An automatic driving control method that takes into account the convergence of lateral error and heading error

Abstract

The present invention relates to an autonomous driving control method that balances the convergence of lateral and heading errors. The method comprises the following steps: trajectory predictor design; error blender design; linear delay design; linear estimator design; and linear feedback controller design. This method effectively addresses interference such as model parameter uncertainty and employs a Lyapunov function to design an error blender, verifying the stability of the autonomous driving control system. The method demonstrates high robustness in the presence of model parameter uncertainty. By accelerating the response of the front wheel angle, tracking error can be reduced, achieving synchronous convergence of lateral and heading errors. This results in a more accurate and reliable tracked trajectory, improving the smoothness and safety of autonomous vehicle operation.

Description

Technical Field

[0001] The present invention relates to an autonomous driving motion control method, and in particular to an autonomous driving control method that takes into account the convergence of lateral error and heading error. Background Art

[0002] Autonomous driving technology is booming. Its implementation can enhance road safety management, significantly alleviate traffic pressure, save energy, and reduce emissions. The current development of autonomous driving technology is also benefiting from interdisciplinary research, including biology, materials, transportation, intelligence, telecommunications, imaging, and sensors. As one of the core technologies for autonomous driving, motion control methods are increasingly being researched. Currently, the design of motion controllers is quite difficult due to the uncertainty and time-varying nature of system model parameters and the underactuated nature of the system.

[0003] Therefore, how to provide a method to resist uncertainty and simultaneously make the lateral error and heading error converge synchronously is a problem that needs to be solved urgently by those skilled in the art. Summary of the Invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide an automatic driving control method that takes into account the convergence of lateral error and heading error.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] An automatic driving control method that takes into account the convergence of lateral error and heading error includes the following steps:

[0007] 1) The trajectory previewer calculates the trajectory tracking error;

[0008] 2) The error fusion unit fuses the trajectory tracking error into a comprehensive tracking error;

[0009] 3) Design a differential error system model based on the comprehensive tracking error;

[0010] 4) Design a linear estimator based on the differential error system model to estimate the system state;

[0011] 5) The linear delay device calculates the desired system state;

[0012] 6) Calculate the system error based on the estimated system state and the expected system state;

[0013] 7) Based on the system error, the linear feedback controller calculates the desired control quantity;

[0014] 8) Calculating the extended state control quantity based on the estimated system state;

[0015] 9) The desired control quantity and the expanded state control quantity are integrated to obtain the final system control quantity.

[0016] The step 1) comprises the following steps:

[0017] 11) Calculate vehicle heading error The calculation method is Where θ is the heading of the tracked point on the reference trajectory, The actual heading of the reference vehicle;

[0018] 12) Vehicle lateral error y e The track previewer is designed to Among them, y e is the vehicle lateral error, v d is the expected speed of the tracked point on the reference trajectory, is the vehicle heading error, l s is the distance from the preview point to the center of mass of the vehicle, L is the wheelbase, δ is the front wheel turning angle of the vehicle, y e The dots on the parameters are the first-order differentials of the parameters.

[0019] In step 2), the error fusion device is designed as Where z is the comprehensive tracking error, tanh(.) is the hyperbolic tangent function, and the system parameters c0>0, c1>0, c2>0.

[0020] In step 3), the differential error system model designed by the error fusion model is

[0021] in, is the uncertain part in the differential error system model, coefficient b0=-c2v d / L, u is the final system control quantity, and the dot on the z parameter is the first-order differential of z.

[0022] The dots on the f(.) function are the first-order differentials of the f(.) function, v is the actual speed of the vehicle, and w is the system disturbance.

[0023] In step 4), the state space model of the designed linear estimator is shown in formula (1),

[0024]

[0025] Among them, state z1 is the estimated value of z, state z2 is the uncertain part in the differential error system model, η is the output of the linear estimator, the circle on the z1 parameter is the first-order differential of z1, and the circle on the z2 parameter is the first-order differential of z2.

[0026] In step 4), the designed linear estimator is discretized as shown in formula (2):

[0027]

[0028] Among them, k is the discrete moment, T1 is the period of the linear estimator, and w o >0 is the gain of the linear estimator, e z (k) is the error of the linear estimator after discretization at the k-th moment; z1(k) is the value of parameter z1 at the k-th moment; z(k) is the value of parameter z at the k-th moment; z1(k+1) is the value of parameter z1 at the k+1-th moment; z2(k) is the value of parameter z2 at the k-th moment; u(k) is the value of the final system control quantity u at the k-th moment; z2(k+1) is the value of parameter z2 at the k+1-th moment.

[0029] In step 5), the designed linear delay device is first-order, the reference input r of the linear delay device is always 0, and the output v1 thereof is also always 0.

[0030] In step 6), the system error is calculated as e1 = v1 - z1.

[0031] In step 7), the linear feedback controller is designed as a proportional controller, and the proportional controller gain is w c / b0,

[0032] The linear feedback controller output u0 is calculated as u0 = -w c z1 / b0, parameter w c >0 is the gain coefficient of the linear estimator.

[0033] In step 8), the expansion state control amount is calculated as u1=z2 / b0.

[0034] In step 9), the final system control quantity calculation method is u=u0+u1.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] 1. The error fusion device of the present invention can fuse the lateral error and heading error into a single error pattern, transforming the system into a fully driven system. This greatly reduces the difficulty of controller design and facilitates stable convergence of the control system.

[0037] 2. The autonomous driving control method of the present invention does not rely on an accurate vehicle dynamics model, can overcome disturbances caused by model uncertainty, greatly simplify the design of controller parameters, quickly stabilize the system, improve system robustness, and achieve efficient control operation;

[0038] 3. The automatic driving control strategy of the present invention, proven by the Lyapunov function, can achieve synchronous convergence of lateral error and heading error, has good stability, and realizes robust operation of the system;

[0039] 4. The trajectory tracking method proposed in the present invention can reduce tracking errors by accelerating the response of the front wheel angle, making the tracked trajectory more accurate and reliable, thereby improving the stability and safety of the autonomous vehicle operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of an automatic driving control method for balancing lateral error and heading error convergence according to the present invention;

[0041] Figure 2 Schematic diagram of the trajectory tracking model of the present invention;

[0042] Figure 3 Schematic diagram of trajectory tracking using a pure tracking method and the method of the present invention;

[0043] Figure 4 Schematic diagram of the turning angle of the pure tracking method and the trajectory tracking method of the present invention;

[0044] Figure 5 Schematic diagram of the error between the pure tracking method and the trajectory tracking method of the present invention. DETAILED DESCRIPTION

[0045] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0046] Example:

[0047] like Figure 1 As shown, Figure 1 This is a schematic diagram of an automatic driving control method of the present invention that takes into account the convergence of lateral error and heading error. The specific steps of this strategy include:

[0048] Step 1: If Figure 2 As shown, a reference trajectory tracking model is established. local xy is the local coordinate system, OglobalXY is the absolute coordinate system, P preview is the preview point, P track is the tracked point on the reference trajectory, v d is the vehicle speed, v y is the lateral speed, v x is the longitudinal speed, l s Preview distance, β is the vehicle's center of mass side slip angle, is the heading of the vehicle, θ is the heading angle of the tracked point, δ is the front wheel turning angle, y e is the lateral error, is the heading error;

[0049] Step 2: Establish the kinematic model between the preview point P preview and the tracked point P track The kinematic model is where y track is the lateral coordinate of the tracked point, and y preview is the lateral coordinate of the preview point;<(

[0050] Step 3: Consider the vehicle as a rigid body. Then, the motion relationship in the two-dimensional plane is shown in Equation (1);

[0051]

[0052] Step 4: Considering that there is no slip in the vehicle motion, then the sideslip angle of the center of mass β≈0. Then, there is a trajectory tracking error model

[0053] Step 5: Use an error fuser to fuse the trajectory tracking error into a comprehensive tracking error. The comprehensive tracking error z is where z is the comprehensive tracking error, tanh(.) is the hyperbolic tangent function, and the system parameters c0>0, c1>0, c2>0;

[0054] Step 6: Design a Lyapunov function to prove that when the comprehensive error converges, the lateral error and the heading error can converge simultaneously. The proof is as follows:

[0055] 1) Design the Lyapunov function as V = y 2 e / 2.

[0056] [[ID=?]]2) Take the derivative of the designed Lyapunov function, which is

[0057] 3) For a short sampling period, it can be considered that the vehicle maintains a constant motion attitude. Then, there is That is

[0058] 4) When z = 0, there is

[0059] 5) If 0 < c0 / c2 < π, there is: [[ID=?]]

[0060] 51) When y e > 0, there is

[0061] 52) When y e > 0, there is

[0062] 53) When y e = 0, there is It should be noted that there are some unclear or incorrect tags in the original text (such as <( which should probably be and some question marks in the translation where the original text seems to be incomplete or unclear). Please check and correct the original text for a more accurate translation.

[0063] 6) In summary, This shows that when the comprehensive error z converges to 0, the lateral error y e and heading error can converge to 0 at the same time.

[0064] Step 7: Design the differential error system model. The differential error system model designed by the error fusion model is

[0065] in, is the uncertain part in the differential error system model, coefficient b0=-c2v d / L, u is the final system control quantity.

[0066] Step 8: Design a linear estimator based on the differential error system model. The state space model of the designed linear estimator is shown in formula (2).

[0067]

[0068] Among them, state z1 is the estimated value of z, state z2 is the uncertain part in the differential error system model, and η is the output of the linear estimator.

[0069] Step 9: The linear estimator designed by formula (2) is discretized as shown in formula (3). According to formula (3), the state z1 and z2 output of the system can be estimated in real time.

[0070]

[0071] Among them, k is the discrete moment, T1 is the period of the linear estimator, and w o >0 is the gain of the linear estimator.

[0072] Step 10: Calculate the desired system state using the linear delay circuit. The designed linear delay circuit is first order, and its reference input r is always 0, and its output v1 is also always 0.

[0073] Step 11: Calculate the system error based on the estimated system state and the expected system state. The system error is calculated as e1 = v1 – z1.

[0074] Step 12: Based on the system error, the linear feedback controller calculates the desired control quantity. The linear feedback controller is designed as a proportional controller with a gain of w c / b0, the linear feedback controller output u0 is calculated as u0 = -w c z1 / b0.

[0075] Step 13: Calculate the extended state control quantity based on the estimated system state. The extended state control quantity calculation method is:

[0076] u1=z2 / b0.

[0077] Step 14: The desired control variable and the expanded state control variable are combined to obtain the final system control variable. The final system control variable is calculated as u = u0 + u1.

[0078] One embodiment of the present invention is Figure 3 As shown in FIG, a scene of lane change trajectory tracking is described. The dotted line in the scene represents the reference trajectory, that is, the trajectory to be tracked. The thick solid line represents the tracking effect of the pure tracking method, and the thick dotted line represents the tracking effect of the method of the present invention. Figure 4 As shown in FIG, when the method of the present invention is used to track the trajectory, the response of the front wheel steering angle is faster and larger than that of the pure tracking method. Figure 5 As shown, the peak tracking error is within 0.25m when using the proposed method, while the peak tracking error is close to 0.5m when using the pure tracking method. These data demonstrate that the proposed trajectory tracking method can reduce tracking error by accelerating the response of the front wheel angle, allowing for simultaneous convergence of lateral and heading errors.

[0079] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. An automatic driving control method that takes into account the convergence of lateral error and heading error, characterized in that: The following steps are involved: 1) The trajectory previewer calculates the trajectory tracking error; 2) The error fusion unit fuses the trajectory tracking error into a comprehensive tracking error; 3) Design a differential error system model based on the comprehensive tracking error; 4) Based on the differential error system model, design a linear estimator to estimate the system state; 5) The linear delay device calculates the desired system state; 6) Calculate the system error based on the estimated system state and the expected system state; 7) Based on the system error, the linear feedback controller calculates the desired control quantity; 8) Calculating the extended state control quantity based on the estimated system state; 9) The desired control quantity and the expanded state control quantity are integrated to obtain the final system control quantity; In the step 1), the track previewer is designed as Among them, y e is the vehicle lateral error, v d is the expected speed of the tracked point on the reference trajectory, is the vehicle heading error, l s is the distance from the preview point to the center of mass of the vehicle, L is the wheelbase, δ is the front wheel turning angle of the vehicle, y e The dots on the parameters are the first-order differentials of the parameters; The vehicle heading error The calculation method is Where θ is the heading of the tracked point on the reference trajectory, is the actual heading of the reference vehicle; In step 2), the error fusion device is designed as Where z is the integrated tracking error, tanh(.) is the hyperbolic tangent function, and the system parameters c0>0, c1>0, c2>0; In step 3), the differential error system model designed by the error fusion model is in, is the uncertain part in the differential error system model, coefficient b0=-c2v d / L, u is the final system control variable, the dot on the z parameter is the first-order differential of z, the dot on the f(.) function is the first-order differential of the f(.) function, v is the actual vehicle speed, and w is the system disturbance; In step 4), the state space model of the designed linear estimator is shown in formula (1), Wherein, state z1 is the estimated value of z, state z2 is the uncertain part in the differential error system model, η is the output of the linear estimator, the dot on the z1 parameter is the first-order differential of z1, and the dot on the z2 parameter is the first-order differential of z2; In step 4), the designed linear estimator is discretized as shown in formula (2): Among them, k is the discrete moment, T1 is the period of the linear estimator, and w o >0 is the gain of the linear estimator, e z (k) is the error of the linear estimator after discretization at the k-th moment; z1(k) is the value of parameter z1 at the k-th moment; z(k) is the value of parameter z at the k-th moment; z1(k+1) is the value of parameter z1 at the k+1-th moment; z2(k) is the value of parameter z2 at the k-th moment; u(k) is the value of the final system control quantity u at the k-th moment; z2(k+1) is the value of parameter z2 at the k+1-th moment; In the step 5), the designed linear delay device is first order; The reference input r of the linear delay device is always 0, and its output v1 is also always 0; In step 6), the system error is calculated as e1=v1–z1; In step 7), the linear feedback controller is designed as a proportional controller; The proportional controller gain is w c / b0, the linear feedback controller output u0 is calculated as u0 = -w c z1 / b0, Parameter w c >0 is the gain coefficient of the linear estimator; In step 8), the expansion state control amount is calculated as u1=z2 / b0; In step 9), the final system control quantity calculation method is u=u0+u1.

Images