CN115598983B - Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay - Google Patents
Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay Download PDFInfo
- Publication number
- CN115598983B CN115598983B CN202211333752.0A CN202211333752A CN115598983B CN 115598983 B CN115598983 B CN 115598983B CN 202211333752 A CN202211333752 A CN 202211333752A CN 115598983 B CN115598983 B CN 115598983B
- Authority
- CN
- China
- Prior art keywords
- time
- vehicle
- longitudinal
- delay
- transverse
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000004891 communication Methods 0.000 claims abstract description 37
- 238000011156 evaluation Methods 0.000 claims abstract description 26
- 238000005096 rolling process Methods 0.000 claims abstract description 7
- 230000001133 acceleration Effects 0.000 claims description 24
- 230000003416 augmentation Effects 0.000 claims description 22
- 239000013598 vector Substances 0.000 claims description 20
- 230000003044 adaptive effect Effects 0.000 claims description 18
- 238000001914 filtration Methods 0.000 claims description 16
- 239000011159 matrix material Substances 0.000 claims description 14
- 238000005070 sampling Methods 0.000 claims description 12
- 230000006641 stabilisation Effects 0.000 claims description 11
- 238000011105 stabilization Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000007704 transition Effects 0.000 claims description 5
- 230000006870 function Effects 0.000 description 9
- 230000008878 coupling Effects 0.000 description 8
- 238000010168 coupling process Methods 0.000 description 8
- 238000005859 coupling reaction Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 7
- 230000014509 gene expression Effects 0.000 description 7
- 230000004044 response Effects 0.000 description 7
- 230000003111 delayed effect Effects 0.000 description 5
- 238000013461 design Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 108010076504 Protein Sorting Signals Proteins 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 230000002441 reversible effect Effects 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000000528 statistical test Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a kind ofThe unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay comprise the following steps: step 1, acquiring a time-varying CAN communication delay estimated value tau of a bottom layer on line CAN The method comprises the steps of carrying out a first treatment on the surface of the Step 2, according to the found prediction time domain N p And (3) carrying out rolling solution on the optimal control problem by using a predictive model and combining a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. and outputting the control quantity of the vehicle. According to the invention, by combining the self-adaptive time delay estimator and the MPC transverse and longitudinal cooperative control algorithm considering time-varying time delay, the problem of vehicle control instability under the limit working condition caused by the characteristic of the unmanned vehicle when the bottom layer is ignored is solved, and the stability of the vehicle is improved while the transverse and longitudinal control accuracy is ensured.
Description
Technical Field
The invention relates to the technical field of intelligent vehicles, in particular to a transverse and longitudinal cooperative control method and device for an unmanned vehicle in consideration of time-varying time delay.
Background
The intelligent vehicle is an important link of an intelligent traffic system, can effectively reduce problems such as traffic accidents, traffic jams and environmental pollution, and therefore becomes a research hot spot in recent years. Motion control is one of the core technologies of intelligent vehicles. The motion control is to generate a control command of a vehicle actuator (such as a steering wheel, an electronic accelerator, a brake, a gear shifting mechanism, etc.) according to a reference track input and a control law, and generate a force or moment influencing the motion of the vehicle so that the vehicle can finally converge on the reference track.
The intelligent vehicle motion control comprises longitudinal control and transverse control, wherein the longitudinal control realizes that the vehicle cruises at a preset speed or keeps a certain distance from a front dynamic target, and the transverse control realizes that the vehicle runs along a planned path and ensures the running safety, stability and riding comfort of the vehicle. The following mainly researches and develops the prior automatic driving vehicle track tracking control:
firstly, horizontal and longitudinal layering control is performed. The control mode realizes decoupling of the dynamic model by neglecting the coupling characteristic of transverse and longitudinal dynamics, reduces the complexity of a single problem and is beneficial to quick solution of transverse and longitudinal control laws. The existing transverse control method can be divided into model-free control and model-based control, wherein the model-free control only depends on errors to calculate the wheel rotation angle, such as control algorithms of pure tracking, stanley, PID and the like. The model-based control method designs an explicit control rate according to the dynamic characteristics of the system, such as control algorithms of LQR, MPC and the like. The existing longitudinal control method can be divided into: the direct control directly generates a desired braking pressure or throttle opening according to a vehicle model and a reference speed trajectory, and the hierarchical control includes an upper speed control and a lower actuator control.
The other is horizontal and vertical coupling control. The control mode fully considers the coupling correlation characteristic between transverse and longitudinal dynamics of the intelligent vehicle, and obtains a transverse and longitudinal movement control law by directly controlling and solving a transverse and longitudinal integrated dynamics model of the vehicle. The transverse and longitudinal coupling control carries out transverse and longitudinal coupling vehicle modeling in the aspect of a model, further carries out more accurate description on dynamic characteristics of the vehicle, realizes overall coordination on two unidirectional tracking performances by adopting transverse and longitudinal comprehensive performance evaluation in the aspect of a performance evaluation function, and realizes more complete construction on a control quantity feasible set by adopting a transverse and longitudinal combined constraint mode in the aspect of control constraint.
One key factor that is often ignored by existing trajectory tracking control algorithms is vehicle floor latency, including CAN communication latency and actuator time lag. The delay is mainly from the communication between the control module and the actuator, wherein a large number of intermediate links exist, such as CAN bus communication delay. The time lag is mainly from the time spent by the actuator in eventually responding to the upper level control command, such as steering and braking execution lag. The delay and the time lag are mainly influenced by hardware performance and bottom control design, and the communication delay and the executor time lag are ignored to cause mismatching of a control model and performance deterioration, so that transient response and stability of the system are reduced, and further the phenomena of steering oscillation, even instability and the like of the vehicle occur. The unmanned vehicle platforms used by many academic research institutions today reduce the underlying latency by installing wire control modules or new actuators, yet the unequally available underlying latency still poses challenges to the stability of current motion control systems.
Disclosure of Invention
The invention aims to provide a transverse and longitudinal cooperative control method and device for an unmanned vehicle, which are used for solving the problem of vehicle control instability of the unmanned vehicle under a limit working condition caused by neglecting a bottom layer time delay characteristic by combining a self-adaptive time delay estimator and an MPC transverse and longitudinal cooperative control algorithm which is used for considering time delay, and improving the vehicle stability while ensuring transverse and longitudinal control precision.
In order to achieve the above purpose, the invention provides a transverse and longitudinal cooperative control method of an unmanned vehicle, which considers time-varying time delay, and comprises the following steps:
Step 2, according to the found prediction time domain N p And (3) carrying out rolling solution on the optimal control problem by using a predictive model and combining a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. and outputting the control quantity of the vehicle.
Further, in step 2, the optimal control problem is described as follows:
where i is the index of the predicted state, k is the current time,
respectively N p Transverse deviation cost term, course angle deviation cost term, longitudinal speed deviation, longitudinal control increment and transverse control increment cost term of vehicle at time k+i, N c To control the time domain, e y (k+i)、
Respectively N p Lateral deviation, heading angle deviation, longitudinal speed deviation, deltaa of the vehicle at time k+i x (k+i)、Δδ f (k+i) is N respectively p Longitudinal acceleration increment, front wheel steering angle increment, Q of vehicle at time k+i 1 、Q 2 、Q 3 、R 1 、R 2 Are all weight coefficients, χ 0 (k) Chi (k+1), chi (k+i-1) are the state variables of the vehicle at the times k, k+1, k+i-1, respectively, u 0 (k) U (k+i-1) is the control amount of the vehicle at the time k and k+i-1, a x (k+j)δ f (k+j) is the transverse and longitudinal control quantity of the vehicle at the time k+j, delta f,max 、a x,max Respectively, the maximum allowable transverse and longitudinal control amounts of the vehicle, delta f,max 、Δa x,max The maximum horizontal and longitudinal control increment is respectively allowed by the vehicle at adjacent time, delta f (k+j)、Δa x (k+j) is the transverse and longitudinal control increment of the vehicle at the time of k+j, f (χ) 0 (k),u 0 (k) And f (χ (k+i-1), u (k+i-1)) are the results of calculation of the variable-dimension time delay augmentation model by using the state quantity and the initial control quantity of the vehicle at the time of k, k+i-1, respectively.
Further, in step 2, the prediction model is a variable-dimension time delay augmentation model constructed by using a transverse and longitudinal time delay link according to the bottom layer time-varying CAN communication time delay estimated value and the time delay characteristic of the actuator:
wherein ,
N τ representing the cause of the signal τ CAN And delay N τ The sampling time steps dt, χ (k) and χ (k+1) are the prediction time domain N p The state variables of the vehicle at times k, k+1, χ '(k) and χ' (k+1) are the first derivatives of χ (k) and χ (k+1), respectively, with respect to time, a x(k) and ax (k+1) are respectively the prediction time domains N p Longitudinal acceleration, delta, of the vehicle at times k, k+1 f(k) and δf (k+1) are respectively the prediction time domains N p The front wheel steering angle of the vehicle at times k and k +1,
respectively, prediction time domain N p Inner k-N τ 、k-N τ +1、k-N τ Desired front wheel steering angle of the vehicle at +2, k-2, k-1, k moment,/->
To predict time domain N p The desired longitudinal acceleration of the vehicle at time k, dt being the sampling time step, < >>
The time lag constant of the first-order inertia link of the transverse actuator and the longitudinal actuator respectively, O is zero matrix, and +.>
Corresponding to the transition matrix A, B for the following simplified formula 1 、B 2 Is a discretization result of (a):
B 1 ＝[1 0 0 0 0] T
wherein ,Ccf 、C cr Respectively the cornering stiffness of the front and rear wheels of the vehicle, l f 、l r The distances from the center of mass of the vehicle to the centers of the front axle and the rear axle, m and v x 、I z The mass, longitudinal speed, moment of inertia of the vehicle, respectively, and κ is the road curvature at the tracking target point.
Further, the step 1 specifically includes:
taking a real-time input expected control instruction as a real-time signal, taking a bottom actual control instruction as a delay signal, and obtaining tau by utilizing an adaptive all-pass filtering delay estimator based on FIR and a delay estimation stabilization strategy based on a mean square error evaluation index CAN 。
Further, the FIR-based adaptive all-pass filter delay estimator is expressed as formula (8) or (9):
wherein: x (k) is a real-time signal, x (k-N) τ ) In order to delay the signal in time,
expressed as vectors
Expressed as vector +.>
X + (k-N τ )＝[x(k+1-N τ ),...,x(k+n max -N τ )] T As a forward vector, X - (k)＝[x(k-1),...,x(k-n max )] T Is a backward vector.
Further, a delay estimation stabilization strategy is obtained based on a mean square error evaluation index by adopting the formula (18):
where τ (k) represents the underlying delay estimate at time k, thr is the update threshold, MSE last Bottom layer delay estimation value representing k-1 moment
MSE index, MSE new Bottom layer delay estimate representing time k>
According to MSE index of (2)
Obtaining τ CAN 。
The invention also provides an unmanned vehicle transverse and longitudinal cooperative control device considering time-varying time delay, which comprises:
an adaptive all-pass filtering time delay estimator for acquiring an underlying time-varying CAN communication time delay estimated value tau on line CAN ；
A horizontal-vertical cooperative controller for predicting the time domain N according to the found p And (3) carrying out optimal control problem rolling solving by using a predictive model and combining a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. and outputting the control quantity of the vehicle, wherein: the optimal control problem is described as:
where i is the index of the predicted state, k is the current time,
respectively N p Transverse deviation cost term, course angle deviation cost term, longitudinal speed deviation, longitudinal control increment and transverse control increment cost term of vehicle at time k+i, N c To control the time domain, e y (k+i)、
Respectively N p Lateral deviation, heading angle deviation, longitudinal speed deviation, deltaa of the vehicle at time k+i x (k+i)、Δδ f (k+i) is N respectively p Longitudinal acceleration increment, front wheel steering angle increment, Q of vehicle at time k+i 1 、Q 2 、Q 3 、R 1 、R 2 Are all weight coefficients, χ 0 (k) Chi (k+1), chi (k+i-1) are the state variables of the vehicle at the times k, k+1, k+i-1, respectively, u 0 (k) U (k+i-1) is the control amount of the vehicle at the time k and k+i-1, a x (k+j)δ f (k+j) is the transverse and longitudinal control quantity of the vehicle at the time k+j, delta f,max 、a x,max Respectively, the maximum allowable transverse and longitudinal control amounts of the vehicle, delta f,max 、Δa x,max The maximum horizontal and longitudinal control increment is respectively allowed by the vehicle at adjacent time, delta f (k+j)、Δa x (k+j) is the transverse and longitudinal control increment of the vehicle at the time of k+j, f (χ) 0 (k),u 0 (k) And f (χ (k+i-1), u (k+i-1)) are the results of calculation of the variable-dimension time delay augmentation model by using the state quantity and the initial control quantity of the vehicle at the time of k, k+i-1, respectively.
Further, the prediction model is a variable-dimension time delay augmentation model constructed by utilizing a transverse and longitudinal time delay link according to a bottom layer time-varying CAN communication time delay estimated value and an actuator time delay characteristic:
wherein ,
N τ representing the cause of the signal τ CAN And delay N τ The sampling time steps dt, χ (k) and χ (k+1) are the prediction time domain N p The state variables of the vehicle at times k, k+1, χ '(k) and χ' (k+1) are the first derivatives of χ (k) and χ (k+1), respectively, with respect to time, a x(k) and ax (k+1) are respectively the prediction time domains N p Longitudinal acceleration, delta, of the vehicle at times k, k+1 f(k) and δf (k+1) are respectively the prediction time domains N p Front wheel steering angle of vehicle at times k, k+1,>
respectively, prediction time domain N p Inner k-N τ 、k-N τ +1、k-N τ Desired front wheel steering angle of the vehicle at +2, k-2, k-1, k moment,/->
To predict time domain N p The desired longitudinal acceleration of the vehicle at time k, dt being the sampling time step, < >>
The time lag constant of the first-order inertia link of the transverse actuator and the longitudinal actuator respectively, O is zero matrix, and +.>
Corresponding to the transition matrix A, B for the following simplified formula 1 、B 2 Is a discretization result of (a):
B 1 ＝[1 0 0 0 0] T
wherein ,Ccf 、C cr Respectively the cornering stiffness of the front and rear wheels of the vehicle, l f 、l r The distances from the center of mass of the vehicle to the centers of the front axle and the rear axle, m and v x 、I z The mass, longitudinal speed, moment of inertia of the vehicle, respectively, and κ is the road curvature at the tracking target point.
Further, the bottom layer time-varying CAN communication delay estimated value tau CAN The acquisition method of the method specifically comprises the following steps:
and taking the real-time input expected control instruction as a real-time signal, taking the bottom actual control instruction as a delay signal, carrying out on-line estimation of bottom time-varying CAN communication delay by utilizing a self-adaptive all-pass filtering delay estimator based on FIR and a delay estimation stabilization strategy based on a mean square error evaluation index, and outputting a bottom time-varying CAN communication delay estimation value.
Further, the FIR-based adaptive all-pass filter delay estimator is expressed as formula (8) or (9):
wherein: x (k) is a real-time signal, x (k-N) τ ) In order to delay the signal in time,
expressed as vectors
Expressed as vector +.>
X + (k-N τ )＝[x(k+1-N τ ),...,x(k+n max -N τ )] T As a forward vector, X - (k)＝[x(k-1),...,x(k-n max )] T Is a backward vector; />
The delay estimation stability strategy based on the mean square error evaluation index is obtained by adopting the formula (18):
where τ (k) represents the underlying delay estimate at time k, thr is the update threshold, MSE last Bottom layer delay estimation value representing k-1 moment
MSE index, MSE new Bottom layer delay estimate representing time k>
According to MSE index of (2)
Obtaining τ CAN 。
Due to the adoption of the technical scheme, the invention has the following advantages:
1. compared with the prior art that the vehicle delay parameter is obtained only through off-line calibration and is constant all the time in operation, the self-adaptive all-pass filtering delay estimator based on the FIR CAN estimate the time-varying CAN communication delay of the bottom layer on line.
2. According to the mean square error evaluation index, the delay estimation stabilization strategy provided by combining the stored historical control signals and the delay estimation value can eliminate the error estimation value of the self-adaptive all-pass filter caused by noise interference;
3. compared with the MPC model-based predictive control framework in the prior art, the method only considers the time delay of the actuator in the predictive model, and based on the on-line estimated value of the bottom layer time delay, the method constructs a transverse and longitudinal time delay link which considers the communication time delay of the bottom layer and the time delay characteristic of the actuator, and establishes a variable-dimension time delay augmentation model based on a two-degree-of-freedom dynamic model of the vehicle, and the problem of control performance reduction of a control algorithm caused by mismatching between the used vehicle model and the dynamic time delay characteristic of the real vehicle CAN be independently solved through the variable-dimension time delay augmentation model;
4. compared with the prior art that a longitudinal controller is independently designed from the aspect of horizontal and longitudinal separation so as to realize unmanned vehicle speed tracking, the method takes a variable dimension time delay augmentation model as a prediction model, establishes a horizontal and longitudinal integrated evaluation function and joint constraint based on horizontal and longitudinal coupling characteristics, designs an MPC horizontal and longitudinal cooperative controller considering time-varying time delay, and comprehensively ensures horizontal and longitudinal control precision and vehicle stability through integrated optimal control.
Drawings
Fig. 1 is a schematic structural diagram of a framework of an unmanned vehicle transverse and longitudinal cooperative control method considering time-varying time delay according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a bottom layer delay related to lateral control according to an embodiment of the present invention.
Fig. 3 is a schematic structural diagram of an unmanned vehicle transverse and longitudinal cooperative control device considering time-varying time delay according to an embodiment of the present invention.
Detailed Description
The present invention will be described in detail with reference to the accompanying drawings and examples.
As shown in fig. 1, the transverse and longitudinal cooperative control device for the unmanned vehicle, which is provided by the embodiment of the invention and takes time-varying time delay into consideration, comprises an adaptive all-pass filtering time delay estimator and a transverse and longitudinal cooperative controller which takes time-varying time delay into consideration. Wherein:
the adaptive all-pass filtering time delay estimator is used for online acquisitionBottom layer time-varying CAN communication delay estimated value tau CAN 。
The transverse and longitudinal cooperative controller is used for carrying out optimal control problem rolling solving by utilizing a prediction model and combining a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. according to a series of reference point information in the found prediction time domain, and outputting the control quantity of the vehicle.
According to the invention, by combining the self-adaptive time delay estimator and the MPC transverse and longitudinal cooperative control algorithm considering time-varying time delay, the problem of vehicle control instability under the limit working condition caused by the characteristic of the unmanned vehicle when the bottom layer is ignored is solved, and the stability of the vehicle is improved while the transverse and longitudinal control accuracy is ensured.
In one embodiment, as shown in fig. 1, the adaptive all-pass filtered delay estimator includes an adaptive all-pass filter and a delay estimation stabilization strategy. The time delay estimator is used for taking a real-time input expected control instruction as a real-time signal, taking a bottom actual control instruction as a time delay signal, carrying out on-line estimation of bottom time-varying CAN communication time delay by utilizing an adaptive all-pass filtering time delay estimator based on FIR (English is fully called as 'Finite Impulse Response', chinese is fully called as 'finite impulse response filter') and a time delay estimation stabilization strategy based on a mean square error evaluation index, and outputting a bottom time-varying CAN communication time delay estimation value.
The method for acquiring the adaptive all-pass filtering time delay estimator based on the FIR comprises the following steps:
constructing a signal channel model:
H(ω)＝e -jτw (1)
wherein: h is fourier transform of an all-pass filter H, i.e., |h (ω) |=1, j is an imaginary number, ω represents frequency, τ is bottom layer delay, i.e., a result of mutual coupling of bottom layer time-varying CAN communication delay and actuator time lag. With a real-time signal (e.g., an unmanned vehicle desired control command) as input and a delayed signal (e.g., an underlying actual control command) as output, the filter h effects a change in signal phase without changing amplitude. Since the phase change depends on the delay, the delay value τ can be obtained by estimating the filter h.
All-pass filter H (ω) is constructed based on FIR:
wherein ：P(ejω ) Is the forward frequency response of the FIR filter P, P (e -jω ) Is the backward frequency response of the FIR filter p.
Let N τ ＝τ/dt，N τ Indicating that the signal is delayed N due to delay characteristics τ The sampling time step dt, dt is the sampling time step, k represents the current time, then based on equation (2), the signal at the current time, i.e. the real-time signal x (k) and the delayed signal x (k-N) τ ) Linear treatment gives formula (3), wherein: * For the convolution operator,
meaning that the formulas on both sides of the arrow are equivalent, p (k) and p (-k) are forward and reverse expressions of the FIR, respectively, forward, i.e., using future information, and reverse, i.e., using past information:
setting a support n max FIR, n of individual filter coefficients max For the largest index number of the filter, this value also defines the upper bound τ of the delay estimate max ＝n max * dt, N τ ≤n max Let n represent the index of the filter coefficients, n e [0, n max ]。
The FIR filter response p (n) can be determined by the filter coefficients a n Described as formula (4):
thus, the all-pass filter of equation (3) can be converted to linear expression (5) based on FIR:
in the formula ,
the method is characterized in that a convolution-based all-pass filter expression formula is converted into a linear expression based on FIR, so that a time delay estimated value can be obtained by solving a linear coefficient.
When the filter coefficients
When the equation (1) and the equation (2) are derived at ω0, dH (ω)/dω is calculated, and the values of the equation (1) and the equation (2) are equal, equation (6) is obtained, so that the delay estimation value τ can be calculated:
because τ is the result of the mutual coupling of CAN communication delay and actuator time lag, the present embodiment decouples both by: taking bottom layer time delay in a transverse control model as an example, real vehicle data is utilized to calibrate a first-order inertial link time lag constant of a transverse actuator offline
And as a fixed value, finally outputting a time-varying bottom layer time-varying CAN communication delay estimated value tau CAN ：
The above equations (1) - (6) convert the estimate of the delay value τ into the filter coefficient a n Is a function of the estimate of (2). Setting x (k) as a real-time control signal issued by a current controller, and x (k-N) τ ) For the delay control signal fed back by the bottom layer, let the filter coefficient a 0 =1, then the linear expression of formula (5) is formula (8):
wherein: x (k) is a real-time signal, x (k-N) τ ) In order to delay the signal in time,
expressed as vectors
Expressed as vector +.>
X + (k-N τ )＝[x(k+1-N τ ),...,x(k+n max -N τ )] T As a forward vector, X - (k)＝[x(k-1),...,x(k-n max )] T Is a backward vector.
Then the real-time signal x (k) and the delayed signal x (k-N) τ ) Can be further expressed as the following linear formula (9):
due to the real-time signal x (k) and the delayed signal x (k-N) τ ) Let d (k) =x (k-N τ )-x(k)，
The delay estimation problem turns into an optimization problem: by minimizing the measurement samples d (k) and by the current filter coefficients +.>
The deviation between y (k) is calculated, so that an optimal filter coefficient a is obtained, and then the optimal filter coefficient a is substituted into the formula (6), so that a time delay estimated value tau can be obtained.
The present embodiment iteratively updates the filter coefficient a (k) based on the gradient descent method, a (k) representing the filter coefficient value at the current time, i.e., at time k, whereas a is the overall expression of the filter coefficient without distinguishing which time is specific. The value of the latest filter coefficient a (k) is used to calculate the delay estimate τ. This embodiment defines the deviation between d (k) and y (k) as:
the defined performance index function is expressed as the following formula (11):
J(k)＝|e(k)| 2 (11)
the current gradients may be:
the updated filter coefficients are expressed as equation (13):
wherein: μ is a learning rate parameter of the gradient descent method.
In one embodiment, a time-varying learning rate parameter mu is obtained based on the LMS, so that error convergence in the optimization process is guaranteed, and finally, the learning rate self-adaptive all-pass filtering time delay estimator is realized. The set adaptive learning rate is formula (14):
in combination with the adaptive learning rate μ, equation (12) becomes:
wherein: ρ is an adaptive learning rate constant coefficient, 0 < ρ < 2/3; epsilon is a very small positive integer used to ensure that the denominator is not 0.
The manner in which the delay estimation stability strategy based on the mean square error index is obtained is described below.
When the delay estimator is simulated and verified based on real-time and delay signal data, the phenomenon that the estimation result is changed between different values at high frequency due to noise interference exists, and even false estimation values such as negative numbers appear, so that the TDE is unstable. Considering the characteristics that the time delay value of the bottom layer of the unmanned vehicle is constant in a short time and time-varying in a long time, a time delay estimation stability strategy is designed based on a mean square error evaluation index, and the error estimation value is eliminated while the TDE estimation stability is improved.
The main idea of the delay estimation stabilization strategy is to use statistical tests. Setting: there is a real-time signal sequence x containing all signal values within M sample time steps dt 1 (t) and delay signal sequence x 2 (t) the range of values of the delay estimate is [ tau ] min ,τ max ]I.e. the upper bound is τ max The lower bound is τ min . Time delay estimated value based on last moment
And the delay estimate at the current moment +.>
Calculating MSE using (16) last Bottom layer delay estimate representing time k-1 +.>
MSE is calculated using equation (17) based on the MSE index of the mean square new Bottom layer delay estimate representing time k>
Mean square MSE index of (2): />
In essence, the MSE index may be obtained using the mean square error provided by equations (16) - (17), or may be calculated using other evaluation criteria such as variance or standard deviation.
Based on the MSE index, the present embodiment provides a delay estimation update strategy as shown in the following formula (18):
wherein: τ (k) represents the bottom delay estimate at time k, thr is the update threshold, and thr=0.2 is set in this embodiment. And comparing the variances of the real-time and delay signals in the same window by the estimated values of adjacent moments, and if the variances are not large, assuming that the TDE result is unreliable, retaining the estimated value of the last moment, otherwise, updating the delay result. According to
Obtaining τ CAN 。
In one embodiment, the method for acquiring the transverse-longitudinal cooperative controller considering time-varying time delay comprises the following steps:
firstly, constructing a variable-dimension time delay augmentation model:
in one embodiment, as shown in fig. 1, the prediction model is a variable dimension time delay augmentation model which considers the time delay of the underlying CAN communication and the time delay characteristic of the actuator by combining the online time delay estimation value, so as to more accurately describe the dynamic characteristic of the vehicle and predict the state of the future time.
The three degree of freedom vehicle dynamics model of the vehicle is as follows:
wherein ,ye For the lateral deviation between the center of the rear axle of the vehicle and the reference point,
epsilon for the rate of change of the lateral deviation of the vehicle e For the heading angle deviation between the longitudinal axis of the vehicle and the reference point, < >>
For the change rate of the course angle deviation of the vehicle, +.>
Respectively the longitudinal speed v of the vehicle x And transverse speed, +.>
For longitudinal acceleration a of the vehicle along the x-axis x ，
For the lateral acceleration of the vehicle along the y-axis, +.>
For the yaw angle of the vehicle->
For vehicle yaw rate, +.>
For the rate of change of the yaw rate of the vehicle over time, < >>
For the vehicle longitudinal acceleration is desired +.>
For the desired front wheel turning angle of the vehicle, κ is the road curvature at the tracking target point, m is the vehicle mass, l f Distance from centroid to front axle center, l r C is the distance from the center of mass to the center of the rear axle cf C is the cornering stiffness of the front wheel cr For the cornering stiffness of the rear wheels, I z Is the moment of inertia of the vehicle.
Considering the vehicle bottom layer CAN communication delay and the actuator time-lag characteristic (to simplify the model complexity, only the actuator time-lag characteristic is considered in the longitudinal direction), taking the horizontal control as an example of the bottom layer delay, as shown in fig. 2 below:
1) Delay tau for CAN communication CAN
Setting a pure hysteresis link, and issuing an expected control quantity by the current controller as follows
The desired control amount actually acting on the vehicle is +.>
2) Regarding actuator time lag
Setting a first-order inertia link, and issuing an expected control quantity by a current controller as follows
The current actual control quantity is delta f (t)、a x (t) the desired control amount actually applied to the vehicle is
Based on the above analysis, the construction of the horizontal and vertical delay links is represented by the following formula (20):
based on a three-degree-of-freedom vehicle dynamics model and a transverse and longitudinal delay link, a delay augmentation prediction model (21) for MPC state prediction is established:
to simplifyThe expression form of the model is set as the following matrix A and matrix B 1 Sum matrix B 2 And discretizing the time-augmented predictive model (21) to sample the time step dt to obtain the following
B 1 ＝[1 0 0 0 0] T
At this time, the optimal transverse control quantity to be solved in the delay augmentation prediction model is as follows
Further conversion is required so that the control quantity to be solved is still +.>
Thus, a lateral control amount is expected based on a known history, i.e
Further obtaining a variable-dimension time delay augmentation prediction model:
in the formula ,
N τ representing the cause of the signal τ CAN And delay N τ The sampling time steps dt, χ (k) and χ (k+1) are the prediction time domain N p The state variables of the vehicle at times k, k+1, χ '(k) and χ' (k+1) are the first derivatives of χ (k) and χ (k+1), respectively, with respect to time, a x(k) and ax (k+1) are respectively the prediction time domains N p Longitudinal acceleration, delta, of the vehicle at times k, k+1 f(k) and δf (k+1) are respectively the prediction time domains N p Front wheel steering angle of vehicle at times k, k+1,>
respectively, prediction time domain N p Inner k-N τ 、k-N τ +1、k-N τ Desired front wheel steering angle of the vehicle at +2, k-2, k-1, k moment,/->
To predict time domain N p The desired longitudinal acceleration of the vehicle at time k, dt being the sampling time step, < >>
The time lag constants of the first-order inertial links of the transverse actuator and the longitudinal actuator are respectively zero matrix, namely, the elements in the matrix are 0,/respectively>
Corresponding to a transition matrix A for simplifying the formula,B 1 、B 2 Discretization result of->
The state quantity of the three-degree-of-freedom vehicle dynamics model shown in formula (19) is part of a formula (22) χ ->
u (k) is the control quantity, +.>
χ (k) is a state quantity, and is specifically expressed as follows:
the final variable dimension delay augmentation model may be expressed as: chi (k+1) =f (chi (k), u (k)), the dimension of the state quantity chi is based on the delay estimate τ CAN (N after discretization) τ ) The dynamic change ensures more accurate description of the vehicle model.
Then, a horizontal-vertical cooperative optimal controller based on MPC is constructed:
the invention adopts comprehensive performance evaluation of transverse and longitudinal to realize overall coordination of two unidirectional tracking performances, and minimizes a transverse and longitudinal integrated evaluation function (23) consisting of transverse deviation, course angle deviation, longitudinal speed deviation, front wheel steering angle increment and longitudinal acceleration increment in a prediction time domain:
wherein ：Np To predict the time domain; n (N) c To control the time domain; q (Q) 1 、Q 2 、Q 3 、R 1 、R 2 Is a weight coefficient；e y (k) To predict the center (x (k), y (k)) of the rear axle of the vehicle at time k in the time domain to a reference point (x r (k),y r (k) A) the vertical distance of the tangent line, i.e. the lateral deviation,
desired heading angle for reference point->
The deviation between the vehicle tracking performance and the current course angle, namely course angle deviation, reflects the tracking performance of the vehicle on the expected path, and is used for ensuring the transverse tracking precision of the vehicle;
To predict the expected longitudinal speed v of the vehicle at time k in the time domain xr (k) Velocity v at the current moment of the vehicle x (k) The deviation of the expected speed curve, namely the longitudinal speed deviation, reflects the tracking performance of the vehicle to the expected speed curve and is used for ensuring the longitudinal tracking precision of the vehicle; Δa x (k) Delta for longitudinal acceleration f (k) For the increment of the front wheel steering angle, the two reactions are used for restraining the control increment, so that the large change of the wheel steering angle and the acceleration is avoided, and the stable control action is ensured.
Constructing a transverse and longitudinal joint constraint condition of an optimal control problem:
wherein: equation (24) is a variable dimension time delay augmentation dynamics model constraint, and the vehicle transverse and longitudinal state quantity and the control quantity are time-varying in a prediction time domain; equation (25) is the mechanical response characteristic constraint of the transverse and longitudinal actuators, including extreme value constraint of wheel rotation angle and longitudinal acceleration and increment constraint.
Finally, the nonlinear programming problem for the transversal and longitudinal cooperative control can be constructed as follows:
wherein J is MPC transverse-longitudinal cooperative controller, i is index of prediction state, and the value range is [ 1N ] p ]K is the current time,
respectively N p Transverse deviation cost term, course angle deviation cost term, longitudinal speed deviation, longitudinal control increment and transverse control increment cost term of vehicle at time k+i, N c To control the time domain, e y (k+i)、
Respectively N p Lateral deviation, heading angle deviation, longitudinal speed deviation, deltaa of the vehicle at time k+i x (k+i)、Δδ f (k+i) is N respectively p Longitudinal acceleration increment, front wheel steering angle increment, Q of vehicle at time k+i 1 、Q 2 、Q 3 、R 1 、R 2 Are all weight coefficients, χ 0 (k) Chi (k+1), chi (k+i-1) are the state variables of the vehicle at the times k, k+1, k+i-1, respectively, u 0 (k) U (k+i-1) is the control amount of the vehicle at the time k and k+i-1, a x (k+j)δ f (k+j) is the transverse and longitudinal control quantity of the vehicle at the time k+j, delta f,max 、a x,max Respectively, the maximum allowable transverse and longitudinal control amounts of the vehicle, delta f,max 、Δa x,max The maximum horizontal and longitudinal control increment is respectively allowed by the vehicle at adjacent time, delta f (k+j)、Δa x (k+j) is the transverse and longitudinal control increment of the vehicle at the time of k+j, f (χ) 0 (k),u 0 (k) F (χ (k+i-1), u (k+i-1)) are the state quantity and the initial of the vehicle at the time of k, k+i-1, respectively, for the variable dimension time delay augmentation modelResults of calculation of the control amount, δ f,max 、Δδ f,max 、a x,max 、Δa x,max The value of (2) is determined by the characteristics of the vehicle itself. />
Solving the constructed multi-constraint nonlinear optimization problem to obtain an optimal control sequence:
U * (k)＝[δ f * (k|k),...,δ f * (k|k+N c -1),a x * (k|k),a x * (k+2|k),...,a x * (k|k+N c -1)] (27)
taking delta f * (k|k) is sent to the bottom steering actuator as the current desired front wheel angle. Taking a x * And (k|k) is taken as the current expected longitudinal acceleration, drive/brake mode switching judgment is carried out based on the set dead zone, and finally a longitudinal control instruction is acquired by an expected acceleration-accelerator opening/brake pressure MAP meter and sent to a longitudinal actuator. The judgment logic is as follows:
wherein: mode (k) represents a k time control Mode, a x * For the desired longitudinal acceleration, Δa is the dead zone offset amount, and the dead zone formed by Δa can ensure smooth switching of the drive/brake mode.
The embodiment of the invention also provides a transverse and longitudinal cooperative control method of the unmanned vehicle, which takes time-varying time delay into consideration, and comprises the following steps:
and step 1, acquiring a time-varying CAN communication delay estimated value of the bottom layer on line. Wherein the bottom layer time-varying CAN communication delay estimated value is denoted as the bottom layer time-varying CAN communication delay estimated value tau at k moment CAN (k)。
Step 2, according to the found prediction time domain N p Within a series of reference point information (e.g. reference point information in the figure contains the desired abscissa x of the reference point r Desired ordinate y r Desired heading angle
The expected longitudinal speed and the path curvature kappa of the point) are combined with a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. to perform optimal control problem rolling solution by using a prediction model, and the control quantity of the vehicle is output.
In one embodiment, in step 2, the optimal control problem may be described as equation (26) in the above embodiment.
In one embodiment, in step 2, the prediction model is a variable-dimension time delay augmentation model constructed by using a transverse and longitudinal time delay link according to the bottom layer time-varying CAN communication time delay estimated value and the actuator time delay characteristic, and the variable-dimension time delay augmentation model may be represented by the formula (22) in the above embodiment. Of course, the predictive model may also be a conventional vehicle dynamics model to obtain control amounts. However, it should be noted that, by adopting the variable-dimension time delay augmentation model, a more accurate MPC horizontal-vertical cooperative controller can be obtained, thereby providing favorable conditions for improving the overall control effect.
In one embodiment, step 1 may be implemented as follows:
and taking the real-time input expected control instruction as a real-time signal, taking the bottom actual control instruction as a delay signal, carrying out on-line estimation of bottom time-varying CAN communication delay by utilizing a self-adaptive all-pass filtering delay estimator based on FIR and a delay estimation stabilization strategy based on a mean square error evaluation index, and outputting a bottom time-varying CAN communication delay estimation value.
It should be noted that, in step 1, the time-varying CAN communication delay estimated value of the bottom layer may be obtained online by using an existing method, such as estimation based on a linear regression method, estimation based on a least square method, estimation based on a least mean square error, and the like.
In one embodiment, in step 2, the FIR-based adaptive all-pass filtering time delay estimator is represented as equation (8) or (9) in the above embodiment.
In one embodiment, in step 2, the delay estimation stability strategy obtained based on the mean square error evaluation index is obtained using equation (18) in the above embodiment.
According to the invention, by combining the self-adaptive time delay estimator and the MPC transverse and longitudinal cooperative control algorithm considering time-varying time delay, the problem of vehicle control instability under the limit working condition caused by the characteristic of the unmanned vehicle when the bottom layer is ignored is solved, and the stability of the vehicle is improved while the transverse and longitudinal control accuracy is ensured.
As shown in fig. 3, the vehicle control system includes an environment sensing unit, a decision planning unit, a floor execution unit, and a floor delay estimation unit and a horizontal-vertical cooperative control unit established by the present invention. The environment sensing unit is used for acquiring environment information, and sending the environment information to the decision planning unit after processing the environment information; the decision planning unit performs global track planning according to the environment information and the vehicle state information and outputs reference track information to the transverse and longitudinal control unit; the bottom layer delay estimation unit uses the method in the invention to estimate the delay value online and outputs the delay value to the transverse and longitudinal control units; after the transverse and longitudinal control unit receives the reference track and the time delay estimated value, the transverse and longitudinal control instructions are calculated by using the method of the invention and finally sent to the bottom execution unit to control the vehicle, so that the accurate and stable transverse and longitudinal control of the unmanned vehicle is realized.
Finally, it should be pointed out that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting. Those of ordinary skill in the art will appreciate that: the technical schemes described in the foregoing embodiments may be modified or some of the technical features may be replaced equivalently; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (8)
1. The unmanned vehicle transverse and longitudinal cooperative control method taking time-varying time delay into consideration is characterized by comprising the following steps of:
step 1, acquiring a time-varying CAN communication delay estimated value tau of a bottom layer on line CAN : taking a real-time input expected control instruction as a real-time signal, taking a bottom actual control instruction as a delay signal, and obtaining tau by utilizing an adaptive all-pass filtering delay estimator based on FIR and a delay estimation stabilization strategy based on a mean square error evaluation index CAN ；
Step (a)2, according to the found prediction time domain N p The inner series of reference point information is utilized to perform optimal control problem rolling solution by combining a prediction model with a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. and output the control quantity of the vehicle;
in step 2, the optimal control problem is described as follows:
where i is the index of the predicted state, k is the current time,
the transverse deviation cost term, the course angle deviation cost term, the longitudinal speed deviation, the longitudinal control increment and the transverse control increment cost term of the vehicle at the moment k+i are respectively defined as N c To control the time domain, e y (k+i)、
The lateral deviation, the course angle deviation and the longitudinal speed deviation of the vehicle at the moment k+i are respectively, and the delta a is x (k+i)、Δδ f (k+i) is the longitudinal acceleration increment and the front wheel steering angle increment of the vehicle at the moment k+i, Q 1 、Q 2 、Q 3 、R 1 、R 2 Are all weight coefficients, χ 0 (k) Chi (k+1), chi (k+i-1) are the state variables of the vehicle at the times k, k+1, k+i-1, respectively, u 0 (k) U (k+i-1) is the control amount of the vehicle at the time k and k+i-1, a x (k+j)、δ f (k+j) is the transverse and longitudinal control amounts of the vehicle at the time k+j，δ f,max 、a x,max Respectively, the maximum allowable transverse and longitudinal control amounts of the vehicle, delta f,max 、Δa x,max The maximum horizontal and longitudinal control increment is respectively allowed by the vehicle at adjacent time, delta f (k+j)、Δa x (k+j) is the transverse and longitudinal control increment of the vehicle at the time of k+j, f (χ) 0 (k),u 0 (k) And f (χ (k+i-1), u (k+i-1)) are the results of calculation of the variable-dimension time delay augmentation model by using the state quantity and the initial control quantity of the vehicle at the time of k, k+i-1, respectively.
2. The unmanned vehicle transverse and longitudinal cooperative control method considering time-varying time delay according to claim 1, wherein in the step 2, the prediction model is a variable-dimension time delay augmentation model constructed by using transverse and longitudinal time delay links according to the bottom layer time-varying CAN communication time delay estimated value and the actuator time delay characteristic:
wherein ,
N τ representing the cause of the signal τ CAN And delay N τ The sampling time steps dt, χ (k) and χ (k+1) are the state quantities of the vehicle at times k and k+1, respectively, χ '(k) and χ' (k+1) are the first derivatives of χ (k) and χ (k+1) with respect to time, respectively, a x(k) and ax (k+1) is the longitudinal acceleration of the vehicle at times k and k+1, respectively, delta f(k) and δf (k+1) is the front wheel steering angle of the vehicle at the times k, k+1, respectively,
Respectively, vehicles are in k-N τ 、k-N τ +1、k-N τ +2, k-2, k-1, time kIs>
For the desired longitudinal acceleration of the vehicle at time k, dt is the sampling time step, +.>
The time lag constant of the first-order inertia link of the transverse actuator and the longitudinal actuator respectively, O is zero matrix, and +.>
Corresponding to the transition matrix A, B for the following simplified formula 1 、B 2 Is a discretization result of (a):
B 1 ＝[1 0 0 0 0] T
wherein ,Ccf 、C cr 、l f 、l r 、m、v x 、I z The rigidity of the front wheel side deflection, the rigidity of the rear wheel side deflection, the distance from the center of mass to the center of the front shaft, the distance from the center of mass to the center of the rear shaft, the mass, the longitudinal speed and the rotational inertia of the vehicle are respectively shown, and kappa is the curvature of the road at the tracking target point.
3. The unmanned vehicle transverse and longitudinal cooperative control method considering time-varying delay according to claim 1 or 2, wherein the FIR-based adaptive all-pass filtering delay estimator is expressed as formula (8) or (9):
wherein: x (k) is a real-time signal, x (k-N) τ ) In order to delay the signal in time,
expressed as vector +.>
Expressed as vector +.>
X + (k-N τ )＝[x(k+1-N τ ),...,x(k+n max -N τ )] T For the forward vector, X- (k) = [ X (k-1),.. max )] T Is a backward vector.
4. The unmanned vehicle transverse and longitudinal cooperative control method considering time-varying time delay according to claim 1 or 2, wherein the time delay estimation stabilization strategy obtained based on the mean square error evaluation index is obtained by adopting the formula (18):
where τ (k) represents the underlying delay estimate at time k, thr is the update threshold, MSE last Bottom layer delay estimation value representing k-1 moment
MSE index, MSE new Bottom layer delay estimate representing time k>
According to MSE index of (2)
Obtaining τ CAN 。
5. The utility model provides a take into account unmanned vehicles horizontal longitudinal cooperative control device of time-varying time delay which characterized in that includes:
an adaptive all-pass filtering time delay estimator for acquiring an underlying time-varying CAN communication time delay estimated value tau on line CAN ；
A horizontal-vertical cooperative controller for predicting the time domain N according to the found p And (3) carrying out optimal control problem rolling solving by utilizing a prediction model and combining a transverse and longitudinal integrated evaluation function J and transverse and longitudinal combined constraint s.t. and outputting vehicle control quantity, wherein: the optimal control problem is described as:
where i is the index of the predicted state, k is the current time,
the transverse deviation cost term, the course angle deviation cost term, the longitudinal speed deviation, the longitudinal control increment and the transverse control increment of the vehicle at the moment k+i are respectively replacedPrice item, N c To control the time domain, e y (k+i)、
The lateral deviation, the course angle deviation and the longitudinal speed deviation of the vehicle at the moment k+i are respectively, and the delta a is x (k+i)、Δδ f (k+i) is the longitudinal acceleration increment and the front wheel steering angle increment of the vehicle at the moment k+i, Q 1 、Q 2 、Q 3 、R 1 、R 2 Are all weight coefficients, χ 0 (k) Chi (k+1), chi (k+i-1) are the state variables of the vehicle at the times k, k+1, k+i-1, respectively, u 0 (k) U (k+i-1) is the control amount of the vehicle at the time k and k+i-1, a x (k+j)、δ f (k+j) is the transverse and longitudinal control quantity of the vehicle at the time k+j, delta f,max 、a x,max Respectively, the maximum allowable transverse and longitudinal control amounts of the vehicle, delta f,max 、Δa x,max The maximum horizontal and longitudinal control increment is respectively allowed by the vehicle at adjacent time, delta f (k+j)、Δa x (k+j) is the transverse and longitudinal control increment of the vehicle at the time of k+j, f (χ) 0 (k),u 0 (k) And f (χ (k+i-1), u (k+i-1)) are the results of calculation of the variable-dimension time delay augmentation model by using the state quantity and the initial control quantity of the vehicle at the time of k, k+i-1, respectively.
6. The unmanned vehicle transverse and longitudinal cooperative control device considering time-varying time delay according to claim 5, wherein the prediction model is a variable-dimension time delay augmentation model constructed by using a transverse and longitudinal time delay link according to an underlying time-varying CAN communication time delay estimated value and an actuator time delay characteristic:
wherein ,
N τ representing the cause of the signalτ CAN And delay N τ The sampling time steps dt, χ (k) and χ (k+1) are the state quantities of the vehicle at times k and k+1, respectively, χ '(k) and χ' (k+1) are the first derivatives of χ (k) and χ (k+1) with respect to time, respectively, a x(k) and ax (k+1) is the longitudinal acceleration of the vehicle at times k and k+1, respectively, delta f(k) and δf (k+1) is the front wheel steering angle of the vehicle at the times k, k+1, respectively,
Respectively, vehicles are in k-N τ 、k-N τ +1、k-N τ Desired front wheel steering angle at +2, k-2, k-1, k moment,/->
For the desired longitudinal acceleration of the vehicle at time k, dt is the sampling time step, +.>
The time lag constant of the first-order inertia link of the transverse actuator and the longitudinal actuator respectively, O is zero matrix, and +.>
Corresponding to the transition matrix A, B for the following simplified formula 1 、B 2 Is a discretization result of (a):
B 1 ＝[1 0 0 0 0] T
wherein ,Ccf 、C cr 、l f 、l r 、m、v x 、I z The rigidity of the front wheel side deflection, the rigidity of the rear wheel side deflection, the distance from the center of mass to the center of the front shaft, the distance from the center of mass to the center of the rear shaft, the mass, the longitudinal speed and the rotational inertia of the vehicle are respectively shown, and kappa is the curvature of the road at the tracking target point.
7. The unmanned vehicle transverse and longitudinal cooperative control device considering time-varying time delay as claimed in claim 5 or 6, wherein the bottom layer time-varying CAN communication time delay estimated value tau CAN The acquisition method of the method specifically comprises the following steps:
and taking the real-time input expected control instruction as a real-time signal, taking the bottom actual control instruction as a delay signal, carrying out on-line estimation of bottom time-varying CAN communication delay by utilizing a self-adaptive all-pass filtering delay estimator based on FIR and a delay estimation stabilization strategy based on a mean square error evaluation index, and outputting a bottom time-varying CAN communication delay estimation value.
8. The unmanned vehicle transverse and longitudinal cooperative control apparatus considering time-varying delay according to claim 7, wherein the FIR-based adaptive all-pass filtering delay estimator is expressed as formula (8) or (9):
wherein: x (k) is a real-time signal, x (k-N) τ ) In order to delay the signal in time,
expressed as vector +.>
Expressed as vector +.>
X + (k-N τ )＝[x(k+1-N τ ),...,x(k+n max -N τ )] T As forward vector, X- (k) = [ X (k-1), …, X (k-n) max )] T Is a backward vector;
the delay estimation stability strategy based on the mean square error evaluation index is obtained by adopting the formula (18):
where τ (k) represents the underlying delay estimate at time k, thr is the update threshold, MSE last Bottom layer delay estimation value representing k-1 moment
MSE index, MSE new Bottom layer delay estimate representing time k>
According to MSE index of (2)
Obtaining τ CAN 。/>
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211333752.0A CN115598983B (en) | 2022-10-28 | 2022-10-28 | Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay |
| PCT/CN2023/109408 WO2024087766A1 (en) | 2022-10-28 | 2023-07-26 | Unmanned vehicle horizontal and longitudinal cooperative control method and apparatus taking time-varying delay into account |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211333752.0A CN115598983B (en) | 2022-10-28 | 2022-10-28 | Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115598983A CN115598983A (en) | 2023-01-13 |
| CN115598983B true CN115598983B (en) | 2023-06-13 |
Family
ID=84851646
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211333752.0A Active CN115598983B (en) | 2022-10-28 | 2022-10-28 | Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN115598983B (en) |
| WO (1) | WO2024087766A1 (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115598983B (en) * | 2022-10-28 | 2023-06-13 | 湖南大学无锡智能控制研究院 | Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay |
| CN117234215B (en) * | 2023-11-09 | 2024-03-05 | 新石器中研(上海)科技有限公司 | Unmanned vehicle longitudinal control method and device, electronic equipment and unmanned vehicle |
| CN117970804B (en) * | 2024-01-31 | 2024-10-18 | 北京理工大学 | Fixed wing unmanned aerial vehicle cluster decision-making method under communication time delay |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108388115A (en) * | 2018-02-08 | 2018-08-10 | 南京邮电大学 | NCS method for compensating network delay based on generalized predictive control |
| CN111142534A (en) * | 2020-01-03 | 2020-05-12 | 江苏大学 | Intelligent vehicle transverse and longitudinal comprehensive track tracking method and control system |
| CN114216693A (en) * | 2021-12-02 | 2022-03-22 | 清华大学 | Dynamic load simulation method and test bench for vehicle composite braking system |
| CN114791743A (en) * | 2022-04-26 | 2022-07-26 | 北京理工大学 | Unmanned aerial vehicle cluster collaborative flight path planning method considering communication time delay |
| CN115042810A (en) * | 2022-06-13 | 2022-09-13 | 西南交通大学 | Transverse and longitudinal comprehensive control method and transverse and longitudinal comprehensive controller for vehicle |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3973922B2 (en) * | 2002-02-15 | 2007-09-12 | 本田技研工業株式会社 | Control device |
| CN110775154B (en) * | 2019-10-25 | 2021-02-05 | 南京航空航天大学 | Electric truck steer-by-wire system and network uncertain control method thereof |
| CN113759707B (en) * | 2020-12-10 | 2024-06-18 | 北京京东乾石科技有限公司 | Predictive control optimization method and device for automatic driving vehicle |
| CN112693449B (en) * | 2021-01-26 | 2022-02-15 | 湖南大学 | Transverse and longitudinal coupling control method under limit working condition of unmanned vehicle |
| CN113076641B (en) * | 2021-03-31 | 2022-09-20 | 同济大学 | Intelligent vehicle-to-vehicle and computer-to-vehicle cooperative steering control parallel computing method based on risk assessment |
| CN114967475B (en) * | 2022-08-01 | 2022-11-15 | 北京理工大学 | Unmanned vehicle trajectory tracking and stability robust control method and system |
| CN115598983B (en) * | 2022-10-28 | 2023-06-13 | 湖南大学无锡智能控制研究院 | Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay |
-
2022
- 2022-10-28 CN CN202211333752.0A patent/CN115598983B/en active Active
-
2023
- 2023-07-26 WO PCT/CN2023/109408 patent/WO2024087766A1/en unknown
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108388115A (en) * | 2018-02-08 | 2018-08-10 | 南京邮电大学 | NCS method for compensating network delay based on generalized predictive control |
| CN111142534A (en) * | 2020-01-03 | 2020-05-12 | 江苏大学 | Intelligent vehicle transverse and longitudinal comprehensive track tracking method and control system |
| CN114216693A (en) * | 2021-12-02 | 2022-03-22 | 清华大学 | Dynamic load simulation method and test bench for vehicle composite braking system |
| CN114791743A (en) * | 2022-04-26 | 2022-07-26 | 北京理工大学 | Unmanned aerial vehicle cluster collaborative flight path planning method considering communication time delay |
| CN115042810A (en) * | 2022-06-13 | 2022-09-13 | 西南交通大学 | Transverse and longitudinal comprehensive control method and transverse and longitudinal comprehensive controller for vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115598983A (en) | 2023-01-13 |
| WO2024087766A1 (en) | 2024-05-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115598983B (en) | Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay | |
| CN112733270B (en) | System and method for predicting vehicle running track and evaluating risk degree of track deviation | |
| CN110356404B (en) | Intelligent driving system with autonomous lane changing function and improved lateral safety | |
| CN110568758B (en) | Parameter self-adaptive transverse motion LQR control method for automatically driving automobile | |
| CN111240187B (en) | Vehicle track tracking control algorithm based on vehicle error model | |
| JP6541878B2 (en) | Vehicle driving support device and vehicle driving support method | |
| CN108791290B (en) | Double-vehicle cooperative adaptive cruise control method based on online incremental DHP | |
| CN112141101B (en) | Method and system for pre-aiming safety path based on CNN and LSTM | |
| CN114945885A (en) | Adaptive control of autonomous or semi-autonomous vehicles | |
| CN110267866B (en) | Apparatus for controlling vehicle trajectory | |
| CN113608530B (en) | Parameter self-tuning LQR path tracking method with PID corner compensation | |
| CN114502450B (en) | Dead time compensation technique in lateral and longitudinal guidance of motor vehicle | |
| CN112298354B (en) | State estimation method for steering wheel and front wheel corner of steering system of unmanned automobile | |
| CN114942642A (en) | Unmanned automobile track planning method | |
| JP2023017703A (en) | Method and device for optimally parameterizing vehicle dynamic control system of vehicle | |
| CN114879698B (en) | Robot driving vehicle obstacle avoidance method based on improved artificial potential field and MPC | |
| CN116834754A (en) | Transverse and longitudinal cooperative control method for self-adaptive speed regulation of automatic driving vehicle | |
| CN115675459A (en) | Intelligent vehicle and control method, device and storage medium for path tracking of intelligent vehicle | |
| CN115432009A (en) | Automatic driving vehicle trajectory tracking control system | |
| CN114779641A (en) | Environment self-adaptive MPC path tracking control method based on new course error definition | |
| CN116691668A (en) | Transverse control method for large intelligent vehicle | |
| CN117389276B (en) | Unmanned vehicle driving path tracking control method based on driving risk prediction | |
| CN114625002A (en) | Vehicle transverse and longitudinal integrated control method based on model predictive control | |
| JP2001134320A (en) | Lane follow-up controller | |
| CN111413979B (en) | Automobile track tracking control method based on rapid model prediction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |