Longitudinal control system and method for autonomous driving vehicle based on feedforward-fuzzy PI

Abstract

The present invention discloses a feedforward-fuzzy PI-based longitudinal control system and method for an autonomous vehicle. The system includes a vehicle information acquisition module, a host controller module, and a host controller module. The present invention is divided into a host controller module and a host controller module based on the control process. The host controller module proposes feedforward fuzzy PI control, which includes feedforward control and fuzzy PI feedback control. The host controller module includes logic switching control and an inverse longitudinal vehicle model. When the control system of the present invention is operated, when the desired speed of the autonomous vehicle changes, its control accuracy is higher than that of traditional control methods, thereby better achieving longitudinal motion control of the vehicle.

Description

Technical Field

[0001] The present invention belongs to the field of longitudinal control of autonomous vehicles, and specifically relates to a longitudinal control system and method of autonomous vehicles based on feedforward-fuzzy PI. Background Art

[0002] In recent years, autonomous driving has become a hot topic in both academia and engineering. Autonomous driving systems encompass multiple disciplines, including perception, high-precision mapping, prediction, decision-making and planning, and motion control. Research on these systems is challenging, and therefore typically implemented in a modular fashion.

[0003] Motion control is one of the key modules of an autonomous driving system. It controls the vehicle to follow the trajectory information planned by the decision-making and planning module. Motion control can generally be divided into lateral motion control and longitudinal motion control. Among them, longitudinal motion control of autonomous vehicles refers to accurately controlling the vehicle speed by adjusting the vehicle's drive actuators and brake actuators. Traditional longitudinal control usually uses the PI algorithm. However, due to the nonlinearity and parameter uncertainty of the vehicle's drive and braking systems, and the fact that the target vehicle speed may change drastically when the driving environment changes, the longitudinal motion control of autonomous vehicles using the PI algorithm is not accurate enough. Summary of the Invention

[0004] One of the objectives of the present invention is to provide an autonomous driving vehicle longitudinal control system based on feedforward-fuzzy PI to solve the problem of inaccurate speed control of the autonomous driving vehicle longitudinal control when the driving environment changes and the target vehicle speed changes.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] A feedforward-fuzzy PI-based longitudinal control system for an autonomous driving vehicle, comprising: a vehicle information acquisition module, an upper controller module, and a lower controller module, wherein:

[0007] The vehicle information acquisition module is used to obtain the target trajectory information and current state information of the autonomous driving vehicle, and output the target speed and current speed to the upper controller module, and output the current speed, transmission ratio and engine speed to the lower controller module;

[0008] The upper controller module is used to perform feedforward control and fuzzy PI feedback control according to the target speed and the current speed, superimpose the planned accelerations output by the feedforward control and the fuzzy PI feedback control respectively as the expected acceleration, and output the expected acceleration to the lower controller module;

[0009] The lower controller module is used to switch the driving and braking logic and calculate the vehicle longitudinal and inverse dynamics model according to the expected acceleration, current speed, transmission ratio and engine speed, and control the autonomous driving vehicle according to the calculation results.

[0010] Several optional methods are also provided below, but they are not intended to be additional limitations on the above-mentioned overall solution. They are merely further supplements or optimizations. Under the premise that there are no technical or logical contradictions, each optional method can be combined separately for the above-mentioned overall solution, or multiple optional methods can be combined.

[0011] Preferably, the feedforward control and fuzzy PI feedback control are performed according to the target speed and the current speed, and the following operations are performed:

[0012] Perform feedforward control based on the target speed to obtain the planned acceleration of the feedforward control output;

[0013] Fuzzy PI feedback control is performed based on the target speed and current speed to obtain the planned acceleration output by the fuzzy PI feedback control.

[0014] Preferably, the feedforward control is performed based on the target speed to obtain the planned acceleration output by the feedforward control, and the following operations are performed:

[0015] a forward =K f v ref

[0016] Where a forward is the planned acceleration of the feedforward control output, K f is the feedforward coefficient, v ref is the target speed;

[0017] The fuzzy PI feedback control is performed based on the target speed and the current speed to obtain the planned acceleration output by the fuzzy PI feedback control, and the following operations are performed:

[0018] Calculate the error value e between the vehicle's target speed and current speed, as well as the deviation change rate ec;

[0019] The error value e and the deviation change rate ec are transformed into variables E and EC corresponding to the fuzzy domain through fuzzification;

[0020] According to variables E and EC, based on fuzzy reasoning, fuzzy rules and defuzzification, the correction value ΔK of the proportional coefficient is output p And the correction value of the integral coefficient ΔK i ;

[0021] Calculate the proportionality factor as:

[0022] K p =Kp0 +ΔK p

[0023] Where K p is the proportionality coefficient, K p0 is the initial value of the proportional coefficient;

[0024] Calculate the integral coefficient as:

[0025] K i =K i0 +ΔK i

[0026] Where K i is the integral coefficient, K i0 is the initial value of the integral coefficient;

[0027] Calculate the planned acceleration as:

[0028] a back =K p e+K i ∫edt

[0029] e=v ref -v

[0030] Where a back is the planned acceleration output by the fuzzy PI feedback control, and v is the current speed.

[0031] Preferably, the driving and braking logic switching and the vehicle longitudinal inverse dynamics model calculation are performed by performing the following operations:

[0032] If the expected acceleration a des If it is positive, the drive control is applied and the desired throttle opening signal is calculated; if the desired acceleration a des If it is negative, the brake control is applied and the expected brake master cylinder pressure signal is calculated;

[0033] Wherein, the calculation of the expected throttle opening signal includes: if the expected acceleration a des If the value is less than the preset buffer zone Δh, the expected throttle opening signal is set to 0; otherwise, the expected throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model;

[0034] Wherein, the calculation of the expected brake master cylinder pressure signal includes: if the expected acceleration a des If the pressure is less than the preset buffer zone -Δh, the expected master cylinder pressure signal is taken as 0; otherwise, the expected master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model.

[0035] Preferably, the desired throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model, and the following operations are performed:

[0036]

[0037] Where, T is the engine output torque, C D is the air resistance coefficient, ρ is the air density, A is the frontal area, v is the current speed, m is the vehicle weight, g is the acceleration of gravity, f is the rolling friction coefficient, δ1 and δ2 are preset constants, i g is the transmission ratio, R is the wheel radius, i0 is the main reducer transmission ratio, η is the transmission efficiency, ω e is the engine speed, γ(T,ω e ) is the engine output torque T and engine speed ω e The relational expression is obtained by looking up the vehicle's engine output torque MAP diagram, α thdes is the expected throttle opening signal;

[0038] The desired brake master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model, and the following operations are performed:

[0039]

[0040] Where a1 is the acceleration without throttle driving force and braking force, k f is the front wheel braking torque, k r is the rear wheel braking torque, p thdes is the expected brake master cylinder pressure signal.

[0041] The feedforward-fuzzy PI-based longitudinal control system for an autonomous vehicle provided by the present invention is divided into an upper controller module and a lower controller module according to the control process. The upper controller module proposes a feedforward fuzzy PI controller, which includes feedforward control and fuzzy PI feedback control. The lower controller module includes logic switching control and an inverse longitudinal vehicle model. In order to verify the effectiveness of the proposed controller, a collaborative simulation platform was constructed using Simulink and CarSim. The simulation results show that when the desired speed of the autonomous vehicle changes, the control accuracy of the controller of the present invention is higher than that of the traditional controller, thereby better realizing the longitudinal motion control of the vehicle.

[0042] A second object of the present invention is to provide a longitudinal control method for an autonomous driving vehicle based on feedforward-fuzzy PI to solve the problem of inaccurate speed control of the autonomous driving vehicle longitudinal control when the driving environment changes and the target vehicle speed changes.

[0043] To achieve the above object, the technical solution adopted by the present invention is:

[0044] The present invention also provides a feedforward-fuzzy PI-based longitudinal control method for an autonomous driving vehicle, the feedforward-fuzzy PI-based longitudinal control method for an autonomous driving vehicle comprising:

[0045] Obtaining target trajectory information and current state information of the autonomous driving vehicle, wherein the target trajectory information includes a target speed, and the current state information includes a current speed, a transmission ratio, and an engine speed;

[0046] Perform feedforward control and fuzzy PI feedback control according to the target speed and current speed, and superimpose the planned acceleration output by the feedforward control and fuzzy PI feedback control as the expected acceleration;

[0047] According to the expected acceleration, current speed, transmission ratio and engine speed, the driving and braking logic are switched and the longitudinal and inverse dynamics models of the vehicle are calculated. The autonomous driving vehicle is controlled according to the calculation results.

[0048] Preferably, the feedforward control and fuzzy PI feedback control are performed according to the target speed and the current speed, including:

[0049] Perform feedforward control based on the target speed to obtain the planned acceleration of the feedforward control output;

[0050] Fuzzy PI feedback control is performed based on the target speed and current speed to obtain the planned acceleration output by the fuzzy PI feedback control.

[0051] Preferably, performing feedforward control based on the target speed to obtain a planned acceleration output by the feedforward control includes:

[0052] a forward =K f v ref

[0053] Where a forward is the planned acceleration of the feedforward control output, K f is the feedforward coefficient, v ref is the target speed;

[0054] The fuzzy PI feedback control is performed based on the target speed and the current speed to obtain the planned acceleration output by the fuzzy PI feedback control, including:

[0055] Calculate the error value e between the vehicle's target speed and current speed, as well as the deviation change rate ec;

[0056] The error value e and the deviation change rate ec are transformed into variables E and EC corresponding to the fuzzy domain through fuzzification;

[0057] According to variables E and EC, based on fuzzy reasoning, fuzzy rules and defuzzification, the correction value ΔK of the proportional coefficient is outputp And the correction value of the integral coefficient ΔK i ;

[0058] Calculate the proportionality factor as:

[0059] K p =K p0 +ΔK p

[0060] Where K p is the proportionality coefficient, K p0 is the initial value of the proportional coefficient;

[0061] Calculate the integral coefficient as:

[0062] K i =K i0 +ΔK i

[0063] Where K i is the integral coefficient, K i0 is the initial value of the integral coefficient;

[0064] Calculate the planned acceleration as:

[0065] a back =K p e+K i ∫edt

[0066] e=v ref -v

[0067] Where a back is the planned acceleration output by the fuzzy PI feedback control, and v is the current speed.

[0068] Preferably, the switching of driving and braking logic and calculation of the vehicle longitudinal and reverse dynamics model include:

[0069] If the expected acceleration a des If it is positive, the drive control is applied and the desired throttle opening signal is calculated; if the desired acceleration a des If it is negative, the brake control is applied and the expected brake master cylinder pressure signal is calculated;

[0070] Wherein, the calculation of the expected throttle opening signal includes: if the expected acceleration a des If the value is less than the preset buffer zone Δh, the expected throttle opening signal is set to 0; otherwise, the expected throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model;

[0071] Wherein, the calculation of the expected brake master cylinder pressure signal includes: if the expected acceleration a desIf the pressure is less than the preset buffer zone -Δh, the expected master cylinder pressure signal is taken as 0; otherwise, the expected master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model.

[0072] Preferably, the calculating of the expected throttle opening signal according to the vehicle longitudinal inverse dynamics model includes:

[0073]

[0074] Where, T is the engine output torque, C D is the air resistance coefficient, ρ is the air density, A is the frontal area, v is the current speed, m is the vehicle weight, g is the acceleration of gravity, f is the rolling friction coefficient, δ1 and δ2 are preset constants, i g is the transmission ratio, R is the wheel radius, i0 is the main reducer transmission ratio, η is the transmission efficiency, ω e is the engine speed, γ(T,ω e ) is the engine output torque T and engine speed ω e The relational expression is obtained by looking up the vehicle's engine output torque MAP diagram, α thdes is the expected throttle opening signal;

[0075] The calculating of the expected brake master cylinder pressure signal according to the vehicle longitudinal inverse dynamics model includes:

[0076]

[0077] Where a1 is the acceleration without throttle driving force and braking force, k f is the front wheel braking torque, k r is the rear wheel braking torque, p thdes is the expected brake master cylinder pressure signal.

[0078] The feedforward-fuzzy PI-based longitudinal control method for autonomous vehicles, provided by this invention, is divided into feedforward fuzzy PI control, including feedforward control and fuzzy PI feedback control, as well as logic switching control and an inverse longitudinal vehicle model. This method integrates feedforward and fuzzy control into the traditional PI algorithm, addressing the issue of inaccurate longitudinal speed control for autonomous vehicles when the target vehicle speed fluctuates dramatically due to changes in the driving environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 This is a structural diagram of the longitudinal control system of an autonomous driving vehicle based on feedforward-fuzzy PI according to the present invention;

[0080] Figure 2 This is a flow chart of the longitudinal control method of an autonomous driving vehicle based on feedforward-fuzzy PI according to the present invention;

[0081] Figure 3 Schematic diagram of the control simulation model of the present invention. DETAILED DESCRIPTION

[0082] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0083] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0084] In order to solve the problem in the prior art of inaccurate speed control in the longitudinal control of an autonomous vehicle when the driving environment changes and the target vehicle speed changes, this embodiment provides a longitudinal control system for an autonomous vehicle based on feedforward-fuzzy PI.

[0085] like Figure 1 As shown, the feedforward-fuzzy PI-based autonomous driving vehicle longitudinal control system of this embodiment includes: a vehicle information acquisition module, an upper controller module and a lower controller module.

[0086] 1) Vehicle information acquisition module.

[0087] This module is used to obtain the target trajectory information and current state information of the autonomous driving vehicle, and output the target speed and current speed to the upper controller module, and output the current speed, transmission ratio and engine speed to the lower controller module.

[0088] This embodiment obtains target trajectory information from a decision-making planning system. The decision-making planning system here can be a path planning system based on program operation, or it can be an interactive system that receives human input to form a planned path. It will not be described in detail in this embodiment.

[0089] The target trajectory information obtained in this embodiment includes target speed, specifically an array, and the array includes the target speed at a predetermined position.

[0090] The current status information is obtained from the vehicle, for example, the vehicle's OBD information is obtained based on the OBD interface provided by the vehicle, the vehicle's location information is obtained according to the vehicle's GPS, etc., which will not be described in detail in this embodiment.

[0091] 2) Upper controller module.

[0092] This module is used to perform feedforward control and fuzzy PI feedback control according to the target speed and current speed, superimpose the planned acceleration output by the feedforward control and fuzzy PI feedback control as the expected acceleration, and output the expected acceleration to the lower controller module.

[0093] The upper controller module calculates the expected acceleration required by the lower controller module based on the speed. In the expected acceleration calculation, feedforward control is performed based on the target speed to obtain the planned acceleration output by the feedforward control; fuzzy PI feedback control is performed based on the target speed and the current speed to obtain the planned acceleration output by the fuzzy PI feedback control.

[0094] Feedforward control dynamically tracks input signals. This embodiment adds feedforward control to traditional feedback control to ensure optimal dynamic tracking and interference rejection for the entire control system. This embodiment further employs fuzzy PI control, which combines fuzzy control with traditional PI control. This combines the dynamics and robustness of fuzzy control with the steady-state accuracy of PI control.

[0095] Its feedforward control calculation is as follows:

[0096] a forward =K f v ref

[0097] Where a forward is the planned acceleration of the feedforward control output, K f is the feedforward coefficient, representing the feedforward depth, v ref is the target speed.

[0098] Its fuzzy PI feedback control calculation is as follows:

[0099] Calculate the error value e between the vehicle's target speed and current speed, as well as the deviation change rate ec.

[0100] The error value e and the deviation change rate ec are transformed into variables E and EC corresponding to the fuzzy domain through fuzzification.

[0101] According to variables E and EC, based on fuzzy reasoning, fuzzy rules and defuzzification, the correction value ΔK of the proportional coefficient is output p And the correction value of the integral coefficient ΔK i .

[0102] Calculate the proportionality factor as:

[0103] K p =K p0 +ΔK p

[0104] Where K p is the proportionality coefficient, K p0 is the initial value of the proportional coefficient.

[0105] Calculate the integral coefficient as:

[0106] K i =K i0 +ΔK i

[0107] Where K i is the integral coefficient, K i0 is the initial value of the integral coefficient.

[0108] Calculate the planned acceleration as:

[0109] a back =K p e+K i ∫edt

[0110] e=v ref -v

[0111] Where a back is the planned acceleration output by the fuzzy PI feedback control, and v is the current speed.

[0112] In feedforward control and fuzzy PI feedback control, an appropriate feedforward coefficient is first determined using a software simulation module. Then, based on the determined feedforward coefficient, a fuzzy algorithm is used to adjust the proportional and integral coefficients in the fuzzy PI feedback control. To reduce coefficient correction effort and improve control accuracy, the initial values ​​of the proportional and integral coefficients in the fuzzy PI feedback control of this embodiment are determined based on those of conventional PI control.

[0113] 3) Lower controller module.

[0114] This module is used to switch the driving and braking logic and calculate the vehicle's longitudinal and inverse dynamics models based on the expected acceleration, current speed, transmission ratio and engine speed, and control the autonomous vehicle based on the calculation results.

[0115] In vehicle longitudinal reverse control, how to control the switching of drive and brake logic has a critical impact on vehicle stability and comfort. Therefore, this embodiment introduces a buffer zone based on the expected positive and negative acceleration to prevent frequent switching between drive and brake. The specific control logic is as follows:

[0116] If the expected acceleration a des Is positive, indicating that the vehicle needs to accelerate, then the drive control is applied and the expected throttle opening signal is calculated; if the expected acceleration a desIf it is negative, it means the vehicle needs to decelerate, so the braking control is applied and the expected brake master cylinder pressure signal is calculated.

[0117] A. Calculate the expected throttle opening signal, including: If the expected acceleration a des If the value is less than the preset buffer zone Δh, the expected throttle opening signal is set to 0; otherwise, the expected throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model.

[0118] In this embodiment, the expected acceleration a des When it is less than the preset buffer zone Δh (Δh is a positive number), the expected throttle opening signal will be 0. At this time, no control force is applied to the vehicle, and preparation is made in advance for the next possible braking of the vehicle. This not only avoids frequent switching between driving and braking of the vehicle, but also helps the vehicle to switch between driving and braking more smoothly, thereby improving the stability of vehicle driving and the user's riding experience.

[0119] Specifically, the expected throttle opening signal is calculated based on the vehicle longitudinal inverse dynamics model, and the following operations are performed:

[0120]

[0121] Where, T is the engine output torque, C D is the air resistance coefficient, ρ is the air density, A is the frontal area, v is the current speed, m is the vehicle weight, g is the acceleration of gravity, f is the rolling friction coefficient, δ1 and δ2 are preset constants, i g is the transmission ratio, R is the wheel radius, i0 is the main reducer transmission ratio, η is the transmission efficiency, ω e is the engine speed, γ(T,ω e ) is the engine output torque T and engine speed ω e The relational expression is obtained by looking up the vehicle's engine output torque MAP diagram, α thdes The desired throttle opening signal is obtained by reversely deriving the desired throttle angle α based on the vehicle dynamic balance during acceleration. thdes and the expected acceleration a des The above mathematical relationship between .

[0122] B. Calculate the expected brake master cylinder pressure signal, including: if the expected acceleration a des If the pressure is less than the preset buffer zone -Δh, the expected master cylinder pressure signal is taken as 0; otherwise, the expected master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model.

[0123] Similarly, in this embodiment, the expected acceleration a desWhen it is less than the preset buffer zone -Δh, the brake master cylinder pressure signal is expected to be 0. At this time, no control force is applied to the vehicle, and preparation is made in advance for the next possible driving of the vehicle. This not only avoids frequent switching between driving and braking, but also helps the vehicle to switch between driving and braking more smoothly, thereby improving the stability of vehicle driving and the user's riding experience.

[0124] Specifically, the expected brake master cylinder pressure signal is calculated based on the vehicle longitudinal inverse dynamics model, and the following operations are performed:

[0125]

[0126] Where a1 is the acceleration without throttle driving force and braking force, k f is the front wheel braking torque, k r is the rear wheel braking torque, p thdes is the expected brake master cylinder pressure signal. According to the vehicle dynamic balance during braking, the embodiment can obtain the expected brake master cylinder pressure p thdes and the expected acceleration a des The above relationship.

[0127] When calculating the desired throttle opening signal and the desired master cylinder pressure signal, the engine output torque, drag coefficient, air density, frontal area, vehicle weight, gravitational acceleration, rolling friction coefficient, wheel radius, final drive ratio, transmission efficiency, front wheel braking torque, and rear wheel braking torque are all known quantities and can be obtained based on the actual vehicle and current environment. The preset constants δ1 and δ2 can be obtained through software simulation and are commonly set constants in simulations.

[0128] This embodiment applies a fuzzy PI algorithm to the longitudinal motion control of an autonomous vehicle, achieving faster response and less overshoot. It also incorporates feedforward control to further improve vehicle speed tracking accuracy. Each module in the aforementioned feedforward-fuzzy PI-based longitudinal control system for autonomous vehicles can be implemented in whole or in part through software, hardware, or a combination thereof. This can be implemented as hardware embedded in or independent of a processor within a computer device, or stored as software within the computer device's memory, enabling the processor to call and execute the corresponding operations of each module.

[0129] In another embodiment, Figure 2 As shown, a longitudinal control method for an autonomous driving vehicle based on feedforward-fuzzy PI is also provided, comprising:

[0130] Obtaining target trajectory information and current state information of the autonomous driving vehicle, wherein the target trajectory information includes a target speed, and the current state information includes a current speed, a transmission ratio, and an engine speed;

[0131] Perform feedforward control and fuzzy PI feedback control according to the target speed and current speed, and superimpose the planned acceleration output by the feedforward control and fuzzy PI feedback control as the expected acceleration;

[0132] According to the expected acceleration, current speed, transmission ratio and engine speed, the driving and braking logic are switched and the longitudinal and inverse dynamics models of the vehicle are calculated. The autonomous driving vehicle is controlled according to the calculation results.

[0133] For the specific limitations of the longitudinal control method of an autonomous driving vehicle based on feedforward-fuzzy PI, please refer to the limitations of the longitudinal control system of an autonomous driving vehicle based on feedforward-fuzzy PI above, which will not be repeated here.

[0134] In a specific embodiment, performing feedforward control and fuzzy PI feedback control according to the target speed and the current speed includes:

[0135] Perform feedforward control based on the target speed to obtain the planned acceleration of the feedforward control output;

[0136] Fuzzy PI feedback control is performed based on the target speed and current speed to obtain the planned acceleration output by the fuzzy PI feedback control.

[0137] In a specific embodiment, performing feedforward control based on the target speed to obtain a planned acceleration output by the feedforward control includes:

[0138] a forward =K f v ref

[0139] Where a forward is the planned acceleration of the feedforward control output, K f is the feedforward coefficient, v ref is the target speed;

[0140] The fuzzy PI feedback control is performed based on the target speed and the current speed to obtain the planned acceleration output by the fuzzy PI feedback control, including:

[0141] Calculate the error value e between the vehicle's target speed and current speed, as well as the deviation change rate ec;

[0142] The error value e and the deviation change rate ec are transformed into variables E and EC corresponding to the fuzzy domain through fuzzification;

[0143] According to variables E and EC, based on fuzzy reasoning, fuzzy rules and defuzzification, the correction value ΔK of the proportional coefficient is output p And the correction value of the integral coefficient ΔK i ;

[0144] Calculate the proportionality factor as:

[0145] K p =K p0 +ΔK p

[0146] Where K p is the proportionality coefficient, K p0 is the initial value of the proportional coefficient;

[0147] Calculate the integral coefficient as:

[0148] K i =K i0 +ΔK i

[0149] Where K i is the integral coefficient, K i0 is the initial value of the integral coefficient;

[0150] Calculate the planned acceleration as:

[0151] a back =K p e+K i ∫edt

[0152] e=v ref -v

[0153] Where a back is the planned acceleration output by the fuzzy PI feedback control, and v is the current speed.

[0154] In a specific embodiment, the switching of driving and braking logic and the calculation of the vehicle longitudinal inverse dynamics model include:

[0155] If the expected acceleration a des If it is positive, the drive control is applied and the desired throttle opening signal is calculated; if the desired acceleration a des If it is negative, the brake control is applied and the expected brake master cylinder pressure signal is calculated;

[0156] Wherein, the calculation of the expected throttle opening signal includes: if the expected acceleration a des If the value is less than the preset buffer zone Δh, the expected throttle opening signal is set to 0; otherwise, the expected throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model;

[0157] Wherein, the calculation of the expected brake master cylinder pressure signal includes: if the expected acceleration a des If the pressure is less than the preset buffer zone -Δh, the expected master cylinder pressure signal is taken as 0; otherwise, the expected master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model.

[0158] In a specific embodiment, the calculating of the expected throttle opening signal according to the vehicle longitudinal inverse dynamics model includes:

[0159]

[0160] Where, T is the engine output torque, C D is the air resistance coefficient, ρ is the air density, A is the frontal area, v is the current speed, m is the vehicle weight, g is the acceleration of gravity, f is the rolling friction coefficient, δ1 and δ2 are preset constants, i g is the transmission ratio, R is the wheel radius, i0 is the main reducer transmission ratio, η is the transmission efficiency, ω e is the engine speed, γ(T,ω e ) is the engine output torque T and engine speed ω e The relational expression is obtained by looking up the vehicle's engine output torque MAP diagram, α thdes is the expected throttle opening signal;

[0161] The calculating of the expected brake master cylinder pressure signal according to the vehicle longitudinal inverse dynamics model includes:

[0162]

[0163] Where a1 is the acceleration without throttle driving force and braking force, k f is the front wheel braking torque, k r is the rear wheel braking torque, p thdes is the expected brake master cylinder pressure signal.

[0164] In another embodiment, in order to more intuitively demonstrate the advantages of the feedforward-fuzzy PI-based autonomous driving vehicle longitudinal control system and method of the present application, this embodiment provides the following specific simulation experiments.

[0165] A collaborative simulation platform using Matlab / Simulink and CarSim was built to verify the designed controller. A D-class sedan in CarSim was selected as the simulation target. The relevant vehicle parameters are shown in Table 1. The control rules of the fuzzy controller are shown in Table 2.

[0166] Table 1 Vehicle parameters

[0167] parameter Value parameter Value <![CDATA[C D ]]> 0.3 A <![CDATA[2.51m 2 ]]> m 1370kg g <![CDATA[9.8m / s 2 ]]> f 0.004 <![CDATA[δ1]]> 0.04 <![CDATA[δ2]]> 0.06 R 0.335m <![CDATA[i0]]> 4.1 η 0.9 <![CDATA[k f ]]> 350N·m / MPa <![CDATA[k r ]]> 200N·m / MPa

[0168] Table 2 Fuzzy controller control rules

[0169]

[0170] In Table 2, Z, P1, P2, P3, N1, N2, and N3 are fuzzy variable states, representing “zero”, “positive small”, “positive medium”, “positive large”, “negative small”, “negative medium”, and “negative large”, respectively. E and EC are transformed by fuzzifying the error e and the deviation change rate ec.

[0171] Figure 3 is a schematic diagram of the control simulation model, where K p0 =1.5, K i0 =0.0001, K f =0.001, Δh=0.1.

[0172] To simulate changes in required speed due to changing driving conditions, the required speed was set to a sinusoidal curve with an amplitude range of 35 km / h to 50 km / h and periods of 5, 10, and 15 seconds. Experiments were conducted using conventional PI control (PI), fuzzy PI control (Fuzzy PI), and the feedforward-fuzzy PI control proposed in this paper. The simulation results for the speed tracking errors of the three control methods are shown in Tables 3 and 4.

[0173] Table 3 Maximum error of speed control

[0174]

[0175] Table 4 Root mean square error of speed control

[0176]

[0177] Conditions 1, 2, and 3 have the desired sinusoidal speed, with an amplitude range of 35 km / h to 50 km / h and periods of 5 seconds, 10 seconds, and 15 seconds, respectively. Tables 3 and 4 show that the maximum error and root mean square error of the feedforward-fuzzy PI control designed in the present invention are smaller than those of the traditional PI control and fuzzy PI control in all three cases, indicating that the control method of the present application can better achieve longitudinal motion control, and the greater the acceleration, the better the control effect.

[0178] Simulation results show that when the desired speed of an autonomous vehicle changes, the control system and method of this application achieves higher control accuracy than traditional control methods. This means that the proposed control method can better achieve longitudinal motion control of autonomous vehicles (unmanned vehicles). The proposed control method can be applied to the motion control of electric unmanned vehicles in the next stage and has great market application value.

[0179] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0180] The above-described embodiments merely illustrate several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that a person skilled in the art would be able to make numerous modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A longitudinal control system for an autonomous vehicle based on feedforward-fuzzy PI, characterized in that: The feedforward-fuzzy PI-based autonomous driving vehicle longitudinal control system includes: a vehicle information acquisition module, an upper controller module and a lower controller module, wherein: The vehicle information acquisition module is used to obtain the target trajectory information and current state information of the autonomous driving vehicle, and output the target speed and current speed to the upper controller module, and output the current speed, transmission ratio and engine speed to the lower controller module; The upper controller module is used to perform feedforward control and fuzzy PI feedback control according to the target speed and the current speed, superimpose the planned accelerations output by the feedforward control and the fuzzy PI feedback control respectively as the expected acceleration, and output the expected acceleration to the lower controller module; The lower controller module is used to perform drive and brake logic switching and vehicle longitudinal and inverse dynamics model calculation according to the expected acceleration, current speed, transmission ratio and engine speed, and control the autonomous driving vehicle according to the calculation results; The feedforward control and fuzzy PI feedback control are performed according to the target speed and the current speed, and the following operations are performed: Perform feedforward control based on the target speed to obtain the planned acceleration of the feedforward control output; Perform fuzzy PI feedback control based on the target speed and current speed to obtain the planned acceleration output by the fuzzy PI feedback control; The feedforward control is performed based on the target speed to obtain the planned acceleration output by the feedforward control, and the following operations are performed: and forward =K f in ref Where a forward is the planned acceleration of the feedforward control output, K f is the feedforward coefficient, v ref is the target speed; The fuzzy PI feedback control is performed based on the target speed and the current speed to obtain the planned acceleration output by the fuzzy PI feedback control, and the following operations are performed: Calculate the error value e between the vehicle's target speed and current speed, as well as the deviation change rate ec; The error value e and the deviation change rate ec are transformed into variables E and EC corresponding to the fuzzy domain through fuzzification; According to variables E and EC, based on fuzzy reasoning, fuzzy rules and defuzzification, the correction value ΔK of the proportional coefficient is output p And the correction value of the integral coefficient ΔK i ; Calculate the proportionality factor as: K p =K p0 +ΔK p Where K p is the proportionality coefficient, K p0 is the initial value of the proportional coefficient; Calculate the integral coefficient as: K i =K i0 +ΔK i Where K i is the integral coefficient, K i0 is the initial value of the integral coefficient; Calculate the planned acceleration as: a back =K p e+K i ∫edt e=v ref -v Where a back is the planned acceleration output by the fuzzy PI feedback control, and v is the current speed.

2. The feedforward-fuzzy PI-based autonomous driving vehicle longitudinal control system according to claim 1, characterized in that: The driving and braking logic switching and the vehicle longitudinal inverse dynamics model calculation are performed as follows: If the expected acceleration a des If it is positive, the drive control is applied and the desired throttle opening signal is calculated; if the desired acceleration a des If it is negative, the brake control is applied and the expected brake master cylinder pressure signal is calculated; Wherein, the calculation of the expected throttle opening signal includes: if the expected acceleration a des If the value is less than the preset buffer zone Δh, the expected throttle opening signal is set to 0; otherwise, the expected throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model; Wherein, the calculation of the expected brake master cylinder pressure signal includes: if the expected acceleration a des If the pressure is less than the preset buffer zone -Δh, the expected master cylinder pressure signal is taken as 0; otherwise, the expected master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model.

3. The feedforward-fuzzy PI-based autonomous driving vehicle longitudinal control system according to claim 2, characterized in that: The desired throttle opening signal is calculated based on the vehicle longitudinal inverse dynamics model, and the following operations are performed: Where, T is the engine output torque, C D is the air resistance coefficient, ρ is the air density, A is the frontal area, v is the current speed, m is the vehicle weight, g is the acceleration of gravity, f is the rolling friction coefficient, δ1 and δ2 are preset constants, i g is the transmission ratio, R is the wheel radius, i0 is the main reducer transmission ratio, η is the transmission efficiency, ω e is the engine speed, γ(T,ω e ) is the engine output torque T and engine speed ω e The relational expression is obtained by looking up the vehicle's engine output torque MAP diagram, α thdes is the expected throttle opening signal; The desired brake master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model, and the following operations are performed: Where a1 is the acceleration without throttle driving force and braking force, k f is the front wheel braking torque, k r is the rear wheel braking torque, p thdes is the expected brake master cylinder pressure signal.

4. A longitudinal control method for an autonomous driving vehicle based on feedforward-fuzzy PI, characterized in that: The feedforward-fuzzy PI-based longitudinal control method for an autonomous driving vehicle includes: Obtaining target trajectory information and current state information of the autonomous driving vehicle, wherein the target trajectory information includes a target speed, and the current state information includes a current speed, a transmission ratio, and an engine speed; Perform feedforward control and fuzzy PI feedback control according to the target speed and current speed, and superimpose the planned acceleration output by the feedforward control and fuzzy PI feedback control as the expected acceleration; Based on the expected acceleration, current speed, transmission ratio, and engine speed, the system performs drive and brake logic switching and calculates the vehicle's longitudinal and inverse dynamics models, and controls the autonomous vehicle based on the calculation results. The feedforward control and fuzzy PI feedback control are performed according to the target speed and the current speed, including: Perform feedforward control based on the target speed to obtain the planned acceleration of the feedforward control output; Perform fuzzy PI feedback control based on the target speed and current speed to obtain the planned acceleration output by the fuzzy PI feedback control; The feedforward control is performed based on the target speed to obtain the planned acceleration output by the feedforward control, including: and forward =K f in ref Where a forward is the planned acceleration of the feedforward control output, K f is the feedforward coefficient, v ref is the target speed; The fuzzy PI feedback control is performed based on the target speed and the current speed to obtain the planned acceleration output by the fuzzy PI feedback control, including: Calculate the error value e between the vehicle's target speed and current speed, as well as the deviation change rate ec; The error value e and the deviation change rate ec are transformed into variables E and EC corresponding to the fuzzy domain through fuzzification; According to variables E and EC, based on fuzzy reasoning, fuzzy rules and defuzzification, the correction value ΔK of the proportional coefficient is output p And the correction value of the integral coefficient ΔK i ; Calculate the proportionality factor as: K p =K p0 +ΔK p Where K p is the proportionality coefficient, K p0 is the initial value of the proportional coefficient; Calculate the integral coefficient as: K i =K i0 +ΔK i Where K i is the integral coefficient, K i0 is the initial value of the integral coefficient; Calculate the planned acceleration as: a back =K p e+K i ∫edt e=v ref -v Where a back is the planned acceleration output by the fuzzy PI feedback control, and v is the current speed.

5. The feedforward-fuzzy PI based longitudinal control method for an autonomous driving vehicle according to claim 4, characterized in that: The driving and braking logic switching and the vehicle longitudinal and reverse dynamics model calculation include: If the expected acceleration a des If it is positive, the drive control is applied and the desired throttle opening signal is calculated; if the desired acceleration a des If it is negative, the brake control is applied and the expected brake master cylinder pressure signal is calculated; Wherein, the calculation of the expected throttle opening signal includes: if the expected acceleration a des If the value is less than the preset buffer zone Δh, the expected throttle opening signal is set to 0; otherwise, the expected throttle opening signal is calculated according to the vehicle longitudinal inverse dynamics model; Wherein, the calculation of the expected brake master cylinder pressure signal includes: if the expected acceleration a des If the pressure is less than the preset buffer zone -Δh, the expected master cylinder pressure signal is taken as 0; otherwise, the expected master cylinder pressure signal is calculated according to the vehicle longitudinal inverse dynamics model.

6. The feedforward-fuzzy PI based longitudinal control method for an autonomous driving vehicle according to claim 5, wherein: The calculating of the expected throttle opening signal according to the vehicle longitudinal inverse dynamics model includes: Where, T is the engine output torque, C D is the air resistance coefficient, ρ is the air density, A is the frontal area, v is the current speed, m is the vehicle weight, g is the acceleration of gravity, f is the rolling friction coefficient, δ1 and δ2 are preset constants, i g is the transmission ratio, R is the wheel radius, i0 is the main reducer transmission ratio, η is the transmission efficiency, ω e is the engine speed, γ(T,ω e ) is the engine output torque T and engine speed ω e The relational expression is obtained by looking up the vehicle's engine output torque MAP diagram, α thdes is the expected throttle opening signal; The calculating of the expected brake master cylinder pressure signal according to the vehicle longitudinal inverse dynamics model includes: Where a1 is the acceleration without throttle driving force and braking force, k f is the front wheel braking torque, k r is the rear wheel braking torque, p thdes is the expected brake master cylinder pressure signal.

Images