Vehicle control method, device, electronic device and storage medium

Abstract

The present application discloses a vehicle control method, device, electronic device and storage medium, which relates to the field of autonomous driving technology. A specific implementation of the method includes: determining a risk monitoring area around the vehicle according to the state information of the vehicle; determining the interaction risk between the vehicle and the obstacle vehicle in the risk monitoring area, and screening the risk vehicle according to the determined interaction risk; inputting the collision information related to the risk vehicle into the strain model predictive control cMPC model, and the cMPC model outputs the control amount for controlling the longitudinal speed of the vehicle; planning the longitudinal speed of the vehicle according to the control amount, and controlling the driving of the vehicle according to the planned longitudinal speed; wherein the collision information at least includes information related to the interaction risk of the risk vehicle.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a vehicle control method, device, electronic device and storage medium. Background Art

[0002] Autonomous driving technology involves multiple aspects such as perception layer, planning layer and control layer. Among them, decision planning is the core of autonomous driving, which makes the best decision and performs path / speed planning based on the environmental perception results, and generates planning results that meet dynamic constraints, driving safety and comfort.

[0003] In order to avoid collisions with obstacles, the distance between the autonomous driving vehicle and the obstacle is generally constrained to ensure a safe space, or the objective function is optimized to ensure the generation of safe and reliable path / speed planning. However, in some complex and risky dynamic scenarios, how to ensure that the autonomous driving vehicle makes accurate, efficient and safe planning remains a research focus and difficulty in the field of autonomous driving. Summary of the invention

[0004] The embodiments of the present application provide a vehicle control method, device, electronic device and storage medium.

[0005] In a first aspect, an embodiment of the present application provides a vehicle control method, the method comprising: determining a risk monitoring area around the vehicle based on status information of the vehicle; determining the interaction risk between the vehicle and obstacle vehicles in the risk monitoring area, and screening risky vehicles based on the determined interaction risk; inputting collision information related to the risky vehicle into a strain-based model predictive control cMPC model, the cMPC model outputting a control quantity for controlling the longitudinal speed of the vehicle; planning the longitudinal speed of the vehicle based on the control quantity, and controlling the driving of the vehicle based on the planned longitudinal speed; wherein the collision information includes at least information of the risky vehicle related to the interaction risk.

[0006] In a second aspect, an embodiment of the present application provides a vehicle control device, which includes: a monitoring area determination module, configured to determine a risk monitoring area around the vehicle based on status information of the vehicle; a risk assessment module, configured to determine the interaction risk between the vehicle and an obstacle vehicle in the risk monitoring area, and screen the risky vehicle based on the determined interaction risk; a model solving module, configured to input collision information related to the risky vehicle into a strain-based model predictive control (cMPC) model, the cMPC model outputting a control quantity for controlling the longitudinal speed of the vehicle, wherein the collision information at least includes information of the risky vehicle related to the interaction risk; a planning module, configured to plan the longitudinal speed of the vehicle based on the control quantity; and a control module, configured to control the driving of the vehicle based on the longitudinal speed planned by the planning module.

[0007] According to a third aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described in any implementation manner in the first aspect.

[0008] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, where the computer instructions are used to cause a computer to execute the method described in any implementation manner of the first aspect.

[0009] The vehicle control method, device, electronic device and storage medium provided in the embodiments of the present application can perform contingency planning based on the determined interaction risk information between the vehicle and the obstacle vehicle. By providing a reliable motion planning solution, the risk of close-range collision of autonomous driving vehicles in risk uncertain scenarios can be effectively reduced.

[0010] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0012] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0013] Figure 2 is a flow chart of an embodiment of a vehicle control method according to the present application;

[0014] Figure 3 is a flow chart of another embodiment of a vehicle control method according to the present application;

[0015] Figure 4 is a schematic diagram of a conflict area according to an example of the present application;

[0016] Figure 5 is a schematic diagram of an application scenario of the vehicle control method according to the present application;

[0017] Figure 6a is a schematic diagram of robust planning according to related techniques;

[0018] Figure 6b is a schematic diagram of contingency planning according to the present application;

[0019] Figure 7 is a schematic structural diagram of an embodiment of a vehicle control device according to the present application;

[0020] Figure 8 It is a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present application. DETAILED DESCRIPTION

[0021] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It should also be noted that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.

[0022] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0023] Figure 1 An exemplary system architecture 100 is shown to which embodiments of the vehicle control method and apparatus of the present application can be applied.

[0024] like Figure 1 As shown, the system architecture 100 may include an autonomous driving vehicle 101, an autonomous driving vehicle 102, a network 103 and a server 104. The network 103 is used to provide a medium for a communication link between the autonomous driving vehicles 101 and 102 and the server 104. The autonomous driving vehicles 101 and 102 may interact with the server 104 through the network 103 using an on-board terminal device to receive or send messages, etc. The on-board terminal device may also be connected to the cloud through the network 103, and may receive information from the cloud, or upload local information data to the cloud. The network 103 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0025] The autonomous driving vehicles 101 and 102 can be driven manually by the user or can be driven automatically with the help of the vehicle's intelligent driving system. Various sensors can be installed on the autonomous driving vehicles 101 and 102, such as cameras, camera sensors, laser radar sensors, etc. During the process of manual driving or autonomous driving by the user, the vehicle-mounted terminal device can perceive the surrounding environment information in real time based on camera sensors, laser radars, cameras, navigation / positioning systems, high-precision maps, etc., and provide some decision-making basis information for vehicle control.

[0026] In addition, the vehicle-mounted terminal device can also realize wireless communication between vehicles or realize communication with other external objects based on V2V (Vehicle to Vehicle) or V2I (Vehicle to Infrastructure) technology. For example, the vehicle-mounted terminal device on the autonomous driving vehicle 101 can obtain driving data information of the autonomous driving vehicle 102 based on the vehicle's sensors, such as cameras, laser radars, or global navigation satellite systems (Global Navigation Satellite System). The driving data information may include but is not limited to the geographic location, driving speed, acceleration, driving direction, path, and other information of the autonomous driving vehicle 102.

[0027] Among them, the vehicle-mounted terminal device can be hardware or software. When the vehicle-mounted terminal device is hardware, it can be an intelligent vehicle-mounted device / module configured on the autonomous driving vehicle, or a portable electronic device carried by the user, such as a laptop, smart phone, tablet computer or smart wearable device. When the vehicle-mounted terminal device is software, it can be installed in the electronic devices listed above, and it can be implemented as multiple software or software modules, or as a single software or software module, which is not specifically limited here.

[0028] The vehicle terminal device can provide various services through various built-in applications. For example, the vehicle terminal device can provide the following vehicle control method: first, determine the risk monitoring area around the vehicle based on the vehicle status information; then, determine the interaction risk between the vehicle and the obstacle vehicle in the risk monitoring area, and screen the risk vehicle based on the determined interaction risk; then, input the collision information related to the risk vehicle into the cMPC (contingency Model PredictControl) model, and the cMPC model outputs the control amount used to control the longitudinal speed of the vehicle; finally, plan the longitudinal speed of the vehicle based on the control amount, and control the vehicle according to the planned longitudinal speed.

[0029] It should be noted that the vehicle control method provided in the embodiment of the present application is generally executed by the vehicle-mounted terminal device of the autonomous driving vehicle. Accordingly, the vehicle control device is generally provided in the vehicle-mounted terminal device, or in the intelligent driving system of the autonomous driving vehicle.

[0030] Among them, the intelligent driving system can rely on artificial intelligence, visual computing, radar, monitoring devices and global positioning systems to assist the driver or fully automatically operate the autonomous driving vehicle. The intelligent driving system may include, for example, an assisted driving system (L2), a high-speed autonomous driving system that requires human supervision (L3), and a highly / fully autonomous driving system (L4 / L5). The vehicle control method provided in accordance with the embodiments of the present application can be applied to various traffic scenarios such as driving navigation, logistics transportation, food delivery, and online car-hailing.

[0031] It should be understood that Figure 1 The number of autonomous driving vehicles, networks, and servers in the embodiment is only for illustration. Any number of autonomous driving vehicles, vehicle-mounted terminals, networks, and servers may be provided as required.

[0032] like Figure 2 , a flow chart 200 of an embodiment of a vehicle control method according to the present application is shown, and the vehicle control method may include the following steps:

[0033] Step 201, determining a risk monitoring area around the vehicle according to the vehicle status information.

[0034] In this embodiment, the execution subject of the vehicle control method (for example Figure 1 The on-board terminal device in the autonomous driving vehicle 101 shown in the figure) can obtain the status information of the vehicle through the vehicle, and the status information may include the current speed, acceleration and other information of the vehicle.

[0035] When determining the risk monitoring area around the ego vehicle, the following two aspects need to be considered: (1) When other vehicles outside the risk monitoring area enter, the ego vehicle can fully respond; (2) Under the premise of ensuring the first aspect, the risk monitoring area should be made as small as possible to ensure efficiency.

[0036] In one example, the state information of the own vehicle may include the current speed and maximum emergency deceleration of the own vehicle, and the risk monitoring area centered on the own vehicle may be determined according to the current speed and maximum emergency deceleration of the own vehicle.

[0037] Step 202, determining the interaction risk between the vehicle and the obstacle vehicle in the risk monitoring area, and screening risky vehicles according to the determined interaction risk.

[0038] In this embodiment, the above-mentioned execution subject can obtain the driving data information of all obstacle vehicles in the risk monitoring area through its own vehicle or roadside equipment.

[0039] In one example, the self-vehicle may be a vehicle equipped with an intelligent driving system, and may sense historical driving data of obstacle vehicles through technologies such as an intelligent transportation system or V2V.

[0040] In one example, the roadside equipment may include data from a sensing device (such as a roadside camera), such as pictures and videos, so as to perform image and video processing and data calculation to obtain historical driving data of the obstructing vehicle. The historical driving data may be driving data of the obstructing vehicle at one or more times in the past, such as speed, positive acceleration, negative acceleration, predicted path, etc.

[0041] In an exemplary embodiment, the conflict area of ​​the interacting vehicles can be defined, and the following time indicators can be used: the time it takes for the vehicle to reach the conflict area TTZ (Time to Zone), the time it takes for the vehicle to decelerate to a safe state TTS (Time to Stop) and determine the risk assessment result of the interaction risk between the ego vehicle and the obstacle vehicle. Here, the conflict area refers to the area where the ego vehicle and the obstacle vehicle may collide; the time when the vehicle reaches the conflict area TTZ Used to directly characterize the interaction risk between the ego vehicle and the obstacle vehicle in the risk uncertainty scenario; the driving time of the ego vehicle to slow down to a safe state TTS It is used to indirectly characterize the interaction risk between the ego vehicle and the obstacle vehicle in the risk uncertainty scenario. More details about the conflict area and time indicators will be elaborated in the following.

[0042] Furthermore, if the risk monitoring area includes multiple obstacle vehicles, the obstacle vehicles in the risk monitoring area can be evaluated one by one, and then based on the preset risk threshold, the obstacle vehicles whose risk assessment results of the interactive risk reach or exceed the risk threshold can be determined as risk vehicles, so that all risk vehicles in the risk monitoring area are screened out. Among them, the preset risk threshold can be used to screen out risk vehicles that may collide with the vehicle. For obstacle vehicles whose risk assessment results of the interactive risk are less than the risk threshold, they are considered relatively safe and can be temporarily not processed to improve efficiency.

[0043] Step 203: input collision information related to the risk vehicle into the cMPC model, and the cMPC model outputs a control variable for controlling the longitudinal speed of the vehicle.

[0044] In this embodiment, the above-mentioned execution entity can construct a cMPC model by introducing a conflict horizon, and establish an objective function with the goal of minimizing the expected planning cost of each minimum planning cycle, and obtain the output of the cMPC model by optimizing and solving the objective function. Exemplarily, the expected planning cost in each minimum planning cycle can be defined as the sum of the planning cost of the normal horizon and the planning cost of the conflict horizon. Among them, the normal horizon focuses on timeliness and does not consider the collision problem with obstacles (such as risky vehicles); while the conflict horizon believes that there will be conflicts with obstacles and focuses on how to avoid collisions with risky vehicles in the scene.

[0045] The collision information related to the risky vehicle at least includes information related to the interaction risk of the risky vehicle, such as the risk assessment result of the interaction risk. In some examples, the collision information may also include information about the conflict area, and driving data information such as the driving speed, acceleration, and position of the ego vehicle and the obstacle vehicle. By inputting the collision information into the cMPC model as a collision constraint condition, and solving the control quantity output by the model based on the established convex objective function, the control quantity can be used to control the longitudinal speed of the ego vehicle.

[0046] Step 204 , planning the longitudinal speed of the vehicle according to the control amount, and controlling the vehicle to travel according to the planned longitudinal speed.

[0047] In this embodiment, the execution subject plans the longitudinal speed of the vehicle based on the control amount obtained in step 203, and controls the vehicle to travel based on the planned longitudinal speed, so as to avoid collision between the vehicle and the risk vehicle, thereby ensuring the safety of the vehicle.

[0048] The vehicle control method provided in this embodiment determines the surrounding obstacles that need attention by dividing the risk monitoring area, and screens out risky vehicles by determining the interaction risk between the vehicle and the obstacle vehicle; then, the collision information related to the risky vehicle is input into the cMPC model, and the longitudinal speed of the vehicle is controlled according to the control amount output by the cMPC model, so that the movement behavior of the vehicle in the future period of time conforms to the expected speed curve, thereby improving the driving safety and stability of the vehicle.

[0049] like Figure 3 , a flowchart 300 of another embodiment of a vehicle control method according to the present application is shown, and the vehicle control method may include the following steps:

[0050] Step 301, determine the risk monitoring area of ​​the vehicle.

[0051] In this embodiment, the risk monitoring area is defined as the vehicle as the center and the radius The area where the vehicle decelerates to a stop at the maximum emergency deceleration, i.e., the radius of risk monitoring as follows:

[0052]

[0053] in, is the current speed of the vehicle, is the maximum emergency deceleration of the vehicle in response to risk, is the adjustment coefficient of vehicle deceleration response, The acceptable experience value is 0.8.

[0054] Step 302, determining the time difference between the ego vehicle and the obstacle vehicle arriving at the conflict area.

[0055] In this embodiment, the ego vehicle may sense or obtain the driving data of the ego vehicle and the obstacle vehicle in the risk monitoring area; and then determine the conflict area between the ego vehicle and the obstacle vehicle based on the driving data of the ego vehicle and the obstacle vehicle.

[0056] Combination Figure 4 As shown in FIG. 1 , the ego vehicle goes straight along the A->B track, and the obstacle vehicle (other vehicle) turns right along the D->C track. When passing through the same intersection, a collision may occur. For example, the conflict area can be determined based on the overlapping area of ​​the ego vehicle's current path (A->B) and the predicted path of the obstacle vehicle (D->C), for example, it can be determined based on the projection position of the overlapping area on the respective driving paths of the ego vehicle and the obstacle vehicle, such as Figure 4 The area shown in the dashed box.

[0057] by Figure 4 The scene shown is an example, and the conflict area can be an irregular polygonal area, which may include the following boundaries: the pre-conflict boundary A of the ego vehicle, which indicates the boundary where the ego vehicle reaches the conflict area; the post-conflict boundary B of the ego vehicle, which indicates the boundary where the ego vehicle leaves the conflict area; the pre-conflict boundary D of the obstacle vehicle, which indicates the boundary where the obstacle vehicle reaches the conflict area; and the post-conflict boundary C of the obstacle vehicle, which indicates the boundary where the obstacle vehicle leaves the conflict area.

[0058] Based on the above conflict areas, determine the time when the vehicle reaches the conflict area and the time it takes for the obstacle vehicle to reach the conflict area :

[0059]

[0060]

[0061] in, is the distance between the ego vehicles before the collision boundary, for t The speed of the car at any moment, is the distance between the obstacle vehicles before the collision. for t Always block the speed of vehicles.

[0062] Then, determine the time difference between the ego vehicle and the obstacle vehicle arriving at the conflict area , the time difference .

[0063] Step 303, determining the driving time for the vehicle to decelerate to a safe state.

[0064] As mentioned above, the time it takes for the vehicle to decelerate to a safe state can be defined TTS Indirect characterization of the interaction risk between vehicles:

[0065]

[0066] in, for t The speed of the car at any moment, Issued for planning t The deceleration of the vehicle at any moment, is the braking coefficient. Usually, the braking deceleration instructions issued according to the plan, but the actual braking effect will be affected by the road environment and vehicle execution. For this reason, the braking coefficient can be increased. , and the experience value can be taken as 1.1.

[0067] In this embodiment, considering that the safety of the scene is inversely proportional to the planned braking amount of the autonomous driving vehicle, the different braking levels of the vehicle can be described according to the pre-set braking deceleration threshold; and the driving time for the vehicle to decelerate to a safe state under different braking levels can be determined respectively. TTS .

[0068] As an exemplary embodiment, according to the three braking deceleration thresholds set: vehicle emergency braking threshold , vehicle moderate braking threshold and vehicle soft braking threshold , the braking level of the vehicle is divided into the following three levels: emergency braking, moderate braking and gentle braking. The set braking deceleration threshold The description is as follows:

[0069]

[0070] The constraints are ,

[0071] Then there is

[0072]

[0073] in, is the braking deceleration threshold, The value of (Vehicle emergency braking threshold), (Vehicle Moderate Braking Threshold) and (vehicle soft braking threshold) one; Choose a time to slow down the ego vehicle; is the braking level of the vehicle, The value of (Vehicle emergency braking), (moderate braking of the vehicle) and (vehicle gentle braking) one; is the time it takes for the vehicle to decelerate to a safe state under different braking levels, that is, , ,and The driving time required to decelerate to a stop.

[0074] Here, for risk assessment in uncertain risk scenarios, the division of different braking levels can correspond to different driving styles of obstacle vehicles, so that the degree of interaction risk between the vehicle and other interacting vehicles can be more refinedly distinguished and evaluated.

[0075] Step 304 , assessing the interaction risk between the ego vehicle and the obstacle vehicle to obtain a risk assessment value.

[0076] In this embodiment, the safety time of the vehicle under different braking levels is defined as as follows:

[0077]

[0078] That is, the safety time of the vehicle The time difference between the vehicle and the obstacle vehicle arriving at the conflict area determined in step 302 , and the driving time for the vehicle to decelerate to a safe state under different braking levels determined in step 303 Determined.

[0079] In this embodiment, three braking deceleration thresholds can also be set: vehicle emergency braking threshold , vehicle moderate braking threshold and vehicle soft braking threshold , the interaction risk between the ego vehicle and the obstacle vehicle is divided into the dangerous degree D , the degree of attention required A and safety S .

[0080] Then the safety time of the vehicle under different braking levels can be calculated. , respectively define the risk assessment value between the ego vehicle and the obstacle vehicle as follows:

[0081]

[0082]

[0083]

[0084] In the formula, is the driving time for the vehicle to decelerate to a safe state under different braking levels, , is the time it takes for the vehicle to reach the conflict area, is the time it takes for the obstacle vehicle to reach the conflict area, , σ is the adjustment parameter of the interaction risk, and σ can be a hyperparameter.

[0085] Among them When , the interaction risk between the ego vehicle and the obstacle vehicle is absolutely dangerous.

[0086]

[0087]

[0088] for In other cases, the obstacle vehicle is considered safe for the ego vehicle, that is, there is no collision risk between the ego vehicle and the obstacle vehicle.

[0089] Step 305 , screening risky vehicles according to the determined interaction risk.

[0090] In this embodiment, the risk assessment results of the interaction risk between the ego vehicle and the obstacle vehicle at different braking levels (corresponding to different interaction risk levels D, A, S) can be estimated based on the conditional random field. :

[0091]

[0092] in, .

[0093] Furthermore, based on the preset risk threshold Screen out risky vehicles, which must meet the following requirements:

[0094]

[0095] That is, the degree of danger is determined by the degree of interaction risk. D(or vehicle emergency braking level) under the risk assessment results If the value is greater than the preset risk threshold , then the obstacle vehicle is determined as a risk vehicle. Among them, the risk threshold The acceptable experience value is 0.9.

[0096] Step 306 : Input the collision information related to the risk vehicle into the cMPC model.

[0097] This application introduces a conflict horizon to construct a cMPC model to perform contingency planning for autonomous driving vehicles, and inputs collision information related to risky vehicles into the constructed cMPC model to construct collision constraints. Among them, the role of the cMPC model is to predict the output of the system in a future time period, and this future time period is called the predictive horizon. Exemplarily, contingency planning refers to solving the optimization problem in the predictive horizon at each sampling moment, obtaining an optimal control sequence, and using the first element in this control sequence as the control amount at the current moment; then repeating this process continuously to obtain the control amount for each subsequent step, so as to be able to respond to changes in the dynamic environment in real time, such as changes in the position and speed of risky vehicles.

[0098] by Figure 5 Taking the simple collision scenario shown as an example, at a road intersection, there is a probability of a collision between the vehicle and other vehicles.

[0099] Figure 6a and Figure 6b The schematic diagrams of robust planning and contingency planning are shown respectively. Figure 6a and Figure 6b It can be seen that the planning action sequence of robust planning is [a0, a1, a2, a3, ...], which cannot cope with the dynamic changes of the external environment. Compared with robust planning, the characteristic of contingency planning is that it always maintains two horizons and optimizes the two horizons at the same time at each sampling moment. Among them, the conflict horizon pursues safety and avoids collisions with risk vehicles in the risk monitoring area. Its corresponding planning action sequence is [a0, a1c, a2c, a3c, ...]; while the normal horizon pursues efficiency, its corresponding planning action sequence is [a0, a1n, a2n, a3n, ...], which can ensure that the vehicle travels at a faster speed while meeting safety and stability, so as to reach the destination in a shorter time.

[0100] In this embodiment, the collision information related to the risk vehicle includes at least the risk assessment result of the interaction risk between the vehicle and the risk vehicle (hereinafter referred to as the risk assessment result of the risk vehicle), and may also include information on the conflict area, driving data information of the risk vehicle, and driving data information of the vehicle, etc. The driving data information may include vehicle position, speed, acceleration, and path information (such as a planned path or a predicted path), etc.

[0101] Step 307, establishing a convex objective function.

[0102] In this embodiment, a convex objective function is established to solve the control amount output by the cMPC model. The output predicted by the cMPC model is compared with the expected output to minimize the difference between the two, which is the objective function. Certain constraints are imposed on the established convex objective function, and the convex objective function is optimized to obtain the optimal solution, that is, the control amount output by the cMPC model.

[0103] First, define the state quantity of the cMPC model as , the input is , represents a controlled system with state and input dimensions n and m respectively, as follows:

[0104]

[0105]

[0106] in, k is the minimum planning period.

[0107] Then, for each minimum planning period k The expected planning cost is minimized and a convex objective function is established. as follows:

[0108]

[0109] in,

[0110]

[0111] In the formula, represents the expected value, is the risk probability in the standard time domain, is the risk probability in the conflict time domain, and ; is the minimum planning period, N To plan the number of frames, Minimum planning period The cost function in the internal standard time domain, Minimum planning period The cost function of the inner conflict time domain,x is the state quantity of the cMPC model, u is the input of the cMPC model.

[0112] As an exemplary embodiment, the risk probability of the conflict time domain It can be defined as follows:

[0113]

[0114] in, is the risk assessment result of the risk vehicle, is the distance between the risk vehicle and the obstacle vehicle before the collision. is the collision time when the risk vehicle reaches the conflict area, is the preset safety collision avoidance time threshold, is the preset minimum safety distance threshold, is the experience factor. , ,and is a hyperparameter, and its empirical values ​​can be 0.15, 0.5 and 6 respectively.

[0115] For the risk probability The above definition takes into account the following three aspects: (1) the degree of danger D The risk assessment results ; (2) The collision time when the risk vehicle reaches the conflict area ; (3) The distance between the risk vehicle and the obstacle vehicle before the collision .

[0116] And the collision time of the risk vehicle arriving at the conflict area Determine as follows:

[0117]

[0118] In the formula, is the current speed of the risk vehicle.

[0119] In addition, plan the number of frames N This can be determined by:

[0120]

[0121] in, is the time it takes for the ego vehicle to pass through the rear boundary of the ego vehicle conflict, It can be defined as the distance between the rear boundary of the ego vehicle and the ego vehicle divided by the current speed of the ego vehicle. is the time step, i.e. the length of a single planning frame (minimum planning period).

[0122] Step 308: Obtain the control quantity output by the cMPC model based on the established convex objective function.

[0123] In this embodiment, taking longitudinal speed planning as an example, the process of obtaining the control amount output by the cMPC model based on the established convex objective function may include the following sub-steps:

[0124] Step 3081, for longitudinal speed planning, only the stability of the control quantity is considered, and the convex objective function is determined as the minimum planning period k The input amount on u The weighted two norm of , that is:

[0125]

[0126] in, represents the expected value, R For input u The weight matrix of is a positive semidefinite matrix.

[0127] The optimization problem of convex objective function is stated as follows:

[0128]

[0129] The constraints are ,

[0130] in is the input quantity in the conflict time domain, It is the input quantity in standard time domain.

[0131] The inequality constraints are as follows:

[0132]

[0133]

[0134] in, A is a block diagonal input matrix, B is the dynamics matrix, C is the offset matrix, G and H are the inequality constraint matrices, b is the offset vector.

[0135] Step 3082, define the control amount as the minimum planning period k The longitudinal second-order acceleration or braking amount within is , then the optimization problem is expressed as:

[0136]

[0137] in,

[0138]

[0139]

[0140] in, , , for The transposed matrix of is the identity matrix, is a 0 matrix.

[0141] Step 3083, optimize the convex objective function according to the set constraints.

[0142] In this embodiment, is the constraint condition, among which, is the distance between the ego vehicle and the planned end point in the conflict time domain, is the distance between the ego vehicles before the collision.

[0143] In this embodiment, the convex objective function can be optimized based on the equation of motion, as follows:

[0144] Since the planned frame number in the conflict time domain is N , where the distance advanced in each frame is , the current planning starting point , then:

[0145]

[0146]

[0147] There are also:

[0148]

[0149] Then there is,

[0150]

[0151] in, Plan the distance to move forward for the first frame. Plan the distance to move forward for the second frame. is the current speed, Plan the forward speed for frame 1, is the time step, The acceleration planned for the first frame, The acceleration planned for frame 2, ..., and so on.

[0152] According to the above results, the optimization process follows the following motion equation:

[0153]

[0154]

[0155] in, .

[0156] Step 3084, solving the convex objective function, and using the obtained optimal solution as the control variable output by the cMPC model.

[0157] In this embodiment, the Lagrange multiplier method can be used to solve the convex objective function, as follows:

[0158]

[0159]

[0160]

[0161] have:

[0162]

[0163] in,

[0164]

[0165]

[0166]

[0167] …

[0168]

[0169] Also,

[0170]

[0171] Substitution , ..., , get the control amount at the current moment :

[0172]

[0173] in, is the Lagrange multiplier, is the Lagrange coefficient, is the current speed of the vehicle, is the distance between the ego vehicles before the collision boundary, k is the minimum planning period, N To plan the number of frames, is the time step.

[0174] At each time step, the cMPC model is used to obtain an optimal control sequence in the prediction time domain, and the first element of the optimal control sequence is used as the current moment. , and then the next optimization is performed. In this way, the optimization is iterated continuously, and the next optimization will be performed at the previous time step. Converge before expiration.

[0175] Step 309, planning the longitudinal speed of the vehicle according to the control amount.

[0176] The control amount obtained in step 308 As the current minimum planning period k The acceleration or braking amount can be calculated, and the longitudinal speed of the vehicle can be planned according to the speed value obtained by integrating the acceleration or braking amount.

[0177] By selecting the second-order acceleration / braking amount as the control amount, it can be ensured that the generated longitudinal speed is continuous and smooth during the speed planning process.

[0178] Step 310: Control the vehicle to travel according to the planned longitudinal speed.

[0179] For example, for an autonomous vehicle, during the speed planning process, the planning layer can perform speed planning, generate planning results and issue instructions. Then, the control layer can control the underlying controller of the vehicle to generate throttle or brake operations to control the vehicle according to the planning results generated by the planning layer, such as longitudinal speed or acceleration / braking amount.

[0180] According to the vehicle control method provided in this embodiment, by introducing time indicators such as conflict areas and safety time to quantitatively analyze the interaction risks between the ego vehicle and obstacle vehicles, risky vehicles can be effectively screened out; then, the collision information related to the risky vehicles is input into the constructed cMPC model for strain planning, and by adding collision constraints and other dynamic constraints to the designed convex objective function, it is possible to avoid collisions between the autonomous driving vehicle and other vehicles, thereby effectively reducing the risk of close-range collisions of the autonomous driving vehicle in uncertain risk scenarios; in addition, while ensuring driving safety, the stability and timeliness of longitudinal speed planning can also be guaranteed.

[0181] Further references Figure 7 As an implementation of the methods shown in the above figures, the present application discloses an embodiment of a vehicle control device, which is similar to Figure 2 or Figure 3 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0182] like Figure 7 As shown, an embodiment of the present application provides a vehicle control device 700, which includes: a monitoring area determination module 701, a risk assessment module 702, a model solution module, a planning module 704 and a control module 705. Among them, the monitoring area determination module 701 is used to determine the risk monitoring area around the self-vehicle according to the state information of the self-vehicle; the risk assessment module 702 is used to determine the interaction risk between the self-vehicle and the obstacle vehicle in the risk monitoring area, and screen the risk vehicle according to the determined interaction risk; the model solution module 703 is used to input the collision information related to the risk vehicle into the strain model predictive control cMPC model, and the cMPC model outputs the control amount for controlling the longitudinal speed of the self-vehicle, wherein the collision information at least includes the information of the risk vehicle related to the interaction risk; the planning module 704 is used to plan the longitudinal speed of the self-vehicle according to the control amount; the control module 705 is used to control the self-vehicle to travel according to the longitudinal speed planned by the planning module.

[0183] In the present embodiment, in the vehicle control device 700, the specific processing of the monitoring area determination module 701, the risk assessment module 702, the model solving module 703, the planning module 704 and the control module 705 and the technical effects thereof can be referred to respectively. Figure 2 Corresponding to step 201 to step 204 in the embodiment, or Figure 3 Corresponding to step 301 to step 310 in the embodiment.

[0184] like Figure 8 , is a block diagram of an electronic device according to a control method for a vehicle according to an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0185] like Figure 8As shown, the electronic device includes: one or more processors 801, a memory 802, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are interconnected using different buses and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 8 A processor 801 is taken as an example.

[0186] The memory 802 is a non-transitory computer-readable storage medium provided in the present application. The memory stores instructions executable by at least one processor to enable at least one processor to perform the vehicle control method provided in the present application. The non-transitory computer-readable storage medium of the present application stores computer instructions, which are used to enable a computer to perform the vehicle control method provided in the present application.

[0187] The memory 802 is a non-transient computer-readable storage medium that can be used to store non-transient software programs, non-transient computer executable programs and modules, such as program instructions / modules corresponding to the vehicle control method in the embodiment of the present application (for example, the attached Figure 7 The processor 801 executes various functional applications and data processing of the server by running the non-transient software programs, instructions and modules stored in the memory 802, that is, the vehicle control method in the above method embodiment is implemented.

[0188] The memory 802 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the blockchain-based information processing electronic device, etc. In addition, the memory 802 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory 802 may optionally include a memory remotely arranged relative to the processor 801, and these remote memories may be connected to the blockchain-based information processing electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0189] The electronic device of the vehicle control method may further include: an input device 803 and an output device 804. The processor 801, the memory 802, the input device 803 and the output device 804 may be connected via a bus or other means. Figure 8 The example of connecting through bus is taken in the following.

[0190] The input device 803 can receive input digital or character information, and generate key signal input related to user settings and function control of blockchain-based information processing electronic devices, such as touch screens, keypads, mice, trackpads, touchpads, indicator rods, one or more mouse buttons, trackballs, joysticks and other input devices. The output device 804 may include a display device, an auxiliary lighting device (e.g., LED), and a tactile feedback device (e.g., a vibration motor). The display device may include, but is not limited to, a liquid crystal display (LCD), a light emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0191] Various implementations of the systems and techniques described herein can be realized in digital electronic circuit systems, integrated circuit systems, dedicated ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0192] These computer programs (also referred to as programs, software, software applications, or code) include machine instructions for programmable processors and can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, device, and / or means (e.g., disk, optical disk, memory, programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as machine-readable signals. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0193] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0194] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0195] A computer system may include clients and servers. Clients and servers are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship to each other.

[0196] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

[0197] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. A vehicle control method, comprising: Determine a risk monitoring area around the vehicle according to the vehicle status information; Determining the interaction risk between the ego vehicle and the obstacle vehicle in the risk monitoring area, and screening risky vehicles according to the determined interaction risk; Inputting collision information related to the risk vehicle into a strain-based model predictive control (cMPC) model, wherein the cMPC model outputs a control variable for controlling the longitudinal speed of the vehicle; Planning the longitudinal speed of the vehicle according to the control amount, and controlling the vehicle to travel according to the planned longitudinal speed; wherein the collision information at least includes a risk assessment result of the interaction risk between the ego vehicle and the risk vehicle, a collision time when the risk vehicle reaches the conflict area between the ego vehicle and the risk vehicle, and a distance between the risk vehicle and the front boundary of the conflict area; and wherein, The control quantity output by the cMPC model is obtained by the following method, including: Establishing a convex objective function of the cMPC model with the goal of minimizing the expected planning cost of each minimum planning cycle, constructing collision constraints based on the collision information, optimizing and solving the convex objective function, and obtaining the control amount; The prediction time domain of the cMPC model includes a standard time domain and a conflict time domain, and the expected planning cost within each minimum planning cycle is the sum of the planning cost within the standard time domain and the planning cost within the conflict time domain.

2. The method according to claim 1, wherein: The state information of the vehicle includes the current speed and the maximum emergency deceleration of the vehicle. Wherein, determining the risk monitoring area around the vehicle according to the status information of the vehicle includes: According to the current speed and maximum emergency deceleration of the vehicle, an area centered on the vehicle and with a distance at which the vehicle decelerates to a stop at the maximum emergency deceleration as a radius is determined as the risk monitoring area.

3. The method according to claim 1, wherein: Determining the interaction risk between the ego vehicle and the obstacle vehicle in the risk monitoring area includes: Determining the time difference between the ego vehicle and the obstacle vehicle arriving at the conflict area; and According to the time difference between the ego vehicle and the obstacle vehicle when they arrive at the conflict area and the driving time for the ego vehicle to decelerate to a safe state at different braking levels, a risk assessment value between the ego vehicle and the obstacle vehicle at different braking levels is determined; Among them, the braking level of the vehicle is divided into vehicle emergency braking, vehicle moderate braking and vehicle gentle braking according to the set braking deceleration threshold, and the driving time for the vehicle to decelerate to a safe state under the different braking levels is determined respectively.

4. The method according to claim 3, wherein: Screen risky vehicles based on identified interaction risks, including: estimating risk assessment results of the vehicle and the obstacle vehicle at different braking levels based on a conditional random field; In response to the risk assessment result at the vehicle emergency braking level being greater than a preset risk threshold, the obstacle vehicle is determined as the risk vehicle.

5. The method according to claim 3, wherein: The conflict area is determined based on an overlapping area between the current path of the ego vehicle and the predicted path of the obstacle vehicle; The conflict area includes the following boundaries: a pre-collision boundary of the ego vehicle, a post-collision boundary of the ego vehicle, a pre-collision boundary of the obstacle vehicle, and a post-collision boundary of the obstacle vehicle.

6. The method according to claim 5, wherein: The collision information also includes information of the conflict area, driving data information of the risk vehicle and driving data information of the own vehicle, wherein the driving data information includes at least one of vehicle position, speed, acceleration and path information.

7. The method according to claim 5, wherein: Optimizing and solving the convex objective function to obtain the control quantity includes: Determine the convex objective function as the minimum planning period k The input amount on u The weighted bi-norm of ; by Optimizing the convex objective function for the collision constraint; and Solving the optimized convex objective function to obtain the control quantity; The control quantity is defined as the minimum planning period k The longitudinal second-order acceleration or braking amount within ,Right now ; is the distance between the vehicle and the planned end point of the conflict time domain, is the pre-collision boundary of the ego vehicle.

8. The method according to claim 7, wherein: Optimizing the convex objective function includes: optimizing the convex objective function based on a motion equation.

9. The method according to claim 7, wherein: Solving the optimized convex objective function to obtain the control quantity includes: The Lagrange multiplier method is used to solve the optimized convex objective function to obtain the control quantity.

10. The method according to any one of claims 1 to 9, wherein: The longitudinal speed of the vehicle is planned according to the control amount, including: The control amount is used as the acceleration or braking amount of each minimum planning cycle, and the acceleration or the braking amount is integrated to obtain a speed value, and the longitudinal speed of the vehicle is planned according to the obtained speed value.

11. A vehicle control device, comprising: A monitoring area determination module, configured to determine a risk monitoring area around the vehicle according to the vehicle status information; a risk assessment module configured to determine the interaction risk between the ego vehicle and the obstacle vehicle in the risk monitoring area, and screen risky vehicles according to the determined interaction risk; A model solving module is configured to input collision information related to the risk vehicle into a strain-based model predictive control (cMPC) model, wherein the cMPC model outputs a control variable for controlling the longitudinal speed of the ego vehicle, wherein the collision information at least includes a risk assessment result of the interaction risk between the ego vehicle and the risk vehicle, a collision time when the risk vehicle reaches a conflict area between the ego vehicle and the risk vehicle, and a distance between the risk vehicle and a front boundary of the conflict area; A planning module, configured to plan the longitudinal speed of the vehicle according to the control amount; as well as A control module, configured to control the vehicle to travel according to the longitudinal speed planned by the planning module; The model solving module obtains the control quantity output by the cMPC model in the following manner, including: Establishing a convex objective function of the cMPC model with the goal of minimizing the expected planning cost of each minimum planning cycle, constructing collision constraints based on the collision information, optimizing and solving the convex objective function, and obtaining the control amount; The prediction time domain of the cMPC model includes a standard time domain and a conflict time domain, and the expected planning cost within each minimum planning cycle is the sum of the planning cost within the standard time domain and the planning cost within the conflict time domain.

12. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

13. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 10.

Images