CN113386781A - Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model - Google Patents
Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model Download PDFInfo
- Publication number
- CN113386781A CN113386781A CN202110563037.5A CN202110563037A CN113386781A CN 113386781 A CN113386781 A CN 113386781A CN 202110563037 A CN202110563037 A CN 202110563037A CN 113386781 A CN113386781 A CN 113386781A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- model
- data
- vehicle dynamics
- layer
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000013528 artificial neural network Methods 0.000 claims abstract description 33
- 238000004364 calculation method Methods 0.000 claims abstract description 19
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 9
- 238000012360 testing method Methods 0.000 claims description 24
- 238000012549 training Methods 0.000 claims description 24
- 238000005457 optimization Methods 0.000 claims description 12
- 238000013461 design Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 5
- 230000006399 behavior Effects 0.000 claims description 4
- 230000009021 linear effect Effects 0.000 claims description 4
- 230000010355 oscillation Effects 0.000 claims description 4
- 238000012545 processing Methods 0.000 claims description 4
- 238000004088 simulation Methods 0.000 claims description 4
- 230000009286 beneficial effect Effects 0.000 claims description 3
- 230000003111 delayed effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 2
- 241000270295 Serpentes Species 0.000 claims description 2
- 230000001133 acceleration Effects 0.000 claims description 2
- 230000003044 adaptive effect Effects 0.000 claims description 2
- 239000010426 asphalt Substances 0.000 claims description 2
- 238000009499 grossing Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 239000000725 suspension Substances 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000013480 data collection Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009022 nonlinear effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000013049 sediment Substances 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units, or advanced driver assistance systems for ensuring comfort, stability and safety or drive control systems for propelling or retarding the vehicle
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/09—Taking automatic action to avoid collision, e.g. braking and steering
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M17/00—Testing of vehicles
- G01M17/007—Wheeled or endless-tracked vehicles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/15—Vehicle, aircraft or watercraft design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W2050/0001—Details of the control system
- B60W2050/0019—Control system elements or transfer functions
- B60W2050/0028—Mathematical models, e.g. for simulation
- B60W2050/0031—Mathematical model of the vehicle
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
- B60W40/06—Road conditions
- B60W40/064—Degree of grip
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
- B60W40/06—Road conditions
- B60W40/068—Road friction coefficient
Abstract
The invention discloses an intelligent vehicle track tracking control method based on a data-driven vehicle dynamics model. In the model-based control, after finishing the two-stage learning, extracting the weight parameters of the obtained neural network vehicle dynamics model to perform forward calculation in a subsequent trajectory tracking control algorithm. Based on the assumption of steady-state turning of the unmanned vehicle, the feedforward front wheel rotation angle and the feedforward steady-state mass center slip angle are obtained by utilizing the learned neural network vehicle dynamics model, the feedforward steady-state mass center slip angle of the vehicle is included in the steering feedback control based on the path, and the tracking control of the reference track is realized.
Description
Technical Field
The invention belongs to the field of intelligent vehicle dynamics control, and particularly relates to a trajectory tracking control method of a vehicle dynamics prediction model based on a data driving technology.
Background
With the continuous improvement of the requirements of drivers on the safety, the maneuverability and the riding comfort of vehicles and the gradual maturity of control theories, the research of automobile intelligent technology is concerned widely. The control technology based on the vehicle dynamics model of the unmanned vehicle can realize better road utilization rate and higher safety, but also needs to adapt to various complex driving environments, such as driving on roads with different road adhesion coefficients and curvature changes or realizing safe and stable emergency obstacle avoidance operation under emergency working conditions.
Most of the existing control methods calculate physical quantities describing vehicle motion such as vehicle yaw angular velocity and the like through a vehicle dynamic mathematical model based on physical derivation, and then design a feedback control system for tracking. However, the vehicle dynamic mathematical model based on physical derivation usually makes certain idealized assumptions during modeling to simplify the vehicle model, which results in that the real dynamic response of the vehicle during driving cannot be accurately calculated, and particularly under extreme conditions, the vehicle system and related subsystems can show high nonlinearity and strong coupling characteristics. Furthermore, model-based trajectory tracking control methods are typically affected by model parameter perturbations, uncertainty interference, time lag, actuator saturation constraints, and the like, and if lateral tire forces in a non-linear region are considered linear or the driving environment changes suddenly, the behavior of the vehicle may become uncontrollable, resulting in an unmanned vehicle that will lose path tracking capability and stability. Therefore, how to establish the unmanned vehicle dynamics prediction model under the condition of simultaneously considering the calculation real-time performance and the fidelity of the model and develop a trajectory tracking control technology based on the model becomes an important problem which needs to be solved urgently at present.
Disclosure of Invention
An intelligent vehicle track tracking control method based on a data-driven vehicle dynamics model comprises model training learning and model-based control. In model training and learning, vehicle dynamics data sets are respectively obtained through a virtual high-fidelity vehicle test and an actual unmanned vehicle test, a neural network vehicle dynamics model is designed based on the time-lag feedback idea and the neural network nonlinear modeling principle, and the obtained vehicle dynamics data sets are used for performing two-stage training and learning on the model. In the model-based control, after finishing the two-stage learning, extracting the weight parameters of the obtained neural network vehicle dynamics model to perform forward calculation in a subsequent trajectory tracking control algorithm. Based on the assumption of steady-state turning of the unmanned vehicle, the feedforward front wheel rotation angle and the feedforward steady-state mass center slip angle are obtained by utilizing the learned neural network vehicle dynamics model, the feedforward steady-state mass center slip angle of the vehicle is included in the steering feedback control based on the path, and the tracking control of the reference track is realized.
The invention has the beneficial effects that:
(1) the invention provides a virtual and actual data collection method for vehicle dynamics with multiple road adhesion coefficients, which lays a data foundation for building a model for vehicle dynamics. The vehicle dynamics virtual data set is high in fidelity, high in confidence coefficient and low in acquisition cost, the demand of real vehicle data can be reduced, the vehicle dynamics real data set can optimize weight parameters for the model again, the accuracy of actual vehicle dynamics prediction response is improved, and the sensors in the provided real vehicle data acquisition device are low in cost and simple in arrangement, and can be widely applied to various vehicle types. .
(2) The vehicle dynamics prediction model based on the time-lag feedback neural network designed by the data driving technology can identify various complex dynamics behaviors in the vehicle running process, including limit cycle, chaos, bifurcation, high nonlinearity and strong coupling characteristics, and can make appropriate prediction on the road surface on which the vehicle runs without clear road surface friction estimation.
(3) The method is based on the assumption of steady-state turning of the unmanned vehicle, the feedforward front wheel rotation angle and the feedforward steady-state mass center slip angle are obtained by utilizing the learned neural network vehicle dynamic model, the feedforward steady-state mass center slip angle of the vehicle is brought into steering feedback control based on the path, tracking control of an expected reference track can be realized under any road conditions and driving conditions, the path tracking precision is ensured, and meanwhile, the transverse stability can be well considered.
Drawings
FIG. 1 is a flow chart of an intelligent vehicle trajectory tracking control based on a data-driven vehicle dynamics model;
FIG. 2 is a flow chart of a high fidelity vehicle dynamics model multi-time step virtual data acquisition;
FIG. 3 is a flow chart of actual unmanned vehicle dynamics real data acquisition;
FIG. 4 is a vehicle dynamics prediction model of a time-lag feedback neural network designed based on data-driven technology;
FIG. 5 is a two-stage training structure diagram of a vehicle dynamics prediction model based on a time-lag feedback neural network;
FIG. 6 is a schematic diagram of a monorail model and trajectory tracking for an unmanned vehicle;
FIG. 7 is a logic flow diagram of a model-based intelligent vehicle trajectory tracking control algorithm;
Detailed Description
The invention will be further explained with reference to the drawings.
Fig. 1 is a general flow chart of intelligent vehicle trajectory tracking control based on a data-driven vehicle dynamics model, including model training learning and model-based trajectory tracking control, as follows:
model training and learning: the method comprises the steps of respectively obtaining vehicle dynamics data sets through a virtual high-fidelity vehicle test and an actual unmanned vehicle test, designing a neural network vehicle dynamics model based on the time-lag feedback idea and the neural network nonlinear modeling principle, and carrying out two-stage training learning on the model by using the obtained vehicle dynamics data sets.
Model-based trajectory tracking control: and after finishing the two-stage learning, extracting the weight parameters of the obtained neural network vehicle dynamics model so as to perform forward calculation in a follow-up trajectory tracking control algorithm. Based on the assumption of steady-state turning of the unmanned vehicle, the feedforward front wheel rotation angle and the feedforward steady-state mass center slip angle are obtained by utilizing the learned neural network vehicle dynamics model, the feedforward steady-state mass center slip angle of the vehicle is included in the steering feedback control based on the path, and the tracking control of the reference track is realized.
FIG. 2 is a flow chart of a high fidelity vehicle dynamics model multi-time step virtual data acquisition. According to the whole vehicle parameter configuration of an actual unmanned vehicle object, vehicle parameters in CarSim high-fidelity vehicle dynamics simulation software are modified, such as vehicle type grade selection, vehicle body length, width, height, wheel base, minimum ground clearance, finishing quality, whole vehicle rotational inertia and the like, and a data acquisition joint simulation model is established in Matlab/Simulink.
Configuring different road transverse and longitudinal sections and adhesion coefficients, and performing combined sinusoidal input on the vehicle speed and the front wheel steering angle in the combined simulation model of CarSim-Matlab/Simulink according to the limit relation of the mutual influence between the vehicle longitudinal speed and the front wheel steering angle:
where d is the total number of sinusoids, amIs the maximum amplitude, ω, of the input longitudinal vehicle speed and front wheel rotation anglemIs the maximum frequency of the input, #jIs the initial phase angle of the sinusoid, j represents the jth sinusoid.
After obtaining the continuous output response signal, the input and output data are processed in multiple time steps (the multiple time step processing means that a complete vehicle dynamics data time series is processed into a plurality of vehicle dynamics data series consisting of 4 time steps, for example, 12345678 is processed as: 1234/2345/3456 … …), and a vehicle dynamics virtual data set is obtained.
Fig. 3 is a flow chart of actual unmanned vehicle dynamics real data acquisition. The unmanned vehicle need only be equipped with an integrated navigation system (GPS global positioning system + inertial navigation) and receive data from the vehicle CAN bus. The method is characterized in that an unmanned vehicle is operated to perform various vehicle tests such as a drift test, a snake test, a single-shift test, a sine frequency sweep steering test, a steady-state rotation test, a quasi-static linear acceleration and deceleration test, an ISO double-shift running test and the like on a dry asphalt pavement, a wet and slippery sediment pavement and an ice and snow pavement, and actual unmanned vehicle dynamic data under high, medium and low pavement adhesion coefficients are obtained. And smoothing the collected data by using a second-order Butterworth low-pass filter with the cut-off frequency of 3Hz to filter the influence of high-frequency behaviors such as suspension vibration on vehicle dynamics. And finally, carrying out data synchronization and multi-time-step processing to obtain an actual vehicle dynamics real data set.
FIG. 4 is a vehicle dynamics prediction model of a time-lag feedback neural network designed based on data-driven technology. Designing a neural network vehicle dynamics model with delayed output feedback by utilizing the nonlinear modeling capability of a neural network; a global neural network vehicle dynamics model is primarily learned that, by including all unknown or unmodeled vehicle dynamics variation effects in the data set, the model can learn potential unknown dynamics state variations including tire non-linear effects, load shifts, etc. of the vehicle.
The neural network vehicle dynamics model specifically adopts the following structure: the first layer is an input layer which has 7 characteristic inputs, namely a yaw angular velocity r and a transverse velocity UyLongitudinal speed UxFront wheel steering angle delta, vehicle longitudinal force FxThe output of the sum network model becomes the derivative output value of the yaw rate and the transverse speed of the input information through time delay feedback
The data for each input feature contains vehicle dynamics information for a total of 4 time steps. The second layer is the LSTM1 network layer, and the hidden layer is designed to have 64 hidden units. The third layer is an activation layer, and the activation function is selected to be a Relu function. The fourth layer is FC1 full link layer, and the hidden layer design has 64 hidden units. The fourth layer is the LSTM2 network layer, the hidden layer is designed to have 64 hidden units, and only outputs the information after the last LSTM-CELL calculation. The fifth layer is an FC2 output regression full link layer, and the hidden layer design has 2 hidden units.
The model predicts the current vehicle yaw rate and lateral velocity derivatives by including vehicle control and state information with 4 time-step states in the input data, particularly state output parameters including time-lapse feedback in the state input data.
The forward calculation method of the learned neural network vehicle dynamics model is as follows:
h＝{xt,…,xt-T}
a_l＝max(0,z1)
wherein x istRepresenting delayed feedback status output, control and status input information in a single time step, h representing x comprising a plurality of historical time step informationtAnd (4) data.
Wlstm{1,2}∈(wi,wf,wg,wo),blstm{1，2}∈(bi,bf,bg,bo) Represents the learned network weights and bias parameters, w, of the 2 LSTM network layersi,wf,wg,wo,bi,bf,bg,boRespectively expressed as weight matrix of an input gate layer, a forgetting gate layer, a Tanh layer and an output gate layer in the LSTM network and bias matrix of the input gate layer, the forgetting gate layer, the Tanh layer and the output gate layer.
Transpose matrix for 2 full-link layer weights, bFC1,bFC2A bias matrix of 2 fully connected layers. Wherein in the equation of the network model, FlstmIs an abbreviation of LSTM network model. a _ l represents an active layer, ziAnd i is 1,2 and 3, which represent weighted outputs of different network layers. Predicted output of network
And
is defined as:
and
where Δ t is 10ms, which is the sampling frequency of the signal.
FIG. 5 is a diagram of a two-stage training architecture of a vehicle dynamics prediction model based on a time-lag feedback neural network. In the first stage, a vehicle dynamics virtual data training set is used for learning to obtain a pre-training model, and the overall error of the virtual data testing set is tested. And carrying out model weight optimization of the second stage by using the vehicle dynamics real data set. Details of network training are as follows: the Loss function is selected as mean square error MSE, the optimizer is selected as Adam, the batch size is set to be 1000, the learning rate is set to be 0.0001, the network model is subjected to learning training based on a learning framework of Pythroch, and an optimization training algorithm is as follows:
wherein, r, UyAnd
the mean square error MSE calculation method is described as follows: derivative output values of predicted vehicle yaw rate and lateral velocity using derived network
Actual vehicle yaw rate and lateral velocity measurements r, UyCarrying out Euler integral calculation of time step delta t being 10ms to obtain the predicted values of the vehicle yaw velocity and the transverse velocity of the next time step predicted by the network
Will predict what is obtained
And label truth value r, UyThe mean square error is calculated, where N is the total number of samples.
And extracting model weight parameters after training is finished, wherein the obtained weight parameters are used for forward calculation to obtain a time-lag feedback neural network-based vehicle dynamics prediction model of the unmanned vehicle.
FIG. 6 is a schematic diagram of a monorail model and trajectory tracking for an unmanned vehicle, where U is the velocity at the vehicle's center of mass; u shapex、UyThe speeds of the vehicle mass center along the x and y directions of the vehicle body coordinate system are respectively; alpha is alphaf,αrRespectively are front and rear wheel side deflection angles; beta is the vehicle mass center slip angle; r is the vehicle yaw rate; l ═ a + b is the wheelbase of the vehicle; a and b are the distances from the mass center of the vehicle to the front and rear axes; m is the vehicle mass, IzThe moment of inertia of the vehicle around the z-axis of the center of mass; fyf,FyrThe lateral resultant force of the front axle tire and the rear axle tire is respectively; fxfThe longitudinal resultant force received by the front axle tire; delta is the front wheel corner, e is the lateral deviation; x is the number ofLAThe longitudinal pre-aiming distance is adopted; e.g. of the typeLAThe transverse deviation at the pre-aiming point is obtained; Δ ψ is the heading bias.
FIG. 7 is a logic flow diagram of a model-based intelligent vehicle trajectory tracking control algorithm. In order to obtain the feedforward steering angle and the centroid slip angle, a second-order nonlinear optimization method is used for solving a balance point of a neural network vehicle dynamic model. Curvature kappa of reference path s and vehicle longitudinal speed U measured by lane line detection modulexAnd as input to an optimization solver to calculate the correct feed-forward steering command.
Based on the vehicle steady-state turning conditions, the following assumptions are set:
wherein the content of the first and second substances,
the model output under the steady-state assumption is subjected to time delay feedback to become the derivative of the yaw velocity and the transverse velocity of the input information;
the model output is fed back with time delay to become the derivative of the yaw rate and lateral velocity of the input information.
Combining with a kinematic model, calculating to obtain a steady state yaw velocity rssAnd steady-state longitudinal force Fx,ss：
Fx,ss＝-mrssUy
Constraining the control input, the delay state and the derivative thereof of the model in multiple time steps to be the same, and calculating the formula as follows:
wherein the content of the first and second substances,
for instantaneous measurement of longitudinal speed along the coordinate system of the body at the center of mass of the vehicle, Uy,ss,δssAnd the transverse speed and the front wheel rotation angle of the vehicle mass center needing to be solved in the steady state along the vehicle body coordinate system.
The method of finding stable points in a neural network vehicle dynamics model is to find points where the derivative of the output state is zero for a given set of controls, states and their derivative inputs. Measuring the speed of the vehicle in the actual running process and the curvature of the reference path, and calculating the obtained steady-state value rss,Fx,ssWill be
rss,Fx,ssAs the input of NNVM, solving the nonlinear optimization problem by a constrained second-order inner point method, and taking the feedforward front wheel steering angle based on kinematics as deltaffwIs optimized and iterated, so as to calculate the maximum and minimum limit constraints (delta) in the steeringmin,δmax) Feed-forward front wheel steering angle delta at steady state equilibrium within rangeffwAnd feed forward lateral velocity Uy,ffwThe nonlinear optimization solution problem is described as follows:
subject toδ≤δmax
δ≥δmin
calculating to obtain steady-state feedforward mass center slip angle beta by utilizing steady-state feedforward transverse velocity and longitudinal velocityffwThe calculation formula is as follows:
βffw＝arctan(Uy,ffw/Ux)
to compensate for errors and disturbances caused by feed forward steering, a path-based steering feedback controller is used to compensate for tracking the desired trajectory. The steady-state centroid slip angle of the vehicle is included in the feedback control, the path tracking effect can be improved, and the robust stability of the preview steering feedback is kept.
The feedback-controlled steering angle deltafbBased on the steady-state feedforward mass center slip angle, measuring the transverse deviation e of the vehicle at the aiming point deviating from the expected track through a lane line detection moduleLAAnd the heading deviation delta psi of the vehicle deviating from the expected track, and the calculation formula is as follows:
δfb＝-kp(eLA+xLAsin(Δψ+βffw))
in the formula kpFor the controller gain, the choice should ensure stability of the control system with the pre-aiming distance xLAThe control system may be more sensitive to heading errors and producing a fast response to yaw motion of the vehicle may be beneficial to improve stability. But if the pre-aiming distance xLAToo long, the control system steering commands will be too sensitive to yaw motion, potentially creating yaw oscillations. Thus, an appropriate pre-line distance x is selected when the tire is saturatedLAThe balance between yaw stability and yaw oscillation should be considered. Establishing the adaptive rate of the pre-aiming distance along with the curvature of the route and the variable speed of the vehicle as follows:
the range of the vehicle speed U is 0-120 km/h, and the initial pre-aiming distance is set to be 8 m. When U is 120km/h and k is 0, the preview distance is 20m at most. When the vehicle speed is constant, the pre-aiming distance is reduced along with the increase of the curvature of the route. When the curvature of the route is fixed, the larger the vehicle speed is, the longer the pre-aiming distance is.
Finally, the obtained feedforward steering angle and the feedback steering angle are added to obtain a final transverse path tracking steering angle delta which is:
δ＝δffw+δfb
the above-listed series of detailed descriptions are merely specific illustrations of possible embodiments of the present invention, and they are not intended to limit the scope of the present invention, and all equivalent means or modifications that do not depart from the technical spirit of the present invention are intended to be included within the scope of the present invention.
Claims (8)
1. An intelligent vehicle track tracking control method based on a data-driven vehicle dynamics model is characterized by comprising a model design and training method and a track tracking control method based on a model;
the model design and training method comprises the following steps: respectively acquiring vehicle dynamics data sets through a virtual high-fidelity vehicle test and an actual unmanned vehicle test, designing a neural network vehicle dynamics model based on the time-lag feedback thought and the long-and-short time memory neural network principle, and performing two-stage training learning on the model by using the obtained vehicle dynamics data sets;
the model-based trajectory tracking control method comprises the following steps: and after finishing the two-stage training and learning, extracting the weight parameters of the obtained neural network vehicle dynamics model to perform forward calculation in a follow-up track tracking control algorithm, obtaining a feedforward front wheel corner and a feedforward steady-state mass center slip angle by using the learned neural network vehicle dynamics model based on the steady-state turning assumption of the unmanned vehicle, and incorporating the feedforward steady-state mass center slip angle of the vehicle into the path-based steering feedback control to realize the tracking control of the reference track.
2. The intelligent vehicle trajectory tracking control method based on the data-driven vehicle dynamics model according to claim 1, characterized in that the model design adopts a neural network vehicle dynamics model, and the specifically adopted structure is as follows: the first layer is an input layer which has 7 characteristic inputs, namely a yaw angular velocity r and a transverse velocity UyLongitudinal speed UxFront wheel steering angle delta, vehicle longitudinal force FxThe output of the sum network model becomes the derivative output value of the yaw rate and the transverse speed of the input information through time delay feedback
The data of each input feature contains 4 time steps of vehicle dynamics information, and the second layer is LSTM1, a hidden layer is designed to have 64 hidden units, a third layer is an activation layer, the activation function is selected to be a Relu function, a fourth layer is an FC1 full-connection layer, the hidden layer is designed to have 64 hidden units, the fourth layer is an LSTM2 network layer, the hidden layer is designed to have 64 hidden units and only outputs information after the last LSTM-CELL calculation, a fifth layer is an FC2 output regression full-connection layer, and the hidden layer is designed to have 2 hidden units;
the model predicts the current vehicle yaw rate and lateral velocity derivatives by including vehicle control and state information with 4 time-step states in the input data, particularly state output parameters including time-lapse feedback in the state input data.
3. The intelligent vehicle trajectory tracking control method based on the data-driven vehicle dynamics model according to claim 2, characterized in that the forward calculation method of the neural network vehicle dynamics model is as follows:
h＝{xt,…,xt-T}
a_l＝max(0,z1)
wherein x istRepresenting delayed feedback status output, control and status input information in a single time step, h representing x comprising a plurality of historical time step informationtAnd (4) data. (ii) a
Wlstm{1,2}∈(wi,wf,wg,wo),blstm{1，2}∈(bi,bf,bg,bo) Represents the learned network weights and bias parameters, w, of the 2 LSTM network layersi,wf,wg,wo,bi,bf,bg,boRespectively expressed as weight matrixes of an input gate layer, a forgetting gate layer, a Tanh layer and an output gate layer in the LSTM network and bias matrixes of the input gate layer, the forgetting gate layer, the Tanh layer and the output gate layer,
transpose matrix for 2 full-link layer weights, bFC1,bFC2A bias matrix of 2 fully connected layers; wherein in the equation of the network model, FlstmIs an abbreviation of the LSTM network model, a _ l stands for activation layer, ziI-1, 2,3 represents weighted output of different network layers, predicted output of network
And
is defined as:
and
where Δ t is 10ms, which is the sampling frequency of the signal.
4. An intelligent vehicle trajectory tracking control method based on a data-driven vehicle dynamics model according to claim 2 or 3, characterized in that the training learning method of the model comprises two stages:
in the first stage, a vehicle dynamics virtual data set is used for learning to obtain a pre-training model, and the overall error of the virtual data test set is tested; in the second stage, model weight optimization is carried out by utilizing a vehicle dynamics real data set;
the network training is as follows: the Loss function is selected as mean square error MSE, the optimizer is selected as Adam, the batch size is set to be 1000, the learning rate is set to be 0.0001, the network model is subjected to learning training based on a learning framework of Pythroch, and an optimization training algorithm is as follows:
wherein, r, Uy，
The mean square error MSE calculation method is as follows: derivative output values of predicted vehicle yaw rate and lateral velocity using derived network
Actual vehicle yaw rate and lateral velocity measurements r, UyWhen in progressAnd performing Euler integral calculation on the time step delta t of 10ms to obtain predicted values of the vehicle yaw rate and the transverse speed of the next time step predicted by the network
Will predict what is obtained
And label truth value r, UyCalculating a mean square error, wherein N is the total number of samples;
and extracting model weight parameters after training is finished, wherein the obtained weight parameters are used for forward calculation to obtain a time-lag feedback neural network-based vehicle dynamics prediction model of the unmanned vehicle.
5. The intelligent vehicle trajectory tracking control method based on the data-driven vehicle dynamics model according to claim 4, characterized in that the virtual data set is obtained by the following method:
setting vehicle parameters in CarSim high-fidelity vehicle dynamics simulation software, such as vehicle type grade, vehicle body length, width, height, wheel base, minimum ground clearance, servicing mass, finished vehicle rotational inertia and the like, and establishing a data acquisition joint simulation model in Matlab/Simulink;
configuring different road transverse and longitudinal sections and adhesion coefficients, and performing combined sinusoidal input on the vehicle speed and the front wheel steering angle in the combined simulation model of CarSim-Matlab/Simulink according to the limit relation of the mutual influence between the vehicle longitudinal speed and the front wheel steering angle:
where d is the total number of sinusoids, amIs the maximum amplitude, ω, of the input longitudinal vehicle speed and front wheel rotation anglemIs the maximum frequency of the input, #jIs the initial phase angle of the sinusoid, j represents the jth sinusoid;
and after the continuous output response signals are obtained, performing multi-time-step processing on the input and output data to obtain a vehicle dynamics virtual data set.
6. The intelligent vehicle trajectory tracking control method based on the data-driven vehicle dynamics model according to claim 4, characterized in that the real data set is obtained by the following method:
the method comprises the steps of installing an integrated navigation system (a GPS (global positioning system) and inertial navigation) on an unmanned vehicle, receiving data from a vehicle CAN (controller area network) bus, operating the unmanned vehicle to perform various vehicle tests such as a drift test, a snake test, a single-line-shifting test, a sine sweep frequency steering test, a steady-state rotation test, a quasi-static linear acceleration and deceleration test, an ISO double-line-shifting running test and the like on a dry asphalt pavement, a wet silt pavement and an ice and snow pavement, obtaining actual unmanned vehicle dynamics data under high, medium and low pavement adhesion coefficients, smoothing the collected data by using a second-order Butterworth low-pass filter with a cut-off frequency of 3Hz to filter the influence of high-frequency behaviors such as suspension vibration on vehicle dynamics, and finally performing data synchronization and multi-time-step processing to obtain an actual vehicle dynamics data set.
7. The intelligent vehicle trajectory tracking control method based on the data-driven vehicle dynamics model according to claim 1, characterized in that the trajectory tracking control method is to solve the balance point of the neural network vehicle dynamics model by using a second-order nonlinear optimization method to obtain a feedforward steering angle and a centroid slip angle, and then obtain the feedforward steering angle and the centroid slip angle through the measured reference path curvature κ and the vehicle longitudinal speed UxAs input to the optimization solver to calculate the correct feed forward steering command.
8. The intelligent vehicle trajectory tracking control method based on the data-driven vehicle dynamics model according to claim 7, characterized in that the trajectory tracking control method comprises the following specific processes:
based on the vehicle steady-state turning conditions, assume: :
combining with a kinematic model, calculating to obtain a steady state yaw velocity rssAnd steady-state longitudinal force Fx,ss：
Fx,ss＝-mrssUy
Constraining the control input, the delay state and the derivative thereof of the model in multiple time steps to be the same, and calculating the formula as follows:
the method for searching the stable balance point in the neural network vehicle dynamics model comprises the following steps: for a given set of controls, states and their derivative inputs, finding a point where the output state derivative is zero; measuring the actual movement of the vehicleThe speed and the reference path curvature in the course of the line, and the calculated steady state value rss,Fx,ssWill be
rss,Fx,ssAs the input of the model, solving the nonlinear optimization problem by a constrained second-order inner point method, and taking the feedforward front wheel steering angle based on kinematics as deltaffwIs optimized and iterated, so as to calculate the maximum and minimum limit constraints (delta) in the steeringmin,δmax) Feed-forward front wheel steering angle delta at steady state equilibrium within rangeffwAnd feed forward lateral velocity Uy,ffwThe nonlinear optimization solution algorithm is described as follows:
subject to δ≤δmax
δ≥δmin
the steady-state feedforward mass center slip angle is obtained by calculating the steady-state feedforward transverse velocity and the steady-state feedforward longitudinal velocity, and the calculation formula is as follows:
βffw＝arctan(Uy,ffw/Ux)
to compensate for errors and disturbances caused by feed-forward steering, a path-based steering feedback controller is used to compensate for tracking the desired trajectory; the steady-state mass center slip angle of the vehicle is included in feedback control to improve the path tracking effect and keep the robust stability of the preview steering feedback;
the feedback-controlled steering angle is based on a steady-state feed-forward centroid slip angle by measuring a forward looking lateral deviation e of the vehicle from a desired trajectoryLAAnd the heading deviation delta psi of the vehicle deviating from the expected track, and the calculation formula is as follows:
δfb＝-kp(eLA+xLAsin(Δψ+βffw))
in the formula kpFor the controller gain, the choice should ensure stability of the control system with the pre-aiming distance xLATo improve stability, the control system may be more sensitive to heading errors and producing a fast response to yaw motion of the vehicle may be beneficial if the pre-aiming distance xLAIf the length is too long, the control system steering command is too sensitive to the yaw movement, and the yaw oscillation can be generated; thus, an appropriate pre-line distance x is selected when the tire is saturatedLAConsidering the balance between yaw stability and yaw oscillation, the adaptive rate of the variation of the pre-aiming distance with the curvature of the journey and the speed of the vehicle is established as follows:
the range of the vehicle speed U is 0-120 km/h, an initial pre-aiming distance is set to be 8m, when U is 120km/h and k is 0, the maximum pre-aiming distance is 20m, when the vehicle speed is fixed, the pre-aiming distance is reduced along with the increase of the path curvature k, and when the path curvature is fixed, the larger the vehicle speed is, the farther the pre-aiming distance is;
finally, the obtained feedforward steering angle and the feedback steering angle are added to obtain a final transverse path tracking steering angle delta which is:
δ＝δffw+δfb。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110563037.5A CN113386781A (en) | 2021-05-24 | 2021-05-24 | Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110563037.5A CN113386781A (en) | 2021-05-24 | 2021-05-24 | Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113386781A true CN113386781A (en) | 2021-09-14 |
Family
ID=77618757
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110563037.5A Pending CN113386781A (en) | 2021-05-24 | 2021-05-24 | Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113386781A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113408047A (en) * | 2021-08-05 | 2021-09-17 | 江苏大学 | Vehicle dynamics prediction model based on time-lag feedback neural network, training data acquisition method and training method |
| CN114137971A (en) * | 2021-11-25 | 2022-03-04 | 北京轻舟智航科技有限公司 | Off-line identification method for delay of steering system |
| CN114148318A (en) * | 2021-12-22 | 2022-03-08 | 吉林大学 | Vehicle path tracking method based on feedback linearization and LQR (Low-rank response) in ice and snow environment |
| CN114211926A (en) * | 2021-12-31 | 2022-03-22 | 重庆工业职业技术学院 | Automobile suspension control system for bumpy road surface |
| CN114265310A (en) * | 2021-11-26 | 2022-04-01 | 重庆大学 | Vehicle robust performance-guaranteeing coupling control method under unreliable data transmission environment |
| CN114347031A (en) * | 2022-01-18 | 2022-04-15 | 广东省智能机器人研究院 | Mobile mechanical arm collision avoidance control method and device considering time-varying time lag |
| CN114379583A (en) * | 2021-12-10 | 2022-04-22 | 江苏大学 | Automatic driving vehicle trajectory tracking system and method based on neural network dynamics model |
| CN114638051A (en) * | 2022-03-08 | 2022-06-17 | 浙江大学 | Intelligent automobile time lag stability analysis method based on system invariants |
| CN114895677A (en) * | 2022-05-17 | 2022-08-12 | 华南理工大学 | Vehicle trajectory tracking method, system, device and storage medium |
| CN115204417A (en) * | 2022-09-13 | 2022-10-18 | 鱼快创领智能科技(南京)有限公司 | Vehicle weight prediction method and system based on ensemble learning and storage medium |
| CN115346366A (en) * | 2022-07-22 | 2022-11-15 | 武汉理工大学 | Intelligent networking fleet control method and system considering road surface adhesion coefficient |
| CN116300410A (en) * | 2023-05-25 | 2023-06-23 | 武汉理工大学 | Corner optimization method and system for data-driven feedforward and feedback compensation |
| CN116588145A (en) * | 2023-07-18 | 2023-08-15 | 上海友道智途科技有限公司 | Semi-trailer robust tracking control method based on predictive LQR |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107298103A (en) * | 2017-07-03 | 2017-10-27 | 厦门大学 | A kind of automatic lane-change hierarchy system of intelligent electric automobile and method |
| CN107561942A (en) * | 2017-09-12 | 2018-01-09 | 重庆邮电大学 | Intelligent vehicle track following model predictive control method based on model compensation |
| CN108622104A (en) * | 2018-05-07 | 2018-10-09 | 湖北汽车工业学院 | A kind of Trajectory Tracking Control method for automatic driving vehicle |
| CN108646763A (en) * | 2018-07-18 | 2018-10-12 | 扬州大学 | A kind of autonomous driving trace tracking and controlling method |
| CN108674414A (en) * | 2018-07-02 | 2018-10-19 | 清华大学 | A kind of intelligent automobile Trajectory Tracking Control method of limiting condition |
| CN109795502A (en) * | 2018-09-27 | 2019-05-24 | 吉林大学 | Intelligent electric automobile path trace model predictive control method |
| CN109866752A (en) * | 2019-03-29 | 2019-06-11 | 合肥工业大学 | Double mode parallel vehicles track following driving system and method based on PREDICTIVE CONTROL |
| CN111806427A (en) * | 2020-06-08 | 2020-10-23 | 北京理工大学 | Comprehensive control method for four-hub motor driven vehicle |
| CN111890951A (en) * | 2020-08-07 | 2020-11-06 | 吉林大学 | Intelligent electric automobile trajectory tracking and motion control method |
| CN112109705A (en) * | 2020-09-23 | 2020-12-22 | 同济大学 | Collision avoidance optimization control system and method for extended-range distributed driving electric vehicle |
-
2021
- 2021-05-24 CN CN202110563037.5A patent/CN113386781A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107298103A (en) * | 2017-07-03 | 2017-10-27 | 厦门大学 | A kind of automatic lane-change hierarchy system of intelligent electric automobile and method |
| CN107561942A (en) * | 2017-09-12 | 2018-01-09 | 重庆邮电大学 | Intelligent vehicle track following model predictive control method based on model compensation |
| CN108622104A (en) * | 2018-05-07 | 2018-10-09 | 湖北汽车工业学院 | A kind of Trajectory Tracking Control method for automatic driving vehicle |
| CN108674414A (en) * | 2018-07-02 | 2018-10-19 | 清华大学 | A kind of intelligent automobile Trajectory Tracking Control method of limiting condition |
| CN108646763A (en) * | 2018-07-18 | 2018-10-12 | 扬州大学 | A kind of autonomous driving trace tracking and controlling method |
| CN109795502A (en) * | 2018-09-27 | 2019-05-24 | 吉林大学 | Intelligent electric automobile path trace model predictive control method |
| CN109866752A (en) * | 2019-03-29 | 2019-06-11 | 合肥工业大学 | Double mode parallel vehicles track following driving system and method based on PREDICTIVE CONTROL |
| CN111806427A (en) * | 2020-06-08 | 2020-10-23 | 北京理工大学 | Comprehensive control method for four-hub motor driven vehicle |
| CN111890951A (en) * | 2020-08-07 | 2020-11-06 | 吉林大学 | Intelligent electric automobile trajectory tracking and motion control method |
| CN112109705A (en) * | 2020-09-23 | 2020-12-22 | 同济大学 | Collision avoidance optimization control system and method for extended-range distributed driving electric vehicle |
Non-Patent Citations (4)
| Title |
|---|
| 徐兴;汤赵;王峰;陈龙;: "基于变权重系数的分布式驱动无人车轨迹跟踪", 中国公路学报, no. 12, 15 December 2019 (2019-12-15), pages 40 - 49 * |
| 柴瑞强;孙涛;田敏杰;: "智能车辆轨迹跟踪控制器设计", 软件导刊, no. 04, 15 April 2020 (2020-04-15), pages 175 - 179 * |
| 潘世举;马康;朱英杰;夏旭;: "基于稳定性的智能车辆综合控制方法", 军事交通学院学报, no. 07, 25 July 2020 (2020-07-25), pages 42 - 49 * |
| 陈龙;解云鹏;蔡英凤;孙晓强;滕成龙;邹凯;: "极限工况下无人驾驶车辆稳定跟踪控制", 汽车工程, no. 08, 25 August 2020 (2020-08-25), pages 23 - 38 * |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113408047A (en) * | 2021-08-05 | 2021-09-17 | 江苏大学 | Vehicle dynamics prediction model based on time-lag feedback neural network, training data acquisition method and training method |
| CN113408047B (en) * | 2021-08-05 | 2024-03-22 | 江苏大学 | Vehicle dynamics prediction model based on time-lag feedback neural network, training data acquisition method and training method |
| CN114137971A (en) * | 2021-11-25 | 2022-03-04 | 北京轻舟智航科技有限公司 | Off-line identification method for delay of steering system |
| CN114137971B (en) * | 2021-11-25 | 2023-06-09 | 北京轻舟智航科技有限公司 | Off-line identification method for steering system delay |
| CN114265310A (en) * | 2021-11-26 | 2022-04-01 | 重庆大学 | Vehicle robust performance-guaranteeing coupling control method under unreliable data transmission environment |
| CN114379583A (en) * | 2021-12-10 | 2022-04-22 | 江苏大学 | Automatic driving vehicle trajectory tracking system and method based on neural network dynamics model |
| CN114148318A (en) * | 2021-12-22 | 2022-03-08 | 吉林大学 | Vehicle path tracking method based on feedback linearization and LQR (Low-rank response) in ice and snow environment |
| CN114148318B (en) * | 2021-12-22 | 2023-10-27 | 吉林大学 | Vehicle path tracking method based on feedback linearization and LQR in ice and snow environment |
| CN114211926B (en) * | 2021-12-31 | 2022-08-05 | 重庆工业职业技术学院 | Automobile suspension control system for bumpy road surface |
| CN114211926A (en) * | 2021-12-31 | 2022-03-22 | 重庆工业职业技术学院 | Automobile suspension control system for bumpy road surface |
| CN114347031A (en) * | 2022-01-18 | 2022-04-15 | 广东省智能机器人研究院 | Mobile mechanical arm collision avoidance control method and device considering time-varying time lag |
| CN114347031B (en) * | 2022-01-18 | 2024-01-16 | 广东省智能机器人研究院 | Mobile mechanical arm collision prevention control method and device considering time-varying time delay |
| CN114638051A (en) * | 2022-03-08 | 2022-06-17 | 浙江大学 | Intelligent automobile time lag stability analysis method based on system invariants |
| CN114638051B (en) * | 2022-03-08 | 2024-02-27 | 浙江大学 | Intelligent automobile time-lag stability analysis method based on system invariants |
| CN114895677A (en) * | 2022-05-17 | 2022-08-12 | 华南理工大学 | Vehicle trajectory tracking method, system, device and storage medium |
| CN115346366B (en) * | 2022-07-22 | 2023-11-28 | 武汉理工大学 | Intelligent network coupled vehicle team control method and system considering road adhesion coefficient |
| CN115346366A (en) * | 2022-07-22 | 2022-11-15 | 武汉理工大学 | Intelligent networking fleet control method and system considering road surface adhesion coefficient |
| CN115204417A (en) * | 2022-09-13 | 2022-10-18 | 鱼快创领智能科技(南京)有限公司 | Vehicle weight prediction method and system based on ensemble learning and storage medium |
| CN116300410B (en) * | 2023-05-25 | 2023-08-22 | 武汉理工大学 | Corner optimization method and system for data-driven feedforward and feedback compensation |
| CN116300410A (en) * | 2023-05-25 | 2023-06-23 | 武汉理工大学 | Corner optimization method and system for data-driven feedforward and feedback compensation |
| CN116588145B (en) * | 2023-07-18 | 2023-10-27 | 上海友道智途科技有限公司 | Semi-trailer robust tracking control method based on predictive LQR |
| CN116588145A (en) * | 2023-07-18 | 2023-08-15 | 上海友道智途科技有限公司 | Semi-trailer robust tracking control method based on predictive LQR |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113386781A (en) | Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model | |
| Tang et al. | An improved kinematic model predictive control for high-speed path tracking of autonomous vehicles | |
| CN109466558B (en) | Road adhesion coefficient estimation method based on EKF (extended Kalman Filter) and BP (Back propagation) neural network | |
| Gao et al. | Robust lateral trajectory following control of unmanned vehicle based on model predictive control | |
| Kapania et al. | Path tracking of highly dynamic autonomous vehicle trajectories via iterative learning control | |
| Ding et al. | Event-triggered vehicle sideslip angle estimation based on low-cost sensors | |
| CN110827535B (en) | Nonlinear vehicle queue cooperative self-adaptive anti-interference longitudinal control method | |
| CN113408047B (en) | Vehicle dynamics prediction model based on time-lag feedback neural network, training data acquisition method and training method | |
| CN110861651B (en) | Method for estimating longitudinal and lateral motion states of front vehicle | |
| CN114379583A (en) | Automatic driving vehicle trajectory tracking system and method based on neural network dynamics model | |
| CN114510063B (en) | Unmanned tracked vehicle and track tracking control method and system thereof | |
| CN115406446A (en) | Multi-axis special vehicle state estimation method based on neural network and unscented Kalman filtering | |
| CN114942642A (en) | Unmanned automobile track planning method | |
| CN105487382A (en) | Micro gyroscope self-adaptive fuzzy sliding mode control method based on dynamic surface | |
| CN114684199A (en) | Vehicle dynamics series hybrid model driven by mechanism analysis and data, intelligent automobile trajectory tracking control method and controller | |
| Song et al. | Reliable estimation of automotive states based on optimized neural networks and moving horizon estimator | |
| Thommyppillai et al. | Advances in the development of a virtual car driver | |
| Kumarawadu et al. | Neuroadaptive output tracking of fully autonomous road vehicles with an observer | |
| Kanarachos | Design of an intelligent feed forward controller system for vehicle obstacle avoidance using neural networks | |
| Yu et al. | Nonlinear model predictive planning and control for high-speed autonomous vehicles on 3D terrains | |
| Huang et al. | Estimation of sideslip angle based on extended Kalman filter | |
| Song et al. | Estimation of Vehicle Sideslip Angle based on Modified Sliding Mode Observer and Recurrent Neural Network | |
| CN115817509A (en) | Multi-axis distributed driving vehicle steering auxiliary track tracking method based on AMPC | |
| CN113060143B (en) | System and method for determining road adhesion coefficient | |
| CN114179818A (en) | Intelligent automobile transverse control method based on adaptive preview time and sliding mode control |
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 |