Search

CN-121973801-A - Safety-priority game incremental man-machine transverse sharing driving control method

CN121973801ACN 121973801 ACN121973801 ACN 121973801ACN-121973801-A

Abstract

The invention discloses a safe priority game incremental man-machine transverse sharing driving control method, which belongs to the technical field of automatic driving of automobiles and the technical field of man-machine co-driving, and comprises the steps of obtaining a driving intention safety factor of a driver, obtaining an expected track of the driver and an intelligent system, calculating a man-machine intention consistency coefficient according to the safety factor and discrete point position information of the expected track of the man-machine, designing a safe priority driving control right distribution strategy according to the safety factor and the man-machine intention consistency coefficient, calculating driving control right coefficients W h , W m ,W h and W m which are weight coefficients of a driver and a machine control torque item in a man-machine sharing driving controller respectively, establishing a touch interaction man-car-road closed loop system pre-aiming error dynamic model, and constructing the man-machine transverse sharing driving control method based on the safe priority driving control right distribution rule by combining a game theory and an incremental model prediction control theory.

Inventors

  • LIU CONGZHI
  • CHEN QITONG

Assignees

  • 重庆大学

Dates

Publication Date
20260505
Application Date
20260318

Claims (7)

  1. 1. The method for controlling the safety-priority game incremental man-machine transverse sharing driving is characterized by comprising the following steps of: Acquiring a driving intention safety factor of a driver; acquiring expected tracks of a driver and an intelligent system; According to the safety factor and the discrete point position information of the human-computer expected track, calculating a human-computer intention consistency coefficient; According to the consistency coefficient of the safety factor and the human-machine intention, designing a driving control right allocation strategy with safety priority, and calculating driving control right coefficients W h , W m ,W h and W m which are weight coefficients of a driver and a machine control torque item in a human-machine shared driving controller respectively; establishing a pre-aiming error dynamics model of a touch interaction human-vehicle-road closed loop system; and combining a game theory and an incremental model predictive control theory to construct the man-machine transverse sharing driving control method based on the driving control right allocation rule of safety priority.
  2. 2. The security-preferred gaming incremental unmanned aerial vehicle lateral sharing driving control method of claim 1, wherein the expression for calculating the unmanned aerial vehicle intention consistency coefficient is: Wherein γ is a driving intention safety factor of a driver, γ=0 represents a driving intention of the driver with high risk, α=1 represents that the machine only takes the driving intention of the machine as a tracking target without considering the driver intention, and the vehicle is ensured to run safely, TIC represents a human-machine target track deviation, and when the driver intention is safe, namely γ=1, the human-machine intention consistency α is inversely related to the target track deviation TIC of the driver and the machine.
  3. 3. The security-preferred game incremental unmanned aerial vehicle lateral sharing driving control method of claim 2, wherein the human-machine target trajectory bias TIC is defined as the average deviation of the human-machine expected trajectory within the N p -step prediction domain: where N p represents the prediction step size of the incremental shared controller, Defining average deviation of human-machine expected tracks in N p prediction steps, c TIC as a track deviation threshold, TIC=1 when the average deviation of the human-machine expected tracks exceeds the threshold, and d hm,i as Euclidean distance between the human-machine expected tracks at the ith time step in the Frenet coordinate system.
  4. 4. The security-priority gaming incremental man-machine transverse sharing driving control method according to claim 1, wherein the specific expression of the driving control authority coefficients W h and W m regulation rules is: ; Wherein W h0 and W m0 are used as basic control weight coefficients, E h and E m are used as correction factors, the basic control weight coefficients W h0 and W m0 determine man-machine driving control authority allocation when alpha=0, the correction factors E h and E m are introduced to adjust the change rate of an exponential function, and the basic control weight and the correction factors are set as follows 、 、 And 。
  5. 5. The security-preferred gaming incremental unmanned aerial vehicle lateral sharing driving control method of claim 1, wherein the haptic interactive human-vehicle-road closed loop system pre-aiming error dynamics model is expressed as: Wherein the state x and the output y are respectively And , And (3) with Respectively representing a lateral position error and a course angle error, v y and ω respectively representing a transverse speed and a yaw rate at the center of mass of the vehicle, θ being a steering wheel angle, τ h and τ m respectively representing control torques kappa r applied to the steering wheel by a driver and an intelligent system as curvatures of reference paths, regarded as environmental interference items, and coefficient matrices sequentially being: Wherein c f and c r respectively represent equivalent cornering stiffness of front and rear tires, L f and L r respectively represent distances from a vehicle center of mass to a front axle and a rear axle, v x represents longitudinal speed at the vehicle center of mass, L p is a pre-aiming distance, K v is total gain when a correction moment is transmitted from the tires to a steering wheel, i sw is steering ratio, J eq is equivalent inertia of a driver-steering wheel interaction system, and b eq is equivalent damping of the system; discretizing the pre-aiming error dynamics model to obtain: ; In the formula, , , , , Ts is the sampling time; Updating the future state of the vehicle in the prediction time domain by adopting an increment MPC, and constructing an augmented state space equation based on a discrete pretightening error dynamics model: Wherein, the augmented state vector xi (k) is expressed as xi (k) = [ x (k) τ h (k)τ m (k)] T ;Δτ h and Deltaτ m are torque control increment of a driver and a machine respectively, eta (k) is an augmented output vector of the kth step, and coefficient matrixes are respectively: and obtaining error prediction output of the shared transverse control system in the future N p steps by iteratively updating a prediction equation: wherein Y (k) is the controlled output in the N p -step prediction time domain and is expressed as: And Control inputs for the driver and the machine in the N c -step control time domain, respectively, are expressed as: wherein N c is the control step size; for the reference path curvature in the N p step prediction domain, we denote: each coefficient matrix in the motion error prediction equation is expressed as: According to a motion error prediction equation, acquiring a transverse position error and a course angle error of the vehicle relative to a reference line in a Frenet coordinate system in a N p -step prediction time domain; designing output coefficient matrices as And At this time, the absolute torque control amounts T h and T m of the driver and the machine are predicted to be output, expressed as: wherein each coefficient is expressed as The method for establishing the man-machine transverse sharing driving control comprises the following steps of: Wherein, the output matrixes Y h and Y m respectively represent the transverse position and course angle errors of the vehicle relative to the target tracking paths of the driver and the machine in the prediction domain, alpha represents the consistency of the intention of the human and the machine, the error weight matrixes Q h and Q m restrict the errors of the vehicle relative to the target tracks of the driver and the machine, the control increment weight matrixes R h and R m restrict the control increment changes of the driver and the machine, the torque weight matrixes W h and W m restrict the control torques of the driver and the machine, and T max 、T min and DeltaT max 、ΔT min respectively represent upper and lower limit constraint sets of the human and the torque increment in the control time domain.
  6. 6. The security-preferred gaming incremental human-machine lateral sharing driving control method according to claim 1, wherein the nash equalization of the incremental human-machine lateral sharing control based on the game theory is expressed as: Wherein, the Optimal control strategy, strategy combination, of driver and machine under each other's optimal strategy, respectively Is Nash equilibrium solution.
  7. 7. The security-preferred gaming incremental unmanned aerial vehicle lateral sharing driving control method of claim 6, wherein solving for nash equalization in real time comprises: The binary agent coupling optimization control problem is expressed as a standard quadratic form, and the relaxation factors epsilon h and epsilon m are introduced, and the standard quadratic form optimization problem is expressed as: Wherein, the And Weight coefficients of the relaxation factor terms respectively; cons h and Cons m are constant terms; 。

Description

Safety-priority game incremental man-machine transverse sharing driving control method Technical Field The invention belongs to the technical field of automatic driving of automobiles and the technical field of man-machine co-driving, and particularly relates to a safe and preferential game incremental man-machine transverse sharing driving control method. Background Autopilot technology presents great potential in improving driving safety and reducing driver burden. However, due to the state of the art and the public acceptance, fully automatic driving techniques are difficult to popularize in a short period of time. The L3 level automatic driving has become an important way for the gradual development of the automatic driving technology, and in the framework of the man-machine co-driving technology, an automatic driving system and a natural driver can independently drive a vehicle, and the automatic driving system and the natural driver can control the vehicle in a time-sharing and weight-sharing manner through information interaction and complete driving tasks. The man-machine sharing driving control can effectively avoid risks caused by incapability of timely taking over the vehicle due to the fact that a driver is separated from a control loop, and the capability of the driver is complementary with the advantages of the intelligent system. However, since the driver and the intelligent system are simultaneously in the control loop, the driver and the intelligent system are mutually coupled and restricted, and control right conflict can be even caused when the intention of the human and the machine is inconsistent. Therefore, a shared control method with driving right conflict resolution capability has become a focus of research in the field of man-machine co-driving. According to the lateral distance and error change of the vehicle center of mass from the lane center line, the Chinese patent (CN 113650609A) adopts a fuzzy logic rule to obtain a man-machine driving weight coefficient, and further establishes a man-machine driving weight sharing model by linearly weighting man-machine control instructions, so as to determine the final steering wheel angle control input of the vehicle. The patent only simulates the driving intention of a driver in the states of distraction, fatigue and the like, and the consistency of the intention of the driver is not fully considered, and the Chinese patent invention (CN 118859951A) discloses a man-machine sharing control method based on multi-target model predictive control, which adjusts a sharing control weight factor based on the time when a vehicle reaches a lane boundary and the collision time, and dynamically adjusts a steering instruction following target and a obstacle avoidance track following target of the driver through the weight factor, but the method is difficult to reflect the dynamic control interaction process of the driver and an intelligent system. The Chinese patent (CN 120245993A) considers the driving risks of different types of drivers, uses fuzzy logic to infer the control weight of the drivers, and further establishes a man-machine non-cooperative game sharing control model. In the method, a driver and an intelligent system only consider own tracking tracks, and when the intention of a man-machine is inconsistent, the continuous fight of the man-machine is easy to happen. Therefore, how to fully consider the consistency of human-machine intention, design the dynamic allocation rule of human-machine driving right, and establish the sharing control method with the capability of resolving driving right conflict is a key problem to be solved. Disclosure of Invention In order to solve the problems of the scheme, the invention provides a safe and preferential game incremental man-machine transverse sharing driving control method, which comprises the following steps: acquiring a driving intention safety factor of a driver, and quantitatively representing the driving intention safety; acquiring expected tracks of a driver and an intelligent system; according to the safety factor and the discrete point position information of the expected human-machine track, calculating a human-machine intention consistency coefficient, and quantitatively representing the degree of difference of human-machine driving intention; Designing a driving control right allocation strategy with safety priority by considering the consistency coefficient of the safety factor and the human-computer intention, and realizing the dynamic adjustment of driving control right coefficients W h and W m; establishing a pre-aiming error dynamics model of a touch interaction human-vehicle-road closed loop system; And combining a game theory and an incremental model predictive control theory, constructing a man-machine transverse sharing driving control method based on a driving control right distribution rule of safety priority, and realizing safety resolution of man-machine driving right con