on the chassis frame attitude at the current moment; then, it performs rolling optimization prediction: at the beginning of the current control cycle, multiple sets of candidate spatial motion commands are preset for the moving chassis, and then each set of candidate spatial motion commands is substituted into the active suspension nonlinear geometric model or local linearization model of the moving chassis. Based on the local control relationship between each active suspension unit and the chassis frame attitude, the state change of the moving chassis in the prediction time domain is calculated, and commands are eliminated according to safety constraints. Then, among the remaining candidate spatial motion commands, the set of candidate spatial motion commands with the smallest spatial evaluation value is selected as the spatial motion command for the current control cycle. S2024 performs linear conversion based on spatial motion commands to obtain the target speeds of the four linear direct drive motors.

6. The control method for a suspension system with multi-dimensional pose adjustment function according to claim 2, characterized in that, In section III, if the following conditions are met: the predicted stroke of any linear direct drive motor is less than the minimum allowable stroke or greater than the maximum allowable stroke; or the predicted distance between any wheel and the chassis frame is less than the preset minimum safety distance; or the predicted angle between any reinforcing rod and the chassis frame or independent steering unit exceeds the preset safety angle range; or the target speed of any hub motor after compensation, the target rotation angle of the servo after optimization, or the target speed of any linear direct drive motor exceeds the allowable range of the corresponding actuator; or the predicted slip risk ratio of any wheel after processing in section II still exceeds the preset safety threshold, then it is determined that there is motion interference.

7. A suspension system with multi-dimensional posture adjustment function, characterized in that, It employs the control method of a suspension system with multi-dimensional pose adjustment function as described in any one of claims 1-6.

8. The suspension system with multi-dimensional pose adjustment function according to claim 7, characterized in that, The suspension system includes: a chassis frame and four independent actuators distributed at the four corners of the chassis frame; The i-th independent execution module is located at the i-th corner of the chassis frame and includes: an i-th independent drive unit, an i-th independent steering unit, and an i-th active suspension unit; i = 1, 2, 3, 4. The i-th independent drive unit independently drives the wheel located at the i-th corner. The i-th independent drive unit is connected to the i-th independent steering unit and supports independent steering within a range of ±90°. The i-th active suspension unit includes: an i-th controllable telescopic component and four reinforcing rods. One end of the i-th controllable telescopic component is hinged to the chassis frame, and the other end is hinged to the i-th independent steering unit. The four reinforcing rods are arranged in parallel, with one end hinged to the chassis frame and the other end hinged to the i-th independent steering unit, forming an acute angle with the i-th controllable telescopic component. The i-th active suspension unit is used to enable the i-th independent execution module to deflect around the corresponding hinge point with the chassis frame. The i-th independent drive unit includes: the i-th hub motor and the wheel at the i-th corner; the output end of the i-th hub motor is connected to the wheel at the i-th corner; the housing of the i-th hub motor is connected to the i-th independent steering unit through a steering knuckle; The i-th independent steering unit includes: the i-th servo motor, whose output is connected to the steering knuckle; The i-th controllable telescopic component is a linear direct drive motor, which is arranged at an angle; the two controllable telescopic components located at the front end of the chassis frame are outward-pointing, and the two controllable telescopic components located at the rear end of the chassis frame are also outward-pointing. In the i-th active suspension unit, the four reinforcing rods are arranged in pairs on both sides of the i-th controllable telescopic component. When viewed along the length of the chassis frame, the reinforcing rods on both sides form two parallel sides of a parallelogram, and the controllable telescopic component is located on the diagonal of the parallelogram.

9. The application of the suspension system with multi-dimensional pose adjustment function as described in claim 7 or 8 in a split-type flying car, characterized in that, The suspension system with multi-dimensional posture adjustment function serves as the mobile chassis of the split-type flying car.

Citation Information

Patent Citations

  • Coordinated control system and method for drive-by-wire chassis of hovercar

    CN120972669A

  • Stair-climbing machine vehicle chassis control system and method based on posture adjustment

    CN122323707A