automatic driving decision method and system
Technical Field
the invention relates to the technical field of automatic driving, in particular to an automatic driving decision method and an automatic driving decision system.
Background
in recent years, for unmanned vehicles, the unmanned vehicles are intelligent vehicles which sense the road environment through a vehicle-mounted sensor system, automatically plan the driving route and control the vehicles to reach the preset targets, the vehicle-mounted sensors are used for sensing the surrounding environment of the vehicles, and the steering and the speed of the vehicles are controlled according to the road, vehicle position and obstacle information obtained through sensing, so that the vehicles can safely and reliably run on the road, and the unmanned vehicles mainly rely on intelligent drivers which mainly use computer systems in the vehicles to achieve the purpose of unmanned driving.
The unmanned vehicle is used as a complex system combining software and hardware, the safe and reliable operation of the unmanned vehicle needs the cooperative cooperation of a plurality of modules such as vehicle-mounted hardware, sensor integration, perception prediction, control planning and the like, and intelligent decision is a key component of the unmanned vehicle and is also one of the research hotspots. In urban environments, due to the fact that driving scenes are complex and changeable, behaviors of traffic participants are difficult to predict, and the behaviors cannot be described by adopting a standard and uniform decision model. The method aims to solve the decision-making and planning problems of unmanned vehicles in complex urban environments.
Disclosure of Invention
This section is for the purpose of summarizing some aspects of embodiments of the invention and to briefly introduce some preferred embodiments. In this section, as well as in the abstract and the title of the invention of this application, simplifications or omissions may be made to avoid obscuring the purpose of the section, the abstract and the title, and such simplifications or omissions are not intended to limit the scope of the invention.
The present invention has been made in view of the above-mentioned conventional problems.
Therefore, one technical problem solved by the present invention is: the invention provides a novel driving behavior decision-making system to complete the motion planning of the unmanned vehicle by learning the decision-making process of the human driver in a complex scene.
In order to solve the technical problems, the invention provides the following technical scheme: an automatic driving decision-making system comprises a sensing module, a driving decision-making module and a driving decision-making module, wherein the sensing module can acquire object information of the surrounding environment of a vehicle and state information of the vehicle; the prediction module is connected with the perception module and can receive perception information and perform information fusion calculation to generate a prediction track; the decision module is connected with the prediction module and can receive track points of the predicted track output decision; and the control module is connected with the decision module on the upper layer, digests the track points and controls the vehicle to execute the track points.
As a preferable aspect of the automatic driving decision system of the present invention, wherein: the sensing module is responsible for detecting and calculating the attributes of objects and vehicles in the surrounding environment from sensor data, and object information is calculated by the prediction module to generate a predicted track and is transmitted to the behavior decision module in the decision planning control system.
as a preferable aspect of the automatic driving decision system of the present invention, wherein: the decision module comprises a behavior decision and trajectory planning module; and the behavior decision receives the routing and routing result, simultaneously receives perception prediction and map information, synthesizes input information, generates and inputs the information into the track planning module for track planning.
as a preferable aspect of the automatic driving decision system of the present invention, wherein: and the behavior decision comprises outputting the current vehicle behavior, a vehicle motion target point and a target vehicle speed, and outputting a planned track of the vehicle motion target point and the target vehicle speed after inputting the behavior decision.
As a preferable aspect of the automatic driving decision system of the present invention, wherein: the environmental state of the sensing module comprises environmental information collected by a laser radar, a millimeter wave radar, an ultrasonic radar and a vision sensor.
as a preferable aspect of the automatic driving decision system of the present invention, wherein: the state information of the vehicle comprises information collected by a positioning/inertial navigation sensor and a wheel speed sensor
as a preferable aspect of the automatic driving decision system of the present invention, wherein: the control module comprises a path tracking system, a gear control, a pedal control, a corner control and an accessory control; and the trajectory planning module outputs a vehicle motion target point and a planned trajectory of a target vehicle speed to the path tracking system and outputs signals of an accelerator/brake, a pedal opening and a steering wheel angle control to the pedal control and the steering wheel angle control to carry out vehicle driving control.
As a preferable aspect of the automatic driving decision method of the present invention, wherein: the method comprises the following steps that a perception module collects object information of the surrounding environment of the vehicle; the object information is input into a prediction module to calculate and generate a predicted track; a decision module receives track points of the predicted track output decision; and the feedback control module digests the trace points and controls the vehicle to execute the trace points.
As a preferable aspect of the automatic driving decision method of the present invention, wherein: the feedback control includes the steps of, during each control cycle, obtaining a series of control input increments in a control time domain:
in the formula: delta Ut *Control increment for time t; according to the basic principle of model predictive control, the first element in the control sequence is used as the actual control input increment to act on the system, namely:
u(t)=u(t-l)+Δut *
in the formula: u (t) is the actual control quantity of the system; the system executes the control quantity until the next moment, at the new moment, the output of the time domain of the next stage is predicted again according to the state information, and a new control increment sequence is obtained through the optimization process; and the steps are repeated in a circulating way until the control process is finished.
As a preferable aspect of the automatic driving decision method of the present invention, wherein: the prediction module also comprises a prediction step of collecting vehicle position information at different moments to form a plurality of complete track sequence data; defining an uncertain moving track with environmental characteristics, a track reference point and a set space of a predicted track point and a track reference point; and establishing a track prediction model and training.
the invention has the beneficial effects that: according to the invention, through the division of the modules, a complex problem of unmanned vehicle decision control planning is effectively solved, and a reasonable segmentation is made from abstraction to concrete according to the calculation logic; by the division, each module can be dedicated to solve the problem of the level, so that the development efficiency of the whole complex software system is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise. Wherein:
fig. 1 is a schematic structural diagram of an automatic driving decision system according to a first embodiment of the present invention;
Fig. 2 is a schematic block diagram of an automatic driving decision system according to a first embodiment of the present invention;
Fig. 3 is a flowchart illustrating an automatic driving decision method according to a second embodiment of the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, specific embodiments accompanied with figures are described in detail below, and it is apparent that the described embodiments are a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making creative efforts based on the embodiments of the present invention, shall fall within the protection scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways than those specifically described and will be readily apparent to those of ordinary skill in the art without departing from the spirit of the present invention, and therefore the present invention is not limited to the specific embodiments disclosed below.
furthermore, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.
The present invention will be described in detail with reference to the drawings, wherein the cross-sectional views illustrating the structure of the device are not enlarged partially in general scale for convenience of illustration, and the drawings are only exemplary and should not be construed as limiting the scope of the present invention. In addition, the three-dimensional dimensions of length, width and depth should be included in the actual fabrication.
Meanwhile, in the description of the present invention, it should be noted that the terms "upper, lower, inner and outer" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of describing the present invention and simplifying the description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation and operate, and thus, cannot be construed as limiting the present invention. Furthermore, the terms first, second, or third are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
The terms "mounted, connected and connected" in the present invention are to be understood broadly, unless otherwise explicitly specified or limited, for example: can be fixedly connected, detachably connected or integrally connected; they may be mechanically, electrically, or directly connected, or indirectly connected through intervening media, or may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
example 1
Referring to the schematic diagrams of fig. 1-2, in the embodiment, an automatic driving decision system is provided, in which an unmanned vehicle is used as a complex system combining software and hardware, and the safe and reliable operation of the system requires the cooperation of multiple modules such as vehicle-mounted hardware, sensor integration, perception prediction, control planning and the like. Specifically, the decision-making system comprises,
The sensing module 100, the sensing module 100 can collect the object information of the surrounding environment of the vehicle and the state information of the vehicle; the environmental status of the sensing module 100 includes environmental information collected by a laser radar, a millimeter wave radar, an ultrasonic radar, and a vision sensor. The vehicle's own state information includes information collected by the positioning/inertial navigation and wheel speed sensors.
The prediction module 200 is connected with the sensing module 100, and can receive sensing information and perform information fusion calculation to generate a predicted track;
The decision module 300 is connected with the prediction module 200 and can receive track points of the predicted track output decision; and the decision module 300 comprises a behavior decision 301 and a trajectory planning module 302; the behavior decision 301 receives the routing result, also receives the perception prediction and the map information, synthesizes the input information, generates and inputs the information to the trajectory planning module 302 for trajectory planning. The action decision 301 comprises outputting the current vehicle action, the vehicle motion target point and the target vehicle speed, and outputting the planned track of the vehicle motion target point and the target vehicle speed after the action decision 301 is input.
The control module 400 is connected with the upper decision module 300, digests the track points, and controls the vehicle to execute the track points. The control module 400 includes a path tracking system, gear control, pedal control, steering angle control, and accessory control;
the trajectory planning module 302 outputs a vehicle motion target point and a planned trajectory of a target vehicle speed to the path tracking system, and outputs control signals of an accelerator/brake, a pedal opening and a steering wheel angle to the pedal control and the steering wheel angle control to control the vehicle driving.
the sensing module 100 is responsible for detecting and calculating the object and vehicle attributes of the surrounding environment from the sensor data, and the object information is calculated by the prediction module 200 to generate a predicted track and is transmitted to the behavior decision module 300 in the decision planning control system.
Example 2
referring to the schematic diagram of fig. 3, based on the system, the embodiment provides an automatic driving decision method, which can be applied to the system, so that each module can individually perform its own functions to concentrate on solving the problem of the current level, thereby improving the development efficiency of the whole complex software system. Specifically, the method comprises the following steps of,
The sensing module 100 collects object information of the surrounding environment of the vehicle;
The object information is input into the prediction module 200 to calculate and generate a predicted track;
The decision module 300 receives track points of a predicted track output decision;
The feedback control module 400 digests the trace points and controls the vehicle to execute the trace points.
Wherein a series of control input increments in the control time domain are obtained during each control cycle:
in the formula: delta Ut *control increment for time t;
According to the basic principle of model predictive control, the first element in the control sequence is used as the actual control input increment to act on the system, namely:
u(t)=u(t-1)+Δut *
in the formula: u (t) is the actual control quantity of the system;
the system executes the control quantity until the next moment, at the new moment, the output of the time domain of the next stage is predicted again according to the state information, and a new control increment sequence is obtained through the optimization process;
And the steps are repeated in a circulating way until the control process is finished.
It should be noted that, in this embodiment, a connection control time model is added, a sampling time period is added, and a problem that a constant sampling period needs to be provided in a control process is solved, and this embodiment is a continuous control for distinguishing interval control, and can implement real-time continuous control on an autonomous vehicle in an autonomous driving decision method, and its specific implementation process is as follows:
establishing a time model which is continuous in time and space, as follows:
Let x (t) be a continuous signal for automatic driving control, and d (t) be 0. At this time the actual discrete output S of the automatic control system0The transfer function between (t) and x (t) is:
Hf(s)=Ko(s)F(s)H(s)[1+S0(s)F(s)G(s)]-1x(t)
wherein Ko(s) is the actual continuous control output, and functions H(s), F(s), and G(s) are continuous functions of the controller. Thus the output H of the continuous driving control systemf(s) can also be written as:
S0(s)=R(s)Hf(S), wherein R (S) ═ 1/S.
The control model for completing the design of automatic driving is as follows:Substituting the formula into a control model, inputting a series of control input increments in a control time domain obtained in each control period into the control model, and continuously updating the control model K after u (t) is substitutedoAnd(s) to realize the process of continuously controlling the automatic driving of the vehicle in space and time.
u(t)=u(t-1)+Δut *x (t) R(s), i.e.:
Hf(s)=Ko(s)F(s)H(s)[1+C0(s)F(s)G(s)]-1u(t)=u(t-1)+Δut *R(s)
s0(s)=R(s)Hf(s)。
in the simultaneous manner, the final control model C is knowno(s) is:
the prediction module 200 further comprises the following prediction steps,
Collecting vehicle position information at different moments to form a plurality of complete track sequence data;
Defining an uncertain moving track with environmental characteristics, a track reference point and a set space of a predicted track point and a track reference point;
and establishing a track prediction model and training.
The specific form is as follows:
Trj=<Ci,{p0,p1,…,pn}>,{pi=<xi,yi,ti>I 1 < n }, wherein Ciis the type of environment in which it is currently located,<xi,yi>is the longitude and latitude, t, of a spatial location pointi(i-0, …, n) is the corresponding timestamp.
the prediction module 200 in this embodiment is a model predictive controller. Model Predictive Control (MPC) is a particular type of control. Its current control action is obtained by solving a finite time domain open loop optimal control problem at each sampling instant. The current state of the process is taken as the initial state of the optimal control problem, and the obtained optimal control sequence only implements the first control action. This is the biggest difference from those algorithms that use pre-calculated control laws. Essentially, model predictive control solves an open-loop optimal control problem. Its idea is independent of the specific model, but its implementation is model dependent.
Another upstream module of the decision-making planning control system is a route routing module, which functions in a simple sense to be understood as navigation inside the unmanned vehicle software system, i.e. guiding the control planning module of the unmanned vehicle software system to run according to what kind of road on a macroscopic level so as to realize the purpose from the starting point to the destination point. It is noted that the routing path finding here is similar to the conventional navigation in some degree, but the details thereof are closely dependent on a high-precision map drawn specifically for unmanned vehicle navigation, and thus is substantially different from the conventional navigation.
generally, routing is implemented as a separate module, and the traffic prediction part can be implemented as a service extension of the sensing module or as a peripheral module of the decision planning control module.
the task of the decision planning control system is to make the most reasonable decision and control on the vehicle by combining the routing intention and the current position of the unmanned vehicle on the basis of the perceived predicted track of the peripheral object.
The whole decision planning control software system can be divided into three modules of behavior decision, action planning and feedback control from top to bottom according to different layers of problem solving.
Wherein, the behavior Decision module (Decision) can be intuitively understood as the 'copilot' of the unmanned vehicle.
the behavior decision receives the result of routing and also receives perception prediction and map information.
By integrating the input information, the behavior decision module macroscopically decides how the unmanned vehicle runs.
The decision at the macro level includes normal car following on the road, waiting for avoidance when encountering traffic lights and pedestrians, and interactive passing between intersections and other vehicles.
for example, when routing requires that an unmanned vehicle keeps driving in the current Lane (Lane), and senses that a vehicle which normally runs ahead is found, the decision of behavior decision is likely to be the following behavior.
And the action planning module solves the problem of planning the action (Motion) of the specific unmanned vehicle in the division of the upper drawing. The function of the method can be understood as that in a small time zone, the problem of how to drive the unmanned vehicle from the point A to the point B is solved. The problem solved here by the action planning module is a more specific step relative to the action decision. The action planning needs to make a plan of the intermediate path points from A to B in a short time t, and comprises selecting specific path points to be passed through, and the speed, the orientation, the acceleration and the like of the unmanned vehicle when reaching each path point. Moreover, action planning needs to guarantee two points: firstly, in the subsequent time, generating a spatio-temporal path from A to B and keeping certain consistency; secondly, these generated path points between a and B, including the velocity heading acceleration to each point, are within the actual operable physical range of the downstream feedback control.
The lowest module of the decision planning control system is a feedback control module. The module is directly butted with a CAN-BUS control interface of an unmanned vehicle bottom layer. The key task of the system is to digest output track points of an upper-layer action planning module, and convert the output track points into signals of an accelerator, a brake and a steering wheel controlled By a vehicle Drive-By-Wire through a series of dynamic calculations combined with vehicle body attributes and external physical factors, so that the vehicle is controlled to actually execute the track points as much as possible.
the feedback control module is primarily concerned with the control of the vehicle itself, as well as the modeling of the interaction with the external physical environment.
the division method of the modules effectively divides the unmanned vehicle decision control planning into a plurality of complex problems from abstraction to concrete according to the calculation logic. By the division, each module can be dedicated to solve the problem of the level, so that the development efficiency of the whole complex software system is improved.
It should be noted that the above-mentioned embodiments are only for illustrating the technical solutions of the present invention and not for limiting, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, which should be covered by the claims of the present invention.