CN117068159B - Adaptive cruise system based on disturbance rejection control - Google Patents

Adaptive cruise system based on disturbance rejection control Download PDF

Info

Publication number
CN117068159B
CN117068159B CN202311102057.8A CN202311102057A CN117068159B CN 117068159 B CN117068159 B CN 117068159B CN 202311102057 A CN202311102057 A CN 202311102057A CN 117068159 B CN117068159 B CN 117068159B
Authority
CN
China
Prior art keywords
acceleration
vehicle
controlled
sequence set
signal value
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
Application number
CN202311102057.8A
Other languages
Chinese (zh)
Other versions
CN117068159A (en
Inventor
许恩永
郑毅
朱纪洪
何水龙
林长波
李慧
朱斌
展新
冯海波
袁夏明
王善超
陈志刚
冯高山
许家毅
邓聚才
李超
鲍家定
郑伟光
胡超凡
陶林
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
Guilin University of Electronic Technology
Dongfeng Liuzhou Motor Co Ltd
Original Assignee
Tsinghua University
Guilin University of Electronic Technology
Dongfeng Liuzhou Motor Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tsinghua University, Guilin University of Electronic Technology, Dongfeng Liuzhou Motor Co Ltd filed Critical Tsinghua University
Priority to CN202311102057.8A priority Critical patent/CN117068159B/en
Publication of CN117068159A publication Critical patent/CN117068159A/en
Application granted granted Critical
Publication of CN117068159B publication Critical patent/CN117068159B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • 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
    • B60W30/14—Adaptive cruise control
    • B60W30/16—Control of distance between vehicles, e.g. keeping a distance to preceding vehicle
    • B60W30/165—Automatically following the path of a preceding lead vehicle, e.g. "electronic tow-bar"
    • 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
    • B60W30/14—Adaptive cruise control
    • B60W30/143—Speed control
    • 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
    • B60W40/10—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 vehicle motion
    • B60W40/105—Speed
    • 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
    • B60W40/10—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 vehicle motion
    • B60W40/107—Longitudinal acceleration
    • 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
    • B60W2520/00—Input parameters relating to overall vehicle dynamics
    • B60W2520/10—Longitudinal speed
    • 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
    • B60W2520/00—Input parameters relating to overall vehicle dynamics
    • B60W2520/10—Longitudinal speed
    • B60W2520/105—Longitudinal acceleration
    • 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
    • B60W2554/00—Input parameters relating to objects
    • B60W2554/40—Dynamic objects, e.g. animals, windblown objects
    • B60W2554/404—Characteristics
    • B60W2554/4042—Longitudinal speed

Landscapes

  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses an adaptive cruise system based on disturbance rejection control, wherein in the system, an environment sensing layer acquires first acceleration of a front vehicle every other preset time period, and an upper-layer controller predicts the first acceleration of the front vehicle in the next preset time period through a front vehicle acceleration predictor and acquires first expected acceleration; the feedforward control module is used for calculating and generating feedforward control quantity of the vehicle through the feedforward control module of the lower controller; and calculating and generating a feedback control quantity of the vehicle to be controlled through a feedback control module of the lower controller, and finally calculating and generating a second expected acceleration of the vehicle to be controlled according to the feedforward control quantity and the feedback control quantity so as to control the acceleration of the vehicle to be controlled through a vehicle executing mechanism. By implementing the method and the device, the acceleration of the automobile to be controlled can be controlled more quickly and accurately in the following mode, so that the following distance is better kept, and the safety is improved.

Description

Adaptive cruise system based on disturbance rejection control
Technical Field
The invention relates to the field of automobile auxiliary driving, in particular to an adaptive cruise system based on disturbance rejection control.
Background
With the development of modern automobiles, the current mass-production automobile types are also provided with auxiliary driving systems of different levels, so that the robot can replace a person to drive to a certain extent. At present, a commonly used cruise control system can enable a vehicle to enter a following mode, enable the vehicle to keep a safe distance from a front vehicle and run along with the front vehicle, and always defaults that the front vehicle runs stably in the following mode, namely the acceleration of the front vehicle is unchanged, but in actual work, the calculated following acceleration is often inaccurate due to disturbance caused by the change of the acceleration of the front vehicle, so that the deviation between the actual following distance and the safe distance is larger, and a certain potential safety hazard exists when the vehicle is in the following mode.
Disclosure of Invention
The invention provides an adaptive cruise system based on disturbance rejection control, which can control acceleration of a car to be controlled in a car following process faster and more accurately so as to better maintain a car following distance and improve safety.
The invention provides an adaptive cruise system based on disturbance rejection control, comprising: the system comprises an environment sensing layer, an upper controller, a lower controller and a vehicle executing mechanism;
The environment sensing layer is used for acquiring first acceleration of a front vehicle at intervals of a preset time through a vehicle radar and sending the acquired first acceleration to the upper controller;
The upper controller is configured to input the first accelerations into a preset front vehicle acceleration predictor, so that the front vehicle acceleration predictor predicts a first acceleration of a front vehicle in a next preset period according to each first acceleration, obtain a first expected acceleration based on the predicted first acceleration, and send the first expected acceleration to the lower controller;
The lower controller includes: a feedforward control module and a feedback control module;
the feedforward control module is used for acquiring the running speed of the vehicle to be controlled through the vehicle executing mechanism, and calculating and generating feedforward control quantity of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first expected acceleration;
The feedback control module is used for acquiring a second acceleration of the vehicle to be controlled through the vehicle executing mechanism, and calculating and generating a feedback control quantity of the vehicle to be controlled according to the second acceleration and the first expected acceleration of the vehicle to be controlled; calculating and generating a second expected acceleration of the vehicle to be controlled according to the feedforward control quantity and the feedback control quantity of the vehicle to be controlled;
the vehicle executing mechanism is used for controlling the acceleration of the vehicle to be controlled according to the second expected acceleration so as to enable the acceleration of the vehicle to be controlled to be consistent with the second expected acceleration.
Further, the front vehicle acceleration predictor predicts, according to each of the first accelerations, a first acceleration of the front vehicle within a preset period of time, including:
Inputting each first acceleration into a preset sequence set through the front vehicle acceleration predictor, generating a first sequence set, and establishing a first accumulation sequence set corresponding to the first sequence set according to the first sequence set;
according to the first accumulation sequence set, constructing an adjacent mean value generation sequence corresponding to the first accumulation sequence set, according to the adjacent mean value generation sequence, establishing a gray Verhulst model, and according to the gray Verhulst model, constructing a whitening equation corresponding to the gray Verhulst model;
And solving the whitening equation to generate a second accumulation sequence set for predicting the acceleration of the front vehicle, and predicting the first acceleration of the front vehicle in a next preset period according to the second accumulation sequence set.
Further, the first set of sequences is represented by the following formula:
In the method, in the process of the invention, Representing a first set of sequences,/>And/>Respectively representing the 1 st to 5 th first accelerations in the first series set;
The first accumulated sequence set corresponding to the first sequence set is represented by the following formula:
In the method, in the process of the invention, Representing a first accumulated sequence set,/>Representing a kth first accumulated acceleration in the first accumulated sequence set, k representing an ordinal number in the sequence set;
The immediately adjacent mean generation sequence is represented by the following formula:
In the method, in the process of the invention, Representing a proximate mean generation sequence corresponding to the first accumulated sequence set,/>Representing a kth immediately adjacent mean in the immediately adjacent mean generation sequence;
The gray Verhulst model is represented by the following formula:
Where α and β are model parameters.
Further, the solving the whitening equation generates a second accumulation sequence set for predicting acceleration of the front vehicle, and predicts the first acceleration of the front vehicle in a preset period according to the second accumulation sequence set, including:
Constructing an equation solving formula and a particle algorithm fitness function of the whitening equation according to the whitening equation, and calculating and generating a second accumulation sequence set for predicting the acceleration of the vehicle in front according to the equation solving formula and the particle algorithm fitness function of the whitening equation; calculating and generating a first acceleration corresponding to the next preset period of the front vehicle according to the second accumulation sequence set;
wherein the equation solution of the whitening equation is represented by:
In the method, in the process of the invention, Representing a (k+1) th second accumulated acceleration in the second accumulated sequence set; e represents a natural constant;
the particle swarm algorithm fitness function is represented by the following formula:
where f fitness denotes a fitness function of the particle swarm algorithm, Representing an ith first accumulated acceleration in the first accumulated sequence set; /(I)Representing a second set of accumulated sequences,/>Representing an ith second accumulated acceleration in the second accumulated sequence set; n represents the number of second accumulated accelerations in the second accumulated sequence set;
Generating a first acceleration corresponding to the following preset period of time of the front vehicle through the following formula calculation:
In the method, in the process of the invention, Representing the predicted first acceleration,/>Representing a second sequence set corresponding to the second accumulated sequence set,/>Indicating the k+1st acceleration in the second set of sequences.
Further, the feedback control module includes: the system comprises a tracking differential module, an expansion state observation module and an error feedback module;
the tracking differential module is used for calculating and generating a first expected acceleration signal value and a first expected acceleration differential signal value of the vehicle to be controlled according to the first expected acceleration;
The extended state observation module is used for acquiring a second acceleration of the vehicle to be controlled through the vehicle execution mechanism, acquiring a historical feedback control quantity of the vehicle to be controlled through the error feedback module, and calculating and generating a second acceleration signal value, a second acceleration differential signal value and a disturbance signal value according to the second acceleration and the historical feedback control quantity of the vehicle to be controlled;
The error feedback module is used for comparing the first expected acceleration signal value with the second acceleration signal value to obtain an acceleration signal difference value; comparing the first expected acceleration differential signal value with the second acceleration differential signal value to obtain an acceleration differential signal difference value; calculating and generating feedback control quantity of the vehicle to be controlled according to the acceleration signal difference value, the acceleration differential signal difference value and the disturbance signal value;
The error feedback module is further configured to obtain a feedforward control amount of the vehicle to be controlled through the feedforward control module, and sum the feedforward control amount and the feedback control amount of the vehicle to be controlled to generate a second desired acceleration of the vehicle to be controlled.
Further, the feedforward control module is further configured to, before calculating the feedforward control amount of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first desired acceleration:
Judging whether the value of the first expected acceleration is positive or negative;
The calculating the feedforward control quantity of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first expected acceleration comprises the following steps:
when it is determined that the first desired acceleration is a positive number, a feedforward control amount of the vehicle to be controlled is calculated by the following formula:
wherein a xdes represents a first expected acceleration, ζ is a rotational mass conversion coefficient, i g is a transmission gear ratio, i 0 is a main reduction ratio, η T represents mechanical efficiency when a transmission system is transmitted, r is a wheel rolling radius, m is a whole vehicle equipment mass of a vehicle to be controlled, g is a gravity coefficient, f is a wheel corresponding rolling resistance coefficient, C D is an air resistance coefficient along a vehicle running direction, A is an automobile windward area, ρ is an air density, and v x is a running speed of the vehicle to be controlled; t VDM denotes a feedforward control amount acting on the vehicle drive;
When it is determined that the first desired acceleration is negative, a feedforward control amount of the vehicle to be controlled is calculated by the following formula:
Where K b is a proportional coefficient of the vehicle braking force and the tire braking pressure, and F VDB is a feedforward control amount acting on the vehicle braking.
Further, a first desired acceleration signal value and a first desired acceleration differential signal value of the vehicle to be controlled are computationally generated by the following formula:
Where x 1 is the first desired acceleration signal value, x 2 is the first desired acceleration differential signal value, r is the tracking speed factor, h is the filter factor, fhan (x 1,x2, r, h) is the fastest control function.
Further, a second acceleration signal value, a second acceleration differential signal value and a disturbance signal value are generated through calculation according to the following formula;
Wherein z 1 represents a second acceleration signal value, z 2 represents a second acceleration differential signal value, z 3 represents a disturbance signal value, λ 01 and λ 02 are weight coefficients, u 1 (k) is a historical feedback control quantity of the vehicle to be controlled, and fal (e, α, δ) is a commonly used nonlinear function.
Further, the generating the feedback control quantity of the vehicle to be controlled according to the acceleration signal difference value, the acceleration differential signal difference value and the disturbance signal value includes:
Generating a preliminary feedback control quantity according to the acceleration signal difference value and the acceleration differential signal difference value, and generating the feedback control quantity of the vehicle to be controlled according to the preliminary feedback control quantity and the disturbance signal value.
Further, the acceleration signal difference value, the acceleration differential signal difference value, the preliminary feedback control amount and the feedback control amount are calculated and generated by the following formulas:
Where e 1 denotes an acceleration signal difference, e 2 denotes an acceleration differential signal difference, λ 01 and λ 02 are weight coefficients, u 0 denotes a preliminary feedback control amount, and u 2 (k) denotes a feedback control amount of the vehicle to be controlled.
The embodiment of the invention has the following beneficial effects:
The invention provides an adaptive cruise system based on disturbance rejection control; according to the system, the first acceleration of the front vehicle is detected through the environment sensing layer, then the upper controller predicts the acceleration of the front vehicle according to the detected first acceleration to obtain the first acceleration of the front vehicle in a preset period, the first expected acceleration is obtained based on the first acceleration, the vehicle to be controlled can better cope with disturbance caused by the acceleration change of the front vehicle according to the first expected acceleration, then the feedforward control quantity is calculated through feedforward control in the lower controller, the calculated feedforward control quantity can carry out certain compensation control on the acceleration of the vehicle to be controlled, in addition, a feedback control module in the lower controller can monitor the error of the vehicle during control in real time according to the second acceleration and the first expected acceleration of the vehicle to be controlled, calculate to generate the feedback control quantity, and finally the second expected acceleration of the vehicle to be controlled is obtained according to the calculated feedforward control and feedback control, so that the vehicle to be controlled can be controlled more quickly and accurately according to the second expected acceleration, the vehicle to be controlled in a following mode, the acceleration of the vehicle to be controlled can be more accurately recognized, and the vehicle to be controlled accurately, and the following distance can be controlled, and the vehicle to be more safely and more controlled.
Drawings
FIG. 1 is a schematic diagram of an adaptive cruise system based on immunity control according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of an RBF neural network according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a working flow of a front vehicle acceleration predictor according to an embodiment of the present invention;
Fig. 4 is a schematic structural diagram of an immunity control module of an underlying controller according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made more apparent and fully hereinafter with reference to the accompanying drawings, in which some, but not all embodiments of the invention are shown. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, an adaptive cruise system based on immunity control according to an embodiment includes: the system comprises an environment sensing layer, an upper controller, a lower controller and a vehicle executing mechanism;
The environment sensing layer is used for acquiring first acceleration of a front vehicle at intervals of a preset time through a vehicle radar and sending the acquired first acceleration to the upper controller;
The upper controller is configured to input the first accelerations into a preset front vehicle acceleration predictor, so that the front vehicle acceleration predictor predicts a first acceleration of a front vehicle in a next preset period according to each first acceleration, obtain a first expected acceleration based on the predicted first acceleration, and send the first expected acceleration to the lower controller;
The lower controller includes: a feedforward control module and a feedback control module;
the feedforward control module is used for acquiring the running speed of the vehicle to be controlled through the vehicle executing mechanism, and calculating and generating feedforward control quantity of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first expected acceleration;
The feedback control module is used for acquiring a second acceleration of the vehicle to be controlled through the vehicle executing mechanism, and calculating and generating a feedback control quantity of the vehicle to be controlled according to the second acceleration and the first expected acceleration of the vehicle to be controlled; calculating and generating a second expected acceleration of the vehicle to be controlled according to the feedforward control quantity and the feedback control quantity of the vehicle to be controlled;
the vehicle executing mechanism is used for controlling the acceleration of the vehicle to be controlled according to the second expected acceleration so as to enable the acceleration of the vehicle to be controlled to be consistent with the second expected acceleration.
Specifically, the environment sensing layer collects first acceleration of a front vehicle every other preset time, then sends the collected first acceleration to an upper-layer controller, then predicts the first acceleration of the front vehicle in a next preset time period through a preset front vehicle acceleration predictor in the upper-layer controller, inputs the predicted first acceleration into a following mode model prediction controller as more accurate information, and can obtain accurate first expected acceleration to be used as a control basis of the vehicle to be controlled through the first expected acceleration; and then a feedforward module in the lower-layer controller calculates feedforward control of the vehicle according to the first expected acceleration, acquires second acceleration of the vehicle to be controlled from the vehicle executing mechanism through a feedback control module, calculates and generates feedback control quantity of the vehicle to be controlled according to the second acceleration and the first expected acceleration of the vehicle to be controlled, and calculates and generates second expected acceleration of the vehicle to be controlled according to the feedforward control quantity and the feedback control quantity of the vehicle to be controlled. And finally, controlling the acceleration of the vehicle to be controlled according to a second expected acceleration based on the vehicle executing mechanism so as to enable the acceleration of the vehicle to be controlled to be consistent with the second expected acceleration. The following mode model predictive controller belongs to a model predictive controller commonly used in the existing following system.
Including but not limited to: the method comprises the steps that a millimeter wave radar and a visual camera are used, an environment sensing layer senses whether other vehicles exist in front of a lane in a sensing range, and if so, the front vehicle speed and first acceleration of the front vehicle are obtained; the acceleration prediction model in the upper controller is mounted in a front vehicle acceleration predictor, and predicts the first acceleration of the front vehicle in a preset period through the acceleration prediction model to obtain the first expected acceleration, and the generated first expected acceleration can be directly used as a control basis of the vehicle to be controlled.
In an alternative embodiment, the invention further compensates for the required acceleration by the lower level controller based on the first desired acceleration, on the one hand by providing as much as possible known vehicle parameters and road information by feed-forward control, and on the other hand by observing disturbances caused by changing driving road conditions by feedback control to compensate for the control strategy. The upper controller works in place of the driver to determine how to operate the vehicle.
An adaptive cruise control system (Adaptive cruise control, ACC), abbreviated ACC system, includes ACC system operating mode switching logic and an ACC system control strategy calculation module. The ACC system control strategy calculation module calculates a desired acceleration signal a xdes of the vehicle motion. The lower controller comprises an anti-interference control module and a vehicle longitudinal inverse dynamics module. The disturbance rejection control module compensates the control quantity according to disturbance on the basis of the control strategy of the upper-layer controller to calculate the actual expected acceleration a rxdes. And calculating the expected throttle opening alpha des or the expected brake pressure F des by using an RBF neural network vehicle inverse longitudinal dynamics model based on EKF online learning, and sending the calculated throttle opening alpha des or the expected brake pressure F des to a vehicle executing mechanism. When the ACC system is in a following mode and a model predictive control algorithm is adopted as an upper controller to calculate a control strategy, a front vehicle acceleration predictor is added under the framework of the model predictive control algorithm, so that the upper controller calculates a better control strategy, namely the expected acceleration a xdes. The disturbance rejection module is added in the lower-layer controller, and the control strategy is compensated on the basis of a xdes to obtain the actual expected acceleration a rxdes. Finally, the vehicle inverse longitudinal dynamics model precisely calculates the throttle opening alpha des or the pressure F des required for tracking the strategy a rxdes according to the expected acceleration a xdes, and inputs the throttle opening alpha des or the pressure F des into an executing mechanism for executing.
It should be noted that the lower layer controller includes an anti-interference control module and a vehicle longitudinal inverse dynamics module, and the feedforward control module and the feedback control module in the invention both belong to the anti-interference control module.
In an alternative embodiment, the RBF neural network architecture is as shown in fig. 2, with inputs x= (x 1,x2) being the current speed and desired acceleration of the vehicle respectively,For output, the throttle opening α des is indicated when the output value is positive, and the brake pressure F des is indicated when the output value is negative. The RBF neural network has only one hidden layer and comprises h neurons and h weights. The output of the network is as follows:
Wherein Φ n (x) is a gaussian function, defined as follows:
Where μ n is the center of the nth hidden layer neuron, σ n is the width of the gaussian function, and iil represents the euclidean norm. The training data (x, y) is obtained by calibrating the vehicle running on a certain road, so that the RBF network is trained to complete the initialization of the network. The RBF network has an on-line learning strategy based on EKF, so that the RBF network has higher tolerance to errors in the calibration process. According to the online learning strategy, new data generated in the running process of the vehicle can be used for carrying out accuracy test on the RBF network in real time, if the accuracy cannot meet the requirement, the number or weight of network nodes can be adjusted, so that the lower control can accurately convert the expected acceleration a des into the expected throttle opening alpha des or the expected braking pressure F des.
In an alternative embodiment, the EKF-based online learning step of the RBF neural network includes: step one, checking whether the requirement of adding nodes in a hidden layer is met:
di＝‖xi-μir‖＞E3
Where μ ir is the center of the neuron in the hidden layer closest to the current input x i. E 1,E2 and E 3 are suitable thresholds. Equation (14) checks whether the output generated by the existing nodes of the current network is within the accuracy requirement. Equation (15) checks whether the square root of the errors of the past M outputs of the network are within the accuracy requirement. Equation (16) ensures that the new node that is added is far from the already existing node. When these conditions are met, then a new neuron is created in the hidden layer in step two, otherwise the network parameters are updated with the EKF in step three.
Step two, adding a new node in a hidden layer of the RBF neural network: after the three conditions in the first step are met, a new hidden layer neuron is added, and the new neuron has the following related parameters:
αh+1＝ei,μh+1＝xi,σh+1＝κ‖xi-μi‖
Setting these parameters by new neurons can eliminate errors, κ as a superposition factor. After adding the neuron, jump to step four.
Updating RBF neural network parameters by using EKF: if the requirement of adding the node in the hidden layer in the first step is not met, the relevant parameter omega * of the node with the center closest to the input x i is updated,The update method using EKF is as follows:
for the Kalman filter gain matrix, the calculation method is as follows:
wherein, As a function f (x i) with respect to the parameter vector w, at/>And (5) obtaining a gradient matrix:
R i is the variance of the measured noise and P i * is the error covariance matrix. The update manner of P i * is as follows:
q is a scalar that determines the random step size allowed in the gradient vector direction. If the number of parameters to be adjusted is z, then P i is a zxz positive definite symmetric matrix. When a new hidden neuron is assigned, the dimension of P i increases:
Where the rows and columns are initialized by p 0, p 0 is an estimate of the uncertainty when the initial value is assigned to the parameter. The dimension z 1 of the identity matrix is equal to the number of new parameters introduced by the new hidden neurons. And then turning to the step four to trim the neural network.
Step four, hidden layer neuron pruning strategy: the last step will build those neurons that contribute less to the network output. Let the matrix o= (O 1,...,oh) represent the output of the hidden layer and a represent the weight matrix a= (α 1,...,αh). Output o n of neurons of the nth hidden layer:
If either α n or σ n in equation (23) is small, then o n will also be small, meaning that the input is far enough from the center of this hidden neuron. To reduce the inconsistency due to the use of the absolute value of the output, the value is normalized to the value of the highest output:
The normalized output r n of each neuron is then observed for N consecutive inputs, and if the output of a certain neuron is below a threshold, that neuron is trimmed off.
The non-linearity of the vehicle can be compensated by training and maintaining an RBF neural network as a vehicle inverse longitudinal dynamics model by using the vehicle driving data. When the vehicle parameters change during running, the vehicle longitudinal dynamics model based on the RBF neural network can also generate accurate expected throttle opening alpha des and expected brake pressure F des according to the expected acceleration a des by learning data generated during running, so that an execution layer can accurately track the expected acceleration.
In a preferred embodiment, the front vehicle acceleration predictor predicts a first acceleration of the front vehicle for a next preset period of time based on each of the first accelerations, including:
In the front vehicle acceleration predictor, firstly, inputting each first acceleration into a preset sequence set to generate a first sequence set, and establishing a first accumulation sequence set corresponding to the first sequence set according to the first sequence set;
According to the first accumulation sequence set, constructing an adjacent mean value generation sequence corresponding to the first accumulation sequence set, according to the adjacent mean value generation sequence, establishing a gray Verhulst model, and according to the gray Verhulst model, constructing a whitening equation corresponding to the gray Verhulst model;
And solving the whitening equation to generate a second accumulation sequence set for predicting the acceleration of the front vehicle, and predicting the first acceleration of the front vehicle in a next preset period according to the second accumulation sequence set.
Specifically, after the first expected acceleration is obtained, the first expected acceleration is added into a preset sequence set, and when the sequence is full of each first expected acceleration, a first sequence set is generated; then, according to the first sequence set, a first accumulation sequence set corresponding to the first sequence set is established, according to the first accumulation sequence set, an immediately adjacent mean value generation sequence corresponding to the first accumulation sequence set is established, and then a corresponding gray Verhulst model and a corresponding whitening equation thereof are established according to the immediately adjacent mean value generation sequence. And solving the whitening equation to generate a second accumulation sequence set for predicting the acceleration of the front vehicle, and acquiring the first acceleration corresponding to the front vehicle in a preset period according to the second accumulation sequence set.
In a preferred embodiment, the first set of sequences is represented by the formula:
In the method, in the process of the invention, Representing a first set of sequences,/>And/>Respectively representing the 1 st to 5 th first accelerations in the first series set;
The first accumulated sequence set corresponding to the first sequence set is represented by the following formula:
In the method, in the process of the invention, Representing a first accumulated sequence set,/>Representing a kth first accumulated acceleration in the first accumulated sequence set, k representing an ordinal number in the sequence set;
The immediately adjacent mean generation sequence is represented by the following formula:
In the method, in the process of the invention, Representing a proximate mean generation sequence corresponding to the first accumulated sequence set,/>Representing a kth immediately adjacent mean in the immediately adjacent mean generation sequence;
The gray Verhulst model is represented by the following formula:
Where α and β are model parameters.
In a preferred embodiment, said solving the whitening equation to generate a second set of accumulated sequences for predicting acceleration of the preceding vehicle, predicting the first acceleration of the preceding vehicle for a next predetermined period of time based on the second set of accumulated sequences, comprises:
Constructing an equation solving formula and a particle algorithm fitness function of the whitening equation according to the whitening equation, and calculating and generating a second accumulation sequence set for predicting the acceleration of the vehicle in front according to the equation solving formula and the particle algorithm fitness function of the whitening equation; calculating and generating a first acceleration of the front vehicle in a next preset period according to the second accumulation sequence set;
wherein the equation solution of the whitening equation is represented by:
In the method, in the process of the invention, Representing a (k+1) th second accumulated acceleration in the second accumulated sequence set; e represents a natural constant;
the particle swarm algorithm fitness function is represented by the following formula:
where f fitness denotes a fitness function of the particle swarm algorithm, Representing an ith first accumulated acceleration in the first accumulated sequence set; /(I)Representing a second set of accumulated sequences,/>Representing an ith second accumulated acceleration in the second accumulated sequence set; n represents the number of second accumulated accelerations in the second accumulated sequence set;
generating a first acceleration of the preceding vehicle for a next preset period of time by the following equation calculation:
In the method, in the process of the invention, Representing the predicted first acceleration,/>Representing a second sequence set corresponding to the second accumulated sequence set,/>Indicating the k+1st acceleration in the second set of sequences.
Specifically, an equation solving formula of the whitening equation is constructed:
In the method, in the process of the invention, Representing a (k+1) th second accumulated acceleration in the second accumulated sequence set; e represents a natural constant;
and constructing a particle algorithm fitness function:
where f fitness denotes a fitness function of the particle swarm algorithm, Representing an ith first accumulated acceleration in the first accumulated sequence set; /(I)Representing a second set of accumulated sequences,/>Representing an ith second accumulated acceleration in the second accumulated sequence set; n represents the number of second accumulated accelerations in the second accumulated sequence set;
and then, calculating and generating a second accumulation sequence set for predicting the acceleration of the front vehicle according to an equation solving formula of the whitening equation and a particle algorithm fitness function, and calculating and generating the first acceleration corresponding to the next preset period of the front vehicle through the following formula based on the second accumulation sequence set:
In the method, in the process of the invention, Representing a first desired acceleration,/>Representing a second sequence set corresponding to the second accumulated sequence set,Indicating the k+1st acceleration in the second set of sequences.
Specifically, in an alternative embodiment, as shown in fig. 3, the front vehicle acceleration predictor based on the gray Verhulst model of the particle swarm optimization algorithm includes the following steps: step one, starting the system, enabling the driver to start the ACC system, and setting the cruising speed v set. The environment sensing layer searches whether a vehicle exists in front of the lane, if the vehicle exists in the searching range, the ACC system enters a following model, and the front vehicle acceleration predictor is started. Step two, acquiring historical data of acceleration of the front vehicle, setting a window n=5 for storing the acceleration of the front vehicle, and initializingThe environmental perception layer samples and updates/>, the acceleration of the front vehicle at time intervals t p =0.1 sStep three, calculating one accumulation generation sequence (1-AGO)/> of front vehicle acceleration historical data Step four, calculating one accumulation generation sequence/>Generated sequence of immediately adjacent means/> And fifthly, establishing a gray Verhulst model and a model solution. The gray Verhulst model is: /(I)The whitening equation for the gray Verhulst model is: /(I)The solution of the whitening equation is used to predict one-time accumulation generated sequence/> Step six, constructing a particle swarm algorithm fitness function f fitness: /(I)Wherein/>A sequence is generated for one accumulation of the front truck acceleration raw data. σ is the weight matrix, and when n=5, σ= (1, 2,3,4, 5). Step seven, solving parameters alpha and beta according to the fitness function in the step six by utilizing a particle swarm algorithm, and carrying into a formula (3) so as to predict a primary accumulation sequence/>, of the acceleration of the front vehicleStep eight, according to/>The front vehicle acceleration is predicted according to the following rule: /(I)The front vehicle predictor is combined with a model control algorithm, and the front vehicle acceleration is predicted by using the values of alpha and beta obtained in the step seven and the prediction equation (5) in the step eight. When the environment sensing layer updates the acceleration history data sequence/>And when the method is used, the values of alpha and beta are redetermined by repeating the steps one to seven so as to ensure the accuracy of the next prediction.
In another alternative embodiment, an instance of the operation of the lead car predictor under the model predictive control framework is provided: assuming that the controller sampling time T s =0.05 s, the sampling time T c =0.25 s of the preceding vehicle acceleration data, the preceding vehicle acceleration prediction period T p in the prediction time domain is the same as the sampling time T c of the preceding vehicle acceleration, and assuming that the prediction time domain n=10 of the model predictive control algorithm. The prediction time domain predicts the future 0.5s of the acceleration change of the front vehicle, and the front vehicle acceleration predictor predicts the future two steps of acceleration in the prediction time domain according to the prediction period. The prediction mode of the front vehicle acceleration in one prediction time domain is as follows:
In a preferred embodiment, the feedback control module comprises: the system comprises a tracking differential module, an expansion state observation module and an error feedback module;
the tracking differential module is used for calculating and generating a first expected acceleration signal value and a first expected acceleration differential signal value of the vehicle to be controlled according to the first expected acceleration;
The extended state observation module is used for acquiring a second acceleration of the vehicle to be controlled through the vehicle execution mechanism, acquiring a historical feedback control quantity of the vehicle to be controlled through the error feedback module, and calculating and generating a second acceleration signal value, a second acceleration differential signal value and a disturbance signal value according to the second acceleration and the historical feedback control quantity of the vehicle to be controlled;
The error feedback module is used for comparing the first expected acceleration signal value with the second acceleration signal value to obtain an acceleration signal difference value; comparing the first expected acceleration differential signal value with the second acceleration differential signal value to obtain an acceleration differential signal difference value; calculating and generating feedback control quantity of the vehicle to be controlled according to the acceleration signal difference value, the acceleration differential signal difference value and the disturbance signal value;
The error feedback module is further configured to obtain a feedforward control amount of the vehicle to be controlled through the feedforward control module, and sum the feedforward control amount and the feedback control amount of the vehicle to be controlled to generate a second desired acceleration of the vehicle to be controlled.
Specifically, as shown in fig. 4, on one side, the tracking differential module calculates and generates a first expected acceleration signal value and a first expected acceleration differential signal value of the vehicle to be controlled according to the first expected acceleration; and the other side is provided with an extended state observation module, a second acceleration of the vehicle to be controlled is obtained through the vehicle executing mechanism, a historical feedback control quantity of the vehicle to be controlled is obtained through the error feedback module, and then a second acceleration signal value, a second acceleration differential signal value and a disturbance signal value are calculated and generated according to the obtained second acceleration and the historical feedback control quantity.
Then, comparing the first expected acceleration signal value with the second acceleration signal value through an error feedback module to obtain an acceleration signal difference value; comparing the first expected acceleration differential signal value with the second acceleration differential signal value to obtain an acceleration differential signal difference value; and calculating and generating a feedback control quantity of the vehicle to be controlled according to the acceleration signal difference value, the acceleration differential signal difference value and the disturbance signal value. And finally, carrying out summation calculation on the feedforward control quantity and the feedback control quantity of the vehicle to be controlled, and generating a second expected acceleration of the vehicle to be controlled.
In a preferred embodiment, the feedforward control module is further configured to, before calculating the feedforward control amount of the vehicle to be controlled based on the traveling speed of the vehicle to be controlled and the first desired acceleration:
Judging whether the value of the first expected acceleration is positive or negative;
The calculating the feedforward control quantity of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first expected acceleration comprises the following steps:
when it is determined that the first desired acceleration is a positive number, a feedforward control amount of the vehicle to be controlled is calculated by the following formula:
wherein a xdes represents a first expected acceleration, ζ is a rotational mass conversion coefficient, i g is a transmission gear ratio, i 0 is a main reduction ratio, η T represents mechanical efficiency when a transmission system is transmitted, r is a wheel rolling radius, m is a whole vehicle equipment mass of a vehicle to be controlled, g is a gravity coefficient, f is a wheel corresponding rolling resistance coefficient, C D is an air resistance coefficient along a vehicle running direction, A is an automobile windward area, ρ is an air density, and v x is a running speed of the vehicle to be controlled; t VDM denotes a feedforward control amount acting on the vehicle drive;
When it is determined that the first desired acceleration is negative, a feedforward control amount of the vehicle to be controlled is calculated by the following formula:
/>
Where K b is a proportional coefficient of the vehicle braking force and the tire braking pressure, and F VDB is a feedforward control amount acting on the vehicle braking.
Specifically, after the feedforward control module acquires the first desired acceleration, it first determines whether the first desired acceleration is a positive number or a negative number, and when it is determined that the first desired acceleration is a positive number, calculates a feedforward control amount for driving the vehicle by the following formula: Upon determining that the first desired acceleration is negative, the first desired acceleration is determined by the formula/> A feedforward control amount for vehicle braking is calculated.
In a preferred embodiment, the generation of the first desired acceleration signal value and the first desired acceleration derivative signal value of the vehicle to be controlled is calculated by the following formula:
Wherein x 1 is a first desired acceleration signal value, T s is a sampling time, x 2 is a first desired acceleration differential signal value, r is a tracking speed factor, h is a filtering factor, fhan (x 1,x2, r, h) is a fastest control function, and sign is a sign function. It should be noted that, all other parameters in fhan () are intermediate values calculated according to the introduced parameters, and have no name or physical meaning.
In a preferred embodiment, the second acceleration signal value, the second acceleration differential signal value and the disturbance signal value are generated by the following formula calculation;
Wherein z 1 represents a second acceleration signal value, z 2 represents a second acceleration differential signal value, z 3 represents a disturbance signal value, λ 01 and λ 02 are weight coefficients, u 1 (k) is a historical feedback control quantity of the vehicle to be controlled, and fal (e, α, δ) is a commonly used nonlinear function.
In a preferred embodiment, the generating the feedback control amount of the vehicle to be controlled according to the acceleration signal difference value, the acceleration differential signal difference value, and the disturbance signal value includes:
Generating a preliminary feedback control quantity according to the acceleration signal difference value and the acceleration differential signal difference value, and generating the feedback control quantity of the vehicle to be controlled according to the preliminary feedback control quantity and the disturbance signal value.
Specifically, a preliminary feedback control amount u 0 is generated according to the acceleration signal difference value and the acceleration differential signal difference value, and then a feedback control amount u 2 (k) of the vehicle to be controlled is generated according to the disturbance signal value z 3 and the preliminary feedback control amount u 0.
In a preferred embodiment, the acceleration signal difference, the acceleration differential signal difference, the preliminary feedback control amount, and the feedback control amount are generated by the following formula calculation:
Where e 1 denotes an acceleration signal difference, e 2 denotes an acceleration differential signal difference, λ 01 and λ 02 are weight coefficients, u 0 denotes a preliminary feedback control amount, and u 2 (k) denotes a feedback control amount of the vehicle to be controlled.
The above embodiments of the present invention have the following effects by implementing the present invention:
1. The gray Verhulst model front vehicle acceleration predictor based on particle swarm optimization is used for predicting the front vehicle acceleration, which is an uncertain scene, and has high instantaneity without a large number of data samples and excessive calculation force. In addition, the model has strong capability of predicting nonlinear data, and improves the calculation accuracy of the safety distance between two vehicles, so that the ACC system can cope with the working condition of large change amplitude of acceleration of the front vehicle in the following mode, and the use experience and safety of the ACC system are improved.
2. On the one hand, known vehicle parameters and road information are provided as much as possible by feed-forward control, and on the other hand, disturbances caused by unknown changing driving road conditions are observed by feedback control so as to compensate the control strategy. Thus, the disturbance rejection control module enables the vehicle to obtain a correct control strategy in a constantly changing driving road environment.
3. The inverse longitudinal vehicle model of the RBF neural network based on EKF online learning not only can compensate the nonlinear characteristics of the vehicle, but also can continuously learn from new data generated in the running process of the vehicle, thereby reducing the influence caused by uncertainty of vehicle parameters, and further enabling a vehicle executing mechanism to accurately track the control strategy of an upper controller. In addition, the RBF neural network only has a hidden layer, and the number of neurons in the hidden layer is controlled by the pruning strategy, so that the network structure cannot be redundant during learning, and the real-time requirement of an automatic driving algorithm is violated.
While the foregoing is directed to the preferred embodiments of the present invention, it will be appreciated by those skilled in the art that changes and modifications may be made without departing from the principles of the invention, such changes and modifications are also deemed to be the invention as set forth in the claims.

Claims (9)

1. An adaptive cruise system based on immunity control, comprising: the system comprises an environment sensing layer, an upper controller, a lower controller and a vehicle executing mechanism;
The environment sensing layer is used for acquiring first acceleration of a front vehicle at intervals of a preset time through a vehicle radar and sending the acquired first acceleration to the upper controller;
The upper controller is configured to input the first accelerations into a preset front vehicle acceleration predictor, so that the front vehicle acceleration predictor predicts a first acceleration of a front vehicle in a next preset period according to each first acceleration, obtain a first expected acceleration based on the predicted first acceleration, and send the first expected acceleration to the lower controller;
The lower controller includes: a feedforward control module and a feedback control module;
the feedforward control module is used for acquiring the running speed of the vehicle to be controlled through the vehicle executing mechanism, and calculating and generating feedforward control quantity of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first expected acceleration;
The feedback control module is used for acquiring a second acceleration of the vehicle to be controlled through the vehicle executing mechanism, and calculating and generating a feedback control quantity of the vehicle to be controlled according to the second acceleration and the first expected acceleration of the vehicle to be controlled; calculating and generating a second expected acceleration of the vehicle to be controlled according to the feedforward control quantity and the feedback control quantity of the vehicle to be controlled;
The vehicle executing mechanism is used for controlling the acceleration of the vehicle to be controlled according to the second expected acceleration so as to enable the acceleration of the vehicle to be controlled to be consistent with the second expected acceleration;
the front vehicle acceleration predictor predicts a first acceleration of a front vehicle within a preset period according to each first acceleration, and the front vehicle acceleration predictor comprises:
Inputting each first acceleration into a preset sequence set through the front vehicle acceleration predictor, generating a first sequence set, and establishing a first accumulation sequence set corresponding to the first sequence set according to the first sequence set;
according to the first accumulation sequence set, constructing an adjacent mean value generation sequence corresponding to the first accumulation sequence set, according to the adjacent mean value generation sequence, establishing a gray Verhulst model, and according to the gray Verhulst model, constructing a whitening equation corresponding to the gray Verhulst model;
And solving the whitening equation to generate a second accumulation sequence set for predicting the acceleration of the front vehicle, and predicting the first acceleration of the front vehicle in a next preset period according to the second accumulation sequence set.
2. The adaptive cruise system based on immunity control of claim 1, wherein the first set of sequences is represented by:
In the method, in the process of the invention, Representing a first set of sequences,/>And/>Respectively representing the 1 st to 5 th first accelerations in the first series set;
The first accumulated sequence set corresponding to the first sequence set is represented by the following formula:
In the method, in the process of the invention, Representing a first accumulated sequence set,/>Representing a kth first accumulated acceleration in the first accumulated sequence set, k representing an ordinal number in the sequence set;
The immediately adjacent mean generation sequence is represented by the following formula:
In the method, in the process of the invention, Representing a proximate mean generation sequence corresponding to the first accumulated sequence set,/>Representing a kth immediately adjacent mean in the immediately adjacent mean generation sequence;
The gray Verhulst model is represented by the following formula:
Where α and β are model parameters.
3. The adaptive cruise system based on disturbance-rejection control of claim 2, wherein said solving said whitening equation to generate a second set of accumulated sequences for predicting acceleration of a preceding vehicle, and predicting a first acceleration of said preceding vehicle for a next predetermined period of time based on said second set of accumulated sequences, comprises:
Constructing an equation solving formula and a particle algorithm fitness function of the whitening equation according to the whitening equation, and calculating and generating a second accumulation sequence set for predicting the acceleration of the vehicle in front according to the equation solving formula and the particle algorithm fitness function of the whitening equation; calculating and generating a first acceleration of the front vehicle in a next preset period according to the second accumulation sequence set;
wherein the equation solution of the whitening equation is represented by:
In the method, in the process of the invention, Representing a (k+1) th second accumulated acceleration in the second accumulated sequence set; e represents a natural constant;
the particle swarm algorithm fitness function is represented by the following formula:
where f fitness denotes a fitness function of the particle swarm algorithm, Representing an ith first accumulated acceleration in the first accumulated sequence set; /(I)Representing a second set of accumulated sequences,/>Representing an ith second accumulated acceleration in the second accumulated sequence set; n represents the number of second accumulated accelerations in the second accumulated sequence set;
generating a first acceleration of the preceding vehicle for a next preset period of time by the following equation calculation:
In the method, in the process of the invention, Representing the predicted first acceleration,/>Representing a second sequence set corresponding to the second accumulated sequence set,Indicating the k+1st acceleration in the second set of sequences.
4. The adaptive cruise system based on immunity control of claim 1, wherein the feedback control module comprises: the system comprises a tracking differential module, an expansion state observation module and an error feedback module;
the tracking differential module is used for calculating and generating a first expected acceleration signal value and a first expected acceleration differential signal value of the vehicle to be controlled according to the first expected acceleration;
The extended state observation module is used for acquiring a second acceleration of the vehicle to be controlled through the vehicle execution mechanism, acquiring a historical feedback control quantity of the vehicle to be controlled through the error feedback module, and calculating and generating a second acceleration signal value, a second acceleration differential signal value and a disturbance signal value according to the second acceleration and the historical feedback control quantity of the vehicle to be controlled;
The error feedback module is used for comparing the first expected acceleration signal value with the second acceleration signal value to obtain an acceleration signal difference value; comparing the first expected acceleration differential signal value with the second acceleration differential signal value to obtain an acceleration differential signal difference value; calculating and generating feedback control quantity of the vehicle to be controlled according to the acceleration signal difference value, the acceleration differential signal difference value and the disturbance signal value;
The error feedback module is further configured to obtain a feedforward control amount of the vehicle to be controlled through the feedforward control module, and sum the feedforward control amount and the feedback control amount of the vehicle to be controlled to generate a second desired acceleration of the vehicle to be controlled.
5. The adaptive cruise control-based adaptive cruise system according to claim 4, characterized in that the feedforward control module, before calculating the feedforward control amount of the vehicle to be controlled from the running speed of the vehicle to be controlled and the first desired acceleration, is further configured to:
Judging whether the value of the first expected acceleration is positive or negative;
The calculating the feedforward control quantity of the vehicle to be controlled according to the running speed of the vehicle to be controlled and the first expected acceleration comprises the following steps:
when it is determined that the first desired acceleration is a positive number, a feedforward control amount of the vehicle to be controlled is calculated by the following formula:
wherein a xdes represents a first expected acceleration, ζ is a rotational mass conversion coefficient, i g is a transmission gear ratio, i 0 is a main reduction ratio, η T represents mechanical efficiency when a transmission system is transmitted, r is a wheel rolling radius, m is a whole vehicle equipment mass of a vehicle to be controlled, g is a gravity coefficient, f is a wheel corresponding rolling resistance coefficient, C D is an air resistance coefficient along a vehicle running direction, A is an automobile windward area, ρ is an air density, and v x is a running speed of the vehicle to be controlled; t VDM denotes a feedforward control amount acting on the vehicle drive;
When it is determined that the first desired acceleration is negative, a feedforward control amount of the vehicle to be controlled is calculated by the following formula:
Where K b is a proportional coefficient of the vehicle braking force and the tire braking pressure, and F VDB is a feedforward control amount acting on the vehicle braking.
6. An adaptive cruise system based on disturbance-rejection control according to claim 5, wherein the first desired acceleration signal value and the first desired acceleration derivative signal value for the vehicle to be controlled are calculated by the following formula:
Wherein x 1 is a first desired acceleration signal value, T s is a sampling time, x 2 is a first desired acceleration differential signal value, r is a tracking speed factor, h is a filtering factor, fhan (x 1,x2, r, h) is a fastest control function, and sign is a sign function.
7. The adaptive cruise system based on disturbance rejection control according to claim 6, wherein the second acceleration signal value, the second acceleration differential signal value, and the disturbance signal value are generated by the following formula calculation;
Wherein z 1 represents a second acceleration signal value, z 2 represents a second acceleration differential signal value, z 3 represents a disturbance signal value, λ 01 and λ 02 are weight coefficients, u 1 (k) is a historical feedback control quantity of the vehicle to be controlled, and fal (e, α, δ) is a commonly used nonlinear function.
8. The adaptive cruise system according to claim 7, wherein the generating the feedback control amount of the vehicle to be controlled based on the acceleration signal difference value, the acceleration differential signal difference value, and the disturbance signal value includes:
Generating a preliminary feedback control quantity according to the acceleration signal difference value and the acceleration differential signal difference value, and generating the feedback control quantity of the vehicle to be controlled according to the preliminary feedback control quantity and the disturbance signal value.
9. The adaptive cruise system according to claim 8, wherein the acceleration signal difference, acceleration differential signal difference, preliminary feedback control amount, and feedback control amount are generated by calculation of the following formulas:
Where e 1 denotes an acceleration signal difference, e 2 denotes an acceleration differential signal difference, λ 01 and λ 02 are weight coefficients, u 0 denotes a preliminary feedback control amount, and u 2 (k) denotes a feedback control amount of the vehicle to be controlled.
CN202311102057.8A 2023-08-30 2023-08-30 Adaptive cruise system based on disturbance rejection control Active CN117068159B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311102057.8A CN117068159B (en) 2023-08-30 2023-08-30 Adaptive cruise system based on disturbance rejection control

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311102057.8A CN117068159B (en) 2023-08-30 2023-08-30 Adaptive cruise system based on disturbance rejection control

Publications (2)

Publication Number Publication Date
CN117068159A CN117068159A (en) 2023-11-17
CN117068159B true CN117068159B (en) 2024-04-19

Family

ID=88709549

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311102057.8A Active CN117068159B (en) 2023-08-30 2023-08-30 Adaptive cruise system based on disturbance rejection control

Country Status (1)

Country Link
CN (1) CN117068159B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107808027A (en) * 2017-09-14 2018-03-16 上海理工大学 It is adaptive with car algorithm based on improved model PREDICTIVE CONTROL
CN108454626A (en) * 2018-01-24 2018-08-28 北京新能源汽车股份有限公司 A kind of the adaptive cruise longitudinally adjust control method and system of vehicle
CN109572695A (en) * 2018-11-08 2019-04-05 湖南汽车工程职业学院 A kind of autonomous driving vehicle Car following control method and system
CN110155052A (en) * 2019-05-29 2019-08-23 台州学院 Improved adaptive cruise lower layer control design case method
CN111332290A (en) * 2020-03-24 2020-06-26 湖南大学 Vehicle formation method and system based on feedforward-feedback control
CN113859236A (en) * 2021-09-07 2021-12-31 中汽创智科技有限公司 Car following control system, car, method, device, equipment and storage medium
CN116552550A (en) * 2023-04-28 2023-08-08 贵州师范大学 Vehicle track tracking control system based on parameter uncertainty and yaw stability

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230182741A1 (en) * 2022-11-14 2023-06-15 Chang'an University Smooth cooperative lane change control method for multi-connected and autonomous vehicle (cav

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107808027A (en) * 2017-09-14 2018-03-16 上海理工大学 It is adaptive with car algorithm based on improved model PREDICTIVE CONTROL
CN108454626A (en) * 2018-01-24 2018-08-28 北京新能源汽车股份有限公司 A kind of the adaptive cruise longitudinally adjust control method and system of vehicle
CN109572695A (en) * 2018-11-08 2019-04-05 湖南汽车工程职业学院 A kind of autonomous driving vehicle Car following control method and system
CN110155052A (en) * 2019-05-29 2019-08-23 台州学院 Improved adaptive cruise lower layer control design case method
CN111332290A (en) * 2020-03-24 2020-06-26 湖南大学 Vehicle formation method and system based on feedforward-feedback control
CN113859236A (en) * 2021-09-07 2021-12-31 中汽创智科技有限公司 Car following control system, car, method, device, equipment and storage medium
CN116552550A (en) * 2023-04-28 2023-08-08 贵州师范大学 Vehicle track tracking control system based on parameter uncertainty and yaw stability

Also Published As

Publication number Publication date
CN117068159A (en) 2023-11-17

Similar Documents

Publication Publication Date Title
US11543787B2 (en) Networked control system time-delay compensation method based on predictive control
CN111624992B (en) Path tracking control method of transfer robot based on neural network
CN114967676A (en) Model prediction control trajectory tracking control system and method based on reinforcement learning
CN110615003B (en) Cruise control system based on strategy gradient online learning algorithm and design method
CN113386781A (en) Intelligent vehicle trajectory tracking control method based on data-driven vehicle dynamics model
Kuutti et al. End-to-end reinforcement learning for autonomous longitudinal control using advantage actor critic with temporal context
CN115598983B (en) Unmanned vehicle transverse and longitudinal cooperative control method and device considering time-varying time delay
CN110758413B (en) Train speed self-adaptive control method based on system parameter identification
CN109191788B (en) Driver fatigue driving judgment method, storage medium, and electronic device
CN115848398B (en) Lane departure early warning system assessment method based on learning and considering driver behavior characteristics
Ma A neural-fuzzy framework for modeling car-following behavior
CN115432009A (en) Automatic driving vehicle trajectory tracking control system
CN117389276B (en) Unmanned vehicle driving path tracking control method based on driving risk prediction
CN117068159B (en) Adaptive cruise system based on disturbance rejection control
CN114253274A (en) Data-driven-based online hybrid vehicle formation rolling optimization control method
CN116755323A (en) Multi-rotor unmanned aerial vehicle PID self-tuning method based on deep reinforcement learning
Li et al. Reinforcement learning based lane change decision-making with imaginary sampling
US20230001940A1 (en) Method and Device for Optimum Parameterization of a Driving Dynamics Control System for Vehicles
Wang et al. A robust design of hybrid fuzzy controller with fuzzy decision tree for autonomous intelligent parking system
CN113705865B (en) Automobile stability factor prediction method based on deep neural network
CN115817509A (en) Multi-axis distributed driving vehicle steering auxiliary track tracking method based on AMPC
Da Rocha et al. Model predictive control of a heavy-duty truck based on Gaussian process
CN115016248A (en) Motor PID control method for optimizing RBF neural network based on PSO algorithm
Topalov et al. Neuro-fuzzy control of antilock braking system using variable-structure-systems-based learning algorithm
Da Lio et al. Robust and sample-efficient estimation of vehicle lateral velocity using neural networks with explainable structure informed by kinematic principles

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