CN-121973579-A - Full-band vibration and noise suppression method for vehicle active suspension

CN121973579ACN 121973579 ACN121973579 ACN 121973579ACN-121973579-A

Abstract

The invention relates to a full-band vibration and noise suppression method of a vehicle active suspension, which comprises the steps of constructing a high-frequency quarter vehicle active suspension model for describing the dynamic relation between suspension vibration and tire acoustic response, designing a model prediction control strategy based on the high-frequency quarter vehicle active suspension model, solving and obtaining model prediction control force meeting system constraint conditions, constructing a reinforcement learning controller, introducing the model prediction control force as expert reference information into a reinforcement learning training process to obtain an active suspension control model, inputting the current system state into the active suspension control model, outputting the active suspension control quantity, and correspondingly controlling the working state of the active suspension. Compared with the prior art, the invention can give consideration to low-frequency comfort and medium-high frequency noise suppression performance, realize cooperative optimization of vehicle vibration and noise in the vehicle, and promote comprehensive sound vibration comfort of the whole vehicle.

Inventors

ZHANG LIJUN
ZHU ZHEHUI
MENG DEJIAN

Assignees

同济大学

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. The full-band vibration and noise suppression method for the vehicle active suspension is characterized by comprising the following steps of: S1, constructing a high-frequency quarter vehicle active suspension model for describing dynamic relation between suspension vibration and acoustic response of a tire; s2, designing a model prediction control strategy based on a high-frequency quarter vehicle active suspension model, and solving to obtain model prediction control force meeting system constraint conditions; s3, constructing a reinforcement learning controller, and introducing model prediction control force serving as expert reference information into a reinforcement learning training process to obtain an active suspension control model; S4, inputting the current system state into an active suspension control model, and outputting to obtain an active suspension control quantity for correspondingly controlling the working state of the active suspension.
2. The full-band vibration and noise suppression method of a vehicle active suspension according to claim 1, wherein S1 is specifically a method for constructing a high-frequency quarter vehicle active suspension model based on suspension structure dynamics, tire rigidity ring characteristics and tire acoustic cavity modal coupling effects.
3. The method for suppressing full-band vibration and noise of an active suspension of a vehicle according to claim 2, wherein S1 comprises the following steps: S11, introducing the characteristic of a rigid ring of the tire into a dynamic model of a suspension structure, describing the structural dynamic behavior of the tire under a medium-high frequency working condition, and establishing a high-frequency structural dynamic model of a tire-wheel system; S12, considering the influence of structural deformation of the tire under the static load on the acoustic cavity mode, equivalently modeling a main acoustic mode of the acoustic cavity of the tire as a simple harmonic vibration unit, decomposing an original degenerate mode into a vertical mode and a transverse mode, selecting the acoustic mode which plays a leading role in acoustic response in a running state of the vehicle as a modeling object, and describing the dynamic behavior of the acoustic cavity of the tire in a coordinate system rotating along with the tire to establish an acoustic cavity model of the tire; S13, based on the tire rigid ring model and the tire acoustic cavity model, establishing a mechanical coupling relation between acoustic cavity vibration and a tire-wheel system, and introducing an equivalent acting force generated by the acoustic cavity vibration into a tire structure dynamics model to describe a bidirectional coupling effect between a tire acoustic cavity mode and a suspension structure; And S14, integrating the tire structure dynamics model, the tire acoustic cavity model, the suspension system and the vehicle body structure dynamics model, and constructing to obtain the high-frequency quarter vehicle active suspension model.
4. The method for suppressing full-band vibration and noise of an active suspension of a vehicle according to claim 3, wherein the high-frequency structural dynamics model of the tire-wheel system in S11 is specifically: where z s 、z u and z b represent sprung mass, unsprung mass, and displacement of the rigid ring, respectively, m s represents sprung mass, m a represents unsprung mass of the non-rigid ring portion, m b represents rigid ring mass, k b and c b represent wheel-ring stiffness and damping connecting the rim and rigid ring, respectively, k ct represents ground stiffness connecting the rigid ring and the ground, k s , and c s represent suspension stiffness and damping, respectively.
5. The method for suppressing full-band vibration and noise of an active suspension of a vehicle according to claim 4, wherein the acoustic cavity model of the tire in S12 is specifically: Wherein X and Z represent the mass center movement of the closed air in the acoustic cavity of the tire, omega 0 is the natural frequency of the acoustic cavity of the tire in an unloaded state, ζ is the damping ratio, f 0 is the static load applied along the Z axis direction, a and b respectively describe the offset of the natural frequencies of the vertical mode and the transverse mode under the action of the load, a <0, b >0, c is the proportionality coefficient, and f r represents the random dynamic component in the interaction force of the tire and the road surface.
6. The method for suppressing full-band vibration and noise of an active suspension of a vehicle according to claim 5, wherein the equivalent acting force generated by the vibration of the acoustic cavity in S13 is specifically: where w e denotes the effective width of the wheel, r w denotes the wheel radius, and ρ 0 denotes the density of the gas in the tire at equilibrium.
7. The full-band vibration and noise suppression method for the vehicle active suspension according to claim 6, wherein the high-frequency quarter vehicle active suspension model in S14 is specifically: Wherein F a is active suspension control force.
8. The method for suppressing full-band vibration and noise of an active suspension of a vehicle according to claim 1, wherein S2 comprises the following steps: s21, acquiring system state information based on a high-frequency quarter vehicle active suspension model, and constructing a predictive control optimization model containing system constraint conditions; S22, solving a predictive control optimization model in a limited predictive time domain to obtain model predictive control force serving as reference control input of the active suspension.
9. The method for full-band vibration and noise suppression of an active suspension of a vehicle according to claim 8, wherein the system state information includes vehicle body acceleration, suspension travel and tire dynamic load, in particular vehicle body vibration response Jerk of vehicle body Speed of vehicle body Suspension relative displacement z 0 -z u and its variation characteristics - 。
10. The method for suppressing full-band vibration and noise of an active suspension of a vehicle according to claim 9, wherein said S3 comprises the steps of: S31, constructing a state observation quantity of the reinforcement learning controller to be S t = [ , z 0 -z u , - The state observables include system state information capable of characterizing low frequency comfort and mid-high frequency vibration and noise characteristics of the vehicle; S32, defining an action space of the reinforcement learning controller as an active suspension control force F a , wherein the value range of the active suspension control force F a is limited by the physical constraints of a suspension structure and an actuator, and F a ∈[F min ,F max ; S33, constructing a composite rewarding function R (S t , a t )=R time (s t , a t )+R freq (s t , a t ) comprising a time domain performance index and a frequency domain performance index, wherein the time domain performance index R time (s t , a t ) is used for restraining low-frequency riding comfort and suspension work state of a vehicle, the frequency domain performance index R freq (s t , a t ) is used for representing vibration response characteristics related to noise in the vehicle in a target frequency band, and the target frequency band comprises 0-20Hz, 20-50Hz and 50-500Hz frequency bands; S34, constructing a strategy network and a value evaluation network based on a reinforcement learning algorithm of a continuous action space, introducing model prediction control force into a reward function, guiding a value evaluation or strategy updating process of reinforcement learning strategy, and training to obtain an active suspension control model.

Description

Full-band vibration and noise suppression method for vehicle active suspension Technical Field The invention relates to the technical field of vehicle active suspension control, in particular to a full-band vibration and noise suppression method for a vehicle active suspension. Background Along with the continuous improvement of the requirements of the vibration comfort and the sound comfort of the automobile, the problems of the vibration and the noise of the automobile gradually become important factors affecting the quality of the whole automobile. Road excitation is transmitted to the vehicle body structure through the tire and suspension system, which easily causes vehicle body vibration and road noise problems in the vehicle, and adversely affects the riding experience. The vehicle vibration and the noise have different action mechanisms in different frequency ranges, namely, the low-frequency vibration mainly affects riding comfort and vehicle operation stability, and the medium-high frequency vibration can be converted into the noise in the vehicle through structural coupling. Therefore, from the system perspective, realizing full-band cooperative control of vehicle vibration and noise has become an important direction of vehicle chassis technology development. At present, an electronic control suspension system provides an effective means for controlling vehicle vibration by actively adjusting suspension force, but the existing suspension control research is mostly based on the development of a traditional vehicle dynamics model, and focuses on the vehicle body vibration suppression effect in a low-frequency range, however, the traditional vehicle dynamics model usually has insufficient consideration on the high-frequency structural characteristics of tires and the acoustic effects of the tires, and is difficult to accurately describe the transmission process of medium-high frequency vibration and noise in the vehicle. In addition, with the application of the intelligent control method in a suspension system, part of control strategies possibly introduce control force with high frequency change under the condition of lacking high-frequency dynamic constraint, so that medium-high frequency vibration of the suspension and a tire system is excited, and the problem of noise in a vehicle is further aggravated. Disclosure of Invention The invention aims to overcome the defects of the prior art and provide the full-frequency-band vibration and noise suppression method for the vehicle active suspension, which can give consideration to low-frequency comfort and medium-high-frequency noise suppression performance, realize cooperative optimization of vehicle vibration and in-vehicle noise, and promote comprehensive sound vibration comfort of the whole vehicle. The invention aims at realizing the following technical scheme that the full-band vibration and noise suppression method for the vehicle active suspension comprises the following steps: S1, constructing a high-frequency quarter vehicle active suspension model for describing dynamic relation between suspension vibration and acoustic response of a tire; s2, designing a model prediction control strategy based on a high-frequency quarter vehicle active suspension model, and solving to obtain model prediction control force meeting system constraint conditions; s3, constructing a reinforcement learning controller, and introducing model prediction control force serving as expert reference information into a reinforcement learning training process to obtain an active suspension control model; S4, inputting the current system state into an active suspension control model, and outputting to obtain an active suspension control quantity for correspondingly controlling the working state of the active suspension. Further, the step S1 is specifically to construct a high-frequency quarter vehicle active suspension model based on suspension structure dynamics, tire rigidity ring characteristics and tire acoustic cavity modal coupling effect. Further, the step S1 includes the following steps: S11, introducing the characteristic of a rigid ring of the tire into a dynamic model of a suspension structure, describing the structural dynamic behavior of the tire under a medium-high frequency working condition, and establishing a high-frequency structural dynamic model of a tire-wheel system; S12, considering the influence of structural deformation of the tire under the static load on the acoustic cavity mode, equivalently modeling a main acoustic mode of the acoustic cavity of the tire as a simple harmonic vibration unit, decomposing an original degenerate mode into a vertical mode and a transverse mode, selecting the acoustic mode which plays a leading role in acoustic response in a running state of the vehicle as a modeling object, and describing the dynamic behavior of the acoustic cavity of the tire in a coordinate system rotating along with the tire to establish an acoust