WO2021175313A1 - 自动驾驶控制方法、装置、车辆及存储介质 - Google Patents

自动驾驶控制方法、装置、车辆及存储介质 Download PDF

Info

Publication number
WO2021175313A1
WO2021175313A1 PCT/CN2021/079289 CN2021079289W WO2021175313A1 WO 2021175313 A1 WO2021175313 A1 WO 2021175313A1 CN 2021079289 W CN2021079289 W CN 2021079289W WO 2021175313 A1 WO2021175313 A1 WO 2021175313A1
Authority
WO
WIPO (PCT)
Prior art keywords
actual
objective function
automatic driving
vehicle
speed
Prior art date
Application number
PCT/CN2021/079289
Other languages
English (en)
French (fr)
Inventor
杨斯琦
吕颖
崔茂源
Original Assignee
中国第一汽车股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国第一汽车股份有限公司 filed Critical 中国第一汽车股份有限公司
Publication of WO2021175313A1 publication Critical patent/WO2021175313A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W2050/0001—Details of the control system
    • B60W2050/0002—Automatic control, details of type of controller or control system architecture
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W2050/0001—Details of the control system
    • B60W2050/0002—Automatic control, details of type of controller or control system architecture
    • B60W2050/0004—In digital systems, e.g. discrete-time systems involving sampling
    • B60W2050/0005—Processor details or data handling, e.g. memory registers or chip architecture

Definitions

  • the embodiments of the present application relate to the field of vehicle control technology, for example, to an automatic driving control method, device, vehicle, and storage medium.
  • Automobile is an indispensable means of transportation in people's life and work. It has increasingly stronger intelligence, especially the automatic driving function, which brings great convenience to people.
  • the parameters in the vehicle kinematics and dynamics model are uncertain, and in a complex and random road environment, there will be other motor vehicles, pedestrians, etc., the behavior of these agents is random It brings great difficulty and challenges to the autonomous decision-making of vehicles.
  • Autonomous driving applies some intelligent decision-making methods, such as the planning and control of autonomous driving based on deep learning.
  • This method can theoretically fit any function, but requires a large number of artificial labels and samples, which cannot be achieved in the short term; for example, based on Reinforcement learning is used to plan and control autonomous driving, set reward and punishment functions based on behaviorism, and optimize decision-making through continuous trial and error.
  • Reinforcement learning is used to plan and control autonomous driving, set reward and punishment functions based on behaviorism, and optimize decision-making through continuous trial and error.
  • some mistakes are unacceptable and untryable. Therefore, the planning and control of autonomous driving are still in relatively simple working conditions, and reliable autonomous decision-making cannot be achieved.
  • the present application provides an automatic driving control method, device, vehicle, and storage medium to improve the reliability of automatic driving.
  • An embodiment of the present application provides an automatic driving control method, including: determining a planned path and a planned speed for automatic driving in a natural coordinate system according to a first objective function; and according to the planned path, the planned speed, and the second objective function Determine the target trajectory of the automatic driving; solve the control sequence according to the target trajectory, and control the vehicle to perform the automatic driving according to the control sequence.
  • the determining the planned path and the planned speed of the automatic driving according to the first objective function in the natural coordinate system includes: setting candidate path points in a lane within a set range, and traversing the connection mode of the candidate path points to generate a path candidate Set; establish the first objective function in the natural coordinate system based on the path candidate set, the first objective function being associated with the actual speed, actual acceleration, actual acceleration derivative, and actual path of the vehicle; solving The first objective function obtains the actual speed, the actual acceleration, the actual acceleration derivative, and the actual path that minimize the first objective function, and the plan is determined according to the obtained actual speed, the actual acceleration, the actual acceleration derivative, and the actual path.
  • the path and the planned speed includes: setting candidate path points in a lane within a set range, and traversing the connection mode of the candidate path points to generate a path candidate Set; establish the first objective function in the natural coordinate system based on the path candidate set, the first objective function being associated with the actual speed, actual acceleration, actual acceleration derivative, and actual path of the vehicle; solving The first objective function obtain
  • the determining the target trajectory of automatic driving according to the planned path, the planned speed, and the second objective function includes: establishing the second objective function in the natural coordinate system based on the path candidate set, and the first The second objective function is associated with the actual state quantity and the state difference quantity of the vehicle, wherein the state difference quantity includes the difference quantity between the actual state quantity and the planned path, and the actual state quantity and the The difference between the planned speeds is calculated; the second objective function is solved to obtain the actual state quantity that minimizes the second objective function, and the target trajectory of the automatic driving is determined according to the obtained actual state quantity.
  • the automatic driving control method further includes: reading the actual state quantity of the vehicle through a sensor; the actual state quantity includes: actual lateral displacement, actual longitudinal displacement, actual speed, actual yaw rate, actual body angle, actual heading Angle and actual acceleration; the actual state quantity is within the range of the road speed limit, start-stop limit, and travel direction limit.
  • the constraint conditions corresponding to the first objective function and the second objective function include: the target trajectory is outside the area of the obstacle vehicle shape, wherein the obstacle vehicle shape is set to take the horizontal axis as the long axis Oval.
  • the solving the control sequence according to the target trajectory includes: constructing a Hamilton function according to the first objective function, the second objective function, and the constraint condition; performing partial differentiation of the Hamilton function to obtain the first function; A continuation-based generalized minimal residual (Continuation/Generalized Minimal Residual, C/GMRES) algorithm is used to solve the first function to obtain the control sequence.
  • constructing a Hamilton function according to the first objective function, the second objective function, and the constraint condition performing partial differentiation of the Hamilton function to obtain the first function
  • a continuation-based generalized minimal residual (Continuation/Generalized Minimal Residual, C/GMRES) algorithm is used to solve the first function to obtain the control sequence.
  • the automatic driving control method further includes: converting a Cartesian coordinate system to the natural coordinate system, wherein the The natural coordinate system takes the centerline of the road as the horizontal axis and the normal of the road as the vertical axis.
  • the embodiment of the present application provides an automatic driving control device, including: a behavior planning module configured to determine a planned path and a planned speed of automatic driving according to a first objective function in a natural coordinate system; a trajectory determination module configured to The planned path, the planned speed, and the second objective function determine the target trajectory of the automatic driving; the sequence solving module is set to solve the control sequence according to the target trajectory; the control module is set to control the vehicle to perform all the tasks according to the control sequence Describe autonomous driving.
  • the embodiment of the present application provides a vehicle, including: one or more processors; a storage device configured to store one or more programs; when the one or more programs are executed by the one or more processors, The one or more processors are caused to implement the automatic driving control method as described above.
  • the embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned automatic driving control method is realized.
  • FIG. 1 is a flowchart of an automatic driving control method provided in Embodiment 1 of this application;
  • Figure 2 is a schematic diagram of establishing a natural coordinate system in the first embodiment of the application
  • FIG. 3 is a flowchart of an automatic driving control method provided by Embodiment 2 of this application.
  • FIG. 4 is a schematic diagram of a path candidate set in Embodiment 2 of this application.
  • FIG. 5 is a schematic diagram of the planned path in the second embodiment of this application.
  • Fig. 6 is a schematic diagram of the planned speed in the second embodiment of the application.
  • FIG. 7 is a schematic diagram of the principle of determining the target trajectory of automatic driving in the second embodiment of the application.
  • FIG. 8 is a schematic diagram of the constraint condition of the vehicle shape in the second embodiment of the application.
  • FIG. 9 is a schematic diagram of the target trajectory in the second embodiment of the application.
  • FIG. 10 is a schematic structural diagram of an automatic driving control device provided in Embodiment 3 of this application.
  • FIG. 11 is a schematic diagram of the hardware structure of a vehicle provided in the fourth embodiment of the application.
  • FIG. 1 is a flowchart of an automatic driving control method provided in Embodiment 1 of this application. This embodiment can be applied to a situation in which automatic driving of the vehicle is realized by predicting and controlling the trajectory and state of the vehicle.
  • the automatic driving control method may be executed by an automatic driving control device, which may be implemented by software and/or hardware, and integrated in the vehicle.
  • the method includes the following steps:
  • S110 Determine the planned path and planned speed of the automatic driving according to the first objective function in the natural coordinate system.
  • path planning calculations are usually based on the Cartesian coordinate system, and this embodiment takes into account that automatic driving is in a structured road scene. It is difficult to establish objective functions, constraints or prediction models using the Cartesian coordinate system, so it is based on natural coordinates.
  • the department predicts and plans the trajectory of autonomous driving.
  • First determine the planned path and planned speed of autonomous driving according to the first objective function. For example, collect road data through cameras, distance sensors, radars, etc.
  • the road data includes lane directions, parking spaces, obstacle positions, etc., and plan out based on road data A suitable path and the speed at which the vehicle travels on that path.
  • the process of determining the planning path and planning speed is essentially the process of optimizing the first objective function.
  • the first objective function can be a maximization optimization problem or a minimization optimization problem.
  • the variables associated with the first objective function include The speed, acceleration, path (or position) of the vehicle, etc., can be solved in an iterative manner to maximize or minimize the value of the first objective function to obtain the planned path and the corresponding planned speed on the path.
  • the basis for establishing the first objective function is to make the vehicle speed change steadily on the basis of meeting the road speed limit requirements and selecting a travelable path without obstacles.
  • the planned path and the planned speed are used as preliminary decisions in the automatic driving control process, and can be used as the initial value and reference value of the hot start of the subsequent target trajectory planning.
  • a speed chart can be drawn according to the location of the vehicle, the location of the surrounding vehicles, and the speed of the surrounding vehicles, and the initial speed planning can be carried out in the chart.
  • the abscissa of the speed chart is time, and the ordinate is the displacement in the longitudinal direction (that is, the direction of the road or the direction of the vehicle).
  • a reference speed can be set as a limiting factor for road speed limits, start and stop traffic lights and other working conditions.
  • the establishment of the first objective function takes into account factors such as speed, acceleration, acceleration derivative, continuity, etc., and can assign different weights to multiple factors, and finally obtain the value of the weighted sum of multiple factors With the largest or smallest solution, the planned path and the planned speed can be obtained.
  • the target trajectory of automatic driving is determined according to the second objective function, that is, the optimization of the planned path and planned speed, the purpose is to make the driving trajectory and speed of the vehicle more consistent with the conditions of continuity. It is convenient for real-time tracking of vehicle driving status.
  • the second objective function takes into account the closeness (or gap) between the actual state of the vehicle and the planning result in step S110, so that the actual trajectory is consistent with the planned path, and also takes into account the vehicle's speed, acceleration, yaw rate, and acceleration derivative. And so on to make the actual speed consistent with the planned speed.
  • the process of determining the target trajectory is essentially a process of optimizing and solving the second objective function.
  • the second objective function can be a maximization optimization problem or a minimization optimization problem in form.
  • the control sequence for the vehicle is solved according to the target trajectory.
  • the control sequence includes a number of parameters that need to be automatically driven to control the vehicle in order to achieve the above-mentioned target trajectory.
  • the trajectory and speed of the lane, lane change process, etc. have been determined), and it takes about 5 minutes to complete the path, then according to the dynamic model and kinematics model of the vehicle, the control sequence can be solved in reverse to obtain the real-time gear of the vehicle within 5 minutes.
  • the position, steering wheel angle, pressure range of the accelerator pedal or brake pedal, engine speed, turn signal switch, etc. can be used to control the vehicle to accurately follow the above-mentioned target trajectory and automatically drive.
  • S140 Control the vehicle to drive automatically according to the control sequence.
  • the vehicle is controlled according to the control sequence, so that the vehicle can accurately drive automatically in accordance with the above-mentioned target trajectory by controlling the corresponding speed, acceleration, yaw rate, heading angle, etc. during the automatic driving process.
  • the method further includes: converting the Cartesian coordinate system to the natural coordinate system, wherein the natural coordinate system is based on the road centerline It is the horizontal axis, and the normal of the road is the vertical axis.
  • FIG. 2 is a schematic diagram of establishing a natural coordinate system in Embodiment 1 of this application.
  • road data such as lane directions, parking spaces, and obstacle positions are collected through cameras, distance sensors, and radars.
  • the Cartesian coordinate system is converted to the natural coordinate system.
  • the natural coordinate system uses the centerline of the road as the reference line as the horizontal axis, and the normal of the road curve as the vertical axis. After the Cartesian coordinate system is converted to the natural coordinate system, the curvature of the road curve can be ignored, which simplifies the planning of the planned path. And forecast calculations.
  • the first embodiment of the present application provides an automatic driving control method that determines a planned path and a planned speed according to a first objective function in a natural coordinate system, and on this basis determines the target trajectory of automatic driving according to the second objective function, and according to the target trajectory Solve the control sequence and control the vehicle for automatic driving, so that the target trajectory is consistent with the planned path and planned speed, accurately control the vehicle for automatic driving, and improve the reliability of automatic driving.
  • FIG. 3 is a flowchart of an automatic driving control method provided in Embodiment 2 of the application. This embodiment is optimized on the basis of the above-mentioned embodiment, and describes the process of solving multiple objective functions and control sequences. For technical details not described in this embodiment, reference may be made to any of the foregoing embodiments.
  • the method includes the following steps:
  • S210 Convert a Cartesian coordinate system to a natural coordinate system, where the natural coordinate system takes the center line of the road as the horizontal axis and the normal line of the road as the vertical axis.
  • FIG. 4 is a schematic diagram of a path candidate set in the second embodiment of this application.
  • the set is used as a reference for the planning path, and the planning path obtained by solving the first objective function is an optimal path in the path candidate set.
  • the first objective function is associated with the actual speed, actual acceleration, actual acceleration derivative, and actual path of the vehicle.
  • a reference speed as the limiting factor of road speed limit, traffic lights start and stop, etc.
  • the establishment of the first objective function takes into account the actual speed, actual acceleration, actual acceleration derivative, continuity and other factors.
  • the factors are given different weights, and the solution that minimizes the weighted sum of multiple factors is finally obtained, and the planning path and planning speed can be obtained.
  • S240 Solve the first objective function to obtain the actual speed, actual acceleration, actual acceleration derivative and actual path that minimize the first objective function, and determine the planned path and plan according to the obtained actual speed, actual acceleration, actual acceleration derivative and actual path speed.
  • the corresponding planning speed can be solved using intelligent optimization algorithms, iterative algorithms, etc.
  • the speed, acceleration, etc. at each moment are all Can be determined.
  • Different planned routes correspond to different planned speeds, and different planned routes have different trajectory distances.
  • Each candidate route corresponds to different planned speeds. The faster the planned speed, the shorter the driving time. Comparing all possible paths (multiple paths in the path candidate set) with the corresponding speeds, the optimal solution can be obtained.
  • the optimal planning path and planning speed can be determined.
  • FIG. 5 is a schematic diagram of the planned path in the second embodiment of this application.
  • -2 to 2 correspond to the first lane
  • 2 to 6 correspond to the second lane
  • the S coordinate refers to the longitudinal distance.
  • Fig. 6 is a schematic diagram of the planned speed in the second embodiment of the application.
  • the planned vehicle speed and planned path for the part where S is 0-40 correspond to the first lane
  • the planned vehicle speed and planned path for the part where S is 40-80 correspond to the second lane.
  • S250 Establish a second objective function in the natural coordinate system based on the path candidate set.
  • the second objective function is associated with the actual state of the vehicle and the state gap, where the state gap includes the gap between the actual state and the planned path, and the gap between the actual state and the planned speed. quantity.
  • the optimization goal of the second objective function is to keep the actual state of the vehicle consistent with the planned path and planned speed, so that the vehicle can drive automatically according to the planned path and planned speed.
  • the actual state quantity is close to the planned path and the planned speed, it needs to meet the continuous condition to facilitate the tracking of the vehicle.
  • the second objective function is, for example: Among them, w 1 , w 2 , w 3 , w 4 , w 5 , w 6 , and w 7 are weights (w 1 , w 2 , w 3 , w 4 , w 5 , w 6 , and w 7 have a weight value greater than or Equal to 0); (x, y) is the actual displacement of the vehicle, (x ref , y ref ) is the planned position of the vehicle according to the planned path and planned speed for automatic driving, (xx ref ) 2 , (yy ref ) 2 are the actual The gap between the position coordinates and the planned position coordinates is controlled within a certain range; (vv ref ) 2 is to control the gap between the actual speed and the planned speed within a certain range; in addition, in order to ensure that the target trajectory is sufficiently smooth, the vehicle can better Be followed and improve the comfort and stability of the ride, control the actual acceleration a of the
  • S260 Solve the second objective function to obtain the actual state quantity that minimizes the second objective function, and determine the target trajectory of the automatic driving according to the obtained actual state quantity.
  • FIG. 7 is a schematic diagram of the principle of determining the target trajectory of automatic driving in the second embodiment of the application.
  • x(t) is the actual state quantity measured by the sensor of the vehicle at time t
  • u'(t) is the optimal control solution at time t
  • y(t) is the output of the model prediction system at time t.
  • the controller part adopts the model prediction method to periodically solve a finite-time open-loop optimization problem based on the current actual state quantity, and act on the vehicle corresponding to the control sequence corresponding to the actual state quantity, so as to adjust the actual state quantity of the vehicle.
  • the vehicle drives automatically according to the target trajectory.
  • the method further includes: reading the actual state of the vehicle through the sensor; the actual state includes: actual lateral displacement, actual longitudinal displacement, actual speed, actual yaw rate, actual body angle, actual The heading angle and actual acceleration; the actual state quantity is within the range of the road speed limit, start-stop limit and driving direction limit.
  • the data is read from the sensor and the actual state of the vehicle is obtained: for example, the actual displacement (x, y), actual speed v, actual body angle ⁇ , actual heading angle ⁇ , actual acceleration a, actual The yaw rate ⁇ and so on.
  • the optimal control sequence of the controlled autonomous vehicle is obtained.
  • the constraint conditions corresponding to the first objective function and the second objective function include: the target trajectory is outside the region of the obstacle vehicle shape, wherein the obstacle vehicle shape is set to an ellipse with the horizontal axis as the long axis.
  • the ability to explicitly handle the constraints is added. This ability comes from the second objective function's prediction of the future dynamic behavior of the system based on the model, that is, adding the constraints to the next moment Input, output or state variable.
  • the constraints can be expressed explicitly in a quadratic programming or nonlinear programming problem that is solved online.
  • the model predictive control can also impose hard constraints on the actual state quantity to be solved, so as to avoid collisions between the target trajectory and the obstacle vehicle.
  • FIG. 8 is a schematic diagram of the constraint condition of the vehicle shape in the second embodiment of the application.
  • hard constraints need to be added. Since the longitudinal speed of the vehicle on the road is often much greater than the lateral speed, it is assumed that the shape of the obstacle vehicle is an ellipse with the long axis of the horizontal axis, so that a certain distance between the vehicle and the obstacle is ensured to ensure safe driving.
  • constraints are: Where, (x obs, y obs) is the center coordinate of the vehicle to the obstacle, r x, r y is a major axis and a minor axis of ellipse; may also be converted to the equality constraint constraints, the introduction of virtual input u d , Then the above formula is converted to:
  • this embodiment uses a non-linear model predictive control algorithm, and selects a non-linear programming method to solve the problem, that is, the C/GMRES algorithm is used to solve the problem.
  • the C/GMRES algorithm uses the continuous method to update the differential equations of the control sequence, and uses the generalized minimum residual method to efficiently solve the linear equations.
  • the C/GMRES algorithm is obtained by using the increment of the control sequence at each sampling time, without using an iterative method to solve the nonlinear optimization problem, and has the advantages of fast calculation speed and high efficiency.
  • the form of the first function is as follows:
  • the nonlinear model predictive control can be solved quickly, so that the entire trajectory planning process can be carried out in real time and iteratively, and finally the control sequence U(t) corresponding to the target trajectory is output, and the processor in the vehicle controls the sequence accordingly Control the automatic driving of the vehicle, and carry out real-time tracking control of the actual state of the vehicle.
  • FIG. 9 is a schematic diagram of the target trajectory in the second embodiment of the application.
  • the vehicle can be controlled to realize automatic driving according to the optimized target trajectory.
  • the solid line represents the vehicle trajectory corresponding to the control sequence solved by the C/GMRES algorithm
  • the dashed line represents the planned path determined during the initial planning process
  • the rectangular box represents the obstacle vehicle that needs to be avoided.
  • the vehicle is in the process of automatic driving. Change lanes.
  • the second embodiment of the present application provides an automatic driving control method, which is described on the basis of the above-mentioned embodiments.
  • the coordinate system is transformed, the candidate path set is generated, and the first objective function is established.
  • determine the planning path and planning speed establish a second objective function and determine the target trajectory, and then use the C/GMRES algorithm to reversely solve the control sequence corresponding to the target trajectory.
  • weights are assigned to multiple influencing factors, which can adapt to different actual needs and flexibly plan the target trajectory; by adding constraint conditions, the safety of the automatic driving trajectory is ensured and conforms to the changing actual road environment.
  • the control method of this embodiment selects a variety of key vehicle dynamics state variables that affect driving decision-making, and proposes different optimal goals hierarchically from the selection of the path candidate set, the planning of the path and speed, and the optimization of the target trajectory.
  • Function to solve optimization problems realize the decoupling of complex path planning problems, enhance the stability and robustness of the system, and adopt a new and efficient numerical algorithm based on the continuous method of generalized minimum residual algorithm (C/GMRES), by using continuous Method to update the differential equation of the control sequence, solve the open-loop optimal solution of the target trajectory model predictive control problem, use the increment of the control sequence at each sampling time to obtain the C/GMRES algorithm sequence, instead of using the traditional iterative method to solve Non-linear optimization problem, so the control method of this embodiment requires short time, high efficiency, and small amount of calculation.
  • C/GMRES generalized minimum residual algorithm
  • FIG. 10 is a schematic structural diagram of an automatic driving control device provided in Embodiment 3 of this application.
  • the automatic driving control device provided in this embodiment includes: a behavior planning module 310 configured to determine a planned path and a planned speed of automatic driving according to a first objective function in a natural coordinate system; a trajectory determining module 320 configured to determine a planned path according to the planned path The planned speed and the second objective function determine the target trajectory of automatic driving; the sequence solving module 330 is configured to solve the control sequence according to the target trajectory; the control module 340 is configured to control the vehicle for automatic driving according to the control sequence.
  • the third embodiment of the present application provides an automatic driving control device, which determines a planned path and a planned speed according to a first objective function in a natural coordinate system, and on this basis determines the target trajectory of automatic driving according to the second objective function, and according to the target trajectory Solve the control sequence and control the vehicle for automatic driving, so that the target trajectory is consistent with the planned path and planned speed, accurately control the vehicle for automatic driving, and improve the reliability of automatic driving.
  • the behavior planning module 310 includes: a candidate set setting unit configured to set candidate path points in a lane within a set range and traverse the connection mode of the path candidate points to generate a path candidate set
  • a first function establishment unit configured to establish a first objective function in a natural coordinate system based on the path candidate set, the first objective function being associated with the actual speed, actual acceleration, actual acceleration derivative, and actual path of the vehicle
  • the first function solving unit is configured to solve the first objective function to obtain the actual speed, actual acceleration, actual acceleration derivative, and actual path that minimizes the first objective function, and according to the obtained actual speed, actual acceleration, and actual acceleration derivative And the actual path determines the planned path and the planned speed.
  • the trajectory determination module 320 includes: a second function establishing unit configured to establish a second objective function in a natural coordinate system based on the path candidate set, and the second objective function is related to the vehicle Is associated with the actual state quantity and the state gap quantity, wherein the state gap quantity includes the gap quantity between the actual state quantity and the planned path and the gap quantity between the actual state quantity and the planned speed
  • the second function solving unit is configured to solve the second objective function to obtain the actual state quantity that minimizes the second objective function, and determine the target trajectory of the automatic driving according to the obtained actual state quantity.
  • the device further includes: a state quantity reading unit configured to read the actual state quantity of the vehicle through a sensor; the actual state quantity includes: actual lateral displacement, actual longitudinal displacement, actual speed, actual yaw rate, The actual body angle, actual heading angle, and actual acceleration; the actual state quantity is within the range of the road speed limit, start-stop limit, and travel direction limit.
  • a state quantity reading unit configured to read the actual state quantity of the vehicle through a sensor; the actual state quantity includes: actual lateral displacement, actual longitudinal displacement, actual speed, actual yaw rate, The actual body angle, actual heading angle, and actual acceleration; the actual state quantity is within the range of the road speed limit, start-stop limit, and travel direction limit.
  • the constraint conditions corresponding to the first objective function and the second objective function include: the target trajectory is outside the area of the obstacle vehicle shape, wherein the obstacle vehicle shape is set to take the horizontal axis as the long axis Oval.
  • the sequence solving module 330 includes: a function construction unit configured to construct a Hamilton function according to the first objective function, the second objective function, and the constraint conditions; a partial differentiation unit configured to perform partial differentiation on the Hamilton function To obtain the first function; the solving unit is set to use the continuity-based generalized minimal residual (C/GMRES) algorithm to solve the first function to obtain the control sequence.
  • a function construction unit configured to construct a Hamilton function according to the first objective function, the second objective function, and the constraint conditions
  • a partial differentiation unit configured to perform partial differentiation on the Hamilton function To obtain the first function
  • the solving unit is set to use the continuity-based generalized minimal residual (C/GMRES) algorithm to solve the first function to obtain the control sequence.
  • C/GMRES continuity-based generalized minimal residual
  • the device further includes: a coordinate system conversion module configured to convert the Cartesian coordinate system to a natural coordinate system before determining the planned path and planned speed of the automatic driving according to the first objective function in the natural coordinate system, wherein the The natural coordinate system takes the centerline of the road as the horizontal axis and the normal of the road as the vertical axis.
  • a coordinate system conversion module configured to convert the Cartesian coordinate system to a natural coordinate system before determining the planned path and planned speed of the automatic driving according to the first objective function in the natural coordinate system, wherein the The natural coordinate system takes the centerline of the road as the horizontal axis and the normal of the road as the vertical axis.
  • the automatic driving control device provided in the third embodiment of the present application can be used to execute the automatic driving control method provided in any of the foregoing embodiments, and has corresponding functions.
  • FIG. 11 is a schematic diagram of the hardware structure of a vehicle provided in the fourth embodiment of the application.
  • Terminals include, but are not limited to: smart terminals such as desktop computers, notebook computers, smart phones, and tablet computers.
  • Servers include, but are not limited to: industrial integration servers, system backend servers, and cloud servers.
  • a vehicle provided in this embodiment includes a processor 410 and a storage device 420. There may be one or more processors in the vehicle. In FIG. 11, one processor 410 is taken as an example. The processor 410 and the storage device 420 in the vehicle may be connected by a bus or other means. Connect as an example.
  • the one or more programs are executed by the one or more processors 410, so that the one or more processors implement the automatic driving control method described in any of the foregoing embodiments.
  • the storage device 420 in the vehicle is used as a computer-readable storage medium and can be used to store one or more programs.
  • the programs can be software programs, computer-executable programs, and modules, such as the automatic driving control method in the embodiment of the present application.
  • Corresponding program instructions/modules (for example, the modules in the automatic driving control device shown in FIG. 10 include: a behavior planning module 310, a trajectory determination module 320, a sequence solving module 330, and a control module 340).
  • the processor 410 executes various functional applications and data processing of the vehicle by running the software programs, instructions, and modules stored in the storage device 420, that is, realizes the automatic driving control method in the foregoing method embodiment.
  • the storage device 420 mainly includes a storage program area and a storage data area.
  • the storage program area can store an operating system and an application program required by at least one function; the storage data area can store data created according to the use of the vehicle, etc.
  • the storage device 420 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices.
  • the storage device 420 may further include memories remotely provided with respect to the processor 410, and these remote memories may be connected to the vehicle through a network. Examples of the aforementioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.
  • the planned path and the planned speed of automatic driving are determined according to the first objective function in the natural coordinate system;
  • the planned path, the planned speed and the second objective function determine the target trajectory of automatic driving; solve the control sequence according to the target trajectory, and control the automatic driving of the vehicle according to the control sequence.
  • the vehicle proposed in this embodiment and the automatic driving control method proposed in the above embodiments are concepts. For technical details not described in detail in this embodiment, please refer to any of the above embodiments.
  • this embodiment also provides a computer-readable storage medium on which a computer program is stored.
  • the automatic driving control method in any of the above-mentioned embodiments of this application is implemented.
  • the method includes: determining a planned path and a planned speed of automatic driving in a natural coordinate system according to a first objective function; determining a target trajectory of automatic driving according to the planned path, the planned speed, and the second objective function; The target trajectory solves the control sequence, and the vehicle is controlled to perform automatic driving according to the control sequence.
  • the storage medium containing computer-executable instructions provided by the embodiments of the present application is not limited to the operation of the automatic driving control method described above, and can also execute the automatic driving provided by any embodiment of the present application. Related operations in the control method.
  • the technical solution of the present application can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), Random Access Memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute this application.
  • a computer device which can be a personal computer, server, or network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Human Computer Interaction (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
  • Traffic Control Systems (AREA)

Abstract

一种自动驾驶控制方法、装置、车辆及存储介质。方法包括：在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度（S110）；根据规划路径、规划速度以及第二目标函数确定自动驾驶的目标轨迹（S120）；根据目标轨迹求解控制序列（S130），并按照控制序列控制车辆进行自动驾驶（S140）。

Description

自动驾驶控制方法、装置、车辆及存储介质
本申请要求在2020年03月05日提交中国专利局、申请号为202010146615.0的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请实施例涉及车辆控制技术领域，例如涉及一种自动驾驶控制方法、装置、车辆及存储介质。
背景技术
汽车是人们生活和工作中必不可少的交通工具，具有越来越强的智能性，尤其是自动驾驶功能，给人们带来了极大便利。在自动驾驶的轨迹规划过程中，车辆运动学和动力学模型中的参数具有不确定性，并且在复杂随机的道路环境中，还会有其他机动车辆、行人等，这些智能体的行为随机性给车辆的自主决策带来极大地难度和挑战。
自动驾驶应用了一些智能决策方法，例如基于深度学习进行自动驾驶的规划和控制，这种方法在理论上能够实现拟合任意的函数，但是需要大量的人为标签和样本，短期无法实现；例如基于强化学习进行自动驾驶的规划和控制，基于行为主义设定奖惩函数，通过不断试错来优化决策，然而在自动驾驶决策上，有些错误是无法接受的和不能尝试的。因此，对于自动驾驶的规划和控制还停留在较为简单的工况下，无法实现可靠的自主决策。
发明内容
本申请提供了一种自动驾驶控制方法、装置、车辆及存储介质，以提高自动驾驶的可靠性。
本申请实施例提供了一种自动驾驶控制方法，包括：在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；根据所述规划路径、所述规划速度以及第二目标函数确定所述自动驾驶的目标轨迹；根据所述目标轨迹求解控制序列，并按照所述控制序列控制车辆进行所述自动驾驶。
所述在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度，包括：在设定范围内的车道内设置候选路径点，并遍历所述候选路径点的连接方式生成路径候选集；基于所述路径候选集在所述自然坐标系下建立所述第一目标函数，所述第一目标函数与所述车辆的实际速度、实际加速度、实际 加速度导数以及实际路径相关联；求解所述第一目标函数得到使所述第一目标函数最小的实际速度、实际加速度、实际加速度导数以及实际路径，并根据求得的实际速度、实际加速度、实际加速度导数以及实际路径确定所述规划路径和所述规划速度。
所述根据所述规划路径、所述规划速度以及第二目标函数确定自动驾驶的目标轨迹，包括：基于所述路径候选集在所述自然坐标系下建立所述第二目标函数，所述第二目标函数与所述车辆的实际状态量以及状态差距量相关联，其中，所述状态差距量包括所述实际状态量与所述规划路径之间的差距量以及所述实际状态量与所述规划速度之间的差距量；求解所述第二目标函数得到使所述第二目标函数最小的实际状态量，并根据求得的实际状态量确定所述自动驾驶的目标轨迹。
所述自动驾驶控制方法还包括：通过传感器读取所述车辆的实际状态量；所述实际状态量包括：实际横向位移、实际纵向位移、实际速度、实际横摆角速度、实际车身角度、实际航向角以及实际加速度；所述实际状态量在道路的速度限制、起停限制以及行驶方向限制的范围内。
所述第一目标函数和所述第二目标函数对应的约束条件包括：所述目标轨迹在障碍物车辆形状的区域之外，其中，所述障碍物车辆形状设定为以横轴为长轴的椭圆形。
所述根据所述目标轨迹求解控制序列，包括：根据所述第一目标函数、所述第二目标函数以及所述约束条件构建汉密尔顿函数；将所述汉密尔顿函数进行偏微分，得到第一函数；采用基于连续性的广义极小残量(Continuation/Generalized Minimal Residual，C/GMRES)算法求解所述第一函数，得到所述控制序列。
在所述在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度之前，所述自动驾驶控制方法还包括：将笛卡尔坐标系转换为所述自然坐标系，其中，所述自然坐标系以道路中心线为横轴，以道路的法线为纵轴。
本申请实施例提供了一种自动驾驶控制装置，包括：行为规划模块，设置为在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；轨迹确定模块，设置为根据所述规划路径、所述规划速度以及第二目标函数确定 所述自动驾驶的目标轨迹；序列求解模块，设置为根据所述目标轨迹求解控制序列；控制模块，设置为按照所述控制序列控制车辆进行所述自动驾驶。
本申请实施例提供了一种车辆，包括：一个或多个处理器；存储装置，设置为存储一个或多个程序；当所述一个或多个程序被所述一个或多个处理器执行，使得所述一个或多个处理器实现如上述的自动驾驶控制方法。
本申请实施例还提供了一种计算机可读存储介质，其上存储有计算机程序，该程序被处理器执行时实现上述自动驾驶控制方法。
附图说明
图1为本申请实施例一提供的一种自动驾驶控制方法的流程图；
图2为本申请实施例一中的建立自然坐标系的示意图；
图3为本申请实施例二提供的一种自动驾驶控制方法的流程图；
图4为本申请实施例二中的路径候选集的示意图；
图5为本申请实施例二中的规划路径的示意图；
图6为本申请实施例二中的规划速度的示意图；
图7为本申请实施例二中的确定自动驾驶的目标轨迹的原理示意图；
图8为本申请实施例二中的车辆形状的约束条件的示意图；
图9为本申请实施例二中的目标轨迹的示意图；
图10为本申请实施例三提供的一种自动驾驶控制装置的结构示意图；
图11为本申请实施例四提供的一种车辆的硬件结构示意图。
具体实施方式
下面结合附图和实施例对本申请作进行说明。可以理解的是，此处所描述的实施例仅仅用于解释本申请，而非对本申请的限定。另外还需要说明的是，为了便于描述，附图中仅示出了与本申请相关的部分而非全部结构。
实施例一
图1为本申请实施例一提供的一种自动驾驶控制方法的流程图。本实施例可适用于通过对车辆的轨迹和状态进行预测和控制以实现车辆的自动驾驶的情况。该自动驾驶控制方法可以由自动驾驶控制装置执行，该自动驾驶控制装置 可以通过软件和/或硬件的方式实现，并集成在车辆中。
如图1所示，该方法包括如下步骤：
S110、在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度。
在路径规划中通常基于笛卡尔坐标系进行计算，而本实施例考虑到自动驾驶是处于结构化道路的场景，采用笛卡尔坐标系难以建立目标函数、约束条件或预测模型等，因此基于自然坐标系对自动驾驶轨迹进行预测和规划。首先，根据第一目标函数确定自动驾驶的规划路径和规划速度，例如，通过摄像头、距离传感器、雷达等采集道路数据，道路数据包括车道方向、车位线、障碍物位置等，根据道路数据规划出一条合适的路径以及车辆在该路径上行驶的速度。确定规划路径和规划速度的过程实质上是对第一目标函数进行优化的过程，第一目标函数可以为最大化的优化问题，也可以为最小化的优化问题，第一目标函数关联的变量包括车辆的速度、加速度、路径(或位置)等，可以通过迭代的方式求解使第一目标函数的值最大或最小的解，即可得到规划路径以及在该路径上对应的规划速度。第一目标函数建立的依据是在符合道路限速要求、选择无障碍物的可行驶路径的基础上，使车辆速度平稳变化。规划路径和规划速度作为自动驾驶控制过程中的初步的决策，可以作为后续目标轨迹规划的热启动的初始值及参考值。
不同的规划路径对于规划速度的确定具有较大影响，车辆所在的不同车道有不同的障碍物，因此车辆可能会采取超车、跟随、让速躲避等不同行为。在规划之前，需要对车辆周围障碍物及周围车辆进行检测，示例性的，可以根据本车位置、周围车辆位置、周围车辆车速绘制出速度图表，在图表中进行速度的初规划。速度图表的横坐标为时间、纵坐标为纵向(即道路方向或车辆行驶方向)位移。在规划速度的过程中，可以设定一个参考速度，作为道路限速、红绿灯起停等工况的限制因素。在此基础上，第一目标函数的建立考虑到速度大小、加速度大小、加速度导数大小、连续性等因素，可以对多个因素赋予不同的权重，最终求得使多个因素的加权和的值最大或最小的解，即可得到规划 路径和规划速度。
S120、根据所述规划路径、所述规划速度以及第二目标函数确定自动驾驶的目标轨迹。
在确定规划路径和规划速度的基础上，根据第二目标函数确定自动驾驶的目标轨迹，即对规划路径和规划速度的优化，目的是使车辆的行驶轨迹和速度其更加符合连续性的条件，便于对车辆行驶状态进行实时跟踪。第二目标函数考虑到车辆的实际状态与步骤S110中规划结果的接近程度(或差距)，使得实际轨迹与规划路径一致，并且还考虑到车辆的速度、加速度、横摆角速度以及加速度导数的大小等，使得实际速度与规划速度一致，此外，在当前路径上的终态时限制车辆的航向角，使车辆与车道线平行，保证行驶方向不会偏离。确定目标轨迹的过程实质上是对第二目标函数进行优化求解的过程，第二目标函数形式上可以为最大化的优化问题，也可以为最小化的优化问题。
S130、根据所述目标轨迹求解控制序列。
根据目标轨迹求解对车辆的控制序列，控制序列包括为了实现上述的目标轨迹所需要对车辆进行自动驾驶控制的多项参数，例如，目标轨迹为在规划路径上按照规划速度行驶(包括何时变道、变道过程的轨迹和速度等都已经确定)，且大约需要5分钟行驶完该路径，则根据车辆的动力学模型、运动学模型可以逆向求解出控制序列在5分钟内车辆实时的挡位、方向盘角度、油门踏板或刹车踏板的压力幅度、发动机转速、转向灯的开关等，据此可以控制车辆准确地按照上述的目标轨迹自动驾驶。
S140、按照所述控制序列控制车辆自动驾驶。
按照控制序列控制车辆，从而在自动驾驶过程中通过控制对应的速度、加速度、横摆角速度、航向角等，使车辆准确地按照上述的目标轨迹自动驾驶。
在在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度之前，所述方法还包括：将笛卡尔坐标系转换为自然坐标系，其中，所述自然坐标系以道路中心线为横轴，以道路的法线为纵轴。
图2为本申请实施例一中的建立自然坐标系的示意图。如图2所示，通过 摄像头、距离传感器、雷达等采集车道方向、车位线、障碍物位置等道路数据，在道路形状已知的情况下，将笛卡尔坐标系转换为自然坐标系。自然坐标系是以道路中心线作为参考线当作横轴，道路曲线的法线作为纵轴，且笛卡尔坐标系转换为自然坐标系后可以忽略道路曲线的曲率，简化了对规划路径的规划和预测的计算。
本申请实施例一提供的一种自动驾驶控制方法，在自然坐标系下根据第一目标函数确定规划路径和规划速度，在此基础上根据第二目标函数确定自动驾驶的目标轨迹，根据目标轨迹求解控制序列并控制车辆进行自动驾驶，从而使目标轨迹与规划路径和规划速度保持一致，准确地控制车辆进行自动驾驶，提高自动驾驶的可靠性。
实施例二
图3为本申请实施例二提供的一种自动驾驶控制方法的流程图，本实施例是在上述实施例的基础上进行优化，对求解多个目标函数以及控制序列的过程进行描述。未在本实施例中描述的技术细节可参见上述任意实施例。
如图3所示，该方法包括如下步骤：
S210、将笛卡尔坐标系转换为自然坐标系，所述自然坐标系以道路中心线为横轴，以道路的法线为纵轴。
S220、在设定范围内的车道内设置候选路径点，并遍历所述候选路径点的连接方式生成路径候选集。
首先综合一定范围内多个车道的情况生成路径候选集。图4为本申请实施例二中的路径候选集的示意图。如图4所示，在车辆前方一定距离(S＝100米)内的多个相邻车道上分段设置候选路径点，遍历候选路径点的所有的连接方式从而生成路径候选集，将路径候选集作为规划路径的参考，求解第一目标函数得到的规划路径即是该路径候选集中的一个最优的路径。
S230、基于路径候选集在自然坐标系下建立第一目标函数。
本实施例中，第一目标函数与车辆的实际速度、实际加速度、实际加速度导数以及实际路径相关联。
设定一个参考速度作为道路限速、红绿灯起停等工况的限制因素，第一目标函数的建立考虑到实际速度大小、实际加速度大小、实际加速度导数大小、连续性等因素，通过对多项因素赋予不同的权重，最终求得使多项因素的加权和的值最小的解，即可得到规划路径和规划速度。以最小化的优化目标为例，第一目标函数可以为：f speed＝w rf r+w af a+w jf j+w cf c，其中，w r、w a、w j、w c为权重(w r、w a、w j、w c的权重值大于或等于0)；f r＝|v limit-v ik|表示将车辆的实际速度与参考速度的差值控制在一定范围内，实际速度不能过大也不能过小，其中v limit表示车辆的参考速度，v ik表示车辆的实际速度；
Figure PCTCN2021079289-appb-000001
表示车辆在当前候选路径点(i,k)时和上一候选路径点(i-1,j)时的速度变化在一定范围内，控制车速稳定，其中v i,k表示车辆在当前候选路径点(i,k)时的速度，v i-1,j表示车辆在上一候选路径点(i-1,j)的速度，Δt表示当前候选路径点(i,k)与上一候选路径点(i-1,j)的时间差值；
Figure PCTCN2021079289-appb-000002
表示实际加速度在一定范围内，其中，v i-2,j表示车辆在候选路径点(i-1,j)的上一候选路径点(i-2,j)时的速度；
Figure PCTCN2021079289-appb-000003
表示实际加速度导数在一定范围内，v t,i,k表示车辆在当前候选路径点(i,k)的时刻t的速度，v t-1表示车辆在上一时刻t-1的速度；其中，(i,k),(i-1,j),(i-2,j)表示在前方一定距离内的相邻车道上设置的不同的候选路径点的序列号，其中i为大于2的整数，k大于等于0，j大于等于0。参见图4所示的候选路径点标号，选取不同的路径点连接生成不同路径，如果发生了变道，则j≠k；t表示时刻且t大于等于0。
S240、求解第一目标函数得到使第一目标函数最小的实际速度、实际加速度、实际加速度导数以及实际路径，并根据求得的实际速度、实际加速度、实 际加速度导数以及实际路径确定规划路径和规划速度。
在候选路径集中找出使得第一目标函数f speed的值最小的路径，即为规划路径，同时可以利用智能优化算法、迭代算法等求解出对应的规划速度，每个时刻的速度、加速度等都可以确定。不同的规划路径对应于不同的规划速度，不同的规划路径的轨迹距离不同，每一条候选路径对应的规划速度也不一样，规划速度越快行驶时长就越短。比较所有可能的路径(路径候选集中的多个路径)和对应的速度，即可得到最优的解。通过结合行驶时长、行驶距离、加速度、加速度导数等因素构建第一目标函数，可以确定最优的规划路径和规划速度。
图5为本申请实施例二中的规划路径的示意图。如图5所示，对于L坐标，-2至2对应于第一条车道，2至6对应于第二条车道，S坐标指纵向距离。图6为本申请实施例二中的规划速度的示意图。如图6所示，S为0-40的部分的规划车速和规划路径对应于第一条车道，S为40-80的部分的规划车速和规划路径是对应于第二条车道。通过采集道路数据监测车道前方存在障碍物后可以规划出变道的轨迹，并根据第一目标函数求解出规划速度。
S250、基于所述路径候选集在自然坐标系下建立第二目标函数。
本实施例中，第二目标函数与车辆的实际状态量以及状态差距量相关联，其中，状态差距量包括实际状态量与规划路径之间的差距量以及实际状态量与规划速度之间的差距量。
第二目标函数的优化目标是使车辆的实际状态量与规划路径、规划速度保持一致，使得车辆按照规划路径和规划速度自动驾驶。实际状态量在接近规划路径和规划速度的前提下，需要符合连续的条件，便于车辆进行跟踪。以最小化的优化目标为例，第二目标函数例如为：
Figure PCTCN2021079289-appb-000004
Figure PCTCN2021079289-appb-000005
其中，w 1、w 2、w 3、w 4、w 5、w 6、w 7为权重(w 1、w 2、w 3、w 4、w 5、w 6、w 7的权重值大于或等于0)；(x，y)是车辆的实际位移，(x ref，y ref)是车辆按照规划 路径和规划速度自动驾驶的规划位置，(x-x ref) 2、(y-y ref) 2是将实际位置坐标与规划位置坐标的差距控制在一定范围内；(v-v ref) 2是将实际速度与规划速度之间的差距控制在一定范围内；此外，为了保证目标轨迹足够光滑，车辆能够更好地被跟随且提高乘坐的舒适度和平稳性，将车辆实际加速度a控制在一定范围内，将实际加速度导数
Figure PCTCN2021079289-appb-000006
控制一定范围内a i-1表示在上一候选路径点的实际加速度，将实际横摆角速度w控制在一定范围内；并且，为保证目标轨迹最终尽量落到道路中心，需要保持车辆的实际航向角与道路中心尽可能接近，θ N表示车辆的实际航向角与道路中心的差值，且应趋向于0。
S260、求解第二目标函数得到使第二目标函数最小的实际状态量，并根据求得的实际状态量确定自动驾驶的目标轨迹。
图7为本申请实施例二中的确定自动驾驶的目标轨迹的原理示意图。如图7所示，x(t)为t时刻车辆的传感器测量的实际状态量，
Figure PCTCN2021079289-appb-000007
为t时刻车辆的状态估计量，u′(t)为t时刻的最优控制解，y(t)为t时刻本模型预测系统的输出。控制器部分采用模型预测方法，周期性地基于当前实际状态量在线求解一个有限时间开环优化问题，并将得到的实际状态量对应的控制序列作用于车辆，调整车辆的实际状态量，从而使车辆准确按照目标轨迹自动驾驶。
在确定目标轨迹的过程中，所述方法还包括：通过传感器读取车辆的实际状态量；实际状态量包括：实际横向位移、实际纵向位移、实际速度、实际横摆角速度、实际车身角度、实际航向角以及实际加速度；实际状态量在道路的速度限制、起停限制以及行驶方向限制的范围内。
在每个控制周期内，从传感器读取数据并得到车辆的实际状态量：例如车辆的实际位移(x，y)、实际速度v、实际车身角度ψ、实际航向角δ、实际加速度a、实际横摆角速度ω等。通过求解开环优化问题得出被控自动驾驶车辆的最 优控制序列。
第一目标函数和第二目标函数对应的约束条件包括：目标轨迹在障碍物车辆形状的区域之外，其中，障碍物车辆形状设定为以横轴为长轴的椭圆形。
在通过第二目标函数计算目标轨迹的过程中，加入显式处理约束条件的能力,这种能力来自第二目标函数基于模型对系统未来动态行为的预测,即，把约束条件加到下一时刻的输入、输出或状态变量上。可以把约束条件显式表示在一个在线求解的二次规划或非线性规划问题中。本实施例中，除第二目标函数外，模型预测控制还可以对待求解的实际状态量进行硬约束，从而避免目标轨迹与障碍物车辆相撞等。
图8为本申请实施例二中的车辆形状的约束条件的示意图。如图8所示，为避免规划的轨迹与障碍车辆相撞，需要增加硬约束条件。由于道路上行驶的车辆纵向速度往往远大于侧向速度，因此假设障碍物车辆的形状为以横轴长轴的椭圆形，使得车辆与障碍物留有一定的距离，保证安全行驶。例如，约束条件为：
Figure PCTCN2021079289-appb-000008
其中，(x obs,y obs)为障碍物车辆的中心点坐标，r x、r y为椭圆的短轴和长轴；也可以将该约束条件转换为等式约束，引入虚拟输入量u d，则上式转换为：
Figure PCTCN2021079289-appb-000009
S270、构建包含模型预测系统的目标函数、模型以及等式约束条件的汉密尔顿函数。
考虑到车辆模型为非线性模型，本实施例采用非线性模型预测控制的算法，并选取非线性规划方法来求解，即，采用C/GMRES算法求解。C/GMRES算法通过使用连续法来更新控制序列的微分方程，并用广义最小残量法高效求解算线性方程。C/GMRES算法是使用控制序列在每个采样时刻的增量获得，而无需使用迭代法求解非线性优化问题，具有计算速度快、效率高的优势。在使用C/GMRES 算法进行迭代求解时，首先根据第一目标函数、第二目标函数、预测模型以及等式约束条件构建汉密尔顿函数，对汉密尔顿函数进行偏微分，从而求得协态变量，汉密尔顿乘子。汉密尔顿函数形式如下：H(X,λ,u,μ)＝L(X,u)+λ Tf(X,u)+μ TC(X,u)，其中X表示车辆模型状态空间方程中选取的车辆状态测量向量，λ表示协态变量，u表示车辆模型状态空间方程中选取的车辆控制输入向量，μ表示与等式约束相关的汉密尔顿乘子，L(X,u)表示车辆系统的目标函数，f(X,u)表示车辆系统的非线性状态向量模型，C(X,u)表示车辆系统的等式约束条件。
S280、对所述汉密尔顿函数进行偏微分，得到第一函数。
通过将汉密尔顿函数对输入变量进行偏微分，结合等式约束条件可以共同构建第一函数F＝0，第一函数的形式如下：
Figure PCTCN2021079289-appb-000010
其中，U(t)＝[u 0(t),μ 0(t),…,u N-1(t),μ N-1(t)]，u(t)＝[u 0(t),…,u N-1(t)]为待求解的车辆系统控制输入向量。
S290、采用基于连续性的广义极小残量(C/GMRES)算法求解第一函数，得到所述控制序列。
在推导出第一函数F＝0后，使用C/GMRES算法将F向0逼近，不断迭代求得输入变量以及汉密尔顿乘子的导数值。在求得输入变量U(t)的导数值后，通过C/GMRES算法可以求得U(t)在实际时域上的值，即为控制序列。通过上述的算法，可以对非线性模型预测控制进行快速求解，使得整个轨迹规划过程可以实时快速的迭代进行，最后输出目标轨迹对应的控制序列U(t)，车辆中的处理器据此控制序列控制车辆自动驾驶，并对车辆的实际状态量进行实时跟踪 控制。
图9为本申请实施例二中的目标轨迹的示意图。按照上述方法求解得到的控制序列控制车辆，即可实现按照优化的目标轨迹自动驾驶。如图9中所示，实线表示根据C/GMRES算法求解控制序列所对应的车辆轨迹，虚线表示初始规划过程中确定的规划路径，矩形框表示需要躲避的障碍物车辆，车辆在自动驾驶过程中变道。
本申请实施例二提供的一种自动驾驶控制方法，在上述实施例的基础上进行说明，基于分层的目标轨迹优化的算法架构，通过变换坐标系、生成候选路径集、建立第一目标函数并确定规划路径和规划速度、建立第二目标函数并确定目标轨迹，然后采用C/GMRES算法逆向求解目标轨迹对应的控制序列。在建立目标函数的过程中，对多项影响因素赋予权重，可以适应不同的实际需求，灵活规划目标轨迹；通过加入约束条件，保证自动驾驶轨迹的安全性，符合多变的实际道路环境。本实施例的控制方法，选取了多样的影响驾驶决策的关键车辆动力学状态量，从路径候选集的选定、路径和速度的规划、目标轨迹的优化，分层提出不同的最优的目标函数进行优化问题求解，实现复杂路径规划问题的解耦，增强系统的稳定性与鲁棒性，并采用新的高效数值算法基于连续法的广义最小残量算法(C/GMRES)，通过使用连续法来更新控制序列的微分方程，求解目标轨迹模型预测控制问题的开环最优解，使用控制序列在每个采样时刻的增量获得C/GMRES算法序列，而不再使用传统的迭代法求解非线性优化问题，所以本实施例的控制方法需要的时间短，效率高，计算量小。
实施例三
图10为本申请实施例三提供的一种自动驾驶控制装置的结构示意图。本实施例提供的自动驾驶控制装置包括：行为规划模块310，设置为在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；轨迹确定模块320，设置为根据所述规划路径、所述规划速度以及第二目标函数确定自动驾驶的目标轨迹；序列求解模块330，设置为根据所述目标轨迹求解控制序列；控制模块340，设置为按照所述控制序列控制车辆进行自动驾驶。
本申请实施例三提供的一种自动驾驶控制装置，在自然坐标系下根据第一目标函数确定规划路径和规划速度，在此基础上根据第二目标函数确定自动驾驶的目标轨迹，根据目标轨迹求解控制序列并控制车辆进行自动驾驶，从而使目标轨迹与规划路径和规划速度保持一致，准确地控制车辆进行自动驾驶，提高自动驾驶的可靠性。
在上述实施例的基础上，所述行为规划模块310，包括：候选集设置单元，设置为在设定范围内的车道内设置候选路径点并遍历所述路径候选点的连接方式生成路径候选集；第一函数建立单元，设置为基于所述路径候选集在自然坐标系下建立第一目标函数，所述第一目标函数与车辆的实际速度、实际加速度、实际加速度导数以及实际路径相关联；第一函数求解单元，设置为求解第一目标函数得到使所述第一目标函数最小的实际速度、实际加速度、实际加速度导数以及实际路径，并根据求得的实际速度、实际加速度、实际加速度导数以及实际路径确定所述规划路径和规划速度。
在上述实施例的基础上，所述轨迹确定模块320，包括：第二函数建立单元，设置为基于所述路径候选集在自然坐标系下建立第二目标函数，所述第二目标函数与车辆的实际状态量以及状态差距量相关联，其中，所述状态差距量包括所述实际状态量与所述规划路径之间的差距量以及所述实际状态量与所述规划速度之间的差距量；第二函数求解单元，设置为求解第二目标函数得到使所述第二目标函数最小的实际状态量，并根据求得的实际状态量确定自动驾驶的目标轨迹。
所述装置还包括：状态量读取单元，设置为通过传感器读取所述车辆的实际状态量；所述实际状态量包括：实际横向位移、实际纵向位移、实际速度、、实际横摆角速度、实际车身角度、实际航向角以及实际加速度；所述实际状态量在道路的速度限制、起停限制以及行驶方向限制的范围内。
所述第一目标函数和所述第二目标函数对应的约束条件包括：所述目标轨迹在障碍物车辆形状的区域之外，其中，所述障碍物车辆形状设定为以横轴为长轴的椭圆形。
序列求解模块330，包括：函数构建单元，设置为根据所述第一目标函数、所述第二目标函数以及所述约束条件构建汉密尔顿函数；偏微分单元，设置为将所述汉密尔顿函数进行偏微分，得到第一函数；求解单元，设置为采用基于连续性的广义极小残量(C/GMRES)算法求解所述第一函数，得到所述控制序列。
所述装置还包括：坐标系转换模块，设置为在在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度之前，将笛卡尔坐标系转换为自然坐标系，其中，所述自然坐标系以道路中心线为横轴，以道路的法线为纵轴。
本申请实施例三提供的自动驾驶控制装置可以用于执行上述任意实施例提供的自动驾驶控制方法，具备相应的功能。
实施例四
图11为本申请实施例四提供的一种车辆的硬件结构示意图。终端包括但不限定于：台式计算机、笔记本电脑、智能手机以及平板电脑等智能终端。服务器包括但不限定于：工业集成服务器、系统后台服务器以及云端服务器。如图11所示，本实施例提供的一种车辆，包括：处理器410和存储装置420。该车辆中的处理器可以是一个或多个，图11中以一个处理器410为例，所述车辆中的处理器410和存储装置420可以通过总线或其他方式连接，图11中以通过总线连接为例。
所述一个或多个程序被所述一个或多个处理器410执行，使得所述一个或多个处理器实现上述实施例中任意所述的自动驾驶控制方法。
该车辆中的存储装置420作为一种计算机可读存储介质，可用于存储一个或多个程序，所述程序可以是软件程序、计算机可执行程序以及模块，如本申请实施例中自动驾驶控制方法对应的程序指令/模块(例如，附图10所示的自动驾驶控制装置中的模块，包括：行为规划模块310、轨迹确定模块320、序列求解模块330以及控制模块340)。处理器410通过运行存储在存储装置420中的软件程序、指令以及模块，从而执行车辆的多种功能应用以及数据处理，即实现上述方法实施例中的自动驾驶控制方法。
存储装置420主要包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序；存储数据区可存储根据车辆的使用所创建的数据等(如上述实施例中的规划路径和规划速度等)。此外，存储装置420可以包括高速随机存取存储器，还可以包括非易失性存储器，例如至少一个磁盘存储器件、闪存器件、或其他非易失性固态存储器件。在一些实例中，存储装置420还可包括相对于处理器410远程设置的存储器，这些远程存储器可以通过网络连接至车辆。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
当上述车辆中所包括一个或者多个程序被所述一个或者多个处理器410执行时，进行如下操作：在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；根据所述规划路径、所述规划速度以及第二目标函数确定自动驾驶的目标轨迹；根据所述目标轨迹求解控制序列，并按照所述控制序列控制车辆自动驾驶。
本实施例提出的车辆与上述实施例提出的自动驾驶控制方法属于构思，未在本实施例中详尽描述的技术细节可参见上述任意实施例。
在上述实施例的基础上，本实施例还提供一种计算机可读存储介质，其上存储有计算机程序，该程序被自动驾驶控制装置执行时实现本申请上述任意实施例中的自动驾驶控制方法，该方法包括：在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；根据所述规划路径、所述规划速度以及第二目标函数确定自动驾驶的目标轨迹；根据所述目标轨迹求解控制序列，并按照所述控制序列控制车辆进行自动驾驶。
当然，本申请实施例所提供的一种包含计算机可执行指令的存储介质，其计算机可执行指令不限于如上所述的自动驾驶控制方法操作，还可以执行本申请任意实施例所提供的自动驾驶控制方法中的相关操作。通过以上关于实施方式的描述，所属领域的技术人员可以清楚地了解到，本申请可借助软件及通用硬件来实现，当然也可以通过硬件实现。基于这样的理解，本申请的技术方案可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存 储介质中，如计算机的软盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、闪存(FLASH)、硬盘或光盘等，包括多个指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请多个实施例所述的自动驾驶控制方法。

Claims (10)

  1. 一种自动驾驶控制方法，包括：
    在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；
    根据所述规划路径、所述规划速度以及第二目标函数确定所述自动驾驶的目标轨迹；
    根据所述目标轨迹求解控制序列，并按照所述控制序列控制车辆进行所述自动驾驶。
  2. 根据权利要求1所述的方法，其中，所述在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度，包括：
    在设定范围内的车道内设置候选路径点，并遍历所述候选路径点的连接方式生成路径候选集；
    基于所述路径候选集在所述自然坐标系下建立所述第一目标函数，所述第一目标函数与所述车辆的实际速度、实际加速度、实际加速度导数以及实际路径相关联；
    求解所述第一目标函数得到使所述第一目标函数最小的实际速度、实际加速度、实际加速度导数以及实际路径，并根据求得的实际速度、实际加速度、实际加速度导数以及实际路径确定所述规划路径和所述规划速度。
  3. 根据权利要求2所述的方法，其中，所述根据所述规划路径、所述规划速度以及第二目标函数确定自动驾驶的目标轨迹，包括：
    基于所述路径候选集在所述自然坐标系下建立所述第二目标函数，所述第二目标函数与所述车辆的实际状态量以及状态差距量相关联，其中，所述状态差距量包括所述实际状态量与所述规划路径之间的差距量以及所述实际状态量与所述规划速度之间的差距量；
    求解所述第二目标函数得到使所述第二目标函数最小的实际状态量，并根据求得的实际状态量确定所述自动驾驶的目标轨迹。
  4. 根据权利要求3所述的方法，还包括：通过传感器读取所述车辆的实际状态量；
    所述实际状态量包括：实际横向位移、实际纵向位移、实际速度、实际横摆角速度、实际车身角度、实际航向角以及实际加速度；
    所述实际状态量在道路的速度限制、起停限制以及行驶方向限制的范围内。
  5. 根据权利要求1所述的方法，其中，所述第一目标函数和所述第二目标 函数对应的约束条件包括：
    所述目标轨迹在障碍物车辆形状的区域之外，其中，所述障碍物车辆形状设定为以横轴为长轴的椭圆形。
  6. 根据权利要求5所述的方法，其中，所述根据所述目标轨迹求解控制序列，包括：
    根据所述第一目标函数、所述第二目标函数以及所述约束条件构建汉密尔顿函数；
    将所述汉密尔顿函数进行偏微分，得到第一函数；
    采用基于连续性的广义极小残量C/GMRES算法求解所述第一函数，得到所述控制序列。
  7. 根据权利要求1-6任一项所述的方法，在所述在自然坐标系下根据第一目标函数确定所述自动驾驶的规划路径和规划速度之前，还包括：
    将笛卡尔坐标系转换为所述自然坐标系，其中，所述自然坐标系以道路中心线为横轴，以道路的法线为纵轴。
  8. 一种自动驾驶控制装置，包括：
    行为规划模块，设置为在自然坐标系下根据第一目标函数确定自动驾驶的规划路径和规划速度；
    轨迹确定模块，设置为根据所述规划路径、所述规划速度以及第二目标函数确定所述自动驾驶的目标轨迹；
    序列求解模块，设置为根据所述目标轨迹求解控制序列；
    控制模块，设置为按照所述控制序列控制车辆进行所述自动驾驶。
  9. 一种车辆，包括：
    至少一个处理器；
    存储装置，设置为存储至少一个程序；
    当所述至少一个程序被所述至少一个处理器执行，使得所述至少一个处理器实现如权利要求1-7中任一所述的自动驾驶控制方法。
  10. 一种计算机可读存储介质，存储有计算机程序，该程序被处理器执行时实现如权利要求1-7中任一所述的自动驾驶控制方法。
PCT/CN2021/079289 2020-03-05 2021-03-05 自动驾驶控制方法、装置、车辆及存储介质 WO2021175313A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010146615.0A CN111338346B (zh) 2020-03-05 2020-03-05 一种自动驾驶控制方法、装置、车辆及存储介质
CN202010146615.0 2020-03-05

Publications (1)

Publication Number Publication Date
WO2021175313A1 true WO2021175313A1 (zh) 2021-09-10

Family

ID=71184235

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/079289 WO2021175313A1 (zh) 2020-03-05 2021-03-05 自动驾驶控制方法、装置、车辆及存储介质

Country Status (2)

Country Link
CN (1) CN111338346B (zh)
WO (1) WO2021175313A1 (zh)

Cited By (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114013450A (zh) * 2021-11-16 2022-02-08 交控科技股份有限公司 车辆运行控制方法、系统和计算机设备
CN114030463A (zh) * 2021-11-23 2022-02-11 上海汽车集团股份有限公司 一种自动泊车系统的路径规划方法及装置
CN114167860A (zh) * 2021-11-24 2022-03-11 东风商用车有限公司 一种自动驾驶最优轨迹生成方法及装置
CN114194217A (zh) * 2022-01-28 2022-03-18 中国第一汽车股份有限公司 车辆自动驾驶方法、装置、电子设备以及存储介质
CN114394113A (zh) * 2022-01-19 2022-04-26 广州小鹏自动驾驶科技有限公司 车辆轨迹重规划方法、装置、电子设备及存储介质
CN114475663A (zh) * 2022-03-08 2022-05-13 北京轻舟智航智能技术有限公司 一种自动驾驶横向控制的处理方法
CN114771551A (zh) * 2022-04-29 2022-07-22 阿波罗智能技术(北京)有限公司 自动驾驶车辆轨迹规划方法、装置和自动驾驶车辆
CN114872711A (zh) * 2022-05-27 2022-08-09 武汉理工大学 基于智能网联车辆的驾驶规划方法、系统、装置及介质
CN115016470A (zh) * 2022-05-30 2022-09-06 东风汽车集团股份有限公司 一种基于学习的自动驾驶局部路径规划优化方法及装置
CN115218902A (zh) * 2022-04-02 2022-10-21 广州汽车集团股份有限公司 一种轨迹规划方法、装置、设备及存储介质
CN116424319A (zh) * 2023-06-12 2023-07-14 上海鉴智其迹科技有限公司 一种车辆控制方法、装置、电子设备及计算机存储介质
WO2023201952A1 (zh) * 2022-04-21 2023-10-26 合众新能源汽车股份有限公司 确定车辆最优行驶轨迹方法和装置
CN117572875A (zh) * 2024-01-15 2024-02-20 上海友道智途科技有限公司 一种基于热启动的实时速度规划方法、系统、设备及介质

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111338346B (zh) * 2020-03-05 2022-10-21 中国第一汽车股份有限公司 一种自动驾驶控制方法、装置、车辆及存储介质
CN112026772B (zh) * 2020-08-14 2021-09-17 清华大学 一种智能网联汽车的路径实时规划与分布式控制方法
CN112406904B (zh) * 2020-08-27 2022-02-01 腾讯科技（深圳）有限公司 自动驾驶策略的训练方法、自动驾驶方法、设备和车辆
CN113778071A (zh) * 2020-09-17 2021-12-10 北京京东乾石科技有限公司 无人车路径规划方法、装置、电子设备、无人车及介质
CN112146667B (zh) * 2020-09-29 2022-10-14 广州小鹏自动驾驶科技有限公司 一种车辆过渡轨迹的生成方法和装置
CN112256037B (zh) * 2020-11-03 2021-07-30 智邮开源通信研究院(北京)有限公司 应用于自动驾驶的控制方法、装置、电子设备及介质
CN114516342B (zh) * 2020-11-19 2024-05-03 上海汽车集团股份有限公司 一种车辆控制方法、装置及车辆
CN113792249A (zh) * 2020-11-25 2021-12-14 北京京东乾石科技有限公司 行驶数据处理方法、装置、存储介质与电子设备
CN112907944A (zh) * 2021-01-18 2021-06-04 陈潇潇 局部交通段的自动驾驶智能交通方法
CN113050627A (zh) * 2021-03-02 2021-06-29 北京旷视机器人技术有限公司 路径规划方法、装置、移动机器人及计算机存储介质
CN112947492B (zh) * 2021-04-14 2023-09-22 北京车和家信息技术有限公司 车辆控制方法、装置、存储介质、电子设备及车辆
CN113110489B (zh) * 2021-04-30 2023-03-10 清华大学 一种轨迹规划方法、装置、电子设备和存储介质
CN113561189B (zh) * 2021-09-27 2021-12-31 深圳市优必选科技股份有限公司 冗余度机器人的关节加速度规划方法、装置、设备及介质
CN113978465A (zh) * 2021-09-28 2022-01-28 阿波罗智能技术(北京)有限公司 一种变道轨迹规划方法、装置、设备以及存储介质
CN114030480B (zh) * 2021-11-03 2023-09-22 重庆理工大学 一种基于避障路径规划的无人车自适应调头控制算法

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120215395A1 (en) * 2011-02-18 2012-08-23 Aznavorian Todd S System and method for automatic guidance control of a vehicle
CN106774329A (zh) * 2016-12-29 2017-05-31 大连理工大学 一种基于椭圆切线构造的机器人路径规划方法
CN109501799A (zh) * 2018-10-29 2019-03-22 江苏大学 一种车联网条件下的动态路径规划方法
CN109712421A (zh) * 2019-02-22 2019-05-03 百度在线网络技术（北京）有限公司 自动驾驶车辆的速度规划方法、装置和存储介质
CN110196590A (zh) * 2019-04-23 2019-09-03 华南理工大学 一种用于机器人路径跟踪的时间最优轨迹规划系统及方法
CN110364009A (zh) * 2019-07-16 2019-10-22 华人运通(上海)自动驾驶科技有限公司 基于路侧设备的行驶规划方法、装置、路侧设备和存储介质
CN111338346A (zh) * 2020-03-05 2020-06-26 中国第一汽车股份有限公司 一种自动驾驶控制方法、装置、车辆及存储介质

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10571921B2 (en) * 2017-09-18 2020-02-25 Baidu Usa Llc Path optimization based on constrained smoothing spline for autonomous driving vehicles
CN109375632B (zh) * 2018-12-17 2020-03-20 清华大学 自动驾驶车辆实时轨迹规划方法
CN109885883B (zh) * 2019-01-21 2023-04-18 江苏大学 一种基于gk聚类算法模型预测的无人车横向运动的控制方法
CN109866752B (zh) * 2019-03-29 2020-06-05 合肥工业大学 基于预测控制的双模式并行车辆轨迹跟踪行驶系统的方法
CN109976355B (zh) * 2019-04-26 2021-12-10 腾讯科技（深圳）有限公司 轨迹规划方法、系统、设备及存储介质
CN110244721B (zh) * 2019-06-05 2022-04-12 杭州飞步科技有限公司 自动驾驶控制方法、装置、设备及存储介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120215395A1 (en) * 2011-02-18 2012-08-23 Aznavorian Todd S System and method for automatic guidance control of a vehicle
CN106774329A (zh) * 2016-12-29 2017-05-31 大连理工大学 一种基于椭圆切线构造的机器人路径规划方法
CN109501799A (zh) * 2018-10-29 2019-03-22 江苏大学 一种车联网条件下的动态路径规划方法
CN109712421A (zh) * 2019-02-22 2019-05-03 百度在线网络技术（北京）有限公司 自动驾驶车辆的速度规划方法、装置和存储介质
CN110196590A (zh) * 2019-04-23 2019-09-03 华南理工大学 一种用于机器人路径跟踪的时间最优轨迹规划系统及方法
CN110364009A (zh) * 2019-07-16 2019-10-22 华人运通(上海)自动驾驶科技有限公司 基于路侧设备的行驶规划方法、装置、路侧设备和存储介质
CN111338346A (zh) * 2020-03-05 2020-06-26 中国第一汽车股份有限公司 一种自动驾驶控制方法、装置、车辆及存储介质

Cited By (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114013450A (zh) * 2021-11-16 2022-02-08 交控科技股份有限公司 车辆运行控制方法、系统和计算机设备
CN114013450B (zh) * 2021-11-16 2023-10-31 交控科技股份有限公司 车辆运行控制方法、系统和计算机设备
CN114030463A (zh) * 2021-11-23 2022-02-11 上海汽车集团股份有限公司 一种自动泊车系统的路径规划方法及装置
CN114030463B (zh) * 2021-11-23 2024-05-14 上海汽车集团股份有限公司 一种自动泊车系统的路径规划方法及装置
CN114167860B (zh) * 2021-11-24 2023-07-07 东风商用车有限公司 一种自动驾驶最优轨迹生成方法及装置
CN114167860A (zh) * 2021-11-24 2022-03-11 东风商用车有限公司 一种自动驾驶最优轨迹生成方法及装置
CN114394113A (zh) * 2022-01-19 2022-04-26 广州小鹏自动驾驶科技有限公司 车辆轨迹重规划方法、装置、电子设备及存储介质
CN114394113B (zh) * 2022-01-19 2023-08-25 广州小鹏自动驾驶科技有限公司 车辆轨迹重规划方法、装置、电子设备及存储介质
CN114194217A (zh) * 2022-01-28 2022-03-18 中国第一汽车股份有限公司 车辆自动驾驶方法、装置、电子设备以及存储介质
CN114194217B (zh) * 2022-01-28 2023-11-28 中国第一汽车股份有限公司 车辆自动驾驶方法、装置、电子设备以及存储介质
CN114475663A (zh) * 2022-03-08 2022-05-13 北京轻舟智航智能技术有限公司 一种自动驾驶横向控制的处理方法
CN114475663B (zh) * 2022-03-08 2024-04-09 北京轻舟智航智能技术有限公司 一种自动驾驶横向控制的处理方法
CN115218902A (zh) * 2022-04-02 2022-10-21 广州汽车集团股份有限公司 一种轨迹规划方法、装置、设备及存储介质
CN115218902B (zh) * 2022-04-02 2024-02-02 广州汽车集团股份有限公司 一种轨迹规划方法、装置、设备及存储介质
WO2023201952A1 (zh) * 2022-04-21 2023-10-26 合众新能源汽车股份有限公司 确定车辆最优行驶轨迹方法和装置
CN114771551B (zh) * 2022-04-29 2023-08-11 阿波罗智能技术(北京)有限公司 自动驾驶车辆轨迹规划方法、装置和自动驾驶车辆
CN114771551A (zh) * 2022-04-29 2022-07-22 阿波罗智能技术(北京)有限公司 自动驾驶车辆轨迹规划方法、装置和自动驾驶车辆
CN114872711A (zh) * 2022-05-27 2022-08-09 武汉理工大学 基于智能网联车辆的驾驶规划方法、系统、装置及介质
CN115016470A (zh) * 2022-05-30 2022-09-06 东风汽车集团股份有限公司 一种基于学习的自动驾驶局部路径规划优化方法及装置
CN116424319A (zh) * 2023-06-12 2023-07-14 上海鉴智其迹科技有限公司 一种车辆控制方法、装置、电子设备及计算机存储介质
CN116424319B (zh) * 2023-06-12 2023-08-29 上海鉴智其迹科技有限公司 一种车辆控制方法、装置、电子设备及计算机存储介质
CN117572875A (zh) * 2024-01-15 2024-02-20 上海友道智途科技有限公司 一种基于热启动的实时速度规划方法、系统、设备及介质
CN117572875B (zh) * 2024-01-15 2024-04-12 上海友道智途科技有限公司 一种基于热启动的实时速度规划方法、系统、设备及介质

Also Published As

Publication number Publication date
CN111338346A (zh) 2020-06-26
CN111338346B (zh) 2022-10-21

Similar Documents

Publication Publication Date Title
WO2021175313A1 (zh) 自动驾驶控制方法、装置、车辆及存储介质
JP7060625B2 (ja) 自動運転車において３ｄｃｎｎネットワークを用いてソリューション推断を行うｌｉｄａｒ測位
CN111780777A (zh) 一种基于改进a*算法和深度强化学习的无人车路径规划方法
CN110597245A (zh) 基于二次型规划和神经网络的自动驾驶换道轨迹规划方法
Shu et al. Autonomous driving at intersections: A critical-turning-point approach for left turns
Li et al. A practical trajectory planning framework for autonomous ground vehicles driving in urban environments
US11958498B2 (en) Data-driven warm start selection for optimization-based trajectory planning
Kebbati et al. Lateral control for autonomous wheeled vehicles: A technical review
CN114879687A (zh) 一种用于无人物流车的智能控制方法
Zhang et al. Structured road-oriented motion planning and tracking framework for active collision avoidance of autonomous vehicles
US11731274B2 (en) Predictive time horizon robotic motion control
Fan et al. A lane changing time point and path tracking framework for autonomous ground vehicle
Kabtoul et al. Proactive and smooth maneuvering for navigation around pedestrians
CN115525054B (zh) 大型工业园区无人清扫车沿边路径跟踪控制方法及系统
Tang et al. Actively learning Gaussian process dynamical systems through global and local explorations
CN114877904A (zh) 一种道路曲率的估计方法、测量装置和计算机设备
Yan et al. Cooperative Steering Control System for Robust Obstacle Avoidance in Cluttered Off-Road Environments
Vijayakumar et al. A holistic safe planner for automated driving considering interaction with human drivers
Li et al. An efficient sampling-based hybrid a* algorithm for intelligent vehicles
Yu et al. Design and Implementation of a Small-scale Autonomous Vehicle for Autonomous Parking
CN111857112A (zh) 一种汽车局部路径规划方法及电子设备
Luo et al. UAV path planning based on the average TD3 algorithm with prioritized experience replay
Li et al. Trajectory Planning for Autonomous Driving in Unstructured Scenarios Based on Deep Learning and Quadratic Optimization
Cheedalla et al. Parallel Parking and Parking Assist System for Autonomous and Semi-Autonomous Vehicles
CN116540747B (zh) 一种避停优先的移动机器人运动规划与避障决策方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21764015

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21764015

Country of ref document: EP

Kind code of ref document: A1