CN114895665A - Horizontal and vertical coordinated control method and system of unmanned vehicle on large curvature curve - Google Patents

Abstract

The invention relates to a transverse and longitudinal cooperative control method and a system for an unmanned vehicle in a large-curvature curve, which comprises the following steps: establishing a kinematics transverse-longitudinal coupling MPC model containing centripetal acceleration constraint, and tracking the track of the unmanned vehicle according to the MPC model; establishing a FNN model based on laser radar point cloud data, and carrying out real-time obstacle avoidance on the unmanned vehicle according to the FNN model; and (3) fusing the MPC model and the FNN model, and carrying out transverse and longitudinal cooperative control on the unmanned vehicle under a large-curvature curve to realize safe driving. The invention can realize safe running of the unmanned vehicle under the condition of track loss according to the planned track state. The invention can be applied in the field of unmanned vehicle control.

Description

Horizontal and vertical coordinated control method and system of unmanned vehicle on large curvature curve

technical field

The invention relates to the field of unmanned vehicle control, in particular to a horizontal and vertical coordinated control method and system for unmanned vehicles on large curvature curves.

Background technique

Ensuring the safety of unmanned vehicles driving on curves is an urgent need for vehicle control. The unmanned vehicle is a non-holonomic motion constraint system with multiple degrees of freedom, highly nonlinear characteristics, and model parameter uncertainty. The underlying motion control problem is a complex nonlinear system control problem with multiple inputs and multiple outputs. At present, the motion control of unmanned vehicles is mostly concentrated on the traditional control algorithm of horizontal and vertical independent control. Although it simplifies the complexity of a single problem and can achieve better results on common roads, it has system nonlinear characteristics, Due to the time delay phenomenon and random uncertainty, it is difficult to achieve a good control effect on some specific roads, such as curves with large curvature and small sections of passable areas. The data-driven intelligent control algorithm can solve the problem of complex model establishment and make up for the shortcomings of modern control theory. The combination of intelligent control algorithm and traditional control algorithm can better solve the problem of model establishment and control accuracy.

As the earliest proposed intelligent control algorithm, fuzzy control has been unanimously recognized by researchers for its robustness. However, due to the extremely high requirements for input rules and experience, fuzzy control does not have a mature set of application scenarios in many fields. The fuzzy rule base needs to be established by researchers according to research and application scenarios. In order to meet the requirements of real-time online optimization control and improve the vehicle stability control performance under multiple constraints, model predictive control, as a new control method with excellent performance, can effectively scroll optimization under dynamic constraint sets. The lateral control algorithm mostly relies on the reliable trajectory given by the planning layer, and does not consider the safety control strategy when the trajectory is lost.

Temporary roads are an indispensable part of traffic roads, and their characteristics such as unpredictability and uncertainty of curve curvature can easily lead to the failure of trajectory planning at the planning layer.

SUMMARY OF THE INVENTION

In view of the above problems, the purpose of the present invention is to provide a lateral and longitudinal coordinated control method and system for an unmanned vehicle with a large curvature curve, which can realize the safety of the unmanned vehicle in the case of the loss of the trajectory according to the planned trajectory state. drive.

In order to achieve the above object, the present invention adopts the following technical solutions: a lateral and longitudinal coordinated control method for an unmanned vehicle on a curve with a large curvature, which includes: establishing a kinematics lateral and longitudinal coupling MPC model containing centripetal acceleration constraints, according to The MPC model performs trajectory tracking of unmanned vehicles; builds an FNN model based on lidar point cloud data, and performs real-time obstacle avoidance for unmanned vehicles according to the FNN model; fuses the MPC model with the FNN model , to perform horizontal and vertical coordinated control of unmanned vehicles under large curvature curves to achieve safe driving.

Further, the establishment of an MPC model of horizontal and vertical coupling of kinematics containing centripetal acceleration constraints includes:

Establish a vehicle kinematics model coupled horizontally and vertically, and add longitudinal speed control parameters to the kinematics model;

Linearizing the kinematics model to obtain a corresponding state space equation;

Set the centripetal acceleration constraint of lateral and longitudinal control to prevent the unmanned vehicle from losing the feasible region during the turning process;

The horizontal and vertical comprehensive evaluation function index is constructed, and the objective function is set. By adjusting the minimum dynamic adjustment coefficient of the gradient, the evaluation function index is smaller than the preset value, and the error between the current control state and the target control state is minimized.

Further, the centripetal acceleration constraint of the lateral and longitudinal control is:

Among them, a _max is the maximum centripetal acceleration, v _x is the longitudinal speed of the vehicle, L is the length of the vehicle, and δ is the target steering angle of the vehicle.

Further, the construction of the horizontal and vertical comprehensive evaluation function index and the setting of the objective function include:

After constructing the horizontal and vertical comprehensive evaluation function indicators, after setting N _p prediction time domains and N _c control time domains, the objective function is set as:

表示控制效率，k表示时域系数。In the formula, Y is the evaluation function index of the MPC model, Q is the gradient minimum dynamic adjustment coefficient of the curve tracking error, R is the gradient minimum dynamic adjustment coefficient of the control efficiency, X(k+i)-X _ref (k+i) means track tracking effect,

represents the control efficiency, and k represents the time domain coefficient.

Further, the establishment of the FNN model, the obstacle avoidance is performed according to the FNN model, including:

Collect data sets including obstacle position information, vehicle speed and corner feedback information as the training set and validation set of the FNN model;

Using lidar as the environment to perceive the road environment, to determine the fuzzy domain of the target perception obstacle information;

Combined with the logical rationality and transparency of fuzzy logic control and the self-adaptation of artificial neural network parameters, a fuzzy neural network model with the position information of obstacles relative to the lidar as input and the target speed and rotation angle as output is set up;

Input the pre-collected obstacle position, speed and turning angle data into the fuzzy neural network model, obtain the actual output value of each group of data network and the expected output value in the data through self-supervised learning, and compare and calculate the two error, the coefficients and parameters of the fuzzy neural network model are corrected based on the error, and the real-time obstacle avoidance of the unmanned vehicle is realized through the trained fuzzy neural network model.

Further, the setting of the fuzzy neural network model includes:

According to the input quantity of m-dimensional obstacle position information, the membership degree of each input variable is calculated according to the fuzzy rules;

Perform fuzzy calculation on each membership degree, and the fuzzy operator adopts the continuous multiplication operator;

Calculate the network output value according to the fuzzy calculation result;

Calculate the error between the network output and the expected output;

According to the error obtained from each input, the coefficients and parameters are corrected, and the corrected fuzzy neural network model is obtained.

Further, the MPC model is integrated with the FNN model to perform horizontal and vertical coordinated control of the unmanned vehicle under large curvature curves, including:

Judging whether there is a successfully planned trajectory, if the time of trajectory planning failure is less than the preset time t _set , it is regarded as the normal jitter of the plan, and the motion state of the previous moment is maintained to continue moving; if the time of trajectory planning failure is greater than the preset time t _set , it is considered that the trajectory planning has failed. At this time, the trajectory tracking mode is converted into the obstacle avoidance mode, and the FNN model is used to avoid obstacles of the unmanned vehicle; if the trajectory planning is successful, the stable trajectory duration is judged. When the stable time parameter is greater than several times the preset time t _set , the MPC model is used to track the trajectory of the unmanned vehicle.

A lateral and longitudinal coordinated control system for an unmanned vehicle on a curve with large curvature, comprising: a trajectory tracking module, establishing an MPC model of horizontal and vertical coupling of kinematics including centripetal acceleration constraints, and performing unmanned driving according to the MPC model Trajectory tracking of vehicles; obstacle avoidance module, establishes an FNN model based on lidar point cloud data, and performs real-time obstacle avoidance of unmanned vehicles according to the FNN model; fusion module, fuses the MPC model and the FNN model, in large curvature curves Under the road, the unmanned vehicles are controlled horizontally and vertically to achieve safe driving.

A computer-readable storage medium storing one or more programs comprising instructions that, when executed by a computing device, cause the computing device to perform any of the above methods.

A computing device comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the The one or more programs include instructions for performing any of the above-described methods.

The present invention has the following advantages due to taking the above technical solutions:

1. The present invention improves the traditional MPC algorithm by establishing a kinematics horizontal and vertical coupling model, and realizes the horizontal and vertical coordinated control of trajectory tracking, so as to ensure that the unmanned vehicle can drive normally when there is a trajectory.

2. The present invention applies the FNN algorithm to the motion control of the unmanned vehicle, and improves the accuracy of the model and the obstacle avoidance performance of the unmanned vehicle through offline training of the model.

3. Aiming at the problem of track loss during curve driving, the present invention sets a safety control strategy of mutual coordination and conversion between the track tracking controller and the obstacle avoidance controller, so that the unmanned vehicle can drive safely even when the track is lost on the temporary road. .

Description of drawings

1 is an overall flow chart of a horizontal and vertical coordinated control method in an embodiment of the present invention;

2 is a detailed flowchart of a horizontal and vertical coordinated control method in an embodiment of the present invention;

FIG. 3 is a fuzzy field of a lidar sensing area in an embodiment of the present invention;

4 is a structural diagram of a fuzzy neural network in an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a computing device in an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, embodiments of the present invention. Based on the described embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art fall within the protection scope of the present invention.

It should be noted that the terminology used herein is for the purpose of describing specific embodiments only, and is not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural as well, furthermore, it is to be understood that when the terms "comprising" and/or "including" are used in this specification, it indicates that There are features, steps, operations, devices, components and/or combinations thereof.

The present invention proposes a lateral and longitudinal coordinated control method and system for an unmanned vehicle on a curve with a large curvature, which is mainly aimed at the safe driving of the unmanned vehicle on a curve with a large curvature of a temporary road.

Aiming at the unpredictability of temporary roads and the uncertainty of curve curvature, which easily lead to the failure of trajectory planning in the planning layer, the present invention adds a safety redundancy method for obstacle avoidance in the control layer, based on the model predictive control algorithm and the fuzzy neural network algorithm. , according to the planned trajectory state, to realize the safe driving of the unmanned vehicle in the case of trajectory loss.

In an embodiment of the present invention, a lateral and longitudinal coordinated control method for an unmanned vehicle on a curve with a large curvature is provided. This embodiment is illustrated by applying the method to a terminal. It can be understood that this method can also Applied to the server, it can also be applied to the system including the terminal and the server, and is realized through the interaction between the terminal and the server. In this embodiment, the present invention is used to realize the safe driving of the unmanned vehicle on the large curvature curve of the temporary road. As shown in FIG. 1 , the method includes the following steps:

1) Establish an MPC model of horizontal and vertical coupling of kinematics including centripetal acceleration constraints, and track the trajectory of the unmanned vehicle according to the MPC model;

2) Establish an FNN model based on lidar point cloud data, and perform real-time obstacle avoidance for unmanned vehicles according to the FNN model;

3) The MPC model is integrated with the FNN model, and the unmanned vehicle is controlled horizontally and vertically under large curvature curves to achieve safe driving.

In the above step 1), the tracking of the curve trajectory is realized by improving the MPC control algorithm. On the basis of the basic dynamic kinematics model of the vehicle, the horizontal and vertical kinematics coupling conditions and the centripetal acceleration constraint conditions are added to establish the horizontal and vertical coupling conditions. vehicle kinematics model. Since the vehicle is a typical complex nonlinear system, it is necessary to transform the nonlinear model into a linear model through Taylor expansion of the established model and then establish the state space equation. The MPC control algorithm is used to predict, feedback correction and rolling optimization of the established model, and then obtain the optimal solution for the established objective function, so as to realize the tracking of the curve trajectory of the unmanned vehicle and improve the tracking accuracy.

Specifically, as shown in Fig. 2, establishing an MPC model of horizontal and vertical coupling of kinematics including centripetal acceleration constraints includes the following steps:

1.1) Establish a vehicle kinematics model of horizontal and vertical coupling, and add longitudinal speed control parameters to the kinematics model;

Since the original kinematics model only includes the lateral control of the vehicle, and the vehicle is a complex horizontal and vertical coupling system, the longitudinal speed control parameters are added on the basis of the lateral kinematics model to realize the horizontal and vertical coordinated control of the unmanned vehicle. . In this embodiment, the vehicle kinematics model of the horizontal and vertical coupling is:

其中(x，y)为目标路径点相对于无人车辆的坐标位置，

为目标路径点相对于车辆的偏航角，v_x为车辆纵向速度；控制量u＝[δ，a]，其中δ为车辆目标转向角度，L是车辆长度，a为加速度指令，后续会通过查表映射到扭矩控制指令。state quantity

where (x, y) is the coordinate position of the target waypoint relative to the unmanned vehicle,

is the yaw angle of the target waypoint relative to the vehicle, v _x is the longitudinal speed of the vehicle; the control variable u=[δ, a], where δ is the vehicle target steering angle, L is the vehicle length, and a is the acceleration command, which will be passed later. The look-up table maps to torque control commands.

1.2) Linearize the kinematics model to obtain the corresponding state space equation;

Since the vehicle system of Ackerman steering is a nonlinear system, the kinematics model cannot be directly transformed into a state-space equation, and a series of mathematical methods need to be transformed to linearize it, and then expressed as a state-space equation. Perform Taylor expansion at (X _r , ur _r ), as shown in equations (2) to (5):

Among them, X _r represents the state quantity at any point in the state space equation, and ur _r represents the control quantity corresponding to any point in the state space equation.

相减，得到式(6)及其新的线性化状态空间方程式(7)：Combine equation (2) with the state variables of random Taylor expansion points

Subtracted to get equation (6) and its new linearized state space equation (7):

再根据前向欧拉法进行离散化，可以计算出下一时刻的预测值

如式(8)所示：make

Then according to the forward Euler method for discretization, the predicted value of the next moment can be calculated

As shown in formula (8):

Among them, T represents the sampling period between two adjacent time domain states, A represents the state matrix of the system, B represents the control matrix of the system, and k represents the time domain coefficient.

1.3) In order to ensure the coordination of speed and turning angle and the loss of the real feasible region, the centripetal acceleration constraints of the lateral and longitudinal control are set to prevent the unmanned vehicle from losing the feasible region in the process of turning;

Among them, the centripetal acceleration constraint of lateral and longitudinal control is:

Among them, a _max is the maximum centripetal acceleration, v _x is the longitudinal speed of the vehicle, L is the length of the vehicle, and δ is the target steering angle of the vehicle.

1.4) Construct the horizontal and vertical comprehensive evaluation function index, set the objective function, and make the evaluation function index smaller than the preset value by adjusting the minimum dynamic adjustment coefficient of the gradient, so as to minimize the error between the current control state and the target control state;

Among them, the horizontal and vertical comprehensive evaluation function indicators are constructed, and after setting N _p prediction time domains and N _c control time domains, the objective function is set as:

表示控制效率，k表示时域系数。In the formula, Y is the evaluation function index of the MPC model, Q is the gradient minimum dynamic adjustment coefficient of the curve tracking error, R is the gradient minimum dynamic adjustment coefficient of the control efficiency, X(k+i)-X _ref (k+i) means track tracking effect,

represents the control efficiency, and k represents the time domain coefficient.

By dynamically adjusting the coefficient Q and the coefficient R with the minimum gradient, as shown in equations (11) and (12), the evaluation function index Y is kept relatively small (that is, less than the preset value), and the difference between the current control state and the target control state is realized. Error is minimal.

In the formula, γ represents the gradient adjustment parameter of the coefficient Q, and δ represents the gradient adjustment parameter of the coefficient R.

In the above step 2), for the situation of trajectory loss, the vehicle obstacle avoidance research is carried out based on the FNN model. Use lidar to perceive the road environment, and construct a fuzzy neural network controller that takes the position (angle and distance) information of obstacles relative to lidar as input and the target speed and rotation angle as output. The neural network mainly includes an input layer, a fuzzification layer, a fuzzy rule layer, a defuzzification layer and an output layer. Through the error between the network output and the actual reference value during each training, the parameters of the weight coefficient and the membership function are corrected. Make it finally close to the true output of the dataset. Finally, the real-time obstacle avoidance of unmanned vehicles is realized through the trained model.

Specifically, as shown in Figure 2, establishing an FNN model and avoiding obstacles according to the FNN model includes the following steps:

2.1) Collect data sets including obstacle position (distance and angle) information, vehicle speed and angle feedback information as the training set and validation set of the FNN model;

In this embodiment, data collection is performed by manual driving.

2.2) Using lidar as the environment to perceive the road environment, to determine the fuzzy domain of target perception obstacle information;

The specific parameter values of the input and output variables are normalized and mapped to appropriate fuzzy sets.

In this embodiment, the area of 180° in front of the lidar is taken as the region of interest, and it is evenly divided into five regions from d ₁ to d ₁₈ to represent the angle of the obstacle relative to the vehicle. The fuzzy set [L8, L7, ..., L1 , L, R, …, R7, R8], the distance is divided into five regions, which are represented by fuzzy sets [DN, DNM, DM, DMF, DF], as shown in Figure 3.

2.3) Combining the logic rationality and transparency of fuzzy logic control and the self-adaptation of artificial neural network parameters, set up a fuzzy neural network model that takes the position information of the obstacle relative to the lidar as the input and the target speed and rotation angle as the output;

The fuzzy neural network model is a five-layer fuzzy neural network including an input layer, a fuzzification layer, a fuzzy rule layer, a defuzzification layer and an output layer.

Among them, setting the fuzzy neural network model includes the following steps:

2.3.1) For the m-dimensional input variable x=[x ₁ , x ₂ ,..., x _m ], calculate the membership degree of each input variable x _j according to the fuzzy rule, j∈m;

采用高斯型，如式(13)所示：Among them, the membership function

The Gaussian type is used, as shown in equation (13):

分别为隶属度函数的中心和宽度；m为输入参数的维数(即特征向量数)；n为模糊子集数。in,

are the center and width of the membership function, respectively; m is the dimension of the input parameter (ie, the number of eigenvectors); n is the number of fuzzy subsets.

2.3.2) Perform fuzzy calculation on each membership degree, and the fuzzy operator ω ⁱ adopts the continuous multiplication operator, as shown in formula (14):

2.3.3) Calculate the network output value _yi according to the fuzzy calculation result;

表示第i次计算时第m个输入量的权重系数。in,

Indicates the weight coefficient of the mth input in the ith calculation.

2.3.4) Calculate the error e between the network output and the expected output as shown in equation (16):

where y _d is the expected output of the network and y _c is the actual output of the network.

2.3.5) Correct the coefficients and parameters according to the error obtained by each input, and obtain the corrected fuzzy neural network model;

As shown in formulas (17) to (19):

表示第i次计算时第j个输入量的权重系数，

表示第i次计算时第j个输入量的隶属度函数的中心，

表示第i次计算时第j个输入量的隶属都函数的宽度。in,

Represents the weight coefficient of the jth input in the ith calculation,

represents the center of the membership function of the jth input in the ith calculation,

Indicates the width of the membership function of the jth input in the ith calculation.

2.4) Input the pre-collected lidar-based obstacle position, unmanned vehicle speed and turning angle data into the fuzzy neural network model (as shown in Figure 4), and obtain the actual data of each group of data networks through self-supervised learning. The output value and the expected output value in the data are compared to calculate the error, and the coefficients and parameters of the fuzzy neural network model are corrected based on the error, and the real-time avoidance of the unmanned vehicle is realized through the trained fuzzy neural network model model. barrier.

In the above step 3), the MPC model is integrated with the FNN model, and the unmanned vehicle is controlled horizontally and vertically under large curvature curves, specifically:

Since large curvature curves easily lead to the failure of trajectory planning, first determine whether there is a successfully planned trajectory. If the time of trajectory planning failure is less than the preset time t _set , it is regarded as the normal jitter of the plan, and the motion state at the previous moment is maintained and continues to move. ; If the time of trajectory planning failure is greater than the preset time t _set , it is determined that the trajectory planning has failed. At this time, the control mode will be switched, the trajectory tracking mode will be converted into the obstacle avoidance mode, and the FNN model will be used for the unmanned vehicle. Obstacle avoidance; if the trajectory planning is successful, determine the duration of the stable trajectory, and if the stable time parameter is greater than several times the preset time t _set , use the MPC model to track the trajectory of the unmanned vehicle.

In an embodiment of the present invention, a lateral and longitudinal coordinated control system for an unmanned vehicle on a curve with large curvature is provided, which includes:

The trajectory tracking module establishes an MPC model of horizontal and vertical coupling of kinematics including centripetal acceleration constraints, and performs trajectory tracking of unmanned vehicles according to the MPC model;

Obstacle avoidance module, establish FNN model based on lidar point cloud data, and perform real-time obstacle avoidance of unmanned vehicles according to the FNN model;

The fusion module integrates the MPC model and the FNN model to perform horizontal and vertical coordinated control of unmanned vehicles under large curvature curves to achieve safe driving.

The system provided in this embodiment is used to execute the foregoing method embodiments. For specific processes and details, please refer to the foregoing embodiments, which will not be repeated here.

As shown in FIG. 5 , which is a schematic structural diagram of a computing device provided in an embodiment of the present invention, the computing device may be a terminal, which may include: a processor, a communications interface, a memory, a display screen and input device. Among them, the processor, the communication interface and the memory complete the communication with each other through the communication bus. The processor is used to provide computing and control capabilities. The memory includes a non-volatile storage medium and an internal memory, the non-volatile storage medium stores an operating system and a computer program, the computer program is executed by the processor to implement a control method; the internal memory is non-volatile The operating system and computer program in the storage medium provide an environment for execution. The communication interface is used for wired or wireless communication with an external terminal, and the wireless communication can be realized by WIFI, a management network, NFC (Near Field Communication) or other technologies. The display screen may be a liquid crystal display screen or an electronic ink display screen, and the input device may be a touch layer covered on the display screen, a button, a trackball or a touchpad set on the casing of the computing device, or an external Keyboard, trackpad or mouse, etc. The processor can call the logic instructions in the memory to execute the following methods: establish an MPC model of horizontal and vertical coupling of kinematics including centripetal acceleration constraints, and track the trajectory of the unmanned vehicle according to the MPC model; FNN model, according to the FNN model for real-time obstacle avoidance of unmanned vehicles; the MPC model and the FNN model are integrated to perform horizontal and vertical coordinated control of unmanned vehicles under large curvature curves to achieve safe driving.

In addition, the above-mentioned logic instructions in the memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes .

Those skilled in the art can understand that the structure shown in FIG. 5 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computing device to which the solution of the present application is applied. The specific computing device may be Include more or fewer components than shown in the figures, or combine certain components, or have a different arrangement of components.

In one embodiment of the present invention, there is provided a computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, when the program instructions When executed by a computer, the computer can execute the methods provided by the above method embodiments, for example, including: establishing an MPC model of kinematics horizontal and vertical coupling including centripetal acceleration constraints, and tracking the trajectory of the unmanned vehicle according to the MPC model; The FNN model is established from the lidar point cloud data, and the real-time obstacle avoidance of the unmanned vehicle is carried out according to the FNN model; the MPC model is integrated with the FNN model, and the unmanned vehicle is controlled horizontally and vertically under large curvature curves to achieve safety. drive.

In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided, where the non-transitory computer-readable storage medium stores server instructions, the computer instructions cause a computer to execute the methods provided in the above embodiments, for example, including : Establish an MPC model with horizontal and vertical coupling of kinematics including centripetal acceleration constraints, and track the trajectory of unmanned vehicles according to the MPC model; build an FNN model based on lidar point cloud data, and perform real-time avoidance of unmanned vehicles according to the FNN model. The MPC model and the FNN model are integrated to perform horizontal and vertical coordinated control of unmanned vehicles under large curvature curves to achieve safe driving.

The implementation principle and technical effect of the computer-readable storage medium provided by the above-mentioned embodiments are similar to those of the above-mentioned method embodiments, and details are not described herein again.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow of a flowchart and/or a block or blocks of a block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory result in an article of manufacture comprising instruction means, the instructions The apparatus implements the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded on a computer or other programmable data processing device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process such that The instructions provide steps for implementing the functions specified in the flow or blocks of the flowcharts and/or the block or blocks of the block diagrams.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. a lateral and longitudinal coordinated control method of an unmanned vehicle of a large curvature curve, is characterized in that, comprising: Establishing an MPC model of horizontal and vertical coupling of kinematics containing centripetal acceleration constraints, and tracking the trajectory of the unmanned vehicle according to the MPC model; Establish an FNN model based on lidar point cloud data, and perform real-time obstacle avoidance of unmanned vehicles according to the FNN model; The MPC model is integrated with the FNN model, and the unmanned vehicle is controlled horizontally and vertically under large curvature curves to achieve safe driving. 2. The lateral and longitudinal coordinated control method of the unmanned vehicle of the large curvature curve as claimed in claim 1, wherein the establishment of the MPC model of the kinematics lateral and longitudinal coupling comprising centripetal acceleration constraints, comprising: Establish a vehicle kinematics model coupled horizontally and vertically, and add longitudinal speed control parameters to the kinematics model; Linearizing the kinematics model to obtain a corresponding state space equation; Set the centripetal acceleration constraint of lateral and longitudinal control to prevent the unmanned vehicle from losing the feasible region during the turning process; The horizontal and vertical comprehensive evaluation function index is constructed, and the objective function is set. By adjusting the minimum dynamic adjustment coefficient of the gradient, the evaluation function index is smaller than the preset value, and the error between the current control state and the target control state is minimized. 3. The lateral and longitudinal coordinated control method of an unmanned vehicle on a large curvature curve according to claim 2, wherein the centripetal acceleration constraint of the lateral and longitudinal control is:

Among them, a _max is the maximum centripetal acceleration, v _x is the longitudinal speed of the vehicle, L is the length of the vehicle, and δ is the target steering angle of the vehicle. 4. The horizontal and vertical coordinated control method for an unmanned vehicle on a large curvature curve as claimed in claim 2, wherein said constructing a horizontal and vertical comprehensive evaluation function index, and setting an objective function, comprising: After constructing the horizontal and vertical comprehensive evaluation function indicators, after setting N _p prediction time domains and N _c control time domains, the objective function is set as:

In the formula, Y is the evaluation function index of the MPC model, Q is the gradient minimum dynamic adjustment coefficient of the curve tracking error, R is the gradient minimum dynamic adjustment coefficient of the control efficiency, X(k+i)-X _ref (k+i) means track tracking effect,

represents the control efficiency, and k represents the time domain coefficient. 5. The horizontal and vertical coordinated control method of the unmanned vehicle of the large curvature curve as claimed in claim 1, it is characterized in that, described establishing FNN model, carry out obstacle avoidance according to described FNN model, comprising: Collect data sets including obstacle position information, vehicle speed and corner feedback information as the training set and validation set of the FNN model; Using lidar as the environment to perceive the road environment, to determine the fuzzy domain of the target perception obstacle information; Combined with the logical rationality and transparency of fuzzy logic control and the self-adaptation of artificial neural network parameters, a fuzzy neural network model with the position information of obstacles relative to the lidar as input and the target speed and rotation angle as output is set up; Input the pre-collected obstacle position, speed and turning angle data into the fuzzy neural network model, obtain the actual output value of each group of data network and the expected output value in the data through self-supervised learning, and compare and calculate the two error, the coefficients and parameters of the fuzzy neural network model are corrected based on the error, and the real-time obstacle avoidance of the unmanned vehicle is realized through the trained fuzzy neural network model. 6. The horizontal and vertical coordinated control method of the unmanned vehicle of the large-curvature curve as claimed in claim 5, wherein the setting of the fuzzy neural network model comprises: According to the input quantity of m-dimensional obstacle position information, the membership degree of each input variable is calculated according to the fuzzy rules; Perform fuzzy calculation on each membership degree, and the fuzzy operator adopts the continuous multiplication operator; Calculate the network output value according to the fuzzy calculation result; Calculate the error between the network output and the expected output; According to the error obtained from each input, the coefficients and parameters are corrected, and the corrected fuzzy neural network model is obtained. 7. The horizontal and vertical coordinated control method of the unmanned vehicle on the large curvature curve according to claim 1, wherein the MPC model is fused with the FNN model, and the unmanned vehicle is horizontally and vertically controlled under the large curvature curve. Vertical coordinated control, including: Judging whether there is a successfully planned trajectory, if the time of trajectory planning failure is less than the preset time t _set , it is regarded as the normal jitter of the plan, and the motion state of the previous moment is maintained to continue moving; if the time of trajectory planning failure is greater than the preset time t _set , it is considered that the trajectory planning has failed. At this time, the trajectory tracking mode is converted into the obstacle avoidance mode, and the FNN model is used to avoid obstacles of the unmanned vehicle; if the trajectory planning is successful, the stable trajectory duration is judged. When the stable time parameter is greater than several times the preset time t _set , the MPC model is used to track the trajectory of the unmanned vehicle. 8. A lateral and longitudinal coordinated control system for an unmanned vehicle on a large curvature curve, characterized in that it comprises: A trajectory tracking module, which establishes an MPC model of horizontal and vertical coupling of kinematics containing centripetal acceleration constraints, and performs trajectory tracking of the unmanned vehicle according to the MPC model; The obstacle avoidance module establishes an FNN model based on the lidar point cloud data, and performs real-time obstacle avoidance of the unmanned vehicle according to the FNN model; The fusion module integrates the MPC model and the FNN model to perform horizontal and vertical coordinated control of unmanned vehicles under large curvature curves to achieve safe driving. 9. A computer-readable storage medium storing one or more programs, characterized in that the one or more programs comprise instructions that, when executed by a computing device, cause the computing device to perform as claimed in the claims Any of the methods described in 1 to 7. 10. A computing device, comprising: one or more processors, a memory, and one or more programs, wherein one or more programs are stored in the memory and configured as the one or more programs A processor executes the one or more programs comprising instructions for performing any of the methods of claims 1 to 7.

Images