CN109445404B - Enhanced in-loop test method for intelligent driving planning decision control system - Google Patents
Enhanced in-loop test method for intelligent driving planning decision control system Download PDFInfo
- Publication number
- CN109445404B CN109445404B CN201811080324.5A CN201811080324A CN109445404B CN 109445404 B CN109445404 B CN 109445404B CN 201811080324 A CN201811080324 A CN 201811080324A CN 109445404 B CN109445404 B CN 109445404B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- control system
- intelligent driving
- planning decision
- driving planning
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
Abstract
The invention discloses an enhanced in-loop test method of an intelligent driving planning decision control system, which comprises the steps of collecting a GPS track of a path by a map collection vehicle provided with an optical fiber combination inertial navigation system, and making a map file; simulating software and hardware environment of the intelligent driving planning decision control system, and installing related software of the intelligent driving planning decision control system in the simulated environment; establishing a vehicle dynamics model, and constructing a closed loop between an intelligent driving planning decision control system and the vehicle dynamics model; an in-loop test communication network is established based on TCP/IP, and a communication link between an intelligent driving planning decision control system and an intelligent driving vehicle is simulated; based on the simulated communication link, the vehicle dynamics model is taken as a controlled object, a test path is simulated by the manufactured map file, and the performance of the intelligent driving planning decision control system in the simulated operation environment is tested. The invention solves the problem that the intelligent driving planning decision control system developed by a code programming mode is difficult to test the model in a loop and the software in a loop.
Description
Technical Field
The invention relates to the technical field of intelligent driving and simulation thereof, in particular to an enhanced in-loop test method for an intelligent driving planning decision control system.
Background
The simulation test is an indispensable stage of forward research and development, particularly aiming at an intelligent driving technology, and a plurality of complex traffic scenes and extreme working conditions have great dangerousness, so that the simulation test has important significance for simulating the actual running condition of an intelligent driving vehicle with high reduction degree.
At present, the intelligent driving simulation technology mainly comprises in-loop test methods such as model in-loop, software in-loop, hardware in-loop and the like. However, the existing on-loop test methods are basically proposed Based on the MBD (Model Based Design) development mode. The code programming-based mode does not go through the process from model design to code generation, so if the in-loop test is carried out by using the method of model in-loop and software in-loop, the code needs to be firstly converted into the model, C codes are generated, and the C codes are imported into MATLAB to carry out simulation experiments of the model in-loop and the software in-loop. Therefore, the source program is separated from the original development environment and running environment, and the code transplantation and running environment are difficult to ensure to have no influence on the test. In addition, the programming languages adopted by the code programming mode of the intelligent driving system comprise C/C + +/C #, Java, Python and other languages, the code amount of the system is large, and the code transplantation reduces the research and development efficiency.
Disclosure of Invention
The invention aims to solve the technical problem that aiming at the defects of the prior art, an intelligent driving planning decision control system enhanced in-loop test method is provided, and the problem that the intelligent driving planning decision control system developed in a code programming mode is difficult to perform model in-loop and software in-loop tests and the problem that the traditional model in-loop and software in-loop test methods do not consider the software operation environment are solved.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows: an intelligent driving planning decision control system enhanced in-loop testing method comprises the following steps:
1) a map acquisition vehicle provided with the optical fiber combined inertial navigation system acquires a GPS track of a path and makes a map file;
2) simulating software and hardware environment of the intelligent driving planning decision control system, and installing related software of the intelligent driving planning decision control system in the simulated environment;
3) establishing a vehicle dynamics model, and constructing a closed loop between an intelligent driving planning decision control system and the vehicle dynamics model;
4) an in-loop test communication network is established based on TCP/IP, and a communication link between an intelligent driving planning decision control system and an intelligent driving vehicle is simulated;
5) based on the simulated communication link, the vehicle dynamics model is taken as a controlled object, a test path is simulated by the manufactured map file, and the performance of the intelligent driving planning decision control system in the simulated operation environment is tested.
The enhanced in-loop test method is an in-loop test method for simulating a system operation environment, a communication link and a test path by utilizing an actual operation source program of the system to complete intelligent driving planning and performance evaluation of a decision control system.
In the step 1), the concrete implementation process for making the map file is as follows: and reading the vehicle position and course information acquired by the combined inertial navigation through an RS232 serial port, recording the data and storing the data in a txt/bin file format to obtain the map file.
The map file acquisition mode is as follows: at time intervals or at distance intervals, the straight line is not less than one point every 5m, and the curve is not less than one point every 1 m.
In the step 3), a vehicle dynamic model is established in Matlab/simulink according to a vehicle dynamic equation; alternatively, a vehicle dynamics model is built in Trucksim or Carsim, and then the model is imported into Matlab/simulink.
In the step 4), the specific steps of constructing the ring test communication network based on the TCP/IP are as follows: firstly, a parallel thread is opened up at a server, a TCP/IP communication server is established, a TCP/IP communication client is established in a vehicle dynamics model based on a TCP/IP function and a TCP/IP communication module in a simulink, then an IP address and a port number are configured, the address is a local loopback address, finally, communication connection is established, data interaction is realized, the server sends a front wheel steering angle and a vehicle longitudinal speed to the client, the client receives a control variable sent by the server, analyzes data, updates a vehicle state and sends the vehicle state quantity to the server.
Compared with the prior art, the invention has the beneficial effects that:
1. a simulated operation environment is constructed, and an enhanced ring is introduced on the basis of a model ring and a software ring, so that the problem that an intelligent driving planning decision control system developed in a code programming mode is difficult to perform model ring and software ring testing is solved, and the in-ring testing is more consistent with the actual situation;
2. a system source program is reserved, and related software of the intelligent driving planning decision control system is directly downloaded to a simulation environment, so that the source program is prevented from being transplanted to Matlab, and the system development efficiency is improved;
3. the vehicle dynamic model is used as a controlled object and introduced into the test network, and a software rack for on-loop test is connected in series.
Drawings
FIG. 1 is a schematic diagram of an enhanced in-loop test of an intelligent driving planning decision control system according to the present invention;
FIG. 2(a) a two degree-of-freedom automotive dynamics simulink model; FIG. 2(b) Trucksim vehicle dynamics modeling;
FIG. 3 is a TCP communication model;
FIG. 4 is a global path of an in-loop test output of global planning software;
FIG. 5 is a local expected trajectory of the local planning software at the ring test output;
FIG. 6 is a desired front wheel yaw angle output by the trajectory tracking controller; (a) the lower controller of the small-curvature curve outputs a front wheel steering angle; (b) the lower controller of the large-curvature curve outputs a front wheel steering angle;
FIG. 7 is a vehicle lateral position offset; (a) deviation of vehicle position on a small curvature curve; (b) deviation of vehicle position in a curve with large curvature.
Detailed Description
The invention comprises the following steps:
the method comprises the following steps: a map acquisition vehicle provided with the optical fiber combined inertial navigation system acquires a GPS track of a path and makes a map file;
step two: simulating software and hardware environment of the intelligent driving planning decision control system, and installing related software of the intelligent driving planning decision control system in the simulated environment;
step three: establishing a vehicle dynamics model, and constructing a closed loop between an intelligent driving planning decision control system and the vehicle dynamics model;
step four: an in-loop test communication network is established based on TCP/IP, and a communication link between an intelligent driving planning decision control system and an intelligent driving vehicle is simulated;
step five: based on the simulated communication link, the vehicle dynamics model is taken as a controlled object, a test path is simulated by the manufactured map file, and the performance of the intelligent driving planning decision control system in the simulated operation environment is tested.
The specific implementation method for making the map file in the first step comprises the following steps: and reading the vehicle position and course information acquired by the combined inertial navigation through an RS232 serial port, recording the data and storing the data in a txt/bin file format. There are two main ways for collecting map files, and according to time intervals or distance intervals, considering the requirements of planning algorithm, it is generally required that the straight line is not less than one point per 5m, and the curve is not less than one point per 1 m.
The software and hardware environment for simulating the operation of the intelligent driving planning decision control system in the second step mainly considers that the original in-loop test method is directed at the MDB development process, and is directed at the code programming development mode of C + +/JAVA and the like, and based on the original in-loop test method, the source code needs to be transplanted into the MATLAB, so that the code transplantation and the operation environment are difficult to ensure that the test is not influenced. The method is a testing method mainly aiming at a code programming development mode, a system source program is reserved, a simulation operation environment is constructed, and an on-loop testing method of a simulation operation environment, a virtual controller and a virtual model is designed.
The specific method for simulating the operating environment is to replace an industrial personal computer or other types of platforms actually operated by the system with a PC (personal computer) with the same or similar hardware configuration and software environment as the software and hardware platform for enhancing the in-loop test, wherein the hardware configuration mainly considers a CPU (central processing unit), a display card, a memory, a hard disk and the like, and the software environment comprises an operating system, a third-party operating library and the like.
The vehicle dynamics model establishment in the third step mainly comprises two methods:
1) establishing a dynamic model in Matlab/simulink according to a vehicle dynamic equation;
2) the vehicle dynamics model is built using third party software, such as in Trucksim or Carsim, and then imported into Matlab/simulink.
The specific method for establishing the dynamic model in Matlab/simulink according to the vehicle dynamic equation is as follows:
(A) deriving vehicle dynamics equations
Taking a two-degree-of-freedom automobile dynamic model as an example, the description of each symbol is shown in the following table:
TABLE 1 notation and description of relevant parameters of two-DOF automobile dynamics model
When U is given, the model can only change the motion in two degrees of freedom, lateral and rotational, by controlling δ, and is therefore called a two-degree-of-freedom model.
The kinetic differential equation is listed for the center of mass of the vehicle, and includes two parts, lateral acceleration and rotational acceleration:
wherein m and IZRespectively the mass and the moment of inertia of the vehicle.
Several model assumptions are introduced below:
(a) small angle assumptions, i.e. delta, alphaf、αrAre all relatively small;
(b) the longitudinal speed U is constant, so there is:
(c) the tire operates in the linear region, namely:
Fyf＝Cfαf,Fyr＝Crαr (4)
make the vehicle state quantity
The first three quantities are used for describing the position and the posture of the vehicle, the second two quantities are used for describing the speed and the angular speed information of the vehicle, and then the two-degree-of-freedom vehicle dynamic equation can be listed as follows:
(B) according to the two-degree-of-freedom vehicle dynamics equation shown in the formula (5), a corresponding model can be established in simulink, and vehicle dynamics modeling is completed.
Taking the example of establishing a vehicle dynamics model by using Trucksim, wherein the specific method for establishing the vehicle dynamics model by using third-party software is as follows:
(A) complete vehicle model construction
Trucksim adopts a parameterized dynamics modeling method, and can simplify a vehicle model simply and effectively. When a whole vehicle model is built by utilizing Trucksim, firstly, vehicle dynamic model parameters are configured according to target vehicle model parameters, and the parameters to be configured comprise a wheel base, the length, the width and the height of the whole vehicle, the mass of the whole vehicle, the lateral deflection rigidity coefficients of front wheels and rear wheels and the like.
(B) Simulation environment construction
The simulation environment building mainly comprises vehicle initial state definition and simulation path design, and comprises initial vehicle speed, gear signal output mode, driving/braking model and expected path building, road surface parameter configuration and the like. Because the designed track tracking controller can periodically provide information such as driving speed, braking deceleration, steering wheel angle and the like, when the vehicle state is defined, an external braking mode and a steering mode are selected, namely, closed loop optimization of driving, braking and steering in a model is not needed, an automatic gear is selected in the gear mode, and a road starting point, width, friction coefficient, curve curvature and the like are set according to the condition of an actually measured line, so that a desired path is constructed.
(C) External call interface setting
The external calling interface needs to be configured with a connector and a resolver and select and set an input/output interface, firstly, a connector ' send ' to a simulink ' needs to be selected, then the resolver is set, a Runge-Kutta numerical solution is selected, model simulation step length is set, then a model input/output port is selected, an input port comprises vehicle driving speed, braking deceleration, steering wheel turning angle, steering wheel rotating speed and the like, an output port comprises vehicle speed, vehicle course angle, vehicle position information and the like, finally, Trucksim is led into the simulink model base in the form of S function (Trucksim S-funceton) through a specified path, and the model is called in the simulink.
The specific method for constructing the closed loop between the intelligent driving planning decision control system and the vehicle dynamics model in the third step is to use the control variables such as the steering angle of the front wheels, the longitudinal speed of the vehicle and the like output by the intelligent driving planning decision control system as the input of the vehicle dynamics model, and use the current vehicle state fed back by the vehicle dynamics model as the feedback of the intelligent driving planning decision control system, so as to establish the closed loop control system between the intelligent driving planning decision control system and the controlled object, wherein the communication link between the intelligent driving planning decision control system and the vehicle dynamics model is constructed based on TCP/IP. In order to verify the theoretical feasibility of the method, a function expression of the intelligent driving planning decision control system and the closed loop of the vehicle dynamic model is derived in the following.
Make the vehicle state quantity
Wherein x, y represent vehicle position,
Representing the vehicle yaw angle, v representing the vehicle lateral velocity, ω representing the yaw rate, and the vehicle dynamics equation in f (-) can be obtained:
ξ(k+1)＝f(ξ(k),δ(k),U(k)) (6)
where ξ (k) represents the vehicle state for the kth control cycle, ξ (k +1) represents the vehicle state for the kth control cycle, δ (k) represents the vehicle front wheel steering angle for the kth control cycle, and u (k) represents the vehicle longitudinal speed for the kth control cycle.
And (2) representing the control rate of the trajectory tracking control by eta (·), wherein the expression form of the input and output functions of the available trajectory tracking control software is as follows:
(δ(k),U(k))＝η(A(k),Ue(k),ξ(k)) (7)
where a (k) represents a local expected trajectory of the kth control cycle output by the local planning software, and a (k) ═ an(k),an-1(k),…,a1(k),a0(k)]，an(k),an-1(k),…,a1(k),a0(k) Expressing a polynomial coefficient corresponding to the local expected track, wherein n represents a polynomial order; u shapee(k) The desired vehicle longitudinal speed for the kth control cycle, which is indicative of the local planning software output.
Assuming that the initial state of the vehicle is xi (1), the intelligent driving planning decision control system and the closed loop function expression form of the vehicle dynamic model obtained by combining the formulas (6) and (7) are as follows:
(δ(1),U(1))＝η(A(1),Ue(1),ξ(1))
ξ(2)＝f(ξ(1),δ(1),U(1))
(δ(k),U(k))＝η(A(k),Ue(k),ξ(k))
ξ(k+1)＝f(ξ(k),δ(k),U(k))
(8)
the global planning, the expected track output by the local planning and the expected vehicle speed of the upper layer are used as the input of a track tracking controller of the intelligent driving planning decision control system, the current state of the vehicle updated by the dynamic model is used as feedback, and the output control quantity of the intelligent driving planning decision control system is used as the input of the dynamic model, so that a closed loop between the intelligent driving planning decision control system and the vehicle dynamic model is formed.
The specific method for constructing the ring test communication network based on the TCP/IP in the fourth step is as follows:
because the intelligent driving planning decision control system generally comprises global planning software, local planning software and a trajectory tracking controller, communication links have been established between the global planning software, the local planning software and the trajectory tracking controller on the upper layer based on an ROS message mechanism or ZMQ communication and the like. Therefore, the ring test communication network is constructed mainly by considering the establishment of the communication link between the controller and the vehicle dynamics model to simulate the communication between the trajectory tracking controller and the CAN bus of the whole vehicle controller. Considering that an intelligent driving planning decision control system has multiple development languages such as C/C + +/C #, Java, simulink and the like, the most widely-applied TCP/IP communication is adopted to simulate the CAN bus communication between the controller and the whole vehicle controller. The process for realizing the communication between the controller and the vehicle dynamics model based on the TCP/IP comprises the following steps:
1) a parallel thread is developed in a controller to establish a TCP/IP communication server, and a TCP/IP communication client is established in a vehicle dynamics model based on a TCP/IP function and a TCP/IP communication module in simulink;
2) configuring an IP address and a port number, wherein the address is a local loopback address '127.0.0.1';
3) the method comprises the steps of establishing communication connection, realizing data interaction in a JSON (java server object) or other data formats, sending control variables such as front wheel steering angles and vehicle longitudinal speeds to a client (model) in the JSON or other data formats by a server (controller), receiving the control variables sent by the controller by the model, analyzing data, updating a vehicle state, packaging the vehicle state quantity into the JSON or other data formats and sending the vehicle state quantity to the controller.
The system is developed in C + +, and comprises global planning software, local planning software and trajectory tracking controller software, wherein a communication network between the software is established and completed based on ZMQ communication, and the system runs in a BECKHOFF CX2040 industrial personal computer.
Based on the example, the schematic diagram of the enhanced in-loop test of the intelligent driving planning decision control system is shown in fig. 1, and the specific application flow of the enhanced in-loop test method of the intelligent driving planning decision control system is as follows:
1. a map acquisition vehicle provided with the optical fiber combined inertial navigation system acquires a GPS track of a path and makes a map file;
the satellite network space XW-GI7660 optical fiber inertia/satellite combined navigation system is used as a GPS track acquisition tool, the map acquisition vehicle is formed by refitting a minibus with the length, width and height of 4745, 1810, 1845(mm), and the map acquisition place is in a long-sand intelligent network connection vehicle test area.
The specific implementation method for making the map file is as follows: after the combined inertial navigation system is electrified and initialized, the upper computer reads that the satellite state is 0B, namely the mark with good state and accurate positioning, the map acquisition vehicle slowly runs at a constant speed in a selected test area at the speed of not more than 5m/s, the vehicle position and the vehicle heading information acquired by the combined inertial navigation system are read through the RS232 serial port, and the data are recorded and stored in the file format of txt/bin. There are two main ways for collecting map files, and according to time intervals or distance intervals, considering the requirements of planning algorithm, it is generally required that the straight line is not less than one point per 5m, and the curve is not less than one point per 1 m.
2. Simulating software and hardware environment of the intelligent driving planning decision control system, and installing related software of the intelligent driving planning decision control system in the simulated environment;
the software and hardware environment for simulating the operation of the intelligent driving planning decision control system mainly considers that the original in-loop test method is directed at the MDB development process, and is directed at the code programming development mode of C + +/JAVA and the like, and based on the original in-loop test method, the source code needs to be transplanted into MATLAB, so that the code transplantation and the operation environment are difficult to ensure to have no influence on the test. The method is a testing method mainly aiming at a code programming development mode, a system source program is reserved, a simulation operation environment is constructed, and an on-loop testing method of a simulation operation environment, a virtual controller and a virtual model is designed.
An intelligent driving planning decision control system developed by the university of China and south runs in a BECKHOFF CX2040 industrial personal computer, and the hardware configuration of the running environment is a CPU (Central processing Unit) i 72715 QE 2.1.1 GHz 4 cores, a memory 4GB DDR3 RAM and a hard disk 32GB CFast flash card; the running software environment is an operating system: microsoft Windows 7 flagship edition, managed code programming model: net Framework 4.0, dynamic link library and static library: libzmq.dll, opencv _ core2410d.dll, opencv _ highgui310.dll, opencv _ imgcodescs 310.dll, tcadsdll, libzmq.lib, json _ vc71_ libmt.lib, json _ vc71_ libmtd.lib, etc.
According to a CPU (central processing unit) ladder diagram, the intelligent driving planning decision control system is built by replacing a BECKHOFF CX2040 industrial personal computer with an Ideapad 320c-15 notebook, and a software and hardware platform for enhancing the in-loop test is configured as follows: a CPU:Intel i7 7500U2.7GHz 2cores, a memory of 4GB DDR4RAM and a hard disk of 500GB HDD; in addition, a Microsoft Windows 7 flagship edition operating system is loaded in the associated Ideapad 320c-15 notebook, Microsoft.NET Framework 4.0 is used as a managed code programming model, and the dynamic link library and the static library are configured according to the environment of the industrial personal computer in which the system actually operates, so that the dynamic link library and the static library are used as the software environment in which the intelligent driving planning decision control system operates.
3. Establishing a vehicle dynamics model, and constructing a closed loop between an intelligent driving planning decision control system and the vehicle dynamics model;
the vehicle dynamic model building method mainly comprises two methods:
1) as shown in fig. 2(a), a vehicle dynamics model is built in Matlab/simulink according to a vehicle dynamics equation;
2) as shown in fig. 2(b), a vehicle dynamics model is first built in Trucksim by using a parametric modeling method, and then the model is imported into Matlab/simulink.
Considering that the vehicle dynamics model established by the Trucksim is more accurate, the vehicle dynamics model is established by the Trucksim in the embodiment, the model takes a 12m intelligent driving passenger car as a target car model, and the configuration of main parameters is shown in the table 2. According to the following parameters, the corresponding configuration needs to be completed in Trucksim.
TABLE 2 vehicle dynamics principal parameters
The inputs of the vehicle dynamic model established by the Trucksim are the longitudinal speed of the vehicle and the steering angle of the front wheels, and the outputs are the position (x, y) of the vehicle, the yaw angle of the vehicle, the transverse speed of the vehicle and the yaw speed of the vehicle.
In addition, the method of the invention is used for enhancing the in-loop test, constructing a simulated operation environment, and constructing a closed loop between an intelligent driving planning decision control system and a vehicle dynamics model after completing vehicle dynamics modeling. The method comprises the steps of taking a front wheel steering angle and a vehicle longitudinal speed output by an intelligent driving planning decision control system as input of a vehicle dynamic model, taking a current vehicle state fed back by the vehicle dynamic model as feedback of the intelligent driving planning decision control system, and establishing a closed-loop control system between the intelligent driving planning decision control system and a controlled object, wherein a communication link between the intelligent driving planning decision control system and the vehicle dynamic model is established in a Socket communication mode based on a TCP protocol. In order to verify the theoretical feasibility of the method, a function expression of the intelligent driving planning decision control system and the closed loop of the vehicle dynamics model is deduced, and the deduction process is as follows:
make the vehicle state quantity
Wherein x, y represent vehicle position,
Representing the vehicle yaw angle, v representing the vehicle lateral velocity, ω representing the yaw rate, and the vehicle dynamics equation in f (-) can be obtained:
ξ(k+1)＝f(ξ(k),δ(k),U(k)) (9)
where ξ (k) represents the vehicle state for the kth control cycle, ξ (k +1) represents the vehicle state for the kth control cycle, δ (k) represents the vehicle front wheel steering angle for the kth control cycle, and u (k) represents the vehicle longitudinal speed for the kth control cycle.
And (2) representing the control rate of the trajectory tracking control by eta (·), wherein the expression form of the input and output functions of the available trajectory tracking control software is as follows:
(δ(k),U(k))＝η(A(k),Ue(k),ξ(k)) (10)
where a (k) represents a local expected trajectory of the kth control cycle output by the local planning software, and a (k) ═ an(k),an-1(k),…,a1(k),a0(k)]，an(k),an-1(k),…,a1(k),a0(k) Expressing a polynomial coefficient corresponding to the local expected track, wherein n represents a polynomial order; u shapee(k) The desired vehicle longitudinal speed for the kth control cycle, which is indicative of the local planning software output.
Assuming that the initial state of the vehicle is xi (1), the intelligent driving planning decision control system and the closed loop function expression form of the vehicle dynamic model obtained by combining the formulas (9) and (10) are as follows:
the global planning, the expected track output by the local planning and the expected vehicle speed of the upper layer are used as the input of a track tracking controller of the intelligent driving planning decision control system, the current state of the vehicle updated by the dynamic model is used as feedback, and the output control quantity of the intelligent driving planning decision control system is used as the input of the dynamic model, so that a closed loop between the intelligent driving planning decision control system and the vehicle dynamic model is formed.
4. An in-loop test communication network is established based on TCP/IP, and a communication link between an intelligent driving planning decision control system and an intelligent driving vehicle is simulated;
the specific method for constructing the on-ring test communication network based on the TCP/IP comprises the following steps:
because the intelligent driving planning decision control system generally comprises global planning software, local planning software and a trajectory tracking controller, communication links have been established between the global planning software, the local planning software and the trajectory tracking controller on the upper layer based on an ROS message mechanism or ZMQ communication and the like. Therefore, the ring test communication network is constructed mainly by considering the establishment of the communication link between the controller and the vehicle dynamics model to simulate the communication between the trajectory tracking controller and the CAN bus of the whole vehicle controller. Considering that an intelligent driving planning decision control system has multiple development languages such as C/C + +/C #, Java, simulink and the like, the most widely-applied TCP/IP communication is adopted to simulate the CAN bus communication between the controller and the whole vehicle controller. The process for realizing the communication between the controller and the vehicle dynamics model based on the TCP/IP comprises the following steps:
1) establishing a TCP/IP communication server end by opening a parallel thread in a controller, and establishing a TCP/IP communication client end based on a TCP/IP function and a TCP/IP communication module in simulink in a vehicle dynamics model;
2) configuring an IP address and a port number, wherein the address is a local loopback address '127.0.0.1';
3) the method comprises the steps of establishing communication connection, realizing data interaction in a JSON (java server object) or other data formats, sending control variables such as front wheel steering angles and vehicle longitudinal speeds to a client (model) in the JSON or other data formats by a server (controller), receiving the control variables sent by the controller by the model, analyzing data, updating a vehicle state, packaging the vehicle state quantity into the JSON or other data formats and sending the vehicle state quantity to the controller.
In this example, a Socket communication mode based on a TCP protocol is used to construct a communication link, and a TCP communication model is shown in fig. 3, and includes the following steps:
1) the server side starts a monitoring program firstly, monitors a specified port and waits for receiving a connection request of the client side;
2) starting a client program to request to connect a designated port of a server;
3) after receiving a connection request of a client, a server establishes Socket (Socket) connection with the client;
4) after the connection is successful, the client and the server open respective input streams and output streams, wherein the input stream of the client is connected to the output stream of the server, the input stream of the server is connected to the output stream of the client, and bidirectional communication can be performed after the connection is successful.
5) And after the communication is finished, the client side and the server side are disconnected respectively.
5. Based on the simulated communication link, the vehicle dynamics model is taken as a controlled object, a test path is simulated by the manufactured map file, and the performance of the intelligent driving planning decision control system in the simulated operation environment is tested.
The test results of this example are shown in fig. 4, 5, 6, and 7, which are the global path output by the global planning software in the loop test, the local expected trajectory output by the local planning software in the loop test, the expected front wheel steering angle output by the trajectory tracking controller, and the vehicle position deviation, respectively.
Claims (5)
1. An intelligent driving planning decision control system enhanced in-loop testing method is characterized by comprising the following steps:
1) a map acquisition vehicle provided with the optical fiber combined inertial navigation system acquires a GPS track of a path and makes a map file; the specific implementation process for making the map file is as follows: reading vehicle position and course information obtained by the combined inertial navigation through an RS232 serial port, recording data and storing the data into a txt/bin file format to obtain the map file;
2) simulating software and hardware environment of the intelligent driving planning decision control system, and installing related software of the intelligent driving planning decision control system in the simulated environment;
3) establishing a vehicle dynamics model, and constructing a closed loop between an intelligent driving planning decision control system and the vehicle dynamics model; namely, the control variables which are output by the intelligent driving planning decision control system and comprise the steering angle of the front wheel and the longitudinal speed of the vehicle are used as the input of a vehicle dynamic model, and the current vehicle state fed back by the vehicle dynamic model is used as the feedback of the intelligent driving planning decision control system, so that a closed-loop control system between the intelligent driving planning decision control system and the controlled object is established; the specific implementation process comprises the following steps:
make the vehicle state quantity
Wherein x, y represent vehicle position,
Representing the vehicle yaw angle, v representing the vehicle lateral velocity, ω representing the yaw rate, and f (-) representing the vehicle dynamics equation, given by: ξ (k +1) ═ f (ξ (k), δ (k), u (k)); where ξ (k) represents the vehicle state in the kth control cycle, ξ (k +1) represents the vehicle state in the kth control cycle, δ (k) represents the vehicle front wheel steering angle in the kth control cycle, and u (k) represents the vehicle longitudinal speed in the kth control cycle; and (2) representing the control rate of the trajectory tracking control by eta (·), wherein the expression form of an input and output function of the trajectory tracking control software is as follows: (δ (k), U (k) ═ η (a (k)), Ue(k) ξ (k)); where a (k) represents a local expected trajectory of the kth control cycle output by the local planning software, and a (k) ═ an(k),an-1(k),…,a1(k),a0(k)]，an(k),an-1(k),…,a1(k),a0(k) Expressing a polynomial coefficient corresponding to the local expected track, wherein n represents a polynomial order; u shapee(k) A desired vehicle longitudinal speed representing a kth control cycle of the local planning software output;
the initial state of the vehicle is set as xi (1), and the function expression form of the intelligent driving planning decision control system and the closed loop of the vehicle dynamic model is as follows:
4) an in-loop test communication network is established based on TCP/IP, and a communication link between an intelligent driving planning decision control system and an intelligent driving vehicle is simulated;
5) based on the simulated communication link, the vehicle dynamics model is taken as a controlled object, a test path is simulated by the manufactured map file, and the performance of the intelligent driving planning decision control system in the simulated operation environment is tested.
2. The intelligent driving planning decision-making control system enhanced in-loop testing method of claim 1, wherein the map file is collected in a manner of: at time intervals or at distance intervals, the straight line is not less than one point every 5m, and the curve is not less than one point every 1 m.
3. The intelligent driving planning decision control system enhanced in-loop testing method of claim 1, wherein in step 3), a vehicle dynamics model is built in Matlab/simulink according to a vehicle dynamics equation; alternatively, a vehicle dynamics model is built in Trucksim or Carsim, and then the model is imported into Matlab/simulink.
4. The intelligent driving planning decision control system enhanced in-loop testing method according to claim 1, wherein in the step 4), the specific steps of constructing the in-loop testing communication network based on the TCP/IP are as follows: firstly, a parallel thread is opened up at a server, a TCP/IP communication server is established, a TCP/IP communication client is established in a vehicle dynamics model based on a TCP/IP function and a TCP/IP communication module in a simulink, then an IP address and a port number are configured, the address is a local loopback address, finally, communication connection is established, data interaction is realized, the server sends a front wheel steering angle and a vehicle longitudinal speed to the client, the client receives a control variable sent by the server, analyzes data, updates a vehicle state and sends the vehicle state quantity to the server.
5. The intelligent driving planning decision control system enhanced in-loop testing method of claim 4, wherein data interaction is implemented in JSON format.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811080324.5A CN109445404B (en) | 2018-09-17 | 2018-09-17 | Enhanced in-loop test method for intelligent driving planning decision control system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811080324.5A CN109445404B (en) | 2018-09-17 | 2018-09-17 | Enhanced in-loop test method for intelligent driving planning decision control system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109445404A CN109445404A (en) | 2019-03-08 |
| CN109445404B true CN109445404B (en) | 2021-06-11 |
Family
ID=65532605
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811080324.5A Active CN109445404B (en) | 2018-09-17 | 2018-09-17 | Enhanced in-loop test method for intelligent driving planning decision control system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109445404B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110083142B (en) * | 2019-04-29 | 2020-03-20 | 清华大学 | Test vehicle control method for intelligent network connection test |
| CN111785027B (en) * | 2019-09-17 | 2023-01-06 | 上海森首科技股份有限公司 | Automatic driving closed-loop information system |
| CN111586130B (en) * | 2020-04-29 | 2021-04-02 | 清华大学 | Performance evaluation method and device of longitudinal control algorithm in mixed traffic scene |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2457328B1 (en) * | 2009-07-24 | 2014-03-26 | Technion Research and Development Foundation, Ltd. | Ultra-high-speed photonic-enabled adc based on multi-phase interferometry |
| JP4876159B2 (en) * | 2009-09-04 | 2012-02-15 | クラリオン株式会社 | Car navigation system |
| CN103065627B (en) * | 2012-12-17 | 2015-07-29 | 中南大学 | Special purpose vehicle based on DTW and HMM evidence fusion is blown a whistle sound recognition methods |
| CN105182968B (en) * | 2015-09-24 | 2018-08-10 | 吉林大学 | A kind of hardware in loop property test platform suitable for automobile C-EPS systems |
| CN106227189B (en) * | 2016-07-29 | 2018-11-09 | 中山大学 | A kind of Two axle drive electric vehicle hardware-in―the-loop test system synchronization control method |
| KR101856348B1 (en) * | 2016-08-26 | 2018-06-25 | 현대자동차주식회사 | Method for verifying of Actuator control data |
| CN106292333A (en) * | 2016-09-13 | 2017-01-04 | 浙江吉利控股集团有限公司 | ESC hardware-in-the-loop test system and ESC hardware-in-the-loop test method |
| CN107991898A (en) * | 2016-10-26 | 2018-05-04 | 法乐第(北京)网络科技有限公司 | A kind of automatic driving vehicle simulating test device and electronic equipment |
| CN106873397B (en) * | 2017-01-23 | 2020-02-14 | 同济大学 | Intelligent networking automobile hardware-in-loop acceleration loading simulation test system |
| CN107727417B (en) * | 2017-09-11 | 2020-07-31 | 江苏大学 | Hardware-in-loop simulation test platform for man-machine co-driving steering system |
| CN108267325A (en) * | 2018-01-29 | 2018-07-10 | 上海测迅汽车科技有限公司 | Unmanned vehicle material object is in ring test method |
| CN108509869A (en) * | 2018-03-14 | 2018-09-07 | 中南大学 | Feature set based on OpenBCI optimizes on-line training method |
| CN108319259B (en) * | 2018-03-22 | 2021-09-03 | 上海科梁信息科技股份有限公司 | Test system and test method |
| CN108445885A (en) * | 2018-04-20 | 2018-08-24 | 鹤山东风新能源科技有限公司 | A kind of automated driving system and its control method based on pure electric vehicle logistic car |
-
2018
- 2018-09-17 CN CN201811080324.5A patent/CN109445404B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109445404A (en) | 2019-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109445404B (en) | Enhanced in-loop test method for intelligent driving planning decision control system | |
| CN112100856B (en) | Automatic driving joint simulation method based on multiple platforms | |
| CN108803607B (en) | Multifunctional simulation system for automatic driving | |
| CN111665738A (en) | In-loop simulation system and information processing method and device thereof | |
| JP2021526699A (en) | Data collection methods, devices, devices, programs and computer-readable storage media | |
| CN109656148A (en) | The emulation mode of the Dynamic Traffic Flow scene of automatic Pilot | |
| CN108446463B (en) | Microcosmic traffic flow collaborative simulation platform, simulation method and safety evaluation method | |
| Ma et al. | ROS-based multi-robot system simulator | |
| CN104133379A (en) | Simulation method for four-rotor aircraft | |
| CN102306216A (en) | Multi-rule simulation test system of lunar vehicle | |
| KR20210063701A (en) | the method for configuring and controling real-time distributed autonomous driving simulation framework | |
| CN113515038B (en) | Vehicle lane changing method, device, equipment and readable storage medium | |
| CN113419518B (en) | VIL test platform based on VTS | |
| CN115344503B (en) | Traffic flow simulation system and simulation method for automatic driving planning control test | |
| CN112035951A (en) | Simulation platform and simulation method for automatic driving algorithm verification | |
| CN112925221A (en) | Auxiliary driving closed loop test method based on data reinjection | |
| Bulsara et al. | Obstacle avoidance using model predictive control: An implementation and validation study using scaled vehicles | |
| CN107194108B (en) | Intelligent traffic simulation system and simulation method based on micro swarm robots | |
| Artunedo et al. | Advanced co-simulation framework for cooperative maneuvers among vehicles | |
| CN112597005A (en) | Closed-loop simulation test system and method for automatic driving full-stack algorithm | |
| CN112000026A (en) | Mars GNC system physical model construction method based on information physical fusion | |
| CN114670797B (en) | Vehicle brake test control system, method, device, electronic equipment and storage medium | |
| CN115826583A (en) | Automatic driving vehicle formation method based on point cloud map | |
| CN115098941A (en) | Unmanned aerial vehicle digital twin control method and platform for agile deployment of intelligent algorithm | |
| CN112230659A (en) | Method for accurately planning movement track, intelligent control equipment and automatic driving vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |