Search

CN-121989936-A - Vehicle interaction decision-making method based on unprotected intersection and related device

CN121989936ACN 121989936 ACN121989936 ACN 121989936ACN-121989936-A

Abstract

The application discloses a vehicle interaction decision method and a related device based on an unprotected intersection, wherein the method comprises the steps of obtaining a system state matrix, a vehicle control matrix and a vehicle control matrix in a linear system state equation of a vehicle and the other vehicle, reversely recursing coupling Li-Care equations of a two-vehicle optimal control gain matrix equation set and a first and second intermediate matrix at the k moment to obtain a vehicle optimal control gain matrix sequence, and controlling the vehicle according to a control quantity sequence determined based on the optimal control gain matrix sequence, wherein the first to fourth semi-positive matrices and the first to fourth semi-positive matrices in an optimization target of the vehicle and the other vehicle linear secondary differential game problem, the first to third semi-positive matrices are respectively used as the first intermediate matrix and the second intermediate matrix at the N moment.

Inventors

  • ZHANG JIAXU
  • SUN GANG
  • Teng ting
  • LI WEI
  • WANG TENGFEI

Assignees

  • 魔门塔(苏州)科技有限公司

Dates

Publication Date
20260508
Application Date
20241101

Claims (10)

  1. 1. A vehicle interactive decision method based on unprotected intersections, the method comprising: Acquiring a system state matrix, a vehicle control matrix and a vehicle control matrix in a linear system state equation of a vehicle and the vehicle control matrix, and acquiring a first semi-positive matrix, a second semi-positive matrix and a first positive matrix in an optimization target of a vehicle linear secondary differential game problem under an unprotected intersection scene, and a third semi-positive matrix, a fourth semi-positive matrix and a second positive matrix in the optimization target of the vehicle linear secondary differential game problem, wherein the first semi-positive matrix is a coefficient of an Nth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the second semi-positive matrix is a coefficient of a kth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the first positive matrix is a coefficient of a kth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the third semi-positive matrix is a coefficient of a Nth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the second semi-positive matrix is a coefficient of a k moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, and the first positive matrix is a coefficient of a k moment linear system state quantity in the optimization target of the vehicle linear secondary differential problem, the k moment in the vehicle linear secondary differential game problem is less than or equal to the linear state quantity in the linear system of the vehicle, and the k moment in the linear state of the vehicle is less than or equal to the linear coefficient in the linear state; the first half positive definite matrix and the third half positive definite matrix are respectively used as a first intermediate matrix at the N time and a second intermediate matrix at the N time, wherein the first intermediate matrix is used for describing the mapping relation between the cooperative quantity of the discrete Hamiltonian function of the own vehicle and the linear system state quantity at the same time, and the second intermediate matrix is used for describing the mapping relation between the cooperative quantity of the discrete Hamiltonian function of the own vehicle and the linear system state quantity at the same time; And performing inverse recursion solution on a system of equations related to an optimal control gain matrix of the vehicle and the other vehicle, a coupling Li-Card equation related to a first intermediate matrix at a k-th time, a coupling Li-Card equation related to a second intermediate matrix at a k-th time, obtaining an optimal control gain matrix sequence of the vehicle, and controlling the vehicle according to a non-linear optimal control quantity sequence of the vehicle determined based on the optimal control gain matrix sequence of the vehicle, wherein the system of equations related to the optimal control gain matrix of the vehicle and the other vehicle is a game-type self-control gain matrix expressing a self-balancing gain matrix of the vehicle and a game-1 of which is obtained by calculating an optimization target related to the linear secondary differential problem of the vehicle and an optimization target related to the linear secondary differential problem of the other vehicle by using a Poisson Jin Zuixiao value algorithm.
  2. 2. The method of claim 1, wherein performing a reverse recursion solution on a set of equations for optimal control gain matrices for the host vehicle and the host vehicle, a coupled licarpa's equation for the first intermediate matrix at a kth time, a coupled licarpa's equation for the second intermediate matrix at a kth time, based on the first intermediate matrix at the nth time, the first positive definite matrix, the second positive definite matrix, the system state matrix, the host vehicle control matrix, the second semi-positive definite matrix, and the fourth semi-positive definite matrix to obtain an optimal control gain matrix sequence for the host vehicle, comprising: Substituting the first intermediate matrix at the N time, the second intermediate matrix at the N time, the first positive definite matrix, the second positive definite matrix, the system state matrix, the own vehicle control matrix and the other vehicle control matrix into the equation set of the own vehicle and other vehicle optimal control gain matrixes to obtain an own vehicle optimal control gain matrix at the N-1 time and an other vehicle optimal control gain matrix at the N-1 time; Substituting the second semi-positive definite matrix, the own vehicle optimal control gain matrix at the N-1 time, the first positive definite matrix, the system state matrix, the own vehicle control matrix, the other vehicle control matrix, the first intermediate matrix at the N time and the other vehicle optimal control gain matrix at the N-1 time into the coupling Rickel equation about the first intermediate matrix at the k time to obtain a first intermediate matrix at the N-1 time, and substituting the fourth semi-positive definite matrix, the other vehicle optimal control gain matrix at the N-1 time, the second positive definite matrix, the system state matrix, the own vehicle control matrix, the other vehicle control matrix, the second intermediate matrix at the N time and the own vehicle optimal control gain matrix at the N-1 time into the coupling Rickel equation about the second intermediate matrix at the k time to obtain a second intermediate matrix at the N-1 time; And analogically, until the optimal control gain matrix of the own vehicle at the 0 th moment and the optimal control gain matrix of the other own vehicle at the 0 th moment are obtained, forming the optimal control gain matrix sequence of the own vehicle according to the optimal control gain matrix of the own vehicle from the 0 th moment to the N-1 th moment.
  3. 3. The method of claim 1, wherein controlling the host vehicle according to a host vehicle nonlinear optimal control quantity sequence determined based on the host vehicle optimal control gain matrix sequence comprises: determining the vehicle-mounted linear optimal control quantity at each moment according to the vehicle-mounted optimal control gain matrix at each moment in the vehicle-mounted optimal control gain matrix sequence and the linear system state quantity at the same moment; According to a state equation of a self-vehicle linear system, a state equation of a self-vehicle nonlinear system and a self-vehicle linear optimal control quantity at each moment, determining a self-vehicle nonlinear optimal control quantity from the 0 th moment to the N-1 th moment, and forming a self-vehicle nonlinear optimal control quantity sequence according to a self-vehicle optimal control gain matrix from the 0 th moment to the N-1 th moment; and controlling the self-vehicle according to the self-vehicle nonlinear optimal control quantity sequence.
  4. 4. The method of claim 1, wherein the set of equations for the optimal control gain matrix for the host vehicle and the host vehicle comprises: Wherein K 1 (K) represents an own vehicle optimal control gain matrix at a kth time, R 1 represents the first positive definite matrix, B 1 represents the own vehicle control matrix, P 1 (k+1) represents a first intermediate matrix at a kth+1 time, a represents the system state matrix, B 2 represents the other vehicle control matrix, K 2 (K) represents an other vehicle optimal control gain matrix at a kth time, R 2 represents the second positive definite matrix, and P 2 (k+1) represents a second intermediate matrix at a kth+1 time.
  5. 5. The method according to any of claims 1-4, wherein the coupled licarpa's equation for the first intermediate matrix at time k comprises: P 1 (k)＝Q 1 +K 1 (k) T R 1 K 1 (k)+(A+B 2 K 2 (k)+B 1 K 1 (k)) T P 1 (k+1)(A+B 1 K 1 (k)+B 2 K 2 (k)); And/or, the coupled licarpa's equation for the second intermediate matrix at the kth time, comprising: P 2 (k)＝Q 2 +K 2 (k) T R 2 K 2 (k)+(A+B 2 K 2 (k)+B 1 K 1 (k)) T P 2 (k+1)(A+B 1 K 1 (k)+B 2 K 2 (k)); Wherein, P 1 (K) represents a first intermediate matrix at a kth time, Q 1 represents the second semi-positive definite matrix, K 1 (K) represents a vehicle optimal control gain matrix at a kth time, R 1 represents the first positive definite matrix, a represents the system state matrix, B 2 represents the vehicle control matrix, K 2 (K) represents a vehicle optimal control gain matrix at a kth time, B 1 represents the vehicle control matrix, P 1 (k+1) represents a first intermediate matrix at a kth+1 time, P 2 (K) represents a second intermediate matrix at a kth time, Q 2 represents the fourth semi-positive definite matrix, R 2 represents the second positive definite matrix, and P 2 (k+1) represents a second intermediate matrix at a kth+1 time.
  6. 6. A vehicle interactive decision device based on unprotected intersections, the device comprising: An acquisition unit configured to acquire a system state matrix, a vehicle control matrix, and a vehicle control matrix in a linear system state equation of a vehicle and a vehicle, and acquire a first semi-positive matrix, a second semi-positive matrix, and a first positive matrix in an optimization target of a vehicle linear secondary differential game problem in an unprotected intersection scene, and a third semi-positive matrix, a fourth semi-positive matrix, and a second positive matrix in an optimization target of a vehicle linear secondary differential game problem, wherein the first semi-positive matrix is a coefficient of an nth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the second semi-positive matrix is a coefficient of a kth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the first positive matrix is a coefficient of a kth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the third semi-positive matrix is a coefficient of a linear system state quantity in the optimization target of the vehicle linear secondary differential problem, the second semi-positive matrix is a coefficient of a linear state quantity in the linear secondary differential game problem of the vehicle, and the second positive matrix is a coefficient of a linear state quantity in a k moment in the optimization target of the vehicle linear secondary differential problem, and the k moment linear system state quantity in the optimization target of the vehicle linear secondary differential problem is equal to or less than or equal to the k moment in the linear system state quantity in the linear secondary differential state of the vehicle; The determining unit is used for respectively using the first semi-positive definite matrix and the third semi-positive definite matrix as a first intermediate matrix at the nth moment and a second intermediate matrix at the nth moment, wherein the first intermediate matrix is a matrix for describing the mapping relation between the cooperative quantity of the discrete Hamiltonian function of the own vehicle and the linear system state quantity at the same moment, and the second intermediate matrix is a matrix for describing the mapping relation between the cooperative quantity of the discrete Hamiltonian function of the own vehicle and the linear system state quantity at the same moment; The solving unit is configured to obtain an optimal control gain matrix sequence of the own vehicle according to the first intermediate matrix at the nth time, the second intermediate matrix at the nth time, the first positive definite matrix, the second positive definite matrix, the system state matrix, the own vehicle control matrix, the other vehicle control matrix, the second semi-positive definite matrix and the fourth semi-positive definite matrix, by performing inverse recursion solving an equation set related to an optimal control gain matrix of the own vehicle and the other vehicle, a coupled licarvex equation related to the first intermediate matrix at the kth time, and a coupled licarvex equation related to the second intermediate matrix at the kth time, where the equation set related to the optimal control gain matrix of the own vehicle and the other vehicle is an expression of feedback assorted solution related to the control gain matrix of the own vehicle and the other vehicle obtained by calculating an optimization target of the linear secondary differential problem of the own vehicle and an optimization target of the linear secondary differential problem of the other vehicle by using a pointtri-Jin Zuixiao value algorithm, and the optimal control gain matrix of the own vehicle comprises from the kth optimal control gain matrix at the nth time to the optimal control gain matrix of the own vehicle and the nth time; and the control unit is used for controlling the self-vehicle according to the self-vehicle nonlinear optimal control quantity sequence determined based on the optimal control gain matrix sequence of the self-vehicle.
  7. 7. The apparatus of claim 6, wherein the solving unit comprises: The first calculation module is used for obtaining an own vehicle optimal control gain matrix at the N-1 time and a other vehicle optimal control gain matrix at the N-1 time by substituting the first intermediate matrix at the N time, the second intermediate matrix at the N time, the first positive definite matrix, the second positive definite matrix, the system state matrix, the own vehicle control matrix and the other vehicle control matrix into the equation set related to the own vehicle and the other vehicle optimal control gain matrix; A second calculation module, configured to calculate a first intermediate matrix at an nth-1 time by substituting the second semi-positive definite matrix, the first positive definite matrix, the system state matrix, the own vehicle control matrix, the first intermediate matrix at the nth time and the own vehicle optimal control gain matrix at the nth-1 time into the coupled daycare equation for the first intermediate matrix at the kth time, and calculate a second intermediate matrix at the nth-1 time by substituting the fourth semi-positive definite matrix, the own vehicle optimal control gain matrix at the nth-1 time, the second positive definite matrix, the system state matrix, the own vehicle control matrix, the second intermediate matrix at the nth time and the own vehicle optimal control gain matrix at the nth-1 time into the coupled daycare equation for the second intermediate matrix at the kth time; the generation module is used for analogizing in sequence until the optimal control gain matrix of the own vehicle at the 0 th moment and the optimal control gain matrix of the other own vehicle at the 0 th moment are obtained, and the optimal control gain matrix sequence of the own vehicle is formed according to the optimal control gain matrix of the own vehicle from the 0 th moment to the N-1 th moment.
  8. 8. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the method according to any of claims 1-5.
  9. 9. An electronic device, the electronic device comprising: One or more processors; The processor is coupled with a storage device for storing one or more programs; The one or more programs, when executed by the one or more processors, cause the electronic device to implement the method of any of claims 1-5.
  10. 10. A vehicle comprising the apparatus of any one of claims 6-7 or the electronic device of claim 9.

Description

Vehicle interaction decision-making method based on unprotected intersection and related device Technical Field The application relates to the technical field of intelligent driving, in particular to a vehicle interaction decision method based on an unprotected intersection and a related device. Background Unprotected intersections refer to intersections without traffic lights, the road shape of which is often irregular, and may have only simple stop signs, yield signs, or even no signs. Therefore, when the past vehicle passes through the intersection without restriction or with less restriction, the occurrence of things is abrupt, so that the opponent and the host vehicle have not enough time and space to execute safe avoidance behavior, and the safety accident is extremely easy to cause. Therefore, how to improve the interaction safety of vehicles in the unprotected intersection scene is an important problem to be solved. Disclosure of Invention The application provides a vehicle interaction decision method and a related device based on an unprotected intersection, which can automatically calculate feedback Nash equilibrium solutions of a self-vehicle and a other vehicle in an unprotected intersection scene, so that the self-vehicle and the other vehicle try to optimize independent and conflict targets in a continuous game process, thereby avoiding collision and improving the safety of vehicle interaction in the unprotected intersection scene. The specific technical scheme is as follows: in a first aspect, an embodiment of the present application provides a vehicle interactive decision method based on an unprotected intersection, where the method includes: Acquiring a system state matrix, a vehicle control matrix and a vehicle control matrix in a linear system state equation of a vehicle and the vehicle control matrix, and acquiring a first semi-positive matrix, a second semi-positive matrix and a first positive matrix in an optimization target of a vehicle linear secondary differential game problem under an unprotected intersection scene, and a third semi-positive matrix, a fourth semi-positive matrix and a second positive matrix in the optimization target of the vehicle linear secondary differential game problem, wherein the first semi-positive matrix is a coefficient of an Nth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the second semi-positive matrix is a coefficient of a kth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the first positive matrix is a coefficient of a kth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the third semi-positive matrix is a coefficient of a Nth moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, the second semi-positive matrix is a coefficient of a k moment linear system state quantity in the optimization target of the vehicle linear secondary differential game problem, and the first positive matrix is a coefficient of a k moment linear system state quantity in the optimization target of the vehicle linear secondary differential problem, the k moment in the vehicle linear secondary differential game problem is less than or equal to the linear state quantity in the linear system of the vehicle, and the k moment in the linear state of the vehicle is less than or equal to the linear coefficient in the linear state; the first half positive definite matrix and the third half positive definite matrix are respectively used as a first intermediate matrix at the N time and a second intermediate matrix at the N time, wherein the first intermediate matrix is used for describing the mapping relation between the cooperative quantity of the discrete Hamiltonian function of the own vehicle and the linear system state quantity at the same time, and the second intermediate matrix is used for describing the mapping relation between the cooperative quantity of the discrete Hamiltonian function of the own vehicle and the linear system state quantity at the same time; And performing inverse recursion solution on a system of equations related to an optimal control gain matrix of the vehicle and the other vehicle, a coupling Li-Card equation related to a first intermediate matrix at a k-th time, a coupling Li-Card equation related to a second intermediate matrix at a k-th time, obtaining an optimal control gain matrix sequence of the vehicle, and controlling the vehicle according to a non-linear optimal control quantity sequence of the vehicle determined based on the optimal control gain matrix sequence of the vehicle, wherein the system of equations related to the optimal control gain matrix of the vehicle and the other vehicle is a game-type self-control gain matrix expressing a self-balancing gain