CN117962538A - Control method, device, equipment and storage medium of pre-aiming type semi-active suspension - Google Patents
Control method, device, equipment and storage medium of pre-aiming type semi-active suspension Download PDFInfo
- Publication number
- CN117962538A CN117962538A CN202311810629.8A CN202311810629A CN117962538A CN 117962538 A CN117962538 A CN 117962538A CN 202311810629 A CN202311810629 A CN 202311810629A CN 117962538 A CN117962538 A CN 117962538A
- Authority
- CN
- China
- Prior art keywords
- model
- control
- semi
- active suspension
- prediction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000000725 suspension Substances 0.000 title claims abstract description 132
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000005457 optimization Methods 0.000 claims abstract description 41
- 230000006870 function Effects 0.000 claims description 45
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000012360 testing method Methods 0.000 claims description 19
- 230000008859 change Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000013016 damping Methods 0.000 description 36
- 230000001133 acceleration Effects 0.000 description 21
- 238000010586 diagram Methods 0.000 description 18
- 238000006073 displacement reaction Methods 0.000 description 17
- 230000005284 excitation Effects 0.000 description 12
- 238000011156 evaluation Methods 0.000 description 10
- 230000003068 static effect Effects 0.000 description 10
- 230000000694 effects Effects 0.000 description 9
- 238000013528 artificial neural network Methods 0.000 description 8
- 238000001228 spectrum Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 238000003062 neural network model Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 5
- 210000002569 neuron Anatomy 0.000 description 4
- 238000005096 rolling process Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 239000006096 absorbing agent Substances 0.000 description 3
- 230000009471 action Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000035939 shock Effects 0.000 description 3
- 238000003491 array Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000002542 deteriorative effect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003116 impacting effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
Landscapes
- Vehicle Body Suspensions (AREA)
Abstract
The application relates to the technical field of vehicles, in particular to a control method, a device, equipment and a storage medium of a pre-aiming type semi-active suspension, wherein the method comprises the following steps: collecting state information at the current moment; based on a pre-established model prediction controller, predicting a state output vector of the semi-active suspension control system in a prediction time domain according to state information, and obtaining a target optimization function according to the state output vector; and calculating a control sequence meeting the target optimization function and the preset constraint condition according to the preset reference input vector, and obtaining a target control quantity based on the control sequence so as to control the semi-active suspension control system based on the target control quantity. Therefore, the problem that the prior art does not consider the front road surface information and cannot ensure the smoothness and the comfort of the vehicle is solved, the road surface information is integrated, the prediction control based on the road pre-aiming is realized, and the riding comfort and the running smoothness can be improved.
Description
Technical Field
The application relates to the technical field of vehicles, in particular to a control method, a device and equipment of a pre-aiming type semi-active suspension and a storage medium.
Background
Along with the rapid development of the automobile industry, the intelligent degree of the automobile is higher and higher, in order to ensure the safety and reliability of the unmanned automobile, most of development of related technologies is focused on the research on environmental perception, path planning and decision control, and the magnetorheological damper serving as an actuator of a suspension has the advantages of good controllability, low energy consumption, continuously adjustable damping force, rapid response and the like, but the magnetorheological damper has stronger hysteresis characteristic and has larger modeling and controller design difficulty.
Most of control algorithms in the related art calculate control inputs based on the current vehicle state, and do not consider road surface information in front of the vehicle, and evaluation indexes for semi-active suspension control are relatively single, so that smoothness and comfort of the vehicle during running cannot be effectively guaranteed, and the problem is to be solved.
Disclosure of Invention
The application provides a control method, a device, equipment and a storage medium of a pre-aiming type semi-active suspension, which are used for solving the problem that the prior art does not consider the front road surface information and cannot ensure the smoothness and the comfort of a vehicle.
An embodiment of a first aspect of the present application provides a method for controlling a pre-aiming type semi-active suspension, including the following steps:
Collecting state information at the current moment;
Based on a pre-established model prediction controller, predicting a state output vector of the semi-active suspension control system in a prediction time domain according to the state information, and obtaining a target optimization function according to the state output vector;
And calculating a control sequence meeting the target optimization function and a preset constraint condition according to a preset reference input vector, and obtaining a target control quantity based on the control sequence so as to control the semi-active suspension control system based on the target control quantity.
Optionally, in some embodiments, before predicting a state output vector of the semi-active suspension control system in the prediction horizon based on the pre-established model prediction controller according to the state information, the method includes:
Establishing a state space model of the semi-active suspension;
Performing external characteristic test on the magnetorheological damper to obtain identification data, training to obtain a forward model and a reverse model of the magnetorheological damper based on the identification data, and establishing a random pavement model and a convex hull pavement model;
And obtaining the pre-established model prediction controller based on the state space model of the semi-active suspension, the random road surface model, the convex hull road surface model, the forward model and the reverse model.
Optionally, in some embodiments, the predicting, based on the pre-established model, a state output vector of the semi-active suspension control system in a prediction time domain according to the state information includes:
Discretizing a state space model of the semi-active suspension to obtain a target state space equation;
and acquiring a current state quantity and a current time-varying disturbance, and obtaining a state output vector of the prediction time domain by utilizing the target state space equation based on the current state quantity and the current time-varying disturbance.
Optionally, in some embodiments, the obtaining a target optimization function according to the state output vector includes:
acquiring an input vector of a control time domain;
And obtaining the target optimization function based on the state output vector, the input vector of the control time domain and the current time-varying disturbance.
Optionally, in some embodiments, the prediction output vector is:
Yp,c(k+1|k)＝Sxx(k)+SuU(k)+SDD(k)；
Wherein p is a prediction time domain, Y p,c (k+ 1|k) is a p-step prediction output vector of the system, x (k) is a current state quantity, U (k) is a control input vector, D (k) is a current time-varying disturbance, S x is a coefficient matrix of the state quantity in the prediction time domain, S u is a coefficient matrix of the control input vector in the prediction time domain, and S D is a coefficient matrix of the road surface height change rate in the prediction time domain.
Optionally, in some embodiments, the objective optimization function is:
Wherein J is the target optimization function, Γ y is a weighted weight coefficient matrix of control output, Γ u is a weighted weight coefficient matrix of control input, and E p (k+1) is a constraint vector of a state quantity and a road surface height variation quantity at the next time of a prediction time domain.
An embodiment of a second aspect of the present application provides a control device for a pre-aiming type semi-active suspension, including:
the acquisition module is used for acquiring state information at the current moment;
the prediction module is used for predicting a state output vector of the semi-active suspension control system in a prediction time domain based on a pre-established model prediction controller according to the state information, and obtaining a target optimization function according to the state output vector;
And the control module is used for calculating a control sequence meeting the target optimization function and the preset constraint condition according to the preset reference input vector, and obtaining a target control quantity based on the control sequence so as to control the semi-active suspension control system based on the target control quantity.
Optionally, in some embodiments, before predicting a state output vector of the semi-active suspension control system in the prediction horizon according to the state information based on the pre-established model prediction controller, the prediction module includes:
The building unit is used for building a state space model of the semi-active suspension;
The testing unit is used for carrying out external characteristic testing on the magnetorheological damper to obtain identification data, training the magnetorheological damper based on the identification data to obtain a forward model and a reverse model of the magnetorheological damper, and establishing a random pavement model and a convex hull pavement model;
The first generation unit is used for obtaining the pre-established model prediction controller based on the state space model of the semi-active suspension, the random road surface model, the convex hull road surface model, the forward model and the reverse model.
Optionally, in some embodiments, the prediction module includes:
The discretization unit is used for discretizing the state space model of the semi-active suspension to obtain a target state space equation;
the second generation unit is used for acquiring the current state quantity and the current time-varying disturbance, and obtaining the state output vector of the prediction time domain by utilizing the target state space equation based on the current state quantity and the current time-varying disturbance.
Optionally, in some embodiments, the prediction module includes:
an acquisition unit configured to acquire an input vector of a control time domain;
And the optimizing unit is used for obtaining the target optimizing function based on the state output vector, the input vector of the control time domain and the current time-varying disturbance.
Optionally, in some embodiments, the prediction output vector is:
Yp,c(k+1|k)＝Sxx(k)+SuU(k)+SDD(k)；
Wherein p is a prediction time domain, Y p,c (k+ 1|k) is a p-step prediction output vector of the system, x (k) is a current state quantity, U (k) is a control input vector, D (k) is a current time-varying disturbance, S x is a coefficient matrix of the state quantity in the prediction time domain, S u is a coefficient matrix of the control input vector in the prediction time domain, and S D is a coefficient matrix of the road surface height change rate in the prediction time domain.
Optionally, in some embodiments, the objective optimization function is:
Wherein J is the target optimization function, Γ y is a weighted weight coefficient matrix of control output, Γ u is a weighted weight coefficient matrix of control input, and E p (k+1) is a constraint vector of a state quantity and a road surface height variation quantity at the next time of a prediction time domain.
An embodiment of a third aspect of the present application provides an electronic device, including: the system comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the program to realize the control method of the pre-aiming type semi-active suspension according to the embodiment.
An embodiment of a fourth aspect of the present application provides a computer-readable storage medium having stored thereon a computer program for execution by a processor for implementing the control method of a pre-aiming semi-active suspension as described in the above embodiment.
Therefore, a state space model of the whole vehicle semi-active suspension is established, and the state space model of one quarter of the whole vehicle semi-active suspension is taken as a prediction model designed by the model prediction controller. Secondly, performing external characteristic test on the magneto-rheological damper, and obtaining experimental data of control current, piston movement displacement and speed and damping force of the magneto-rheological damper as neural network model identification data; and building two road surface models of random and convex hulls. And thirdly, respectively obtaining a forward model and a reverse model of the magnetorheological damper by adopting a neural network identification method according to the dynamometer data of the magnetorheological damper. Finally, taking the vertical acceleration of the vehicle body as control output, taking the dynamic travel of the suspension as constraint output, simultaneously taking the upper limit and the lower limit of the output of the magnetorheological damper into consideration, integrating road surface information, taking the road surface information as measurable time-varying interference, and designing a model predictive controller based on road pre-aiming.
Additional aspects and advantages of the application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the application.
Drawings
The foregoing and/or additional aspects and advantages of the application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flow chart of a control method of a pre-aiming type semi-active suspension provided according to an embodiment of the application;
FIG. 2 is a schematic diagram of the construction principle of a preset model predictive controller according to an embodiment of the application;
FIG. 3 is a schematic diagram of a vehicle dynamics model provided according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a characteristic curve of a magnetorheological damper having a piston movement speed of 0.052m/s provided in accordance with one embodiment of the present application;
FIG. 5 is a schematic diagram of a characteristic curve of a magnetorheological damper having a piston movement speed of 0.131m/s according to one embodiment of the present application;
FIG. 6 is a schematic diagram of a characteristic curve of a magnetorheological damper having a piston movement speed of 0.262m/s provided in accordance with one embodiment of the present application;
FIG. 7 is a schematic diagram of a characteristic curve of a magnetorheological damper having a piston movement speed of 0.524m/s provided in accordance with one embodiment of the present application;
fig. 8 is a schematic diagram of a convex hull pavement height variation curve and a random pavement height variation curve according to an embodiment of the present application;
FIG. 9 is a schematic diagram of a forward model provided according to one embodiment of the present application;
FIG. 10 is a schematic diagram of an inverse model controller according to one embodiment of the present application;
fig. 11 is a schematic diagram of a vehicle control principle according to an embodiment of the present application;
FIG. 12 is a schematic view showing the variation of the vertical acceleration of the vehicle body when the vehicle under the C-level random road surface passes through the road surface at 30km/h according to one embodiment of the present application;
FIG. 13 is a schematic representation of the variation of suspension travel when a class C random subsurface vehicle passes over a road at 30km/h in accordance with one embodiment of the application;
FIG. 14 is a schematic illustration of a tire dynamic-static load ratio of a class C random sub-road vehicle passing over a road at 30km/h in accordance with one embodiment of the present application;
FIG. 15 is a schematic diagram of the output damping force of a class C random subsurface vehicle at 30km/h through the road according to one embodiment of the application;
FIG. 16 is a schematic view showing the variation of the vertical acceleration of the vehicle body when the vehicle under the C-level random road surface passes through the road surface at 60km/h according to one embodiment of the present application;
FIG. 17 is a schematic representation of the variation of suspension travel when a class C random subsurface vehicle passes over a road at 60km/h in accordance with one embodiment of the present application;
FIG. 18 is a schematic illustration of a tire dynamic-static load ratio for a class C random sub-road vehicle passing over a road at 60km/h in accordance with one embodiment of the present application;
FIG. 19 is a schematic diagram of the output damping force of a class C random sub-surface vehicle passing over a road at 60km/h in accordance with one embodiment of the present application;
FIG. 20 is a schematic representation of the variation in acceleration of a vehicle body as it passes through a convex hull at a speed of 30km/h, in accordance with one embodiment of the present application;
FIG. 21 is a schematic representation of the variation of suspension travel as a vehicle passes through a convex hull at a speed of 30km/h in accordance with one embodiment of the present application;
FIG. 22 is a graphical representation of the change in dynamic to static load ratio of a vehicle passing through a convex hull at a speed of 30km/h, in accordance with one embodiment of the present application;
FIG. 23 is a schematic representation of the variation in output damping force as a vehicle passes through a convex hull at a speed of 30km/h in accordance with one embodiment of the present application;
FIG. 24 is a block diagram of a control device for a pre-addressed semi-active suspension according to an embodiment of the present application;
Fig. 25 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative and intended to explain the present application and should not be construed as limiting the application.
The following describes a control method, a device, an electronic device and a storage medium of a pre-aiming type semi-active suspension according to an embodiment of the application with reference to the accompanying drawings. Aiming at the problem that the related art mentioned in the background art does not consider the front road surface information and cannot guarantee the smoothness and the comfort of a vehicle, the application provides a control method of a pre-aiming type semi-active suspension, wherein in the method, a state space model of the whole vehicle semi-active suspension is established, and the state space model of a quarter vehicle semi-active suspension is taken as a prediction model designed by a model prediction controller. Secondly, performing external characteristic test on the magneto-rheological damper, and obtaining experimental data of control current, piston movement displacement and speed and damping force of the magneto-rheological damper as neural network model identification data; and building two road surface models of random and convex hulls. And thirdly, respectively obtaining a forward model and a reverse model of the magnetorheological damper by adopting a neural network identification method according to the dynamometer data of the magnetorheological damper. Finally, taking the vertical acceleration of the vehicle body as control output, taking the dynamic travel of the suspension as constraint output, simultaneously taking the upper limit and the lower limit of the output of the magnetorheological damper into consideration, integrating road surface information, taking the road surface information as measurable time-varying interference, and designing a model predictive controller based on road pre-aiming.
Specifically, fig. 1 is a schematic flow chart of a control method of a pre-aiming type semi-active suspension according to an embodiment of the present application.
As shown in fig. 1, the control method of the pre-aiming type semi-active suspension comprises the following steps:
in step S101, state information of the current time is acquired.
Specifically, based on a pre-established model predictive controller, state information x (t) is collected at the current time t, and the state output of the predictive system at the future time t, t+N p
In step S102, based on the pre-established model prediction controller, a state output vector of the semi-active suspension control system in a prediction time domain is predicted according to the state information, and a target optimization function is obtained according to the state output vector.
Optionally, in some embodiments, predicting the state output vector of the semi-active suspension control system in the prediction horizon based on the pre-established model predictive controller according to the state information includes: discretizing a state space model of the semi-active suspension to obtain a target state space equation; and acquiring the current state quantity and the current time-varying disturbance, and obtaining a state output vector of the prediction time domain by utilizing a target state space equation based on the current state quantity and the current time-varying disturbance.
Specifically, the embodiment of the application can take sampling time as 0.01s, and discretize a continuous state space equation established by semi-active suspension modeling to obtain the following components:
x(k+1)＝Ax(k)+Buu(k)+Bdd(k)
yc(k)＝Ccx(k)+Dcu(k)；
wherein:
x (k) is the current state quantity,/> U (k) is suspension dynamic deflection at k moment,/>D (k) is the deformation/>, of the tire at time kControl output y c is sprung mass acceleration: /(I)
Defining a prediction time domain as p, controlling the time domain as m, and predicting the state of the obtained future p steps as follows according to the current state quantity x (k) and the time-varying disturbance (pavement height change rate) d (k+i), i=0, 1 … p-1 in the prediction time domain:
x(k+1)＝Ax(k)+Buu(k)+Bdd(k)
x(k+2)＝A2x(k)+ABuu(k)+Buu(k+1)+ABdd(k)+Bdd(k+1)
……
x(k+p)＝Apx(k)+Ap-1Buu(k)+Ap-2Buu(k+1)+…+Ap-mBuu(k+m+1)
+Ap-1Bdd(k)+Ap-2Bdd(k+1)+…+Bdd(k+p-1)
the controlled output of the future p steps can be deduced from the state prediction process as:
yc(k+1)＝CcAx(k)+CcBuu(k)+CcBdd(k)+Dcu(k+1)
yc(k+2)＝CcA2x(k)+CcABuu(k)+CcBuΔu(k+1)+Dcu(k+2)+CccABdd(k)
+CcBdd(k+1)
……
yc(k+p)＝CcApx(k)+CcAp-1Buu(k)+CcAp-2Buu(k+1)+…+Dcu(k+p)
+CcAp-1Bdd(k)+CcAp-2Bdd(k+1)+…+CcBdd(k+p-1)
The output prediction is written into a matrix form, firstly, a p-step prediction output vector Y p,c (k+ 1|k) of the system is defined, an m-step control input vector U (k) is used, and a p-step time-variable disturbance D (k) is as follows:
the predicted output of the system can be written as:
Yp,c(k+1|k)＝Sxx(k)+SuU(k)+SDD(k)；
wherein:
the pretargeting of embodiments of the present application is embodied in predicting system states and outputs, where the disturbance D (k) is known, and where time-varying disturbances are considered in the rolling optimization.
Optionally, in some embodiments, deriving the target optimization function from the state output vector includes: acquiring an input vector of a control time domain; and obtaining a target optimization function based on the state output vector, the input vector of the control time domain and the current time-varying disturbance.
Specifically, in order to improve comfort to a greater extent, vehicle body acceleration is used as an optimization index, suspension dynamic travel and tire dynamic deformation are used as constraint conditions, and meanwhile, control input is considered, and an objective function taking a predicted output variable and the control input as optimization targets is established as follows:
J＝||Γy(Yp,c(k+1|k))||2+‖ΓuU(k)‖2；
Wherein ,Γy＝diag(Γy1,Γy2,…,Γyp),Γu＝diag(Γu1,Γu2,…,Γum) are weighted weight coefficient matrices of control outputs and control inputs, respectively.
Bringing the predicted output into an objective function to obtain:
Wherein E p(k+1)＝-Sxx(k)-SD d (k).
Ignoring the quantity independent of the control output u yields the final objective function as:
Optionally, in some embodiments, before predicting the state output vector of the semi-active suspension control system in the prediction time domain based on the pre-established model prediction controller according to the state information, the method includes: establishing a state space model of the semi-active suspension; performing external characteristic test on the magnetorheological damper to obtain identification data, training to obtain a forward model and a reverse model of the magnetorheological damper based on the identification data, and establishing a random road surface model and a convex hull road surface model; and obtaining a pre-established model prediction controller based on a state space model, a random road surface model, a convex hull road surface model, a forward model and a reverse model of the semi-active suspension.
Specifically, in combination with the illustration of fig. 2, the embodiment of the application aims at the control problem of the magnetorheological semi-active suspension, integrates road surface information, and realizes the predictive control based on road pre-aiming. The method is characterized in that the method replaces the vehicle body vertical acceleration of the riding comfort of a table to be control output, takes the tire dynamic load representing the safe operation stability and the suspension dynamic travel representing the mechanical limitation to be constraint output, considers the upper and lower output limits of the magnetorheological damper of the actuator, takes the road surface information, namely the road surface height change rate, as the measurable time-varying interference according to the idea of road pre-aiming, and designs the road pre-aiming model prediction controller.
In order to achieve the above object, the embodiment of the present application may pre-establish a model predictive controller.
Firstly, a state space model of a semi-active suspension of the whole vehicle is established, a state space model of a quarter of the semi-active suspension of the whole vehicle is taken as a prediction model designed by a model prediction controller, and secondly, external characteristic tests are carried out on the magnetorheological damper, and experimental data of control current, piston motion displacement, speed and damping force of the magnetorheological damper are obtained and are taken as neural network model identification data; and building two road surface models of random and convex hulls. And thirdly, respectively obtaining a forward model and a reverse model of the magnetorheological damper by adopting a neural network identification method according to the dynamometer data of the magnetorheological damper. And finally, taking the vertical acceleration of the vehicle body as control output, taking the dynamic travel of the suspension as constraint output, simultaneously taking the upper and lower limits of the output of the magnetorheological damper into consideration, integrating road surface information, taking the road surface information as measurable time-varying interference, and designing a road pre-aiming model prediction controller. Through simulation experiments, the road pre-aiming model predictive controller can improve riding comfort under the conditions of ensuring safety and meeting damping force and dynamic travel constraint by comparing the passive suspension, and can further improve suspension performance.
1. Semi-active suspension modeling.
Assuming that the body is considered as undamped rigid spring mass of the concentrated parameters; the parts below the body and above the tire are regarded as rigid unsprung mass; neglecting the tire damping, consider only its stiffness effect. The established semi-active suspension model of the whole vehicle is shown in fig. 3. Taking the static balance point of each variable as the origin of coordinates, according to Newton's second law, the kinetic equation of the vehicle is obtained as follows:
Wherein, m s is sprung mass, m u is unsprung mass, k s is suspension spring stiffness, k t is tire stiffness, x s and x u are displacement of sprung mass and unsprung mass respectively, x r is road displacement input, c is uncontrollable damping of the magnetorheological damper, and F d is controllable damping force generated by the magnetorheological damper.
Select state x asWherein x s-xu is suspension dynamic travel,/>For sprung mass velocity, x u-xr is tire deformation,/>For unsprung mass speed, control output y c is taken to be sprung mass acceleration/>(Also called as vehicle body vertical acceleration), taking constraint output y b as [ x s-xu,xu-xr]T, disturbance d as road surface height change rate/>On the premise of pre-aiming, the pavement height change rate is known in real time. The state space equation of the semi-active suspension is established as follows:
wherein:
wherein, table 1 is selected vehicle model parameters.
TABLE 1
| Parameters (parameters) | Sprung mass | Unsprung mass | Spring rate | Tire stiffness | Uncontrollable damping |
| Sign symbol | ms | mu | ks | kt | c |
| Numerical value | 345kg | 45kg | 22000(N/m) | 200000(N/m) | 1000(N·s/m) |
2. Damping characteristics of a magnetorheological damper.
The damping characteristics of the magnetorheological damper are influenced by the current and the piston speed, and the piston movement speed is consistent with the external excitation speed, so that the piston movement speed and the external excitation speed are one variable. Sinusoidal testing of the shock absorber at different currents and piston speeds is required.
(1) Recommended amplitude: determining a test amplitude value to be +/-20 mm according to the total stroke of the shock absorber;
(2) Test speed: four sets of excitation speeds (0.052 m/s, 0.131m/s, 0.262m/s, 0.524 m/s);
(3) Test direction: in the vertical direction (initial position in the middle of the damper).
And sequentially introducing control current: 0A, 0.1A, …, 0.9A, 1.0A. And obtaining a characteristic curve of the magneto-rheological damper. Wherein, FIG. 4 is a characteristic curve of a magneto-rheological damper with a piston movement speed of 0.052m/s, FIG. 5 is a characteristic curve of a magneto-rheological damper with a piston movement speed of 0.131m/s, FIG. 6 is a characteristic curve of a magneto-rheological damper with a piston movement speed of 0.262m/s, and FIG. 7 is a characteristic curve of a magneto-rheological damper with a piston movement speed of 0.524 m/s.
The F-S curve shows that the indicator curve of the magnetorheological damper is very full no matter in the stretching stroke or the compression stroke, which shows that the magnetorheological damper has good damping dissipation characteristic. At each excitation speed, the damping force increases with increasing current value in the coil. As can be seen from the F-V curve, the damping force of the magnetorheological damper shows obvious hysteresis characteristics along with the change of the speed, and the hysteresis phenomenon is more obvious along with the increase of the current value. At the lowest speed, the damping force varies within [ -100N,1200N ] for positive movement and within [ -1500N, -300N ] for negative movement. At a maximum speed of 0.524m/s, the damping force varies within [0N,2000N ] for positive movement and within [ -2500N,0N ] for negative movement. As can be seen from the F-S and F-V curves, the hysteresis loop has strong left-right and up-down symmetry and certain regularity.
Therefore, the damping force of a magnetorheological damper is mainly influenced by two factors, namely the movement speed of a piston and the current. Influence of piston movement speed on damping characteristics: the damper force of the shock absorber increases with increasing speed, the increasing trend being saturation nonlinearity, i.e. the damper force enters the saturation region when the speed increases to a certain extent. The main effects of current on damping characteristics are: the damping force increases with increasing current and the increasing trend is likewise saturated non-linear.
3. And modeling pavement input.
In the vehicle vertical dynamics research, a reasonable road surface excitation model is as important as a vehicle model, and common road surface excitation models comprise a random road surface excitation model and a convex hull road surface excitation model; the random pavement excitation, the pavement fluctuation condition in the time domain is randomly changed and is closest to the real pavement unevenness condition; bump excitation is used to simulate bumps in real pavement, such as speed bumps, well covers, stones, etc.
(1) And modeling a random pavement.
The measured road surface unevenness data can be processed to obtain an unevenness power spectrum, and the road surface power spectrum density can be expressed as:
when the vehicle runs on a road surface with a space frequency of n at a speed v, converting the space power spectrum density to obtain a corresponding time power spectrum:
1) White noise is integrated.
Road surface irregularities are described by a linear system, with random white noise input and road surface irregularities displacement output. The pavement power spectrum density of the random filtering white noise is
q＝H(jω)ω；
Wherein sigma 2 is the variance of white noise omega, the value is 1, and q is the displacement of road surface unevenness.
This can be achieved by:
The differential time domain representation of the road surface irregularities can be obtained as:
The road surface time domain displacement can be solved by integrating the white noise signal described above, and is therefore referred to as a white noise integration method.
2) White noise is filtered.
The filtering white noise is based on the integral white noise, the condition that the road surface spectrum is approximately horizontal in a low frequency range is considered, the lower cut-off frequency f 0 is added into the road surface model, and the power spectrum density is expressed as follows:
the low pass filter transfer function can also be obtained as:
Its time domain displacement is expressed as:
The lower cut-off frequency f 0 is about 0.01Hz, so that the time domain pavement displacement obtained by the solving is consistent with the actual pavement spectrum as much as possible.
(2) Modeling a convex hull pavement.
Convex hull road excitation refers to a short-lived high-intensity road excitation signal due to road relief, and is also often considered to be a short-lived random signal. Impact mathematical model of raised pavement:
Wherein A is the bump height, v is the vehicle speed, L is the length of the bump on the driving road surface, and t 0 is the time for the vehicle to enter the bump.
(3) Road surface modeling results.
The method comprises the steps of selecting an integral white noise method to establish a random road surface waveform, taking a vehicle running speed of v=30 km/h, taking a road surface grade of C, taking G0 to obtain 128 multiplied by 10 < -6 > m < 3 >, selecting a convex hull height A=10 cm, and a bump length L=5 m, wherein t0=0.6 s. The obtained convex hull height change curve and the pavement height change curve are shown in fig. 8.
4. A dynamic model of a magnetorheological damper.
The dynamics model of a magnetorheological damper can be divided into a forward model and a reverse model. The forward model is used for predicting output damping force according to the input current and the relative motion state of the piston, and has the main functions of revealing the operation mechanism of the magnetorheological damper, replacing a real damper in simulation and serving as a force sensor in actual control. The inverse model is used to predict the required control current based on the desired control force and piston motion state, and acts as a damper controller in actual control.
(1) And (5) a forward model.
A BP neural network has only one input layer and one output layer, and hidden layers can be one or more layers according to the requirements of actual research contents. The number of the nodes of the input layer is determined according to the actual research content, the magnetorheological damper forward model is a damping force model, and the damping force F is mainly determined by input current, piston displacement and piston speed without considering the influence of temperature, and the number of the nodes of the input layer is 3 in the embodiment of the application, namely the current I, the piston displacement S and the piston speed V at the moment. The number of hidden layers and the number of neurons have close relations with the complexity of the problem to be solved, the nature of experimental data, and the like, and the structure of the hidden layers largely determines the training speed, the prediction accuracy, and the like of the BP network. In theory, the higher the number of hidden layers, the higher the precision should be, but practical research shows that the too many hidden layer neurons are easy to be excessively fitted, and the too long learning time and poor generalization and adaptation capability can be caused. In the embodiment of the application, a double hidden layer neural network is selected, and the number of neurons of two hidden layers is 12. The forward model structure diagram is shown in fig. 9.
(2) And (5) a reverse model.
The number of nodes of an input layer of the reverse model of the magnetorheological damper is 3, namely damping force F, piston displacement S and piston speed V at the moment respectively, and the node of an output layer is 1, so that the current I at the moment is distributed at 0-1A, normalization is not needed, and the number of neurons of two hidden layers is 12. The control structure of the specific inverse model is shown in fig. 10:
In step S103, a control sequence satisfying the target optimization function and the preset constraint condition is calculated according to the preset reference input vector, and a target control amount is obtained based on the control sequence, so as to control the semi-active suspension control system based on the target control amount.
The preset constraint conditions which the semi-active suspension needs to meet are as follows:
(1) The control input constraint is that the adjustable damping force has upper and lower limits: it can be seen from the characteristic curves of the magnetorheological damper that the damping force of the output is distributed at [ -2500N,2500N ], so the constraint of the damping force can be expressed as |f|+.f dmax, where F dmax =2500N, and thus the constraint of the control input can be expressed as u min≤u(k+i)≤umax, i=1, 2 …, m-1, where u min＝-2500N,umax =2500N.
(2) Constraint output constraints are that there are constraints on suspension dynamic travel and tire dynamic load: 1) The suspension dynamic travel constraint requires a wall pile foundation limiting block which is expressed as |x s-xu|≤Smax, wherein S max is the upper limit of the suspension dynamic travel; 2) Tire dynamic load constraints require that the tire dynamic load k t(xu-xr) be less than the tire static load f ku＝(ms+mu) g, also expressed as constraints on tire deformation: the constraint output therefore satisfies: y min≤yb(k+i)≤ymax, i=1, 2 …, p, where
Predicting constraint output by using a prediction output deriving method to obtain:
Yp,b(k+1|k)＝Sx,bx(k)+Su,bU(k)+SD,bD(k)；
Substituting constraint output into constraint equation to be arranged into standard form to obtain:
And the problem of optimizing the objective function obtained by the constraint equation is converted into a typical quadratic optimal planning problem. Under the conditions that the vehicle state is measurable and the road surface information is pre-aimed, the solving step of the model predictive controller can be divided into:
1. Time t=t 0 gives an initial value u 0. Calculating dynamic response at the moment k according to an objective function and constraint conditions in the control model, judging whether an optimal solution is met, if the optimal solution is not met, further optimizing and solving the QP problem, and calling a quadratic programming function to calculate an optimal control sequence [ u (k), u (k+1), … u (k+m) ] T;
2. the first element u (k) of the system acts on the system and calculates the next state variable x (k+1);
3. when the true time t is less than t end, updating k=k+1, estimating an initial value u of a new pre-aiming interval, taking x (k+1) calculated in the previous step as a current state, and entering a new optimization cycle.
Specifically, referring to fig. 11, according to the embodiment of the present application, a control sequence u (t+k) satisfying an objective function and constraint conditions is calculated according to a reference input vector R (k) = [ R (k+1) R (k+2) … R (k+p) ] T, and a first element u (y) of the control sequence is selected as a control quantity of a controlled system, and acts on an actual system. At the next time t+1, a new control sequence is calculated in the same way and a new first element u (t+1) is implemented.
As will be appreciated by those skilled in the art, the basic idea of MPC (Model-based Predictive Control, model predictive controller) is to obtain a viable solution for closed-loop optimal control by solving the open-loop optimization problem on-line, essentially consisting of three parts, namely predictive Model, rolling optimization and feedback correction.
(1) Prediction model: the predictive control is based on a model, but does not depend on the form of the model, so long as the model used is capable of predicting the future state of the system. The model prediction commonly used includes a convolution model, a state space model, a transfer function model and the like, and in recent years, a neural network prediction model, namely model-free prediction control, is also proposed by a learner.
(2) And (3) rolling optimization: the predictive control targets future control inputs. Predicting the state of the future N p moment based on the model at the moment t, optimizing according to an objective function and constraint conditions to obtain open loop input in the future N m moment, and applying a first control variable. The above process is repeated with the prediction domain becoming [ t+1, t+N p +1] at time t+1.
(3) Feedback correction: in a rolling optimization process, the control input is obtained by open loop optimization without consideration of closed loop unknown disturbances, which can lead to deviations from optimal control of the system. The model predictive control preferentially detects the actual output of the controlled object in each step, and corrects the actual output in real time, so that the anti-interference capability of the system is improved.
The most obvious characteristic of predictive control is explicit processing constraint, the constraint often exists practically, if the constraint is ignored, the control performance is poor, even the system is unstable, and the open loop optimization problem considers the constraint condition to have the following characteristics:
(1) Explicit form: constraints are expressed in their original form in the open loop optimization problem;
(2) Solving on line, constraint can be increased or reduced at any time;
(3) Active processing constraints: predicting whether a constraint violation occurs in the future and taking control action in advance.
In order to improve riding comfort under the conditions of ensuring safety and meeting damping force and dynamic travel constraints, suspension evaluation indexes can be set, and comfort of a vehicle is improved as much as possible in a range meeting dynamic travel through reasonable selection of control parameters.
Among them, the suspension evaluation index can be evaluated from three aspects of ride comfort, suspension stroke, and steering stability.
(1) Riding comfort.
The evaluation method of the smoothness of the automobile is divided into subjective evaluation and objective evaluation. The subjective evaluation may cause a great difference in the evaluation results due to individual differences, and the evaluation results are difficult to determine. Defining a root mean square planting function as:
where N represents the number of data points in the time domain to be solved, Is the vertical acceleration of the vehicle body.
(2) Suspension travel.
The suspension dynamic travel describes the degree of variation of the suspension displacement relative to the static equilibrium position, and should be limited within the allowable travel range, if the upper and lower travel exceeding the limit causes a "breakdown" phenomenon, severely deteriorating the comfort. Therefore, it is required to limit the suspension stroke to a certain displacement range.
|xs-xu|≤Smax；
Wherein S max is the upper limit of the suspension travel.
(3) Steering stability.
The steering stability of a vehicle refers to the ability of the tire to remain in contact with the road surface, i.e., the ability of the wheels to filter out disturbances from the road surface and maintain running stability. The steering stability is therefore also called tire ground contact. In studying the driving dynamics of a vehicle, tire dynamic deformation or wheel dynamic load is often used as an evaluation index for steering stability. Only if the dynamic load of the tire is smaller than the static load, the grounding performance can be ensured.
kt(xu-xr)≤fku；
Wherein f ku＝(ms+mu) g represents the tire static load.
The goal of suspension system design is to improve vehicle comfort and to keep suspension travel and tire deflection within reasonable limits. Because of the contradiction among the three, the vehicle acceleration is taken as an optimization target, the suspension dynamic travel and the tire dynamic deformation are taken as constraint conditions, and the comfort of the vehicle is improved as much as possible in the range meeting the dynamic travel through reasonable selection of control parameters.
In order to further understand the control method of the pre-aiming type semi-active suspension according to the embodiment of the present application, the following details are described in connection with the specific embodiment.
And detecting the designed controller effect according to the random road surface and the convex hull road surface established by the road surface input modeling, wherein the vehicle parameters are selected from the table 1, and the suspension dynamic range constraint is 0.08m. The weights of the control output and the control input in the objective function of all simulation experiments are respectively Γ y＝diag(1,1,…,1),Γu =diag (0.008,0.008, …, 0.008), the prediction time domain p=60, the control time domain m=p=60, the used step length is 0.01s, and the simulation time is 3s.
Under the C-level random road surface, the vehicle is set to have a lower running speed of 30km/h, and the vehicle speed is kept unchanged when the vehicle passes through the road surface. The control effect of the passive suspension and the road pre-aiming model pre-aiming controller is compared. By comparing the vertical acceleration of the vehicle body, the dynamic travel of the suspension and the dynamic-static load ratio of the tire are shown in the experimental result graphs of different controllers, as shown in fig. 12, 13, 14 and 15. Under the action of the road pre-aiming model predictive controller, the vertical acceleration of the vehicle body is reduced to a certain extent, the suspension dynamic strokes of the passive suspension and the semi-active suspension are within the constraint range of < -0.08m and 0.08m, the dynamic-static load ratio of the tire is smaller than 1, and the output damping force is kept between < -2500N and 2500N >, so that the constraint is satisfied, and therefore, the road pre-aiming model predictive controller can obtain good performance output and ensure that the constraint output is within the range only under the condition that the bottom layer is the feedforward inversion controller.
In order to further examine the control effect of the predicted MPC, the vehicle speed is raised to 60km/h, the vehicle speed is kept unchanged when the vehicle passes through the road surface, the vertical acceleration of the vehicle body is compared, the dynamic and static load ratio of the tire is compared with experimental effect diagrams of different controllers, as shown in fig. 16, 17, 18 and 19, it can be seen that the road pre-aiming model prediction controller can still reduce the acceleration of the vehicle body along with the increase of the vehicle speed, but the effect is not obvious like that of the low speed, the dynamic and static load ratios of the suspension of the passive suspension and the semi-active suspension are increased to a certain extent, but the output damping force is not in the constraint range, so that the road pre-aiming model prediction controller can improve the performance output (the vertical acceleration of the vehicle body) at the relatively high vehicle speed and meet the constraints of the driving safety (the dynamic and static load ratio of the tire) and the mechanical mechanism and the actuator (the damping force).
Next, the effect of the convex hull subsurface controller was verified. The convex hull height a=10 cm and the length l=5m are chosen, and the car enters the convex hull at t=0.6s. The running speed is 30km/h, the speed is kept unchanged when the vehicle passes through the road surface, the prediction time domain=the control time domain=60, and the pre-aiming step length is 60 steps, namely 0.6s. The control effect of the controller is predicted by comparing the passive suspension and the road pre-aiming model, as shown in fig. 20, 21, 22 and 23. In terms of performance output, the semi-active suspension can obviously reduce the acceleration of a vehicle body under the action of the road pre-aiming model predictive controller, better riding comfort can be obtained compared with a passive suspension, the passive suspension exceeds the dynamic range of the suspension aiming at the mechanical structure constraint of the suspension, the phenomenon of impacting a limiting block is obvious, the constraint of the dynamic range of the suspension is violated, in addition, the pre-aiming controller also reduces the dynamic and static load ratio of a tire, and the running safety is improved.
According to the control method of the pre-aiming type semi-active suspension, a state space model of the whole vehicle semi-active suspension is established, and the state space model of the quarter vehicle semi-active suspension is taken as a prediction model designed by a model prediction controller. Secondly, performing external characteristic test on the magneto-rheological damper, and obtaining experimental data of control current, piston movement displacement and speed and damping force of the magneto-rheological damper as neural network model identification data; and building two road surface models of random and convex hulls. And thirdly, respectively obtaining a forward model and a reverse model of the magnetorheological damper by adopting a neural network identification method according to the dynamometer data of the magnetorheological damper. Finally, taking the vertical acceleration of the vehicle body as control output, taking the dynamic travel of the suspension as constraint output, simultaneously taking the upper limit and the lower limit of the output of the magnetorheological damper into consideration, integrating road surface information, taking the road surface information as measurable time-varying interference, and designing a model predictive controller based on road pre-aiming.
Next, a control device of a pre-aiming type semi-active suspension according to an embodiment of the present application will be described with reference to the accompanying drawings.
Fig. 24 is a block schematic diagram of a control device for a pre-addressed semi-active suspension according to an embodiment of the present application.
As shown in fig. 24, the control device 10 of the pre-aiming type semi-active suspension includes: the system comprises an acquisition module 100, a prediction module 200 and a control module 300.
The collection module 100 is configured to collect state information at a current time.
The prediction module 200 is configured to predict a state output vector of the semi-active suspension control system in a prediction time domain according to state information based on a pre-established model prediction controller, and obtain a target optimization function according to the state output vector.
The control module 300 is configured to calculate a control sequence that satisfies a target optimization function and a preset constraint condition according to a preset reference input vector, and obtain a target control amount based on the control sequence, so as to control the semi-active suspension control system based on the target control amount.
Optionally, in some embodiments, before predicting the state output vector of the semi-active suspension control system in the prediction time domain based on the pre-established model prediction controller according to the state information, the prediction module 200 includes: the device comprises a building unit, a testing unit and a first generating unit.
The system comprises a building unit, a control unit and a control unit, wherein the building unit is used for building a state space model of the semi-active suspension.
The testing unit is used for carrying out external characteristic testing on the magnetorheological damper to obtain identification data, training the magnetorheological damper based on the identification data to obtain a forward model and a reverse model of the magnetorheological damper, and establishing a random pavement model and a convex hull pavement model.
The first generation unit is used for obtaining a pre-established model prediction controller based on a state space model, a random road surface model, a convex hull road surface model, a forward model and a reverse model of the semi-active suspension.
Optionally, in some embodiments, the prediction module includes: a discretization unit and a second generation unit.
The discretization unit is used for discretizing the state space model of the semi-active suspension to obtain a target state space equation.
The second generation unit is used for acquiring the current state quantity and the current time-varying disturbance, and obtaining a state output vector of the prediction time domain by utilizing the target state space equation based on the current state quantity and the current time-varying disturbance.
Optionally, in some embodiments, the prediction module includes: an acquisition unit and an optimization unit.
The acquisition unit is used for acquiring the input vector of the control time domain.
And the optimizing unit is used for obtaining a target optimizing function based on the state output vector, the input vector of the control time domain and the current time-varying disturbance.
Optionally, in some embodiments, the predicted output vector is:
Yp,c(k+1|k)＝Sxx(k)+SuU(k)+SDD(k)；
Wherein p is a prediction time domain, Y p,c (k+ 1|k) is a p-step prediction output vector of the system, x (k) is a current state quantity, U (k) is a control input vector, D (k) is a current time-varying disturbance, S x is a coefficient matrix of the state quantity in the prediction time domain, S u is a coefficient matrix of the control input vector in the prediction time domain, and S D is a coefficient matrix of the road surface height change rate in the prediction time domain.
Optionally, in some embodiments, the objective optimization function is:
wherein J is a target optimization function, Γ y is a control output weighting coefficient matrix, Γ u is a control input weighting coefficient matrix, and E p (k+1) is a constraint vector of a state quantity and a road surface height variation quantity at the next time of a prediction time domain.
It should be noted that the foregoing explanation of the embodiment of the control method of the pre-aiming type semi-active suspension is also applicable to the control device of the pre-aiming type semi-active suspension of this embodiment, and will not be repeated here.
According to the control device for the pre-aiming type semi-active suspension, provided by the embodiment of the application, the state space model of the whole vehicle semi-active suspension is established, and the state space model of the quarter vehicle semi-active suspension is taken as a prediction model designed by a model prediction controller. Secondly, performing external characteristic test on the magneto-rheological damper, and obtaining experimental data of control current, piston movement displacement and speed and damping force of the magneto-rheological damper as neural network model identification data; and building two road surface models of random and convex hulls. And thirdly, respectively obtaining a forward model and a reverse model of the magnetorheological damper by adopting a neural network identification method according to the dynamometer data of the magnetorheological damper. Finally, taking the vertical acceleration of the vehicle body as control output, taking the dynamic travel of the suspension as constraint output, simultaneously taking the upper limit and the lower limit of the output of the magnetorheological damper into consideration, integrating road surface information, taking the road surface information as measurable time-varying interference, and designing a model predictive controller based on road pre-aiming.
Fig. 25 is a schematic structural diagram of an electronic device according to an embodiment of the present application. The electronic device may include:
Memory 2501, processor 2502, and computer programs stored on memory 2501 and executable on processor 2502.
The processor 2502, when executing the program, implements the control method of the pre-aiming semi-active suspension provided in the above embodiment.
Further, the electronic device further includes:
a communication interface 2503 for communication between the memory 2501 and the processor 2502.
Memory 2501 for storing computer programs that may be run on the processor 2502.
The memory 2501 may comprise high speed RAM (Random Access Memory ) memory, and may also include non-volatile memory, such as at least one disk memory.
If the memory 2501, the processor 2502, and the communication interface 2503 are implemented independently, the communication interface 2503, the memory 2501, and the processor 2502 may be connected to each other and perform communication with each other through buses. The bus may be an ISA (Industry Standard Architecture ) bus, a PCI (PERIPHERAL COMPONENT, external device interconnect) bus, or EISA (Extended Industry Standard Architecture ) bus, among others. The buses may be divided into address buses, data buses, control buses, etc. For ease of illustration, only one thick line is shown in fig. 25, but not only one bus or one type of bus.
Alternatively, in a specific implementation, if the memory 2501, the processor 2502, and the communication interface 2503 are implemented on a single chip, the memory 2501, the processor 2502, and the communication interface 2503 may communicate with each other through internal interfaces.
The processor 2502 may be a CPU (Central Processing Unit ) or an ASIC (Application SPECIFIC INTEGRATED Circuit, application specific integrated Circuit) or one or more integrated circuits configured to implement embodiments of the present application.
The embodiment of the application also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, implements the control method of the pre-aiming type semi-active suspension.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or N embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Furthermore, the terms "first," "second," and the like, are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In the description of the present application, "N" means at least two, for example, two, three, etc., unless specifically defined otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more N executable instructions for implementing specific logical functions or steps of the process, and further implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present application.
It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. As with the other embodiments, if implemented in hardware, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable gate arrays, field programmable gate arrays, and the like.
Those of ordinary skill in the art will appreciate that all or a portion of the steps carried out in the method of the above-described embodiments may be implemented by a program to instruct related hardware, where the program may be stored in a computer readable storage medium, and where the program, when executed, includes one or a combination of the steps of the method embodiments.
While embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.
Claims (10)
1. The control method of the pre-aiming type semi-active suspension is characterized by comprising the following steps of:
Collecting state information at the current moment;
Based on a pre-established model prediction controller, predicting a state output vector of the semi-active suspension control system in a prediction time domain according to the state information, and obtaining a target optimization function according to the state output vector;
And calculating a control sequence meeting the target optimization function and a preset constraint condition according to a preset reference input vector, and obtaining a target control quantity based on the control sequence so as to control the semi-active suspension control system based on the target control quantity.
2. The method of claim 1, wherein predicting a state output vector of a semi-active suspension control system in the prediction horizon from the state information based on the pre-established model predictive controller comprises:
Establishing a state space model of the semi-active suspension;
Performing external characteristic test on the magnetorheological damper to obtain identification data, training to obtain a forward model and a reverse model of the magnetorheological damper based on the identification data, and establishing a random pavement model and a convex hull pavement model;
And obtaining the pre-established model prediction controller based on the state space model of the semi-active suspension, the random road surface model, the convex hull road surface model, the forward model and the reverse model.
3. The method of claim 2, wherein predicting a state output vector of the semi-active suspension control system in a prediction horizon based on the state information based on the pre-established model predictive controller comprises:
Discretizing a state space model of the semi-active suspension to obtain a target state space equation;
and acquiring a current state quantity and a current time-varying disturbance, and obtaining a state output vector of the prediction time domain by utilizing the target state space equation based on the current state quantity and the current time-varying disturbance.
4. A method according to claim 3, wherein said deriving an objective optimization function from said state output vector comprises:
acquiring an input vector of a control time domain;
And obtaining the target optimization function based on the state output vector, the input vector of the control time domain and the current time-varying disturbance.
5. A method according to claim 3, wherein the predicted output vector is:
Yp,c(k+1|k)＝Sxx(k)+SuU(k)+SDD(k)；
Wherein p is a prediction time domain, Y p,c (k+ 1|k) is a p-step prediction output vector of the system, x (k) is a current state quantity, U (k) is a control input vector, D (k) is a current time-varying disturbance, S x is a coefficient matrix of the state quantity in the prediction time domain, S u is a coefficient matrix of the control input vector in the prediction time domain, and S D is a coefficient matrix of the road surface height change rate in the prediction time domain.
6. The method of claim 5, wherein the objective optimization function is:
Wherein J is the target optimization function, Γ y is a weighted weight coefficient matrix of control output, Γ u is a weighted weight coefficient matrix of control input, and E p (k+1) is a constraint vector of a state quantity and a road surface height variation quantity at the next time of a prediction time domain.
7. A control device for a pre-addressed semi-active suspension, comprising:
the acquisition module is used for acquiring state information at the current moment;
the prediction module is used for predicting a state output vector of the semi-active suspension control system in a prediction time domain based on a pre-established model prediction controller according to the state information, and obtaining a target optimization function according to the state output vector;
And the control module is used for calculating a control sequence meeting the target optimization function and the preset constraint condition according to the preset reference input vector, and obtaining a target control quantity based on the control sequence so as to control the semi-active suspension control system based on the target control quantity.
8. The apparatus of claim 7, wherein the prediction module, prior to predicting a state output vector of a semi-active suspension control system in the prediction horizon based on the pre-established model prediction controller, according to the state information, comprises:
The building unit is used for building a state space model of the semi-active suspension;
The testing unit is used for carrying out external characteristic testing on the magnetorheological damper to obtain identification data, training the magnetorheological damper based on the identification data to obtain a forward model and a reverse model of the magnetorheological damper, and establishing a random pavement model and a convex hull pavement model;
The first generation unit is used for obtaining the pre-established model prediction controller based on the state space model of the semi-active suspension, the random road surface model, the convex hull road surface model, the forward model and the reverse model.
9. An electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor executing the program to implement the method of controlling a pre-addressed semi-active suspension as claimed in any one of claims 1 to 6.
10. A computer-readable storage medium, on which a computer program is stored, characterized in that the program is executed by a processor for implementing a control method of a pre-aiming semi-active suspension as claimed in any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311810629.8A CN117962538A (en) | 2023-12-26 | 2023-12-26 | Control method, device, equipment and storage medium of pre-aiming type semi-active suspension |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311810629.8A CN117962538A (en) | 2023-12-26 | 2023-12-26 | Control method, device, equipment and storage medium of pre-aiming type semi-active suspension |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117962538A true CN117962538A (en) | 2024-05-03 |
Family
ID=90853858
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311810629.8A Pending CN117962538A (en) | 2023-12-26 | 2023-12-26 | Control method, device, equipment and storage medium of pre-aiming type semi-active suspension |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117962538A (en) |
-
2023
- 2023-12-26 CN CN202311810629.8A patent/CN117962538A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Theunissen et al. | Preview-based techniques for vehicle suspension control: A state-of-the-art review | |
| Gandhi et al. | Performance analysis of half car suspension model with 4 DOF using PID, LQR, FUZZY and ANFIS controllers | |
| Ren et al. | State observer-based sliding mode control for semi-active hydro-pneumatic suspension | |
| CN100484788C (en) | Apery intelligent control method for harmonizing auto magnetorheological half-initiative suspension according to posture | |
| Caponetto et al. | A soft computing approach to fuzzy sky-hook control of semiactive suspension | |
| Qin et al. | Adaptive nonlinear active suspension control based on a robust road classifier with a modified super-twisting algorithm | |
| Soleymani et al. | Adaptive fuzzy controller for vehicle active suspension system based on traffic conditions | |
| CN109334378B (en) | Vehicle ISD suspension active control method based on single neuron PID control | |
| CN112776551B (en) | Semi-active control method and system for magneto-rheological suspension based on motion pattern | |
| Rao et al. | Active suspension system of a 3 DOF quarter car using fuzzy logic control for ride comfort | |
| Zin et al. | An LPV/H~~ Active Suspension Control for Global Chassis Technology: Design and Performance Analysis | |
| JP2007233985A (en) | Optimal control method of system | |
| KR102654627B1 (en) | Vehicle control device, vehicle control method and vehicle control system | |
| Muhammed et al. | Linear-control vs. ADRC for automatic management of the handling-comfort contradiction of a quarter-car system | |
| CN113761768A (en) | Integrated optimization design method of magneto-rheological damper for whole vehicle vibration suppression | |
| Al-Ghanim et al. | Modeling, simulation, and control of half car suspension system using Matlab/Simulink | |
| CN117962538A (en) | Control method, device, equipment and storage medium of pre-aiming type semi-active suspension | |
| Aldair | Neurofuzzy controller based full vehicle nonlinear active suspension systems | |
| Gustafsson et al. | Neural network controller for semi-active suspension systems with road preview | |
| Khan et al. | Neuro-fuzzy wavelets based network for full car active suspension system | |
| Ahmed et al. | Adaptive robust gain scheduled control of vehicle semi-active suspension for improved ride comfort and road handling | |
| ASI et al. | Adaptive speed control of an autonomous vehicle with a comfort objective | |
| Sathishkumar et al. | Reduction of axis acceleration of quarter car suspension using pneumatic actuator and active force control technique | |
| CN113525535B (en) | Cab semi-active suspension control method and device based on fuzzy control | |
| Ballard | Semi-active Nonlinear Predictive Suspension Control for Off-road Vehicles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |