Vehicle longitudinal control method, device, electronic device and storage medium

Abstract

The present disclosure relates to a vehicle longitudinal control method, device, electronic device and storage medium. The method is applied to a vehicle longitudinal control device, the device includes a position outer loop system, a speed inner loop system and a tracking differentiator, and the method includes: obtaining the current posture of the vehicle in the global coordinate system, using the position outer loop system to generate a desired speed compensation control amount according to the current reference posture and current posture of the vehicle, the current reference posture is obtained from the reference trajectory sent by the planning system of the vehicle, and then using the tracking differentiator to filter the feedforward signal sent by the planning system, and using the speed inner loop system to generate an acceleration control amount for longitudinal control of the vehicle according to the filtered feedforward signal, the desired speed compensation control amount and the current longitudinal speed of the vehicle. The method provided by the present disclosure reduces the longitudinal speed tracking error of the vehicle by accurately controlling the vehicle longitudinally, so that the vehicle can smoothly track the desired longitudinal acceleration.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to a vehicle longitudinal control method, device, electronic device and storage medium. Background Art

[0002] At present, driverless vehicles are widely used in daily life. Among the key technologies of driverless vehicles, longitudinal control is one of the main functions of the driverless control layer. Longitudinal control mainly refers to the speed, acceleration output by longitudinal planning and the speed, acceleration and other information fed back by the chassis, and the real-time calculation of the longitudinal control amount to control the brakes and throttle of the vehicle, so that the driverless vehicle can automatically complete the desired tracking task. In longitudinal control, the control algorithm is the core, which is particularly important for the key performance of the driverless vehicle, such as tracking accuracy and ride comfort, and also determines the quality of the driverless vehicle in completing the driverless task.

[0003] However, the stability, accuracy, and robustness of the control algorithm are greatly challenged by factors such as the unmanned vehicle itself and the external road environment. When the number of unmanned vehicles increases and it is difficult to maintain them regularly, it is necessary to repeatedly calibrate the relevant components in the unmanned vehicles. In addition, the road environment where the vehicle is traveling is complex, such as potholes, which will cause tracking errors. In summary, the above factors will cause the unmanned vehicle to fail to track the given speed and given acceleration. Summary of the invention

[0004] In order to solve the above technical problems, the present disclosure provides a vehicle longitudinal control method, device, electronic device and storage medium, which reduce the vehicle longitudinal speed tracking error by accurately controlling the vehicle longitudinally, so that the vehicle can smoothly track the desired longitudinal acceleration.

[0005] In a first aspect, an embodiment of the present disclosure provides a vehicle longitudinal control method, which is applied to a vehicle longitudinal control system, wherein the vehicle longitudinal control system includes a position outer loop system, a speed inner loop system, and a tracking differentiator, and the method includes:

[0006] Get the current position of the vehicle in the global coordinate system;

[0007] Generate a desired speed compensation control amount using the position outer loop system according to a current reference posture of the vehicle and the current posture, wherein the current reference posture is obtained from a reference trajectory issued by a planning system of the vehicle;

[0008] Using the tracking differentiator to filter the feedforward signal sent by the planning system;

[0009] The speed inner loop system is used to generate an acceleration control variable for longitudinally controlling the vehicle according to the filtered feedforward signal, the desired speed compensation control variable and the current longitudinal speed of the vehicle.

[0010] In a second aspect, an embodiment of the present disclosure provides a vehicle longitudinal control device, the device is configured in a vehicle longitudinal control system, the vehicle longitudinal control system includes a position outer loop system, a speed inner loop system and a tracking differentiator, the device includes:

[0011] The acquisition module is used to obtain the current position of the vehicle in the global coordinate system;

[0012] A first generating module, configured to generate a desired speed compensation control amount according to a current reference posture of the vehicle and the current posture by using the position outer loop system, wherein the current reference posture is obtained from a reference trajectory issued by a planning system of the vehicle;

[0013] A filtering module, used for filtering the feedforward signal sent by the planning system by using the tracking differentiator;

[0014] The second generating module is used to generate an acceleration control amount for longitudinally controlling the vehicle by using the speed inner loop system according to the filtered feedforward signal, the desired speed compensation control amount and the current longitudinal speed of the vehicle.

[0015] In a third aspect, an embodiment of the present disclosure provides an electronic device, including:

[0016] Memory;

[0017] Processor; and

[0018] Computer programs;

[0019] The computer program is stored in the memory and is configured to be executed by the processor to implement the above-mentioned vehicle longitudinal control method.

[0020] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above-mentioned vehicle longitudinal control method when executed by a processor.

[0021] The present disclosure relates to a vehicle longitudinal control method, device, electronic device and storage medium. The method is applied to a vehicle longitudinal control device, the device includes a position outer loop system, a speed inner loop system and a tracking differentiator, and the method includes: obtaining the current posture of the vehicle in the global coordinate system, using the position outer loop system to generate a desired speed compensation control amount according to the current reference posture and current posture of the vehicle, the current reference posture is obtained from the reference trajectory sent by the planning system of the vehicle, and then using the tracking differentiator to filter the feedforward signal sent by the planning system, and using the speed inner loop system to generate an acceleration control amount for longitudinal control of the vehicle according to the filtered feedforward signal, the desired speed compensation control amount and the current longitudinal speed of the vehicle. The method provided by the present disclosure reduces the longitudinal speed tracking error of the vehicle by accurately controlling the vehicle longitudinally, so that the vehicle can smoothly track the desired longitudinal acceleration. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0024] Figure 1 A schematic diagram of a vehicle longitudinal control system provided by an embodiment of the present disclosure;

[0025] Figure 2 A schematic flow chart of a vehicle longitudinal control method provided by an embodiment of the present disclosure;

[0026] Figure 3 A schematic diagram of a vehicle provided in an embodiment of the present disclosure;

[0027] Figure 4 A trajectory tracking schematic diagram provided for an embodiment of the present disclosure;

[0028] Figure 5 A schematic flow chart of another vehicle longitudinal control method provided by an embodiment of the present disclosure;

[0029] Figure 6 A schematic diagram of the structure of a vehicle longitudinal control device provided by an embodiment of the present disclosure;

[0030] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0031] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0032] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0033] At present, among the key technologies of unmanned driving, longitudinal control is one of the main functions of the unmanned driving control layer. Longitudinal control mainly refers to the real-time calculation of the longitudinal control quantity through the speed, acceleration output by the longitudinal planning and the speed, acceleration and other information fed back by the chassis to control the brakes and throttle of the vehicle (hereinafter referred to as the unmanned vehicle) so that the vehicle can automatically complete the desired tracking task. In longitudinal control, the control algorithm is the core, and its key performance such as tracking accuracy and ride comfort of the unmanned vehicle is particularly important, and it also determines the quality of the vehicle's completion of the unmanned driving task. However, affected by many factors such as the vehicle itself and the external road environment, the stability, accuracy and robustness of the control algorithm have been greatly challenged.

[0034] Specifically, when the number of unmanned vehicles increases and it is difficult to maintain them regularly, the acceleration-brake / throttle calibration table of the vehicle will need to be repeatedly calibrated due to the consistency of the vehicle, and with the aging of the vehicle and the replacement of the chassis firmware, the performance of the chassis will also change, and it also needs to be repeatedly calibrated. In addition, the road environment where the vehicle is traveling is complex. In potholes, uphill and downhill sections, speed bumps and other sections, the vehicle will be unable to climb, slide, and cannot brake downhill, and cannot cross speed bumps and potholes. Based on the above-mentioned many influencing factors, from the perspective of the longitudinal tracking results, it will cause the unmanned vehicle to track the given speed and given acceleration. Tracking control problems. In this case, if the tracking error cannot be effectively processed, there will be a great threat to the tracking performance and even driving safety of the vehicle. In response to the above-mentioned tracking control problem, the large amount of engineering parameter resetting and trial and error work used in the prior art can reduce the impact of the tracking control problem, but it wastes a lot of manpower and material resources, and is not suitable for the quantitative production and maintenance of unmanned vehicles. It is also impossible to ensure that the vehicle can stably and accurately track the given acceleration, resulting in a relatively low completion quality of the longitudinal tracking task.

[0035] In view of the above technical problems, the disclosed embodiment provides a vehicle longitudinal control method based on longitudinal speed planning for various complex situations: the vehicle starts slowly, cannot climb uphill, cannot cross speed bumps and potholes, etc., caused by factors such as inaccurate vehicle calibration table, actuator delay, and slope, speed bumps, potholes, etc. during the driving journey. The vehicle longitudinal control method can also be understood as a longitudinal double closed-loop compensation control method for unmanned vehicles, which can be used to improve the robustness of the basic control algorithm of the system. The longitudinal speed tracking error of the vehicle caused by the calibration error of the vehicle longitudinal transmission mechanism system, longitudinal slope and speed bumps is corrected, so that the unmanned vehicle can keep stable, accurate and robust tracking of the longitudinal desired speed (given speed) and desired acceleration (given acceleration) signals. First, the kinematic model of the position outer loop system and the dynamic model of the speed inner loop system are established; then, in the speed inner loop system, the total disturbance to the longitudinal dynamic model is modeled as a composite disturbance, and the composite disturbance is observed as an extended state of the speed inner loop system together with the vehicle's speed signal through an extended state observer, and compensated in the control channel; in the position outer loop system, a controller based on the kinematic model is designed. In addition, in order to avoid the oscillation of the expected reference signal sent by the planning system, a tracking differentiator is designed to perform filtering processing to reduce the jitter of the feedforward control quantity. Finally, the extended state observer is discretized based on the Euler method and applied to the embedded computer. The longitudinal dual closed-loop control method proposed in the present disclosure enables the final control of the vehicle to have better robustness when facing a complex environment. It is described in detail through one or more of the following embodiments.

[0036] Figure 1 A schematic diagram of a vehicle longitudinal control system provided in an embodiment of the present disclosure is provided, wherein the vehicle longitudinal control system 100 includes a position outer loop system 110, a velocity inner loop system 120 and a tracking differentiator 130, wherein the position outer loop system 110 includes a kinematic model 111 and a kinematic controller 112, the velocity inner loop system 120 includes a dynamic controller 121, a dynamic model 122 and an extended state observer 123, and the tracking differentiator 130 includes two filters, which are respectively denoted as a first filter 131 and a second filter 132.

[0037] Figure 2 A flow chart of a vehicle longitudinal control method provided in an embodiment of the present disclosure is applied to a vehicle longitudinal control system 100. The vehicle longitudinal control system 100 can be configured on a server or a terminal, and specifically includes the following steps: Figure 2 The following steps S210 to S240 are shown:

[0038] It is understandable that before implementing the vehicle longitudinal control method, it is necessary to establish a dual closed-loop system model of an unmanned vehicle with compound disturbances, that is, to establish the position outer loop system 110 and the speed inner loop system 120 constituting the vehicle longitudinal control system 100 .

[0039] Specifically, a kinematic model of the position outer loop system 110 is established to obtain the position and posture of the vehicle in the global coordinate system and the local coordinate system. For example, see Figure 3 , Figure 3 A schematic diagram of a vehicle provided in an embodiment of the present disclosure, Figure 3 The vehicle is shown in two Cartesian coordinate systems, one is the global coordinate system (X, O, Y) and the other is the local coordinate system (x, o, y). Figure 3 The geometric relationship shown can be used to obtain the formula (1) shown below. Formula (1) can be understood as the calculation formula of the kinematic model 111.

[0040]

[0041] In the formula, q(t)=[x(t)y(t)θ(t)] T represents the position and orientation of the vehicle at time t in the global coordinate system. u(t)=[v(t)ω(t)] represents the vehicle control input vector, where the longitudinal velocity of the vehicle at time t is v(t) and the angular velocity is ω(t)= δ f (t) front wheel deflection angle, vehicle wheelbase is l fr .

[0042] S210: Obtain the current position and posture of the vehicle in the global coordinate system.

[0043] It can be understood that the above-mentioned kinematic model 111 is used to generate the current posture of the vehicle in the global coordinate system XOY according to the current reference posture and speed control value (longitudinal speed). The current posture is recorded as the above-mentioned q(t), wherein the current reference posture is obtained from the reference trajectory issued by the planning system of the vehicle. The planning system of the vehicle is used to provide the controller of the vehicle with expected information (expected signal). The expected information can be expected posture information, speed information and acceleration information. The expected information can also be understood as reference information, which is the information to be tracked by the vehicle.

[0044] S220: Generate a desired speed compensation control amount using the position outer loop system according to the current reference posture and the current posture of the vehicle.

[0045] It can be understood that on the basis of the above S210, the pre-constructed position outer loop system 110 is used to generate the expected speed compensation control amount according to the current reference posture issued by the planning system and the current posture obtained, that is, the speed compensation control amount is calculated by the error between the expected posture and the actual posture to reduce the speed tracking error.

[0046] Optionally, the position outer loop system 110 further includes a first operator 113 .

[0047] Optionally, generating the desired speed compensation control amount in the above S220 can be specifically implemented through the following steps:

[0048] A first operator is used to obtain an error state vector of the vehicle in a local coordinate system according to the current reference posture and the current posture of the vehicle, wherein the local coordinate system is a coordinate system constructed with the center point of the rear axle of the vehicle as the origin.

[0049] The kinematic controller is used to generate a desired velocity compensation control amount according to the error state vector.

[0050] It can be understood that in order to achieve trajectory tracking of the vehicle, the first operator 113 is used to calculate the difference between the current reference posture and the current posture of the vehicle to obtain the error state vector of the vehicle in the local coordinate system xoy.

[0051] For example, see Figure 4 , Figure 4 A trajectory tracking schematic diagram provided in an embodiment of the present disclosure, Figure 4 Including vehicle 410 and tracking trajectory 420 issued by the planning system, according to Figure 4 The geometric relationship shown in Figure 2 gives formula (2):

[0052]

[0053] In the formula, q e (t) = [x e (t)y e (t)θ e (t)] T represents the state tracking error vector (error state vector), q r (t) = [x r (t)y r (t)θ r (t)] T represents the position and direction of the vehicle at time t in the reference trajectory signal (tracking trajectory 420) issued by the planning system, and the position is recorded as [x r (t)y r (t)], and the direction is denoted by θ r (t),u r(t) = [v r (t)ω r (t)] T represents the expected tracking signal issued by the planning system, where the local coordinate system (x, o, y) is a coordinate system constructed with the center point of the rear axle of the vehicle as the origin. Figure 4 As shown in 430.

[0054] It can be understood that after formula (2) is obtained, the first-order derivative of formula (2) is calculated to obtain formula (3). Formula (3) can be understood as the calculation formula of the first operator 113. Based on formula (3), the current reference posture, current posture and longitudinal speed of the vehicle are calculated to obtain the error state vector q of the vehicle in the local coordinate system. e (t).

[0055]

[0056] It is understandable that for the longitudinal tracking of the vehicle, only the above formula (3) can be considered. The kinematic controller 112 can then be used according to Get the desired speed compensation control amount.

[0057] It is understandable that after the error state vector is determined, the kinematic controller 112 is used to generate the desired speed compensation control amount according to the error state vector. Exemplarily, the formula for calculating the desired speed compensation control amount by the kinematic controller 112 is shown in formula (4):

[0058] v c (k) = f out (q e (k)) (4)

[0059] In the formula, v c (k) represents the desired speed compensation control amount at time k, q e (k) represents the error state vector at time k, and f(*) represents any control law that can make the system converge.

[0060] S230: Utilize the tracking differentiator to filter the feedforward signal sent by the planning system.

[0061] Wherein, the feedforward signal includes a feedforward velocity signal and a feedforward acceleration signal.

[0062] It is understandable that, based on the above S220, since the expected reference signal may not be smooth, if it is directly transmitted to the chassis actuator of the vehicle as a feedforward control quantity, it may cause the oscillation of the unmanned driving system, the lower high pressure and the system divergence and other problems. Therefore, the tracking differentiator 130 is designed to filter the feedforward signal issued by the planning system. The feedforward signal is the expected reference signal, including the expected reference speed (feedforward speed signal) and the expected reference acceleration (feedforward acceleration signal). The tracking differentiator 130 is designed to filter and smooth the expected reference signal, which can effectively avoid the chassis wear and high pressure caused by the oscillation of the expected reference signal issued by the plan, and can also reduce the sudden acceleration, point braking, emergency braking and other phenomena that occur in the vehicle during operation. After the tracking differentiator 130 is filtered, the saturation problem of the unmanned driving system is also considered, and the upper and lower limits of the control quantity are limited, which to a certain extent avoids the divergence and uncontrollability of the unmanned driving system.

[0063] Specifically, the first filter 131 is used to filter the expected reference speed, and the second filter 132 is used to filter the expected reference acceleration. The entire filtering process adopts the form of "differentiation first, then prediction". The filtering process of the expected reference speed by the first filter 131 is described as an example. The calculation formula of the first filter 131 is shown in formula (5):

[0064]

[0065] In the formula, v ref is the desired reference speed, v ref =x0, x0 is the input signal, x1 and x2 are the output signals, k represents the current moment, k+1 represents the next moment, signal x1 tracks x0 after filtering, signal x2 is used as the differential information of x0, h is the sampling period of the system, r is the fast factor of the tracking differentiator 130, h1 is the prediction time, r and h1 are adjustable parameters of the tracking differentiator 130, preferably, the prediction time h1 is 1 to 1.5 times of the filtering factor h0, wherein the larger the variable r is, the faster the tracking speed is, and the larger the variable h0 is, the better the filtering effect is, fhan(x1,x2,r,h) represents the fastest control comprehensive function, That is the expected reference speed after filtering Similarly, the second filter 132 obtains the filtered expected reference acceleration The formula is the same as the above formula (5) and will not be repeated here.

[0066] Among them, the fastest control comprehensive function fhan(x1,x2,r,h) is shown in formula (6):

[0067]

[0068] In the formula, fsg(y,c)=(sign(x+c)-sign(xc)) / 2.

[0069] S240: Generate an acceleration control variable for longitudinally controlling the vehicle using the speed inner loop system according to the filtered feedforward signal, the desired speed compensation control variable, and the current longitudinal speed of the vehicle.

[0070] It can be understood that, based on the above S220 and S230, the speed inner loop system 120 is used to generate an acceleration control amount for longitudinal control of the vehicle according to the filtered feedforward signal, the desired speed compensation control amount and the current longitudinal speed of the vehicle, wherein the current longitudinal speed of the vehicle can be understood as the speed of the actual vehicle, that is, the real-time speed issued by the chassis system.

[0071] After generating the acceleration control amount, the vehicle longitudinal control system 100 also generates the speed control amount, and calculates the posture at the next moment in real time according to the speed control amount, which is specifically achieved through the following steps:

[0072] The dynamic model is used to generate a speed control variable according to the acceleration control variable and an external disturbance affecting a driving system of the vehicle.

[0073] The kinematic model is used to generate the next moment posture of the vehicle in the global coordinate system according to the current reference posture and the speed control amount.

[0074] It can be understood that in the speed inner loop system, for internal and external factors such as potholes, ups and downs, mechanical wear, etc., the above external factors are modeled as compound disturbances, and combined with the vehicle dynamics model, an unmanned vehicle dynamics model 122 with compound disturbances is established. Compared with the traditional vehicle dynamics model, the dynamics model 122 injects the influence of compound disturbances, further improving the robustness of the final control.

[0075] It can be understood that after the acceleration control amount is generated, the dynamic model 122 is used to calculate the sum of the acceleration control amount and the external disturbance affecting the driving system to obtain the speed control amount, wherein the external disturbance affecting the driving system can be understood as a composite disturbance, including unmodeled dynamics and various external disturbances. The dynamic model 122 is a model with disturbances, as shown in the following formula (7):

[0076] v(t)=a(t)+d(t) (7)

[0077] In the formula, v(t) represents the longitudinal velocity control variable, a(t) represents the longitudinal acceleration control variable, d(t) represents the external disturbance (compound disturbance), and d(t) and its derivative are both bounded.

[0078] It can be understood that the above formula (3) is about The formula of (t) and formula (7) can be regarded as the longitudinal system model of the vehicle.

[0079] The embodiment of the present disclosure provides a vehicle longitudinal control method, which models internal and external factors such as potholes, ups and downs, and mechanical wear as composite disturbances, establishes an unmanned vehicle dynamics model with composite disturbances, and designs an extended state observer based on the model to observe the composite disturbances. It is not limited to any specific disturbance and can process all disturbances with bounded derivatives, effectively avoiding the repetitive work of map fixed-point marking, regular vehicle maintenance, etc. in traditional technologies. Secondly, in the calculation process, in order to avoid the oscillation of the expected reference signal sent by the planning system, the expected reference signal is filtered and smoothed by tracking the differentiator, which to a certain extent avoids the phenomenon of sudden acceleration, point braking, emergency braking, and high pressure during driving. Finally, the method provided by the present disclosure has the ability to generalize in different scenarios, which can not only solve a single disturbance but also resist the adverse effects of multiple disturbances acting at the same time.

[0080] On the basis of the above embodiment, optionally, the speed inner loop system is used to generate an acceleration control amount for longitudinally controlling the vehicle according to the filtered feedforward signal, the desired speed compensation control amount and the current longitudinal speed of the vehicle, specifically including the following: Figure 5 Steps S510 to S520 shown:

[0081] It can be understood that the extended state observer 123 is constructed by the following method:

[0082] An initial model is constructed based on a longitudinal velocity of a vehicle and a composite disturbance of a driving system of the vehicle.

[0083] Based on the initial model, an extended state observer for observing the longitudinal velocity and the composite disturbance is established.

[0084] The extended state observer is discretized to obtain a constructed extended state observer.

[0085] It is understandable that for the speed inner loop system 120, an extended state observer 123 is designed to observe and compensate for the vehicle's speed and composite disturbance. Since the vehicle's speed transmitted by the chassis has certain noise during the actual driving process, the state of the unmanned driving system may not be completely accurate. On the other hand, due to the influence of the composite disturbance, the unmanned driving system cannot complete the expected speed signal tracking action. The composite disturbance may be caused by modeling errors, slopes, potholes, speed bumps, etc. Therefore, based on the above problems, an extended state observer 123 is designed to observe the composite disturbance as an extended state of the unmanned driving system together with the longitudinal speed signal.

[0086] Specifically, the composite disturbance is taken as the expanded state x2 of the unmanned driving system, the longitudinal speed v=x1, the composite disturbance d=x2, and a=τ. The dynamic model is rewritten as formula (8). Formula (8) can be understood as the initial model, that is, the initial model is constructed based on the longitudinal speed v of the vehicle and the composite disturbance d of the vehicle driving system.

[0087]

[0088] Where the variable ε is the derivative of x2. Assuming that the composite disturbance d and its first-order derivative are bounded, ε is also bounded.

[0089] Optionally, based on the initial model shown in formula (8), an extended state observer (ESO) for observing the longitudinal velocity and the composite disturbance is established. Through the second-order extended state observer 123, the observed values ​​of the internal state of the unmanned driving system and the composite disturbance can be obtained, and the compensation control amount can be calculated according to the observed values. The calculation formula of the extended state observer 123 is shown in the following formula (9):

[0090]

[0091] Where z1 and z2 are the observed values ​​of x1 and x2 respectively, z1 represents the observed value of the longitudinal velocity, z2 represents the observed value of the composite disturbance, the error e = z1-x1, β1 and β2 are two positive adjustable gain parameters. fal(e,σ,δ) is a nonlinear non-smooth function, and its specific form is shown in formula (10):

[0092]

[0093] Wherein, σ and δ are related parameters of the function f(e, σ, δ), preferably σ=0.5, δ=h, and h is the sampling period.

[0094] Optionally, in the process of constructing the extended state observer, a controller bandwidth is determined, and a plurality of gain parameters in the constructed extended state observer are adjusted according to the controller bandwidth.

[0095] It is understandable that in the process of constructing the extended state observer, a parameter tuning strategy is formulated. In actual engineering applications, the gain parameters of the extended state observer formula (9) need to be tuned. The actual parameter tuning is performed according to the existing engineering method. Preferably, the engineering method can be the bandwidth method, in which the gain parameters β1 and β2 satisfy the relationship shown in formula (11):

[0096]

[0097] In the formula, ω c For the controller bandwidth, in practical applications, only the controller bandwidth needs to be adjusted to adjust the gain parameter, which greatly improves the efficiency of parameter setting.

[0098] It is understandable that in terms of parameter setting, the "bandwidth method" can be used in the gain parameters of the extended state observer 123 to achieve dimensionality reduction of the adjustable parameters, which can greatly improve the parameter setting efficiency. In terms of parameter setting of the PID controller, the Ziegler-Nichols parameter setting method can be used, which can determine more accurate parameters for the PID controller. After the above two adjustable parameters are set, only fine-tuning is required to achieve the expected effect, reducing the work of repeated parameter adjustment.

[0099] It is understandable that after obtaining the extended state observer shown in formula (9), the extended state observer needs to be discretized before it can be applied in the embedded system. Considering the computational efficiency of the unmanned driving system after discretization, the extended state observer can be discretized by the Euler method, but it is not limited to this method. The bilinear discretization method can also be applied. The specific discretization method is not limited. After discretizing formula (9), the obtained extended state observer is shown in formula (12):

[0100]

[0101] In the formula, k represents the current moment, k+1 represents the next moment, and h is the sampling period.

[0102] S510: Observe the current longitudinal velocity and the composite disturbance of the vehicle using the extended state observer, and generate an observation value of the current longitudinal velocity and an observation value of the composite disturbance.

[0103] It can be understood that the constructed extended state observer 123 is used to observe the current longitudinal speed and the composite disturbance of the vehicle to generate the observation value z1 of the current longitudinal speed and the observation value z2 of the composite disturbance.

[0104] S520, generating an acceleration control variable for longitudinally controlling the vehicle according to the filtered feedforward signal, the desired speed compensation control variable, the observed value of the longitudinal speed, and the observed value of the composite disturbance.

[0105] Optionally, the above S520 generates the acceleration control amount, which can be implemented by the following steps:

[0106] The dynamic controller generates an acceleration compensation control variable based on a speed error obtained from the filtered feedforward speed signal, the desired speed compensation control variable and the observed value of the longitudinal speed.

[0107] An acceleration control variable is generated according to the filtered feedforward acceleration signal, the acceleration compensation control variable and the observation value of the composite disturbance.

[0108] The vehicle longitudinal control system 100 further includes a second operator 124 . The second operator 124 may be directly configured in the vehicle longitudinal control system 100 , or may be configured in the position outer loop system 110 or the speed inner loop system 120 .

[0109] It can be understood that the second operator 124 is used to calculate the sum of the filtered feedforward velocity signal output by the first filter 131 and the desired velocity compensation control amount output by the kinematic controller 112, and calculate the difference between the sum and the observed value of the longitudinal velocity output by the expanded state observer 123 to obtain the velocity error, see the following formula (13) for details.

[0110]

[0111] Where e1(k) represents the speed error, represents the filtered feedforward velocity signal, v c represents the desired speed compensation control amount, and z1 represents the observed value of the longitudinal speed.

[0112] Optionally, after obtaining the speed error, the speed error is controlled by the dynamic controller 121 to obtain the control amount of the nominal system, wherein the dynamic controller 121 can also be called a feedback controller, and the calculation formula in the dynamic controller 121 is shown in formula (14):

[0113] u normal (k) = f inner (e1(k)) (14)

[0114] In the formula, u normal (k) represents the control quantity of the nominal system, and f(*) is any control law that can converge to the nominal system, including but not limited to pure tracking method, Proportional-Integral-Differentiation (PID) method, Model Predictive Control (MPC) method, etc. Preferably, a PID controller can be used to generate the control quantity of the nominal system.

[0115] It can be understood that when the dynamics controller 121 is a PID controller, the parameter tuning method of the PID in the PID controller can use the Ziegler-Nichols parameter tuning method, wherein the PID parameters are as shown in Table 1.

[0116] Table 1:

[0117]

[0118] Among them, the PID parameters are described as follows: K c : k when only proportional control is used and the steady-state error of the control system is minimized p Value. c : The oscillation period of the control system when only proportional control is used. T i : Integration time of the control system. T d : Differential time of the control system. T: PID control sampling calculation cycle. k p ,k i ,k d :The parameter to be adjusted. Through step response and numerical analysis, the appropriate K can be obtained c Value and P c Value, and finally look up the table to get k p ,k i ,k d Gain.

[0119] Optionally, the vehicle longitudinal control system 100 further includes a third operator 125. The third operator 125 may be directly configured in the vehicle longitudinal control system 100 or in the speed inner loop system 120. The specific configuration position is not limited.

[0120] It can be understood that the third operator 125 calculates the sum of the control amount of the nominal system output by the dynamic controller 121, the filtered feedforward acceleration signal output by the second filter 132, and the composite disturbance compensation control amount output by the extended state observer 123 to obtain the control amount of the inner loop system acting on the vehicle, that is, the acceleration control amount. The formula for calculating the acceleration control amount by the third operator 125 is shown in formulas (15) and (16):

[0121] a c (k) = u normal (k) (15)

[0122]

[0123] In the formula, a(k) represents the acceleration control amount, represents the filtered feedforward acceleration signal, u normal (k) and a c (k) all represent the control quantity of the nominal system, a c (k)-z2(k) represents the acceleration compensation control amount, -z2(k) represents the compound disturbance compensation control amount, where z2(k) is the observed value of the compound disturbance. It can be understood that adding the filtered feedforward acceleration signal can improve the response speed of the unmanned driving system.

[0124] It is understandable that, considering the saturation problem of the unmanned driving system, the acceleration control amount output by formula (16) is saturated using a saturation function, and the saturated acceleration control amount is transmitted to the chassis, as shown in formula (17):

[0125] a'(k)=sat(a(k)) (17)

[0126] Where a'(k) represents the acceleration control amount after saturation processing, and sat(*) represents the saturation function.

[0127] It can be understood that after obtaining the acceleration control quantity, the dynamic model 122 is used to obtain the speed control quantity of the vehicle. The specific generation process is not repeated here. The acceleration control quantity and the speed control quantity are used as the control quantities of the vehicle. Subsequently, the control quantity of the vehicle is sent to the underlying system of the vehicle through the communication network to complete the actual action of the vehicle.

[0128] The disclosed embodiment provides a vehicle longitudinal control method, which observes the internal state and compound disturbance of the unmanned driving system based on an extended state observer, and uses the observed value of the compound disturbance as an automatic compensation control quantity to resist the adverse effects of internal and external factors such as potholes, ups and downs, and mechanical structure wear. Compared with traditional operations such as map fixed-point marking and regular vehicle maintenance that require a lot of manpower and material resources, it is more automated. Secondly, the disturbance is compensated in the designed dynamic controller, so that the vehicle has a strong anti-interference ability and can continuously and stably track the desired trajectory.

[0129] Figure 6 The vehicle longitudinal control device provided by the embodiment of the present disclosure can execute the processing flow provided by the vehicle longitudinal control method embodiment, such as Figure 6 As shown, the vehicle longitudinal control device 600 is configured in the vehicle longitudinal control system, the vehicle longitudinal control system includes a position outer loop system, a speed inner loop system and a tracking differentiator, and the vehicle longitudinal control device 600 includes an acquisition module 610, a first generation module 620, a filtering module 630 and a second generation module 640, wherein:

[0130] An acquisition module 610 is used to acquire the current position and posture of the vehicle in the global coordinate system;

[0131] A first generating module 620, configured to generate a desired speed compensation control amount according to a current reference posture of the vehicle and the current posture by using the position outer loop system, wherein the current reference posture is obtained from a reference trajectory issued by a planning system of the vehicle;

[0132] A filtering module 630, configured to filter the feedforward signal sent by the planning system using the tracking differentiator;

[0133] The second generating module 640 is used to generate an acceleration control variable for longitudinally controlling the vehicle by using the speed inner loop system according to the filtered feedforward signal, the desired speed compensation control variable and the current longitudinal speed of the vehicle.

[0134] Optionally, the position outer loop system includes a first operator and a kinematic controller.

[0135] Optionally, the first generating module 620 is used to:

[0136] Obtaining, by a first operator, an error state vector of the vehicle in a local coordinate system according to the current reference posture and the current posture of the vehicle, wherein the local coordinate system is a coordinate system constructed with the center point of the rear axle of the vehicle as the origin;

[0137] The kinematic controller is used to generate a desired velocity compensation control amount according to the error state vector.

[0138] Optionally, the speed inner loop system includes an extended state observer.

[0139] Optionally, the second generating module 640 is used for:

[0140] Observing the current longitudinal velocity and the composite disturbance of the vehicle by using the extended state observer, and generating an observation value of the current longitudinal velocity and an observation value of the composite disturbance;

[0141] An acceleration control variable for longitudinally controlling the vehicle is generated based on the filtered feedforward signal, the desired speed compensation control variable, the observed value of the longitudinal speed, and the observed value of the composite disturbance.

[0142] Optionally, the speed inner loop system further includes a dynamic controller, and the feedforward signal includes a feedforward speed signal and a feedforward acceleration signal.

[0143] Optionally, the second generating module 640 is used for:

[0144] generating an acceleration compensation control variable using the dynamic controller according to a speed error obtained from the filtered feedforward speed signal, the desired speed compensation control variable and the observed value of the longitudinal speed;

[0145] An acceleration control variable is generated according to the filtered feedforward acceleration signal, the acceleration compensation control variable and the observation value of the composite disturbance.

[0146] Optionally, the speed inner loop system also includes a dynamic model, and the position outer loop system also includes a kinematic model.

[0147] Optionally, after generating the acceleration control amount, the device 600 is further used to:

[0148] generating a speed control variable using the dynamic model according to the acceleration control variable and an external disturbance affecting a driving system of the vehicle;

[0149] The kinematic model is used to generate the next moment posture of the vehicle in the global coordinate system according to the current reference posture and the speed control amount.

[0150] Optionally, the extended state observer in the device 600 is constructed by the following method:

[0151] constructing an initial model based on a longitudinal velocity of a vehicle and a composite disturbance of a driving system of the vehicle;

[0152] Based on the initial model, establishing an extended state observer for observing the longitudinal velocity and the composite disturbance;

[0153] The extended state observer is discretized to obtain a constructed extended state observer.

[0154] Optionally, during the process of constructing the extended state observer, the device 600 is further used for:

[0155] A controller bandwidth is determined, and a plurality of gain parameters in the constructed extended state observer are tuned according to the controller bandwidth.

[0156] Figure 6 The vehicle longitudinal control device of the illustrated embodiment can be used to implement the technical solution of the above-mentioned method embodiment, and its implementation principle and technical effect are similar and will not be described in detail here.

[0157] Figure 7 The electronic device provided by the embodiment of the present disclosure may be a server or a machine as described above. The electronic device provided by the embodiment of the present disclosure may execute the processing flow provided by the above embodiment, such as Figure 7 As shown, the electronic device 700 includes: a processor 710, a communication interface 720 and a memory 730; wherein the computer program is stored in the memory 730 and is configured so that the processor 710 executes the above-mentioned vehicle longitudinal control method.

[0158] In addition, an embodiment of the present disclosure further provides a computer-readable storage medium on which a computer program is stored. The computer program is executed by a processor to implement the vehicle longitudinal control method described in the above embodiment.

[0159] In addition, an embodiment of the present disclosure further provides a computer program product, which includes a computer program or instructions, and when the computer program or instructions are executed by a processor, the vehicle longitudinal control method as described above is implemented.

[0160] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0161] The above description is only a specific embodiment of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A vehicle longitudinal control method, characterized in that: Applied to a vehicle longitudinal control system, the vehicle longitudinal control system includes a position outer loop system, a speed inner loop system and a tracking differentiator, the method includes: Acquire the current posture of the vehicle in the global coordinate system; wherein the current posture is generated by the position outer loop system according to the current reference posture and the speed control amount, and the speed control amount is determined by the speed inner loop system according to the previous acceleration control amount; Generate a desired speed compensation control amount using the position outer loop system according to a current reference posture of the vehicle and the current posture, wherein the current reference posture is obtained from a reference trajectory issued by a planning system of the vehicle; Using the tracking differentiator to filter the feedforward signal sent by the planning system; The speed inner loop system is used to generate an acceleration control variable for longitudinally controlling the vehicle according to the filtered feedforward signal, the desired speed compensation control variable and the current longitudinal speed of the vehicle.

2. The method according to claim 1, characterized in that The position outer loop system includes a first operator and a kinematic controller. The outer position loop system is used to generate a desired speed compensation control amount according to the current reference posture and the current posture of the vehicle, including: Obtaining, by a first operator, an error state vector of the vehicle in a local coordinate system according to the current reference posture and the current posture of the vehicle, wherein the local coordinate system is a coordinate system constructed with the center point of the rear axle of the vehicle as the origin; The kinematic controller is used to generate a desired velocity compensation control amount according to the error state vector.

3. The method according to claim 1, characterized in that The speed inner loop system includes an extended state observer, The speed inner loop system generates an acceleration control amount for longitudinally controlling the vehicle according to the filtered feedforward signal, the desired speed compensation control amount, and the current longitudinal speed of the vehicle, including: Observing the current longitudinal velocity and the composite disturbance of the vehicle by using the extended state observer, and generating an observation value of the current longitudinal velocity and an observation value of the composite disturbance; An acceleration control variable for longitudinally controlling the vehicle is generated based on the filtered feedforward signal, the desired speed compensation control variable, the observed value of the longitudinal speed, and the observed value of the composite disturbance.

4. The method according to claim 3, characterized in that The speed inner loop system further includes a dynamic controller, and the feedforward signal includes a feedforward speed signal and a feedforward acceleration signal. The step of generating an acceleration control amount for longitudinally controlling the vehicle according to the filtered feedforward signal, the desired speed compensation control amount, the observed value of the longitudinal speed, and the observed value of the composite disturbance comprises: generating an acceleration compensation control variable using the dynamic controller according to a speed error obtained from the filtered feedforward speed signal, the desired speed compensation control variable and the observed value of the longitudinal speed; An acceleration control variable is generated according to the filtered feedforward acceleration signal, the acceleration compensation control variable and the observation value of the composite disturbance.

5. The method according to claim 1, characterized in that The speed inner loop system also includes a dynamic model, and the position outer loop system also includes a kinematic model. After generating the acceleration control amount, the method further includes: generating a speed control variable using the dynamic model according to the acceleration control variable and an external disturbance affecting a driving system of the vehicle; The kinematic model is used to generate the next moment posture of the vehicle in the global coordinate system according to the current reference posture and the speed control amount.

6. The method according to claim 3, characterized in that: The extended state observer is constructed by the following method: constructing an initial model based on a longitudinal velocity of a vehicle and a composite disturbance of a driving system of the vehicle; Based on the initial model, establishing an extended state observer for observing the longitudinal velocity and the composite disturbance; The extended state observer is discretized to obtain a constructed extended state observer.

7. The method according to claim 6, characterized in that In the process of constructing the extended state observer, the method further includes: A controller bandwidth is determined, and a plurality of gain parameters in the constructed extended state observer are tuned according to the controller bandwidth.

8. A vehicle longitudinal control device, characterized in that: The device is configured in a vehicle longitudinal control system, the vehicle longitudinal control system includes a position outer loop system, a speed inner loop system and a tracking differentiator, and the device includes: An acquisition module, used to acquire the current posture of the vehicle in the global coordinate system; wherein the current posture is generated by the position outer loop system according to the current reference posture and the speed control amount, and the speed control amount is determined by the speed inner loop system according to the previous acceleration control amount; A first generating module, configured to generate a desired speed compensation control amount according to a current reference posture of the vehicle and the current posture by using the position outer loop system, wherein the current reference posture is obtained from a reference trajectory issued by a planning system of the vehicle; A filtering module, used for filtering the feedforward signal sent by the planning system by using the tracking differentiator; The second generating module is used to generate an acceleration control amount for longitudinally controlling the vehicle by using the speed inner loop system according to the filtered feedforward signal, the desired speed compensation control amount and the current longitudinal speed of the vehicle.

9. An electronic device, characterized in that: include: Memory; processor; as well as Computer programs; The computer program is stored in the memory and is configured to be executed by the processor to implement the vehicle longitudinal control method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the vehicle longitudinal control method according to any one of claims 1 to 7 are implemented.

Images