WO2025200580A1 - 智能驾驶方法、装置及车辆 - Google Patents
智能驾驶方法、装置及车辆Info
- Publication number
- WO2025200580A1 WO2025200580A1 PCT/CN2024/138104 CN2024138104W WO2025200580A1 WO 2025200580 A1 WO2025200580 A1 WO 2025200580A1 CN 2024138104 W CN2024138104 W CN 2024138104W WO 2025200580 A1 WO2025200580 A1 WO 2025200580A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- vehicle
- obstacle
- information
- speed
- risk level
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W60/00—Drive control systems specially adapted for autonomous road vehicles
- B60W60/001—Planning or execution of driving tasks
- B60W60/0011—Planning or execution of driving tasks involving control alternatives for a single driving scenario, e.g. planning several paths to avoid obstacles
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/09—Taking automatic action to avoid collision, e.g. braking and steering
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/095—Predicting travel path or likelihood of collision
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/095—Predicting travel path or likelihood of collision
- B60W30/0953—Predicting travel path or likelihood of collision the prediction being responsive to vehicle dynamic parameters
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
- B60W30/095—Predicting travel path or likelihood of collision
- B60W30/0956—Predicting travel path or likelihood of collision the prediction being responsive to traffic or environmental parameters
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W60/00—Drive control systems specially adapted for autonomous road vehicles
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2552/00—Input parameters relating to infrastructure
- B60W2552/50—Barriers
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2554/00—Input parameters relating to objects
- B60W2554/80—Spatial relation or speed relative to objects
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2720/00—Output or target parameters relating to overall vehicle dynamics
- B60W2720/10—Longitudinal speed
Definitions
- the present application relates to the field of intelligent driving, and in particular to an intelligent driving method, device and vehicle.
- the present application provides an intelligent driving method, device and vehicle, which can improve the flexibility of the vehicle in avoiding dynamic obstacles in the automatic driving cruise state, improve the vehicle's driving efficiency and traffic efficiency, and enhance the driving experience.
- the present application provides an intelligent driving method, the method comprising: detecting a first obstacle; the first obstacle being a dynamic obstacle; determining a first risk level based on first information of the first obstacle and first driving information of a vehicle; the first risk level being used to indicate the likelihood of a collision between the first obstacle and the vehicle under the first information and the first driving information; obtaining a first cruising speed that matches the first risk level, the first cruising speed being the minimum driving speed of the vehicle at the first risk level to avoid a collision with the first obstacle; and controlling the vehicle to travel at the first cruising speed.
- the first obstacle is a dynamic obstacle around the vehicle that may cause a collision risk with the vehicle.
- the vehicle makes a prediction based on the first information of the first obstacle and the first driving information of the vehicle to obtain prediction information, and then performs short-term deduction and collision detection based on the prediction information to determine risk information.
- the vehicle processes risk information based on a longitudinal risk assessment algorithm or model and a lateral risk assessment algorithm or model to determine longitudinal and lateral risks.
- the vehicle comprehensively analyzes the longitudinal and lateral risks, considering the combined impact of different risks on driving, and determines the risk level.
- a mapping table between risk information and risk levels is pre-configured in the vehicle, and the vehicle determines the risk level corresponding to the current risk information based on the mapping table.
- the vehicle comprehensively considers obstacle information and vehicle information to accurately determine the risk level of collision between the obstacle and the vehicle, so that the vehicle can make more reasonable decisions based on the risk level, improve the vehicle's traffic efficiency, and enhance driving safety.
- a mapping table between risk levels and cruising speeds is pre-configured in the vehicle.
- the vehicle can determine a first cruising speed that matches the first risk level based on the mapping relationship.
- controlling a vehicle to travel at a first cruising speed includes: generating a speed planning curve based on the current speed of the vehicle and the first cruising speed; adjusting the vehicle's travel speed to the first cruising speed according to the speed planning curve, and controlling the vehicle to travel at the first cruising speed.
- a speed planning curve is generated based on the first cruising speed related to the current vehicle state and obstacle information to guide the vehicle's driving speed, optimize the vehicle's driving efficiency, make the vehicle speed more adapted to the surrounding environment, and improve driving safety and comfort.
- Controlling the vehicle to travel at a first cruising speed includes: if the distance between the current position of the vehicle and the second obstacle is greater than a second threshold, controlling the vehicle to travel at the first cruising speed; the second threshold is the minimum safe distance between the vehicle and the second obstacle.
- the present application provides a vehicle comprising the intelligent driving device as described in the second aspect.
- Exemplary vehicles include cars, trucks, motorcycles, buses, lawn mowers, recreational vehicles, amusement park vehicles, construction equipment, trams, golf carts, trains, etc., which are not particularly limited in this application.
- the power of the above-mentioned vehicles can be provided by gasoline, diesel, electricity, solar energy, or hydrogen energy.
- the present application provides a chip system comprising at least one processor and at least one interface circuit, wherein the at least one interface circuit is used to perform transceiver functions, and the at least one processor is used to execute the method of the first aspect and any one of the embodiments of the first aspect.
- the present application provides a computer-readable storage medium, which includes a computer program.
- the computer program When the computer program is executed by a processor, it implements the method of the first aspect and any one of the embodiments of the first aspect.
- the present application provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, implements the method of the first aspect and any one of the embodiments of the first aspect.
- FIG1 is a schematic diagram of a vehicle driving scenario provided by an embodiment of the present application.
- FIG2 is a schematic structural diagram of a vehicle provided in an embodiment of the present application.
- FIG6 is a schematic diagram of a flow chart of an intelligent driving method provided in an embodiment of the present application.
- FIG8 is a second schematic diagram of an intelligent driving scenario provided by an embodiment of the present application.
- FIG9 is a third schematic diagram of an intelligent driving scenario provided in an embodiment of the present application.
- FIG10 is a fourth schematic diagram of an intelligent driving scenario provided by an embodiment of the present application.
- FIG11 is a schematic diagram of the structure of an intelligent driving device provided in an embodiment of the present application.
- FIG12 is a schematic structural diagram of a chip system provided in an embodiment of the present application.
- first and second are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the quantity of the technical features indicated. Therefore, a feature specified as “first” or “second” may explicitly or implicitly include one or more of the features.
- a cruising vehicle's handling strategy for dynamic obstacles involves actively braking to avoid potential collisions and ensure vehicle safety. This strategy only considers the vehicle's behavior (such as the distance from the current vehicle to the predicted collision point) and does not account for interactions between the vehicle and the dynamic obstacle (such as the dynamic obstacle's proactive yielding). This results in low vehicle traffic efficiency and weak overtaking capabilities.
- vehicle 11 is in autonomous cruising mode and moving forward.
- Vehicle 12 (a dynamic obstacle) is moving to the left.
- Vehicle 11 determines collision point 13 based on information obtained from both vehicle 11 (such as speed and position) and vehicle 12 (such as speed and position).
- Vehicle 11 detects a potential collision with vehicle 12, actively brakes, and waits for vehicle 12 to pass before restarting. Vehicle 11 only considers its own driving safety and actively avoids vehicle 12, without considering vehicle 12's handling method. If vehicle 12 also actively brakes, vehicle 11 needs to reassess the situation after braking.
- Vehicle 11's handling method for avoiding vehicle 12 (a dynamic obstacle) is conservative, rigid, and inflexible. Vehicle 11 brakes as soon as it detects a dynamic obstacle, resulting in a poor driving experience. Moreover, vehicle 11's abrupt braking causes the braking point to be far from the collision point, resulting in low traffic efficiency and weak overtaking ability.
- an embodiment of the present application provides an intelligent driving method, which includes: detecting a first obstacle; determining a first risk level based on first information of the first obstacle and first driving information of the vehicle; the first risk level is used to indicate the possibility of a collision between the first obstacle and the vehicle under the first information and the first driving information; obtaining a first cruising speed that matches the first risk level, the first cruising speed being the minimum driving speed for the vehicle to avoid a collision with the first obstacle under the first risk level; and controlling the vehicle to travel at the first cruising speed.
- the method provided in the embodiment of the present application determines a matching cruising speed based on the risk between the vehicle and the obstacle, controls the vehicle at the cruising speed, optimizes the intelligent driving strategy, improves the flexibility of the vehicle in avoiding dynamic obstacles in the automatic driving cruise state, and while ensuring vehicle safety, enhances the driving experience, improves the vehicle's driving efficiency, traffic efficiency, and overtaking ability, so that the vehicle can better adapt to complex traffic environments.
- obstacle avoidance scenarios include robot obstacle avoidance scenarios, vehicle obstacle avoidance scenarios (such as vehicle obstacle avoidance in autonomous driving scenarios), and autonomous mobile device obstacle avoidance scenarios (such as drone obstacle avoidance, etc.).
- the technical solutions provided in the embodiments of this application can be applied to various mobile smart devices.
- the mobile smart devices may include, but are not limited to, vehicles, artificial intelligence (AI) devices (such as robots), and the like.
- AI artificial intelligence
- the solutions can be applied to other devices (such as servers, mobile phones, etc.) that have the function of controlling the aforementioned mobile smart devices.
- Mobile smart devices or other devices can implement the intelligent driving methods provided in the embodiments of this application through their components (including hardware and software).
- FIG2 is a structural diagram of a vehicle 200 provided in an embodiment of the present application.
- vehicle 200 may include various subsystems, including but not limited to intelligent driving system 210.
- vehicle 200 may include more or fewer subsystems, and each subsystem may include multiple components.
- each subsystem and component of vehicle 200 may be connected via wired or wireless communication technology.
- Intelligent driving system 210 can obtain first obstacle information (including but not limited to the location of the first obstacle) and driving information of vehicle 200 (including but not limited to various vehicle information such as driving speed and acceleration). Intelligent driving system 210 can also determine a risk level based on this information, obtain a cruising speed that matches the risk level, and control vehicle 200 to travel at the cruising speed.
- first obstacle information including but not limited to the location of the first obstacle
- driving information of vehicle 200 including but not limited to various vehicle information such as driving speed and acceleration.
- Intelligent driving system 210 can also determine a risk level based on this information, obtain a cruising speed that matches the risk level, and control vehicle 200 to travel at the cruising speed.
- the above-mentioned vehicle 200 can be a new energy vehicle, an electric vehicle, an intelligent vehicle, a car, a truck, a motorcycle, a bus, a boat, a lawn mower, an entertainment vehicle, an amusement park vehicle, construction equipment, a tram, a golf cart, and a train, etc., and the embodiments of the present application do not make any special limitations.
- Processor 301 can be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits used to control the execution of the program of the present application.
- CPU central processing unit
- ASIC application-specific integrated circuit
- the mobile smart device may include more or fewer components than those shown in Figures 2 and 3, or may combine or separate certain components, replace certain components, or arrange the components differently.
- the components shown in the figures may be implemented in hardware, software, or a combination of software and hardware.
- the intelligent driving system includes a perception system, a planning system, and a control system.
- the perception system, the planning system, and the control system are communicatively connected.
- the planning system and the control system are used to execute the intelligent driving method provided by the present application.
- the planning system includes a prediction module, a decision module, and a planning module.
- the prediction information includes the prediction information of obstacles (such as the future speed, future position, future motion trajectory, etc. of the obstacle) and the prediction information of the vehicle (such as the future speed, future position, future motion trajectory, etc. of the vehicle).
- the decision module is configured to perform calculations based on the prediction information sent by the prediction module to determine risk information.
- the decision module is further configured to integrate various risk information to assess horizontal and vertical risks and determine a risk level.
- the decision module is further configured to determine a corresponding control decision based on the risk level.
- the decision module is further configured to transmit the risk level to the planning module; alternatively, the decision module transmits the risk information, risk level, and control decision to the planning module.
- the risk information includes but is not limited to the collision point, the collision time, the distance from the vehicle to the collision point, and the distance from the first obstacle to the collision point.
- the first obstacle is a dynamic obstacle
- the collision point is the intersection of the future motion trajectory of the dynamic obstacle and the future motion trajectory of the vehicle.
- risk information may also include environmental information, such as weather, road conditions, and traffic conditions. It should be understood that when determining risk information, the vehicle should comprehensively consider the vehicle's surroundings to assist in making a more accurate risk assessment and enabling the vehicle to adopt a more appropriate cruising speed.
- environmental information such as weather, road conditions, and traffic conditions.
- lateral and longitudinal risks include lateral risks and longitudinal risks.
- Lateral risks are risks in the lateral direction faced by the vehicle during driving, that is, risks in the left and right directions of the vehicle, including the possibility of avoiding obstacles or colliding with obstacles. For example, the risk of collision caused by other vehicles changing lanes, overtaking, merging, etc., as well as the threats that lateral obstacles from different sources such as roadside obstacles, pedestrians, and animals may pose to vehicle driving.
- Longitudinal risks are risks in the longitudinal direction faced by the vehicle during driving, that is, risks in the front and rear directions of the vehicle, including the possibility of slowing down and braking or accelerating to avoid collisions. For example, the risk of rear-end collision caused by insufficient following safety distance from the vehicle in front, and the situation where the vehicle needs to brake urgently or change lanes to avoid sudden braking, deceleration or obstacles in front.
- the risk level is used to indicate the probability of collision between the first obstacle and the vehicle, for example, low risk, medium risk, and high risk.
- Control decisions include yielding and overtaking.
- a yielding decision involves slowing down to avoid an obstacle, while an overtaking decision involves accelerating to overtake. For example, when a vehicle faces a dynamic obstacle with a potential collision risk, it typically slows down to yield to the obstacle and then continues driving after the collision risk is resolved. Alternatively, if the vehicle is closer to the collision point and the collision risk is low, it may choose to accelerate and overtake the obstacle.
- the decision module calculates and deduces risk information based on the predicted information. Based on this risk information, the decision module assesses both horizontal and vertical risks. Combining these assessments, the decision module determines the risk level. Based on predefined rules, the decision module then determines the control strategy corresponding to the risk level.
- control strategy corresponding to low risk is the rush-to-pass strategy
- control strategy corresponding to medium risk and high risk is the yield strategy
- the planning module is used to perform speed planning based on the various information sent by the decision module and generate a speed planning result (such as cruising speed).
- the planning module is also used to send the speed planning result (such as cruising speed) to the control system.
- the planning module determines a cruising speed that matches the risk level based on the risk information.
- the planning module determines a cruising speed that matches the risk level based on the control strategy, risk information, and risk level.
- the planning module can also plan the vehicle's driving speed based on the cruising speed and generate a speed planning curve.
- Cruising speed can be understood as the minimum speed to avoid collision with a dynamic obstacle, calculated based on current vehicle information, obstacle information, and collision risk, when the vehicle detects a dynamic obstacle and there is a risk of collision with the vehicle.
- Cruising speed can be understood as the minimum critical speed that the vehicle should maintain based on the current scenario. The vehicle speed should smoothly change from the current speed to the cruising speed, maintaining steady forward movement at the cruising speed.
- the current scenario is an obstacle avoidance scenario, if the control strategy is determined to be a yield strategy, the cruising speed will be a lower value, and the vehicle will be instructed to decelerate based on this cruising speed.
- the vehicle may move forward slowly, dynamically adjusting the cruising speed based on real-time risk detection until the obstacle passes the collision point first, eliminating the collision risk. If the control strategy is determined to be a preemptive strategy, the cruising speed will be a higher value, and the vehicle will be instructed to accelerate based on this cruising speed, increasing its speed to quickly pass the collision point and eliminating the collision risk.
- control system controls the vehicle to travel at the speed planning result (such as cruising speed) according to the speed planning result (such as cruising speed) sent by the planning module, thereby realizing intelligent driving of the vehicle and successfully avoiding obstacles.
- FIG5 shows a schematic diagram of the architecture of another intelligent driving system 50 provided in an embodiment of the present application.
- the intelligent driving system 50 may include a sensing device 51 and an intelligent driving device 52 .
- the sensing device 51 and the intelligent driving device 52 are communicatively connected.
- the sensing device 51 includes a vehicle sensor, which is typically located inside the vehicle.
- the vehicle sensor can be used to sense information about the first obstacle around the vehicle and the vehicle's driving information.
- the vehicle sensor can be an accelerometer, a gyroscope, a wheel speed sensor, an air pressure sensor, an ultrasonic sensor, a camera sensor, a positioning sensor, a radar sensor, or the like.
- the sensing device 51 is a vehicle sensor
- the sensing device 51 moves with the vehicle
- the sensing device 51 and the intelligent driving device 52 can communicate through the wireless communication network.
- sensing device 51 is only for illustrative purposes.
- the intelligent driving device 52 is used to execute the intelligent driving method provided herein. Specifically, the intelligent driving device 52 can be used to determine the current risk level based on obstacle information and vehicle driving information. The risk level indicates the probability of a collision between the obstacle and the vehicle. The intelligent driving device 52 can also be used to obtain a cruising speed that matches the current risk level. The cruising speed is the minimum speed at which the vehicle can avoid colliding with the obstacle under the current risk level. The intelligent driving device 52 is also used to control the vehicle to travel at the cruising speed.
- intelligent driving device 52 may be an intelligent driving computing platform.
- This platform implements intelligent driving, decision-making, planning, and control functions and is a core component of the vehicle.
- the intelligent driving computing platform interacts with various vehicle components, acquiring real-time data from each component and controlling its operation.
- the intelligent driving device 52 may be a server.
- the server may be a Linux server, a Windows server, or other server device that can provide simultaneous access to multiple devices. It may also be a server cluster consisting of multiple regions, multiple computer rooms, and multiple servers.
- the intelligent driving device 52 may be a server of an intelligent transportation system, such as a physical server or a cloud server, which is not limited in this embodiment of the present application.
- the sensing device 51 and intelligent driving device 52 in the intelligent driving system 50 are independent components that interact with each other to achieve intelligent driving.
- the intelligent driving system 50 may also include only the intelligent driving device 52, with other components controlled by other systems in the vehicle, and these systems interact to achieve intelligent driving.
- the above description of the intelligent driving system is only an example.
- the various modules in the above-mentioned intelligent driving system are divided according to functional logic, and other division methods may be used in practice.
- the above-mentioned modules can be named by other names.
- each module can be implemented by hardware, software, or a combination of hardware and software. Whether a specific module is implemented in hardware, software, or a combination of hardware and software depends on the specific application and design constraints of the technical solution. Different modules can be implemented by different hardware, and multiple modules can also be implemented by the same hardware.
- FIG6 shows a flow chart of an intelligent driving method provided by an embodiment of the present application.
- the execution subject of the method can be, for example, the mobile intelligent device described above, or a processor in the mobile intelligent device.
- the execution subject is a mobile intelligent device, and the mobile intelligent device is a vehicle.
- the method includes the following steps:
- S601 A vehicle detects a first obstacle, which is a dynamic obstacle.
- the vehicle is in automatic driving mode or intelligent driving mode, and the vehicle is in cruising state.
- L0 No Automation, which means the vehicle is fully operated by a human driver without any support from any automated systems
- L1 Driver Assistance, which means the vehicle provides driving assistance for one of the following operations: steering and acceleration and deceleration; the rest is handled by the driver
- L2 Partial Automation, which means the vehicle provides driving assistance for several operations: steering and acceleration and deceleration; the rest is handled by the driver (the driver needs to maintain an observation of road conditions and be ready to take over control at any time)
- L3 Conditional Automation, which means the vehicle can complete most driving operations and can drive autonomously under certain conditions.
- L4 High Automation, which means the vehicle can complete all driving operations in most situations and drives independently. The driver only needs to intervene in rare circumstances, such as extreme weather or complex traffic conditions; L5: Full Automation, which means the vehicle can drive completely autonomously on any road and in all conditions without driver intervention.
- the autonomous driving mode is applicable to levels L2-L4.
- the vehicle can assist the driver, controlling acceleration and deceleration to avoid dynamic obstacles.
- the vehicle can perform most driving operations and avoid obstacles through sensors, computer vision, artificial intelligence, and other technologies.
- the vehicle can sense its surroundings and independently perform all driving operations and avoid obstacles.
- Cruise control means that after the vehicle activates cruise control mode, the vehicle maintains a predetermined speed and moves forward without the driver having to step on the accelerator or brake to maintain the speed. In this state, the vehicle automatically adjusts the accelerator and brake to maintain a stable speed.
- the first obstacle may be a dynamic obstacle around the vehicle that could cause a collision risk.
- a dynamic obstacle is an obstacle that remains in motion, such as a moving vehicle, a running pedestrian, a walking pedestrian, or a running animal.
- the first obstacle includes, but is not limited to, moving objects, vehicles, pedestrians, and the like.
- a vehicle detects obstacles around it using various sensors.
- a vehicle can use an onboard camera to capture images of its surroundings and identify obstacles on the road through image processing and computer vision technologies.
- a vehicle can also detect obstacles by using a radar system to emit radio waves and receive reflected waves to detect objects in the surrounding environment.
- a vehicle can also use a lidar system to emit laser beams and determine the location and shape of obstacles based on the reflection time of the laser beams.
- a vehicle can also detect obstacles using ultrasonic sensors.
- a vehicle when a vehicle detects an obstacle around it, it also obtains information about the obstacle (such as the obstacle's location, speed, direction, shape, etc.). Based on the information about each obstacle, the vehicle analyzes whether there is a risk of collision between the obstacle and the vehicle, and whether the obstacle is a dynamic obstacle, and determines dynamic obstacles that pose a collision risk as first obstacles.
- information about the obstacle such as the obstacle's location, speed, direction, shape, etc.
- the vehicle may detect its own vehicle status information (such as driving information, etc.) in real time.
- vehicle status information such as driving information, etc.
- the vehicle uses various sensors to obtain information about surrounding obstacles (such as the obstacle's location, speed, direction, acceleration, shape, size, and other information), as well as vehicle driving information (such as the vehicle's location, speed, acceleration, and direction). Based on the obstacle information and vehicle driving information, the vehicle predicts the future behavior of the obstacle and the vehicle to determine whether there is a risk of collision. Based on the obstacle information, the vehicle determines whether the obstacle is dynamic.
- surrounding obstacles such as the obstacle's location, speed, direction, acceleration, shape, size, and other information
- vehicle driving information such as the vehicle's location, speed, acceleration, and direction
- vehicle 11 obtains information about vehicle 12, and vehicle 12 is traveling to the left at a certain speed. There is a possibility that the future motion trajectory of vehicle 11 and the future motion trajectory of vehicle 12 intersect at the same time point, and vehicle 12 is determined to be the first obstacle.
- the vehicle turns left.
- the future trajectory determined by the vehicle's state is shown by the solid line in Figure 7.
- the vehicle detects a pedestrian walking in front of it on the left, approaching the vehicle.
- the future trajectory of the pedestrian determined by the detected pedestrian information is shown by the dashed line in Figure 7. If the pedestrian's future trajectory intersects the vehicle's future trajectory, the vehicle will determine that the pedestrian is the first obstacle.
- the vehicle uses various sensors to obtain information about obstacles around the vehicle (such as the obstacle's location, speed, direction, acceleration, shape, size, etc.), as well as vehicle driving information (such as the vehicle's location, speed, acceleration, direction, etc.). Based on this information, the vehicle determines whether the current scene is a dynamic obstacle interaction scene. If so, the dynamic obstacle in the scene is determined as the first obstacle.
- information about obstacles around the vehicle such as the obstacle's location, speed, direction, acceleration, shape, size, etc.
- vehicle driving information such as the vehicle's location, speed, acceleration, direction, etc.
- the dynamic obstacle interaction scenario can be understood as a scenario in which the vehicle interacts with dynamic obstacles to avoid obstacles in an autonomous driving or intelligent driving scenario.
- the vehicle detects surrounding obstacles and first determines whether there is a dynamic obstacle; if so, it determines whether the obstacle poses a risk of collision with the vehicle. If so, it determines that the current scene is a dynamic obstacle interaction scene, and determines the obstacle in the scene as the first obstacle.
- a vehicle traveling straight ahead detects a pedestrian running to the right from its left front.
- the pedestrian's future trajectory is likely to intersect with the vehicle's, and the distance between the vehicle and pedestrian is decreasing.
- This scenario is considered a dynamic obstacle interaction scenario, with the pedestrian identified as the primary obstacle.
- Vehicle-pedestrian interactions can occur in the following ways: the vehicle detects the pedestrian and slows down to yield; the pedestrian sees the vehicle and slows down to yield; or the pedestrian sees the vehicle but the distance between them is too far, causing the pedestrian to speed past.
- the vehicle also detects the vehicle's surrounding environment information (such as road information, weather information, traffic light information, etc.) in real time.
- the vehicle screens the obstacles around the vehicle based on the surrounding environment information, obstacle information, and vehicle driving information to determine the first obstacle.
- the above embodiment uses a single dynamic obstacle as an example.
- multiple dynamic obstacles may exist around the vehicle.
- the obstacle closest to the vehicle and posing a collision risk is determined as the primary obstacle.
- the vehicle can treat each dynamic obstacle as a primary obstacle and combine information from multiple primary obstacles for speed planning.
- obstacles around the vehicle are screened and obstacles with which the vehicle has a collision risk are identified as the first obstacle. This allows the vehicle to more accurately identify potential collision threats, subsequently assess the risk level of the obstacle, efficiently adjust the vehicle's speed based on the risk level, flexibly control the vehicle, reduce unnecessary braking or avoidance maneuvers, improve vehicle traffic flow and driving efficiency, and ultimately achieve intelligent driving and an enhanced driving experience.
- the vehicle determines a first risk level based on the first information about the first obstacle and the first driving information of the vehicle.
- the first risk level is used to indicate the likelihood of a collision between the first obstacle and the vehicle based on the first information and the first driving information.
- the first information includes, but is not limited to: the current position of the first obstacle (such as the coordinate information of the first obstacle relative to the vehicle), the current distance of the first obstacle (such as the distance between the first obstacle and the vehicle, so as to evaluate the spatial relationship between the first obstacle and the vehicle), the current speed of the first obstacle (such as determining the movement speed of the first obstacle by continuously detecting the position change of the first obstacle), the current acceleration of the first obstacle, the current direction of the first obstacle, the shape of the first obstacle, the size of the first obstacle, the attributes of the first obstacle (such as vehicle, pedestrian, building, etc.), the current state of the first obstacle (such as stationary, moving, accelerating, decelerating, etc.), etc.
- the current position of the first obstacle such as the coordinate information of the first obstacle relative to the vehicle
- the current distance of the first obstacle such as the distance between the first obstacle and the vehicle, so as to evaluate the spatial relationship between the first obstacle and the vehicle
- the current speed of the first obstacle such as determining the movement speed of the first obstacle by continuously
- the first driving information includes but is not limited to: the current position of the vehicle (such as the coordinate information of the vehicle), the current speed of the vehicle, the current acceleration of the vehicle, the current direction of the vehicle, etc.
- the vehicle predicts risk information based on the first information of the first obstacle and the first driving information of the vehicle, and then evaluates the lateral and longitudinal risks based on the risk information to determine a first risk level.
- Risk information includes, but is not limited to, the first collision point, collision time, the current distance between the vehicle and the first collision point, and the current distance between the first obstacle and the first collision point.
- the first collision point is the intersection of the future trajectory of the first obstacle predicted based on the first information and the future trajectory of the vehicle predicted based on the first driving information.
- Risk information may also include information about the obstacle, the vehicle, and information indicating the relative relationship between the obstacle and the vehicle, such as the obstacle's speed, the vehicle's speed, the speed difference between the obstacle and the vehicle, and the distance between the obstacle and the vehicle.
- the vehicle processes the first information about the first obstacle and the first driving information of the vehicle according to a predefined algorithm or model to determine risk information.
- the vehicle assesses the lateral direction (also described as lateral risk) and the longitudinal direction (also described as longitudinal risk) based on the risk information, and determines a risk level based on the lateral risk and the longitudinal risk.
- lateral risk refers to the risk in the left and right directions of the vehicle, including the possibility of avoiding obstacles or colliding with obstacles.
- Longitudinal risk refers to the risk in the front and back directions of the vehicle, including the possibility of slowing down and braking or accelerating to avoid collisions.
- the predefined algorithms include, but are not limited to, motion prediction algorithms, trajectory generation algorithms, collision detection algorithms, and the like.
- the vehicle processes first information about a first obstacle using a trajectory generation algorithm to determine a future motion trajectory of the first obstacle.
- the vehicle processes first driving information of the vehicle using a trajectory generation algorithm to determine a future motion trajectory of the vehicle.
- the vehicle processes the future motion trajectory of the first obstacle and the future motion trajectory of the vehicle using a collision detection algorithm, analyzes the intersection of the two trajectories, and determines risk information.
- the predefined algorithms may also include machine learning algorithms for predicting obstacle behavior (such as the future motion trajectory of the first obstacle), and Bayesian filtering algorithms for fusing data detected by various sensors to accurately estimate the distance and speed information between the first obstacle and the vehicle.
- the predefined model may be a trained machine learning model or a deep learning model (neural network model).
- the neural network model may be a preconfigured neural network model in the vehicle or a neural network model obtained from an open source platform or cloud service.
- the neural network model may be a convolutional neural network (CNN).
- a predefined model is a CNN model for assessing collision risk.
- the CNN model training process includes: collecting a large amount of data from real driving scenarios as input data, including but not limited to vehicle driving information and obstacle information.
- the collected input data is preprocessed by features extraction and other preprocessing operations.
- the preprocessed data is labeled to obtain labels corresponding to the input data (for example, each set of input data (obstacle information and vehicle information) is labeled with corresponding risk information).
- An appropriate machine learning algorithm or deep learning model is selected to construct the CNN model.
- the labeled dataset is input into the CNN model, and prediction results are obtained through training using the forward propagation algorithm.
- a loss function is calculated based on the prediction results and labels, and a backpropagation algorithm is initiated to adjust the model parameters of the CNN model based on the loss function.
- the CNN model repeats the above forward propagation, loss calculation, backpropagation, and parameter adjustment process until the CNN model converges or reaches the set number of iterations. Training ends and the CNN model is output.
- the trained CNN model can be applied to actual prediction, recognition, and other tasks.
- the vehicle determines risk information through a CNN model. For example, first information about a first obstacle and first driving information of the vehicle are input into the CNN model, and the CNN model inputs the risk information.
- model in the embodiments of the present application can be trained based on a large amount of real driving data, which can be understood as relevant data when the driver responds to obstacles. This allows the vehicle to make decisions more consistent with those of a human driver, thereby improving the vehicle's intelligence.
- the vehicle performs a prediction based on the first information of the first obstacle and the first driving information of the vehicle to obtain prediction information, and then performs short-term deduction and collision detection based on the prediction information to determine risk information.
- Predicted information can be understood as the future behavior of obstacles and vehicles. This includes, but is not limited to, the future speed, position, state, acceleration, and trajectory of obstacles, and the future speed, position, state, acceleration, and trajectory of vehicles.
- the first obstacle is a dynamic obstacle.
- the vehicle determines predicted information about the first obstacle, such as the future trajectory and future speed of the first obstacle, based on the first information about the first obstacle and a predefined algorithm or model.
- the vehicle determines predicted information about the vehicle, such as the future trajectory and future speed of the vehicle, based on the first driving information and a predefined algorithm or model.
- the vehicle determines a first collision point based on the future trajectory of the first obstacle and the future trajectory of the vehicle. Based on the first collision point, the first information, and the first driving information, the vehicle determines risk information such as the collision time, the current distance between the vehicle and the first collision point, and the current distance between the first obstacle and the first collision point.
- Predefined algorithms include, but are not limited to, motion prediction algorithms and trajectory generation algorithms.
- Predefined models include trained machine learning models, etc.
- the vehicle After receiving the prediction information, the vehicle performs a short-term simulation based on the prediction information, simulating the relative motion of the vehicle and obstacles within a short period of time. Based on the short-term simulation, the vehicle performs collision detection, comparing the speed, position, and other information of the vehicle and obstacle to determine the risk.
- collision detection includes but is not limited to geometric collision detection, time arrival analysis, etc.
- the vehicle determines the risk level.
- the vehicle processes the risk information based on a longitudinal risk assessment algorithm or model and a lateral risk assessment algorithm or model to determine the longitudinal and lateral risks.
- the vehicle comprehensively analyzes the longitudinal and lateral risks, considering the combined impact of different risks on driving, to determine the risk level.
- a mapping table of risk information and risk levels is pre-configured in the vehicle.
- the mapping table includes risk information of different dimensions and their corresponding risk levels.
- the vehicle determines the risk level corresponding to the current risk information based on the mapping table. For example, by performing big data analysis and evaluation on historical data and simulation data, the relationship between risk information of different dimensions (such as distance, speed, track intersection, etc.) and risk levels is determined, a mapping table of risk information and risk levels is developed, and the mapping table is integrated into the vehicle. It is understandable that the mapping table can be continuously updated and improved in line with technological development and actual usage to adapt to more traffic environments.
- vehicle 11 executes the intelligent driving method of the present application. While driving, vehicle 11 detects vehicle 12, determines that there is a risk of collision between the vehicle 12 and the vehicle 12, and identifies vehicle 12 as the first obstacle. Vehicle 11 obtains information such as vehicle 12's current speed, position, direction, and acceleration. Vehicle 11 also obtains its own current driving information, such as its own position, speed, acceleration, and direction.
- vehicle 11 determines its own future trajectory, vehicle 12's future trajectory, the collision point, the time it takes for vehicle 11 to reach the collision point to be greater than the time it takes for vehicle 12 to reach the collision point, the distance it takes for vehicle 11 to reach the collision point to be greater than the distance it takes for vehicle 12 to reach the collision point, the speed of vehicle 11 to be less than the speed of vehicle 12, and the distance between vehicle 11 and vehicle 12 to be greater. Based on this information, it can be determined that the time it takes for vehicle 11 to reach the collision point is less than the time it takes for vehicle 12 to reach the collision point, indicating a low risk of collision between the two vehicles. Therefore, the risk level is determined to be medium.
- the vehicle executes the intelligent driving method of the present application, detects pedestrian information during vehicle driving, and determines that there is a risk of collision between the pedestrian and the vehicle. Based on the vehicle information and the pedestrian information, the vehicle determines that the pedestrian's speed is less than the vehicle's speed, and the distance from the pedestrian to the collision point is less than the distance from the vehicle to the collision point. The time it takes for the pedestrian to reach the collision point is less than the time it takes for the vehicle to reach the collision point, but the two are relatively close, and the pedestrian is still walking. Based on the above information, it can be determined that if the pedestrian and vehicle are both moving at their current speeds, the risk of collision between the two is greater, so the risk level is high.
- the vehicle may also determine a control strategy corresponding to the risk level according to preset rules.
- the preset rules include a correspondence between the risk level and the control strategy, and the control strategy includes a yield strategy or an overtaking strategy.
- the vehicle predicts the collision risk based on the information of the first obstacle and the vehicle, determines the corresponding control strategy based on the collision risk, plans the vehicle speed based on the control strategy and collision risk, efficiently adjusts the vehicle's driving speed, controls the vehicle to take a more appropriate approach to avoid collisions, improves the vehicle's traffic rate, reduces unnecessary braking, and improves the driving experience.
- the vehicle comprehensively considers obstacle information and vehicle information to accurately determine the risk level of collision between the obstacle and the vehicle, so that the vehicle can make more reasonable decisions based on the risk level, control the vehicle's driving speed, improve the vehicle's traffic efficiency, and enhance the driving experience and vehicle safety.
- the vehicle acquires a first cruising speed that matches the first risk level; the first cruising speed is the minimum driving speed of the vehicle at the first risk level for avoiding collision with the first obstacle.
- a mapping table between risk levels and cruising speeds is pre-configured in the vehicle.
- the vehicle can determine a first cruising speed that matches the first risk level based on the mapping relationship.
- the vehicle calculates a first cruising speed corresponding to the first risk level through a predefined rule or model based on the first risk level, the first information, and the first driving information.
- the predefined rules can be a set of rules regarding risk level and cruising speed determined through logical reasoning, analysis of historical data, or empirical data.
- the predefined model can be a machine learning model or deep learning model obtained by training historical data on risk level and cruising speed.
- the first cruising speed should be a valid value.
- the first cruising speed is greater than or equal to 0.
- the collision risk between vehicles 11 and 12 in FIG1 is low, with a medium risk level.
- the first cruising speed of vehicle 11 is determined to be speed 1.
- the collision risk between the vehicle and the pedestrian in FIG7 is high, with a high risk level.
- the first cruising speed of the vehicle is determined to be speed 2.
- Speed 1 should be greater than speed 2.
- a higher risk level indicates a greater likelihood of collision between the vehicle and an obstacle. Therefore, the vehicle's speed should be reduced as quickly as possible, allowing the vehicle to cruise at a lower speed.
- S604 The vehicle controls the vehicle to travel at a first cruising speed.
- the first cruising speed can be understood as a target driving speed planned by the vehicle based on the current risk level to avoid collisions between the vehicle and dynamic obstacles in the future. Therefore, the vehicle needs to adjust its current driving speed to the first cruising speed.
- Controlling the vehicle to travel at the first cruising speed should be understood as controlling the vehicle's travel speed to adjust from the current speed to the first cruising speed, and then controlling the vehicle to travel stably at the first cruising speed.
- This control process includes a speed adjustment process and a speed maintenance process.
- the vehicle generates a speed planning curve based on the current speed of the vehicle and the first cruising speed, so that the vehicle adjusts the vehicle's driving speed to the first cruising speed according to the speed planning curve, thereby allowing the vehicle to travel at the first cruising speed.
- the speed planning model incorporates the first cruising speed as a constraint to ensure that the vehicle's speed within the speed planning curve does not fall below the set first cruising speed.
- the speed planning model treats speed planning as an optimization problem, finding the optimal speed curve while satisfying various constraints.
- the speed planning model generates a smooth speed planning curve connecting the current speed and the first cruising speed, ensuring a smooth transition from the current speed to the first cruising speed.
- the vehicle adjusts its speed based on the information on the speed planning curve, gradually adjusting it to the first cruising speed.
- the speed planning curve of the vehicle 11 in FIG1 is speed planning curve 1.
- the speed planned by speed planning curve 1 is: the vehicle 11 smoothly drops from the current speed V1 to speed 1, and then maintains speed 1 to continue driving.
- the speed planning curve of the vehicle in FIG7 is speed planning curve 2.
- the speed planned by speed planning curve 2 is: the vehicle smoothly drops from the current speed V2 to speed 2, and then maintains speed 2 to continue driving.
- the risk level of the vehicle 11 in FIG1 is medium risk, and the risk level of the vehicle in FIG7 is high risk, and speed 1 is greater than speed 2.
- the vehicle 11 in FIG1 does not need to drop to speed 1 quickly, so the vehicle deceleration time is greater than the vehicle deceleration time in FIG7 .
- the vehicle's driving speed is smoothly adjusted to the first cruising speed, which can avoid problems such as vehicle loss of control and reduced comfort caused by sudden changes in vehicle speed, thereby ensuring the safety of the vehicle and driving experience.
- the vehicle before controlling the vehicle to travel at the first cruising speed, the vehicle needs to determine whether the first cruising speed can be used for speed planning.
- the vehicle determines whether the distance between the current vehicle position and the first collision point is greater than a first threshold; if so, the vehicle adopts a first cruising speed to plan the vehicle's driving speed; otherwise, the vehicle does not adopt the first cruising speed to plan the vehicle's driving speed.
- the first threshold is the minimum safe distance between the vehicle and the first obstacle.
- the first threshold is the minimum safe distance between the vehicle and the first obstacle, which is pre-calculated based on dynamic characteristics and safety considerations.
- vehicle 11 and vehicle 12 are traveling toward collision point 13.
- Vehicle 12 may pass the collision point first, but if vehicle 12 suddenly stalls when passing the collision point, the distance between vehicle 11 and collision point 13 is less than the first threshold value. The distance is too short to provide vehicle 11 with sufficient reaction time to deal with the emergency. Even if vehicle 11 brakes suddenly, there is still a risk of collision. Therefore, before the vehicle enters cruise control, it is necessary to ensure that there is enough space around the vehicle to provide sufficient reaction time to respond to various traffic situations in a timely manner.
- the vehicle determines whether there is a second obstacle on the vehicle's future motion trajectory, where the second obstacle is a static obstacle and the distance between the current vehicle position and the second obstacle is greater than the distance between the current vehicle position and the first collision point; if so, it determines whether the distance between the current vehicle position and the second obstacle is greater than a second threshold; if so, the vehicle adopts the first cruising speed to plan the vehicle's driving speed; otherwise, the first cruising speed is not adopted to plan the vehicle's driving speed.
- the second threshold is the minimum safe distance between the vehicle and the second obstacle
- the first threshold is the minimum safe distance between the vehicle and the second obstacle that is pre-calculated based on dynamic characteristics and safety considerations.
- vehicle 91 is traveling forward and vehicle 92 is traveling left.
- Vehicle 91 is executing the intelligent driving method of the present application.
- Vehicle 91 detects vehicle 92 as a dynamic obstacle. Based on information from both vehicles 91 and 92, it determines a risk level of 1 and sets a first cruising speed of 3 for vehicle 91 based on this risk level.
- vehicle 91 also detects that a rock 94 is located after collision point 93 on its predicted future trajectory, and the distance between rock 94 and vehicle 91 is less than a second threshold. In this case, vehicle 91 does not use the planned speed of 3 and actively brakes to a stop.
- the vehicle determines whether the distance between the current vehicle position and the first collision point is greater than a first threshold; if so, it continues to determine whether there is a second obstacle on the vehicle's future motion trajectory, the second obstacle is a static obstacle, and the distance between the current vehicle position and the second obstacle is greater than the distance between the current vehicle position and the first collision point; if so, it determines whether the distance between the current vehicle position and the second obstacle is greater than a second threshold; if so, the vehicle adopts the first cruising speed to plan the vehicle's driving speed; otherwise, the first cruising speed is not adopted to plan the vehicle's driving speed.
- the first threshold is lower than the second threshold. Therefore, before using the first cruising speed to plan the vehicle's driving speed, it is necessary to determine whether there is sufficient space between the current vehicle and the first collision point, as well as between the vehicle and the static obstacle, to avoid the obstacle through cruising. If there is insufficient space, cruising should not be used to avoid the obstacle. Instead, emergency braking can be used to reserve more room for the vehicle to maneuver.
- the vehicle must ensure a sufficient safe distance from the collision point to allow it to stop before the collision point, ensuring sufficient room for maneuver.
- determining whether the vehicle can adopt the first cruising speed to plan the vehicle's driving speed based on the safety distance can better ensure vehicle safety.
- the vehicle changes its driving speed according to the first cruising speed and the speed planning curve of the current speed plan. While the vehicle changes the driving speed, it also collects information about the first obstacle and the vehicle in real time, determines the corresponding risk level based on the information collected in real time, and adaptively adjusts the cruising speed according to the risk level.
- the vehicle determines the cruising speed corresponding to the risk level, it first plans the speed based on the cruising speed, and then determines whether the cruising speed can be adopted (that is, whether a sufficient safety distance is left based on the cruising speed so that the vehicle can respond in time). If it can be adopted, the vehicle's driving speed is adjusted according to the speed planning curve determined by the first cruising speed and the current speed.
- the vehicle uses the first cruising speed and the speed planning curve determined by the current speed to adjust the vehicle's driving speed, if the risk level corresponding to the first obstacle information and vehicle information collected by the vehicle in real time has not changed, the cruising speed will not be changed. After the vehicle's driving speed is adjusted to the first cruising speed, the vehicle travels at the first cruising speed.
- the vehicle when the vehicle is adjusting the vehicle's driving speed using a speed planning curve determined by a first cruising speed and a current speed, if the risk level corresponding to the first obstacle information and vehicle information collected by the vehicle in real time changes, the vehicle determines its matching cruising speed based on the changed risk level, and replans the vehicle's driving speed based on the cruising speed so that the vehicle travels according to the replanned cruising speed.
- the vehicle detects the first obstacle information and the vehicle information in real time. If a second risk level is determined based on the second information of the first obstacle and the second driving information of the vehicle; the second risk level is higher than the first risk level, and the second risk level is used to indicate the possibility of a collision between the first obstacle and the vehicle under the second information and the second driving information; then a second cruising speed matching the second risk level is obtained; the second cruising speed is lower than the first cruising speed, and the second cruising speed is the minimum driving speed of the vehicle to avoid a collision with the first obstacle under the second risk level; and the vehicle is controlled to travel at the second cruising speed.
- the vehicle detects the first obstacle information and the vehicle information in real time. If a third risk level is determined based on the third information of the first obstacle and the third driving information of the vehicle; the third risk level is lower than the first risk level, and the third risk level is used to indicate the possibility of a collision between the first obstacle and the vehicle under the third information and the third driving information; then a third cruising speed matching the third risk level is obtained; the third cruising speed is higher than the first cruising speed, and the third cruising speed is the minimum driving speed for the vehicle to avoid a collision with the first obstacle under the first risk level; and the vehicle is controlled to travel at the third cruising speed.
- the first information described above can be understood as the obstacle information collected at the first moment, and the second/third information can be understood as the obstacle information collected at the Nth moment.
- N is an integer greater than 1.
- the first driving information can be understood as the vehicle driving information collected at the first moment, and the second/third driving information can be understood as the vehicle driving information collected at the Nth moment.
- the second risk level is the collision risk between the first obstacle and the vehicle determined based on the second information and the second driving information.
- the third risk level is the collision risk between the first obstacle and the vehicle determined based on the third information and the third driving information. Both the second risk level and the second risk level are different from the first risk level.
- the second risk level being higher than the first risk level can be understood as indicating that the collision risk between the first obstacle and the vehicle at time N has increased compared to the first time (e.g., if the first obstacle does not slow down and cuts in front of the vehicle), so the vehicle should reduce its cruising speed to avoid the first obstacle and allow it to pass the collision point first. Therefore, the second cruising speed corresponding to time N is lower than the first cruising speed at the first time, and the vehicle uses the newly determined second cruising speed for speed planning, allowing the vehicle to travel at this newly determined second cruising speed.
- the third risk level being higher than the first risk level can be understood as meaning that the collision risk between the first obstacle and the vehicle at time N has decreased compared to the collision risk at time one (e.g., the first obstacle has slowed down and given way). Therefore, the vehicle should be able to increase its cruising speed and pass the collision point first. Therefore, the third cruising speed corresponding to time N is lower than the first cruising speed at time one.
- the vehicle uses the newly determined third cruising speed for speed planning, allowing the vehicle to travel at this newly determined third cruising speed.
- the vehicle after planning the speed of vehicle 11 according to speed 1, the vehicle continues to detect information of vehicle 11 and vehicle 12. If it is detected that vehicle 12 is decelerating, the speed of vehicle 12 is less than the speed of vehicle 11, and the time taken for vehicle 12 to reach the collision point is not much different from the time taken for vehicle 11 to reach the collision point, it is determined that the risk of collision between the two has increased.
- the risk level at this time is high risk, and vehicle 11 should reduce the cruising speed, such as speed 4, and plan the vehicle speed based on speed 4 so that the vehicle can travel according to speed 4.
- the prerequisite for the vehicle to plan the vehicle's driving speed according to the second cruising speed and the vehicle's driving speed according to the third cruising speed should be: the distance between the vehicle position at the Nth moment and the first collision point is greater than the first threshold; or, the distance between the vehicle position at the Nth moment and the first collision point is greater than the first threshold, and there is a static second obstacle on the vehicle's future motion trajectory, the distance between the vehicle position at the Nth moment and the second obstacle is greater than the distance between the vehicle position at the Nth moment and the first collision point, and the distance between the vehicle position at the Nth moment and the second obstacle is also greater than the second threshold.
- the vehicle detects information of the obstacle and the vehicle in real time, and determines a fourth risk level based on fourth information of the first obstacle and fourth driving information of the vehicle; the fourth risk level is higher than the first risk level, and the fourth risk level is used to indicate the possibility of a collision between the first obstacle and the vehicle under the fourth information and the fourth driving information; a fourth cruising speed matching the fourth risk level is obtained; the fourth cruising speed is lower than the first cruising speed, and the fourth cruising speed is the minimum driving speed of the vehicle to avoid a collision with the first obstacle under the fourth risk level; it is determined that the distance between the current position of the vehicle and the first collision point is less than or equal to a first threshold; the first collision point is the intersection of the future motion trajectory of the obstacle and the future motion trajectory of the vehicle predicted based on the first information and the first driving information, and the first threshold is the minimum safe distance between the vehicle and the first obstacle; the driving speed of the vehicle is reset to
- Vehicle 11 After planning the speed of vehicle 11 based on speed 1, the vehicle continues to detect information about vehicles 11 and 12. If vehicle 12 is detected to be decelerating, and its speed is less than that of vehicle 11, and the time it takes for vehicle 12 to reach the collision point is not much different from that of vehicle 11, then the risk of collision between the two vehicles may increase, and the risk level at this time is high. Vehicle 11 should reduce its cruising speed, such as to speed 4, and plan its speed based on speed 4. However, if it is detected that the distance between the current vehicle position and the collision point is less than a first threshold, the vehicle cannot respond to the unexpected situation in a timely manner, so the vehicle should exit the cruising strategy and gradually reduce its driving speed to zero.
- the vehicle detects information of the obstacle and the vehicle in real time, and determines a fifth risk level based on fifth information of the first obstacle and fifth driving information of the vehicle; the fifth risk level is higher than the first risk level, and the fifth risk level is used to indicate the possibility of a collision between the first obstacle and the vehicle under the fifth information and the fifth driving information; obtains a fifth cruising speed that matches the fifth risk level; the fifth cruising speed is lower than the first cruising speed, and the fifth cruising speed is the minimum driving speed for the vehicle to avoid a collision with the first obstacle under the fifth risk level; determines that the distance between the current vehicle position and the second obstacle is less than or equal to a second threshold; the second threshold is the minimum safe distance between the vehicle and the
- the fourth information/fifth information can be understood as obstacle information collected at time N
- the fourth driving information/fifth driving information can be understood as vehicle driving information collected at time N.
- the fourth risk level is the collision risk between the first obstacle and the vehicle determined based on the fourth information and the fourth driving information.
- the fifth risk level is the collision risk between the first obstacle and the vehicle determined based on the fifth information and the fifth driving information.
- the fourth risk level is higher than the first.
- the vehicle should reduce its cruising speed to avoid the first obstacle, allowing it to pass the collision point first.
- the vehicle also detects that the distance between the vehicle's position at time N and the first collision point is less than or equal to the first threshold. This means that the distance between the vehicle's position at time N and the first collision point is less than or equal to the minimum safe distance.
- the vehicle should smoothly exit the cruising strategy, transitioning smoothly from the vehicle's speed at time N to zero, allowing the vehicle to brake before the collision point and leaving sufficient space.
- the fifth risk level is lower than the first risk level.
- the vehicle should increase its cruising speed and pass the obstacle first. However, the vehicle detects a second, static obstacle in its future trajectory.
- the distance between the vehicle's position at moment N and the second obstacle is greater than the distance between the vehicle's position at moment N and the first collision point, and the distance between the vehicle's position at moment N and the second obstacle is less than or equal to the second threshold. In this case, the distance between the vehicle's position at moment N and the second obstacle is less than or equal to the minimum safe distance.
- the vehicle continues to move forward, the distance between the vehicle and the collision point is getting closer and closer, and the distance between the vehicle and the obstacle is also getting closer and closer.
- the vehicle actively exits the cruise state to ensure the safe driving of the vehicle so that the vehicle can respond to various traffic conditions in a timely manner.
- the application scenario of this application is an autonomous driving scenario, where the vehicle autonomously perceives the surrounding environment and makes corresponding decisions.
- the decisions made by the vehicle should be close to the behavior and decisions of human drivers.
- the vehicle when the vehicle avoids obstacles according to the cruising speed corresponding to the current risk level, it determines whether the safety distance is met based on the real-time distance information. When the real-time distance is less than the safety distance, it exits the cruise smoothly to ensure the safety of the vehicle and the comfort of the user.
- a vehicle when a vehicle faces a dynamic obstacle, it collects real-time information of the dynamic obstacle and the vehicle, determines the collision risk between the dynamic obstacle and the vehicle, determines the minimum cruising speed of the vehicle based on the collision risk, plans the vehicle's driving speed based on the minimum cruising speed, and allows the vehicle to travel to the collision point at the minimum cruising speed.
- the vehicle detects the behavior of the dynamic obstacle in real time and conducts a game. If the dynamic obstacle is not avoided, the vehicle can travel at a lower speed, brake at a position close to the collision point, and wait for the dynamic obstacle to pass the collision point before driving again. Alternatively, if the dynamic obstacle is not avoided, the vehicle can travel at an even lower speed.
- the vehicle When the dynamic obstacle passes the collision point, the vehicle is still a certain distance away from the collision point. After the dynamic obstacle passes the collision point, the vehicle can accelerate to pass the collision point. Alternatively, if the dynamic obstacle is avoided, the speed can be gradually increased according to the risk level, and the collision point can be passed first.
- vehicle 1001 detects that vehicle 1002 will cross its future driving path, and the collision point of the two vehicles is 1003.
- vehicle 1001 actively brakes and the braking point is 1004.
- vehicle 1001 adopts a cruising strategy and cruises at a lower speed.
- the final braking point is 1005, which is closer to the collision point.
- the technical solution of the present application reduces or increases the vehicle's cruising speed according to the real-time collision risk when the vehicle faces a dynamic obstacle, and flexibly controls the vehicle based on the cruising speed to negotiate with the dynamic obstacle, avoiding unnecessary braking or braking operations, so that the vehicle outputs a human-like speed planning strategy, increases the vehicle's bargaining ability and human-likeness, improves the vehicle's traffic efficiency and driving efficiency, and enhances the driving experience.
- the first obstacle can be a static obstacle, such as an immovable object or building.
- the vehicle After the vehicle detects the first obstacle, it can perform a risk assessment and slowly approach the obstacle at the minimum cruising speed until it brakes to a stop.
- the vehicle After detecting the first obstacle, the vehicle can slowly exit the cruise state. This adds a method for handling static obstacles while the vehicle is cruising.
- the vehicle determines a matching cruising speed based on the risk between the vehicle and the obstacle, plans the vehicle's driving speed based on the cruising speed, and controls the vehicle to travel at the cruising speed.
- the vehicle does not brake directly, but instead drives slowly at a lower speed, playing a game with the dynamic obstacle.
- the vehicle is flexibly controlled based on real-time risk, optimizing the intelligent driving strategy and increasing the vehicle's flexibility in avoiding dynamic obstacles in the autonomous cruising state. While ensuring vehicle safety, it also improves the vehicle's traffic efficiency and ability to overtake, enhancing the driving experience, and allowing the vehicle to better adapt to complex traffic environments.
- the above mainly introduces the solution provided by the embodiment of the present application from the perspective of method.
- it includes hardware structures and/or software modules corresponding to the execution of each function.
- the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software driven hardware manner depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
- the intelligent driving device 1100 includes a detection module 1101 and a processing module 1102.
- the intelligent driving device 1100 is used to execute the above-mentioned intelligent driving method, for example, for executing the intelligent driving method shown in Figures 6 to 10.
- the intelligent driving device 1100 may also include other modules, or the intelligent driving device 1100 may include fewer modules.
- the embodiments of the present application do not specifically limit the specific form and implementation of the intelligent driving device.
- the detection module 1101 is configured to detect a first obstacle; the first obstacle is a dynamic obstacle.
- the processing module 1102 is used to determine a first risk level based on the first information about the first obstacle and the first driving information of the vehicle; the first risk level is used to indicate the possibility of a collision between the first obstacle and the vehicle under the first information and the first driving information.
- the processing module 1102 is further configured to obtain a first cruising speed that matches the first risk level, where the first cruising speed is the minimum speed at which the vehicle can avoid colliding with the first obstacle under the first risk level.
- the processing module 1102 is further configured to control the vehicle to travel at a first cruising speed.
- each module in the intelligent driving device 1100 is respectively for realizing the corresponding processes of the intelligent driving method described in the above method embodiment. All relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional unit. For the sake of brevity, they will not be repeated here.
- the intelligent driving device 1100 shown in FIG11 may further include a storage module (not shown in FIG11 ) storing a program or instruction.
- the intelligent driving device 1100 shown in FIG11 may perform the intelligent driving method described in the above method embodiment.
- the technical effects of the intelligent driving device 1100 shown in FIG11 may refer to the technical effects of the intelligent driving method described in the above method embodiment, and will not be repeated here.
- the chip system 1200 includes at least one processor 1201 and at least one interface circuit 1202.
- the one processor may be the processor 1201 shown in the solid box in Figure 12 (or the processor 1201 shown in the dotted box)
- the one interface circuit may be the interface circuit 1202 shown in the solid box in Figure 12 (or the interface circuit 1202 shown in the dotted box).
- the two processors include the processor 1201 shown in the solid box in Figure 12 and the processor 1201 shown in the dotted box
- the two interface circuits include the interface circuit 1202 shown in the solid box in Figure 12 and the interface circuit 1202 shown in the dotted box. This is not limited.
- the processor 1201 and the interface circuit 1202 can be interconnected via a line.
- the interface circuit 1202 can be used to receive signals.
- the interface circuit 1202 can be used to send signals to other devices (such as the processor 1201).
- the interface circuit 1202 can read instructions stored in the memory and send the instructions to the processor 1201.
- the intelligent driving device can perform the various steps in the above embodiment.
- the chip system can also include other discrete devices, which is not specifically limited in the embodiment of the present application.
- the chip system can be a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), a system on a chip (SoC), a central processing unit (CPU), a network processor (NP), a digital signal processor (DSP), a microcontroller unit (MCU), a programmable logic device (PLD) or other integrated chips.
- FPGA field programmable gate array
- ASIC application-specific integrated circuit
- SoC system on a chip
- CPU central processing unit
- NP network processor
- DSP digital signal processor
- MCU microcontroller unit
- PLD programmable logic device
- each step in the above method embodiment can be completed by hardware integrated logic circuits in a processor or by software instructions.
- the method steps disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware processor, or by a combination of hardware and software modules in a processor.
- An embodiment of the present application also provides a computer-readable storage medium storing one or more computer programs, wherein the one or more computer programs include instructions that, when executed by a computer, enable the computer to execute the corresponding process of the intelligent driving method in the above embodiment.
- the disclosed methods may be implemented as computer program instructions encoded in a machine-readable format on a computer-readable storage medium or on other non-transitory media or articles of manufacture.
- An embodiment of the present application also provides a computer program product.
- the computer program product When the computer program product is run on a computer, it enables the computer to execute the above-mentioned related steps to implement the intelligent driving method in the above-mentioned embodiment.
- the apparatus, computer-readable storage medium, computer program product, or chip provided in the embodiments of the present application are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects of the corresponding methods provided above, and will not be repeated here.
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Human Computer Interaction (AREA)
- Traffic Control Systems (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
Abstract
本申请提供一种智能驾驶方法、装置及车辆,应用于智能驾驶领域。该方法包括:检测到第一障碍物;该第一障碍物为动态障碍物;根据该第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级;该第一风险等级用于表示在该第一信息和该第一行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第一风险等级匹配的第一巡航速度,该第一巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制车辆以第一巡航速度行驶。本申请能够提高车辆在自动驾驶巡航状态下避让动态障碍物的灵活性,在保障车辆安全的同时,提高车辆的行驶效率、通行效率和抢行能力,提升驾乘体验。
Description
本申请要求于2024年03月27日提交国家知识产权局、申请号为202410366424.3、申请名称为“智能驾驶方法、装置及车辆”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
本申请涉及智能驾驶领域,尤其涉及一种智能驾驶方法、装置及车辆。
当车辆处于自动驾驶的巡航状态行驶时,若检测到车辆前方有动态障碍物(如走动的行人或行驶的车辆等),车辆通常采用较为保守的处理方式,如紧急刹停车辆,主动避让动态障碍物,以减少碰撞的可能性。但该处理方式缺乏灵活性,导致车辆驾乘体验不佳。
因此,如何提高车辆在自动驾驶巡航状态下的避让动态障碍物的灵活性,提升驾乘体验成为亟需解决的技术问题。
本申请提供了一种智能驾驶方法、装置及车辆,能够提高车辆在自动驾驶巡航状态下避让动态障碍物的灵活性,提高车辆的行驶效率和通行效率,提升驾乘体验。
为了实现上述目的,本申请提供了如下技术方案:
第一方面,本申请提供一种智能驾驶方法,该方法包括:检测到第一障碍物;该第一障碍物为动态障碍物;根据该第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级;该第一风险等级用于表示在该第一信息和该第一行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第一风险等级匹配的第一巡航速度,该第一巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制该车辆以第一巡航速度行驶。
本申请中,基于车辆与障碍物之间的风险确定匹配的巡航速度,基于巡航速度规划车辆行驶速度,以便灵活躲避障碍物。提高了车辆在自动驾驶巡航状态下避让动态障碍物的灵活性,在保障车辆安全的同时,提升了驾乘体验,提高了车辆的行驶效率、通行效率和抢行能力,以便车辆更好的适应复杂的交通环境。
在一些实施例中,车辆处于自动驾驶模式或智能驾驶模式,车辆处于巡航状态行驶。
在一些实施例中,第一障碍物是车辆周围与车辆可能发生碰撞风险的动态障碍物。
在一些实施例中,车辆在检测到车辆周围的障碍物时,也会获取该障碍物的信息。车辆根据各障碍物的信息分析各障碍物与车辆是否存在碰撞的风险,以及障碍物是否为动态障碍物,将存在碰撞风险的动态障碍物确定为第一障碍物。
在本申请中,对于车辆周围的障碍物进行筛选确定第一障碍物,可以更准确的识别潜在的碰撞威胁,以便后续评估该障碍物的风险等级,根据风险等级高效调整车辆的速度规划,减少不必要的制动或避让操作,从而提升驾乘体验,提高车辆通行率和行驶效率。
根据第一方面,或者以上第一方面的任意一种实现方式,该方法还包括:根据该第一障碍物的第二信息和车辆的第二行驶信息,确定第二风险等级;该第二风险等级高于该第一风险等级,该第二风险等级用于表示在该第二信息和该第二行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第二风险等级匹配的第二巡航速度;该第二巡航速度低于该第一巡航速度,该第二巡航速度为在该第二风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制该车辆以第二巡航速度行驶。
根据第一方面,或者以上第一方面的任意一种实现方式,该方法还包括:根据该第一障碍物的第三信息和车辆的第三行驶信息,确定第三风险等级;该第三风险等级低于该第一风险等级,该第三风险等级用于表示在该第三信息和该第三行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第三风险等级匹配的第三巡航速度;该第三巡航速度高于该第一巡航速度,该第三巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制车辆以第三巡航速度行驶。
本申请中,车辆实时检测第一障碍物的信息和车辆的行驶信息,根据每一时刻的障碍物信息和车辆信息确定该时刻对应的碰撞风险,基于当前时刻的碰撞风险动态改变巡航速度,使得基于改变后的巡航速度规划的车辆速度更适应车辆周围环境的变化,有助于更大程度地保障车辆安全。车辆根据每一时刻的信息动态调整巡航速度,提高了车辆在自动驾驶巡航状态下避让动态障碍物的灵活性,以便车辆更好的适应复杂的交通环境。
根据第一方面,或者以上第一方面的任意一种实现方式,根据该第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级,包括:根据该第一障碍物的第一信息和车辆的第一行驶信息预测风险信息;该风险信息包括以下一项或多项:第一碰撞点,碰撞时间,当前该车辆位置与该第一碰撞点之间的距离,当前该第一障碍物与该第一碰撞点之间的距离;该第一碰撞点为根据该第一信息和该第一行驶信息预测的第一障碍物的未来运动轨迹和车辆的未来运动轨迹的相交点;根据该风险信息评估横纵向风险,确定该第一风险等级。
在一些实施例中,车辆根据预定义的算法或模型处理第一障碍物的第一信息和车辆的第一行驶信息,确定风险信息。车辆再根据风险信息评估横向方向的方向(也可以描述为横向风险)和纵向方向的风险(也可以描述为纵向风险),根据横向风险和纵向风险,确定风险等级。
其中,预定义的算法包括但不限于运动预测算法、轨迹生成算法、碰撞检测算法等;预定义的模型可以是训练得到的机器学习模型或深度学习模型(神经网络模型)等。例如,神经网络模型是卷积神经网络。
在一些实施例中,车辆根据第一障碍物的第一信息和车辆的第一行驶信息进行预测,得到预测信息,再根据预测信息进行短时推演和碰撞检测确定风险信息。
在一些实施例中,车辆基于纵向风险评估算法或模型、横向风险评估算法或模型处理风险信息,确定纵向风险和横向风险。车辆综合分析纵向风险和横向风险,考虑不同风险对驾驶的综合影响,确定风险等级。
在一些实施例中,车辆中预配置有风险信息和风险等级的映射表。车辆根据该映射表确定当前风险信息对应的风险等级。
本申请中,车辆综合考虑障碍物信息和车辆信息,可以准确确定障碍物与车辆发生碰撞的风险等级,以便车辆根据风险等级做出更合理的决策,提高车辆的通行效率,提升驾驶安全性。
在一些实施例中,车辆中预配置有风险等级和巡航速度的映射表。车辆可以根据该对应关系确定与第一风险等级匹配的第一巡航速度。
在另一些实施例中,车辆根据第一风险等级、第一信息和第一行驶信息,通过预定义的算法或模型计算出第一风险等级对应的第一巡航速度。
根据第一方面,或者以上第一方面的任意一种实现方式,控制车辆以第一巡航速度行驶,包括:根据该车辆的当前速度和该第一巡航速度生成速度规划曲线;根据该速度规划曲线将该车辆的行驶速度调整至第一巡航速度,控制该车辆以第一巡航速度行驶。
在本申请中,根据当前的车辆状态和障碍物信息相关的第一巡航速度,生成速度规划曲线来指导车辆的行驶速度,优化车辆行驶效率,使得车辆速度更适配周围环境,提高驾驶安全性和舒适性。
根据第一方面,或者以上第一方面的任意一种实现方式,控制车辆以第一巡航速度行驶,包括:若当前该车辆位置与该第一碰撞点之间的距离大于第一阈值,控制车辆以第一巡航速度行驶;该第一碰撞点为根据该第一信息和该第一行驶信息预测的第一障碍物的未来运动轨迹和车辆的未来运动轨迹的相交点,该第一阈值为该车辆与该第一障碍物之间的最小安全距离。
根据第一方面,或者以上第一方面的任意一种实现方式,在该车辆的未来运动轨迹上存在第二障碍物,该第二障碍物为静态障碍物,当前该车辆位置与该第二障碍物之间的距离大于当前该车辆位置与该第一碰撞点之间的距离,控制车辆以第一巡航速度行驶,包括:若当前该车辆位置与该第二障碍物之间的距离大于第二阈值,控制车辆以第一巡航速度行驶;该第二阈值为该车辆与该第二障碍物之间的最小安全距离。
在本申请中,在车辆采用第一巡航速度进行速度规划之前,判断车辆周围是否由足够的空间,是否满足安全性保障距离,有助于预防潜在碰撞风险,提高行车安全性。
根据第一方面,或者以上第一方面的任意一种实现方式,该方法还包括:根据该第一障碍物的第四信息和车辆的第四行驶信息,确定第四风险等级;该第四风险等级高于该第一风险等级,该第四风险等级用于表示在该第四信息和该第四行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第四风险等级匹配的第四巡航速度;该第四巡航速度低于该第一巡航速度,该第四巡航速度为在该第四风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;若当前该车辆位置与第一碰撞点之间的距离小于或等于第一阈值,将车辆的行驶速度归零;该第一碰撞点为根据该第一信息和该第一行驶信息预测的第一障碍物的未来运动轨迹和车辆的未来运动轨迹的相交点,该第一阈值为该车辆与该第一障碍物之间的最小安全距离。
根据第一方面,或者以上第一方面的任意一种实现方式,在该车辆的未来运动轨迹上存在第二障碍物,该第二障碍物为静态障碍物,当前该车辆位置与该第二障碍物之间的距离大于当前该车辆位置与该第一碰撞点之间的距离,该方法还包括:根据该第一障碍物的第五信息和车辆的第五行驶信息,确定第五风险等级;该第五风险等级高于该第一风险等级,该第五风险等级用于表示在该第五信息和该第五行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第五风险等级匹配的第五巡航速度;该第五巡航速度低于该第一巡航速度,该第五巡航速度为在该第五风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;若当前该车辆位置与该第二障碍物之间的距离小于或等于第二阈值,将车辆的行驶速度归零;该第二阈值为该车辆与该第二障碍物之间的最小安全距离。
本申请中,在车辆巡航避让障碍物时,根据实时的距离信息判断是否满足安全距离,当实时距离小于安全距离时,平稳退出巡航,保障车辆的安全性,以及用户的舒适性。
第二方面,本申请提供一种智能驾驶装置,该装置包括:处理器和存储器,该存储器与该处理器耦合,该存储器用于存储计算机可读指令,当该处理器从该存储器中读取该计算机可读指令,使得该智能驾驶装置执行如第一方面以及第一方面中任意一种实施方式的方法。
第三方面,本申请提供一种车辆,该车辆包括如第二方面所述的智能驾驶装置。
示例性的,车辆包括轿车、卡车、摩托车、公共汽车、割草机、娱乐车、游乐场车辆、施工设备、电车、高尔夫球车、火车等,本申请不做特别的限定。上述车辆的动力可以由汽油、柴油、电能、太阳能、或氢能等提供。
第四方面,本申请提供一种芯片系统,包括至少一个处理器和至少一个接口电路,该至少一个接口电路用于执行收发功能,该至少一个处理器用于执行如第一方面以及第一方面中任意一种实施方式的方法。
第五方面,本申请提供一种计算机可读存储介质,该计算机可读存储介质包括计算机程序,当该计算机程序被处理器执行时实现如第一方面以及第一方面中任意一种实施方式的方法。
第六方面,本申请提供一种计算机程序产品,该计算机程序产品包括计算机程序,当该计算机程序被处理器执行时实现如第一方面以及第一方面中任意一种实施方式的方法。
第二方面至第六方面以及各方面中任意一种实现方式所对应的技术效果,可参见上述第一方面及第一方面中任意一种实现方式所对应的技术效果,此处不再赘述。
图1为本申请实施例提供的车辆行驶的场景示意图;
图2为本申请实施例提供的车辆的结构示意图;
图3为本申请实施例提供的可移动智能设备的结构示意图;
图4为本申请实施例提供的智能驾驶系统的架构示意图一;
图5为本申请实施例提供的智能驾驶系统的架构示意图二;
图6为本申请实施例提供的智能驾驶方法的流程示意图;
图7为本申请实施例提供的智能驾驶场景示意图一;
图8为本申请实施例提供的智能驾驶场景示意图二;
图9为本申请实施例提供的智能驾驶场景示意图三;
图10为本申请实施例提供的智能驾驶场景示意图四;
图11为本申请实施例提供的智能驾驶装置的结构示意图;
图12为本申请实施例提供的芯片系统的结构示意图。
在本申请实施例的描述中,除非另有说明,“/”表示或的意思,例如,A/B可以表示A或B;本文中的“和/或”仅仅是一种描述关联对象的关联关系,表示可以存在三种关系,例如,A和/或B,可以表示:单独存在A,同时存在A和B,单独存在B这三种情况。
以下,术语“第一”、“第二”仅用于描述目的,而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此,限定有“第一”、“第二”的特征可以明示或者隐含地包括一个或者更多个该特征。
在本申请实施例的描述中,除非另有说明,“多个”的含义是两个或两个以上。在本申请实施例中,“示例性的”或者“例如”等词用于表示作例子、例证或说明。本申请实施例中被描述为“示例性的”或者“例如”的任何实施例或设计方案不应被解释为比其它实施例或设计方案更优选或更具优势。确切而言,使用“示例性的”或者“例如”等词旨在以具体方式呈现相关概念。
在一些示例中,巡航状态的车辆对于动态障碍物的处理策略为主动刹停,以避免潜在的碰撞风险,保障车辆安全。该处理策略仅考虑车辆本身的行为(如当前车辆到预测碰撞点的距离等),未考虑车辆与动态障碍物之间的交互作用(如动态障碍物主动让行),导致车辆通行效率低,抢行能力弱。
例如,如图1所示,车辆11处于自动驾驶巡航状态并向前行驶,车辆12(动态障碍物)向左行驶,车辆11根据获取的车辆11信息(如速度、位置等)和车辆12信息(如速度、位置等)确定碰撞点13,车辆11检测到可能与车辆12发生碰撞,主动刹停车辆,等待车辆12通过后再启动。车辆11仅考虑自车行驶安全,主动避让车辆12,未考虑车辆12的处理方式。若车辆12也主动刹停车辆,车辆11需要根据刹停后的情况重新判断。车辆11避让车辆12(动态障碍物)的处理方式保守僵硬,不灵活,车辆11检测到动态障碍物就刹停,导致驾乘体验差。而且,车辆11紧急刹停,导致车辆刹停点距离碰撞点较远,车辆的通行效率较低,车辆抢行能力弱。
为了解决以上所述的技术问题,本申请实施例提供一种智能驾驶方法,该方法包括:检测到第一障碍物;根据该第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级;该第一风险等级用于表示在该第一信息和该第一行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第一风险等级匹配的第一巡航速度,该第一巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制车辆以第一巡航速度行驶。本申请实施例提供的方法,基于车辆与障碍物之间的风险确定匹配的巡航速度,控制车辆以该巡航速度,优化了智能驾驶策略,提高了车辆在自动驾驶巡航状态下的避让动态障碍物的灵活性,在保障车辆安全的同时,提升了驾乘体验,提高了车辆的行驶效率、通行效率和抢行能力,以便车辆更好的适应复杂的交通环境。
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行描述,显然,所描述的实施例仅仅是本申请一部分实施例,而不是全部的实施例。
本申请实施例中的智能驾驶方法可以应用于任意的避障场景。例如,避障场景包括机器人避障场景、车辆避障场景(如自动驾驶场景中车辆避障)、自主移动设备避障场景(如无人机避障等)等。
本申请实施例提供的技术方案可以应用于各种可移动智能设备中。示例性的,该可移动智能设备可以包括但不限于车辆、人工智能(artificial intelligence,AI)设备(如机器人)等。或者应用于具有控制前述可移动智能设备的功能的其他设备(比如:服务器、手机终端等)中。可移动智能设备或其他设备可以通过其包含的组件(包括硬件和软件),实现本申请实施例提供的智能驾驶方法。
以可移动智能设备为车辆为例,图2为本申请实施例提供的一种车辆200的结构示意图。
在本申请实施例中,车辆200可以包括各种子系统,例如包括但不限于智能驾驶系统210等。可选的,车辆200可以包括更多或更少的子系统,并且每个子系统可包括多个元件。另外,车辆200的每个子系统和元件可以通过有线通信技术或者无线通信技术建立连接。
智能驾驶系统210可以获取第一障碍物信息(如包括但不限于第一障碍物位置等各种信息)、车辆200自身的行驶信息(如包括但不限于行驶速度、加速度等各种车辆信息)。智能驾驶系统210还可以基于这些信息确定风险等级,获取与该风险等级匹配的巡航速度,控制车辆200以该巡航速度行驶。
可选的,该智能驾驶系统210可以包括但不限于高阶驾驶系统(advanced driver system,ADS)、高级驾驶辅助系统(advanced driver assistance system,ADAS)等中的一种或多种。或者,随着未来驾驶技术的演进,该智能驾驶系统210也可以为其他级别的驾驶系统。
上述车辆200可以为新能源车辆、电动车辆、智能车辆、轿车、卡车、摩托车、公共汽车、船、割草机、娱乐车、游乐场车辆、施工设备、电车、高尔夫球车、和火车等,本申请实施例不做特别的限定。
示例性的,上述仅以车辆举例说明本申请实施例中可移动智能设备的结构,但并不构成对可移动智能设备结构、形态的限制。
图3为本申请实施例提供的又一种可移动智能设备的结构示意图。例如,可移动智能设备为智能机器人。该可移动智能设备包括至少一个处理器301,通信线路302,存储器303以及至少一个通信接口304。
处理器301可以是一个通用中央处理器(central processing unit,CPU),微处理器,专用集成电路(application-specific integrated circuit,ASIC),或一个或多个用于控制本申请方案程序执行的集成电路。
通信线路302可包括通路或总线,在上述组件之间传送信息。
通信接口304,用于与其他设备通信。在本申请实施例中,通信接口304可以是模块、电路、总线、接口、收发器或者其它能实现通信功能的装置。可选的,当通信接口是收发器时,该收发器可以为独立设置的发送器,该发送器可用于向其他设备发送信息,该收发器也可以为独立设置的接收器,用于从其他设备接收信息。该收发器也可以是将发送、接收信息功能集成在一起的部件。
存储器303可以是只读存储器(read-only memory,ROM)、随机存取存储器(random access memory,RAM)或者能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质。存储器303可以是独立存在,通过通信线路302与处理器301相连接。存储器303也可以和处理器301集成在一起。
其中,存储器303用于存储用于实现本申请方案的计算机执行指令。处理器301用于执行存储器303中存储的计算机执行指令,从而实现本申请下述实施例提供的方法。
在本申请另一些实施例中,可移动智能设备可以包括比图2、图3所示的更多或更少的部件,或者组合某些部件,或者拆分某些部件,或者替换某些部件,或者不同的部件布置。图示的部件可以以硬件,软件或软件和硬件的组合实现。
以可移动智能设备为车辆,车辆处于自动驾驶场景为例,对本申请的智能驾驶方法进行说明。
参考图4,图4示出了本申请实施例提供的一种智能驾驶系统的系统架构示意图。智能驾驶系统包括感知系统、规划系统和控制系统。其中,感知系统、规划系统和控制系统之间通信连接。
在本申请实施例中,感知系统用于检测第一障碍物并获取第一障碍物的信息(如第一障碍物的当前位置、速度、加速度、形状等)。例如,第一障碍物包括动态障碍物,动态障碍物包括行走的行人、行驶的车辆等。感知系统还用于获取车辆的行驶信息。感知系统还用于感知周围环境信息,如道路信息、交通标志等。感知系统还用于将获取的各类信息发送至规划系统。
在本申请实施例中,规划系统和控制系统用于执行本申请提供的智能驾驶方法。其中,规划系统包括预测模块、决策模块和规划模块。
在一些实施例中,预测模块用于根据感知系统发送的各类信息进行预测,确定预测信息。具体的,预测模块用于根据第一障碍物的信息和车辆的行驶信息,确定预测信息;预测模块还用于将预测信息发送至决策模块。
其中,预测信息包括障碍物的预测信息(如障碍物的未来速度、未来位置、未来运动轨迹等)和车辆的预测信息(如车辆的未来速度、未来位置、未来运动轨迹等)。
例如,第一障碍物为动态障碍物,根据动态障碍物的信息(如位置、速度、加速度、运动方向、形状等)预测动态障碍物的未来运动轨迹,动态障碍物的未来位置、动态障碍物的未来速度等预测信息,根据车辆的行驶信息预测车辆的未来运动轨迹,车辆的未来位置、车辆的未来速度等预测信息。
在一些实施例中,决策模块用于根据预测模块发送的预测信息进行计算,确定风险信息。决策模块还用于综合各个风险信息评估横纵向风险,确定风险等级。决策模块还用于根据风险等级确定其对应的控制决策。决策模块还用于将风险等级发送给规划模块;或者,决策模块将风险信息、风险等级和控制决策发送至规划模块。
其中,风险信息包括但不限于碰撞点,碰撞时间,车辆到碰撞点的距离、以及第一障碍物到碰撞点的距离。
例如,第一障碍物为动态障碍物,碰撞点为动态障碍物的未来运动轨迹和车辆的未来运动轨迹的相交点。
可选的,风险信息还可以包括环境信息等。例如,天气、路面情况、交通环境等。应理解,车辆确定风险信息时,应综合考虑车辆所处的环境信息,辅助车辆做出更准确的风险判断,以便车辆采取更合适的巡航速度。
其中,横纵向风险包括横向风险和纵向风险。横向风险为车辆在行驶过程中所面临的横向方向上的风险,也就是车辆左右方向的风险,包括躲避障碍物或与障碍物碰撞的可能性。例如,其他车辆的变道、超车、并线等行为可能带来的碰撞风险,以及路边障碍物、行人、动物等不同来源的横向障碍物可能对车辆行驶造成的威胁。纵向风险为车辆在行驶过程中所面临的纵向方向上的风险,也就是车辆前后方向的风险,包括减速刹车或加速前进以避免碰撞的可能性。例如,与前车之间的跟车安全距离不足可能引发的追尾风险,以及前方突然刹车、减速或阻挡物等因素导致车辆需要紧急刹车或变道避让的情况。
其中,风险等级用于表示第一障碍物和车辆发生碰撞的可能性大小。例如,低风险、中风险和高风险等。
其中,控制决策包括让行决策或抢行决策。可以理解的是,让行决策就是车辆降速避让障碍物,抢行决策为车辆加速抢行通过。例如,车辆在面对与自车可能发生碰撞的动态障碍物时,采取的措施通常为降速让行障碍物,等碰撞风险解除后再行驶。或者,车辆在面对与自车可能发生碰撞的动态障碍物时,若确定碰撞风险较小,且自车更靠近碰撞点时,也可以选择加速抢行先通过障碍物。
具体的,决策模块根据预测信息进行计算、推演,得到风险信息。决策模块根据风险信息评估横向方向的风险和纵向方向的风险,结合横向方向和纵向方向的风险评估结果,确定风险等级。决策模块根据预定义的规则,确定风险等级对应的控制策略。
例如,低风险对应的控制策略为抢行策略,中风险和高风险对应的控制策略为让行策略。
在一些实施例中,规划模块用于根据决策模块发送的各类信息进行速度规划,生成速度规划结果(如巡航速度)。规划模块还用将速度规划结果(如巡航速度)发送至控制系统。
具体的,规划模块根据风险信息,确定该风险等级匹配的巡航速度;或者,规划模块根据控制策略、风险信息和风险等级,确定该风险等级匹配的巡航速度。规划模块还可以根据该巡航速度规划车辆的行驶速度,生成速度规划曲线。
其中,巡航速度可以理解为在车辆检测到存在动态障碍物且动态障碍物与自车有碰撞风险的场景时,根据当前车辆信息、障碍物信息和碰撞风险规划的避免车辆与障碍物碰撞的最低行驶速度。巡航速度可以理解成基于当前场景确定的车辆行驶速度应保持的最低临界值。车辆速度应该从当前速度平滑改变为巡航速度,并以巡航速度稳定向前行驶。可以理解的是,由于当前场景为避障场景,若确定控制策略为让行策略,则巡航速度为一个较小的数值,车辆根据该巡航速度指导车辆减速行驶,最终车辆可能缓慢向前行驶,根据实时检测的风险,动态改变巡航速度,直至障碍物先通过碰撞点,解除碰撞危机。若确定控制策略为抢行策略,则巡航速度为一个较大的数值,车辆需要根据该巡航速度指导车辆加速行驶,车辆提升速度快速通过碰撞点,解除碰撞危机。
在本申请实施例中,控制系统根据规划模块发送的速度规划结果(如巡航速度),控制车辆以该速度规划结果(如巡航速度)行驶,实现车辆的智能行驶,顺利躲避障碍物。
示例性的,参考图5,图5示出了本申请实施例提供的另一种智能驾驶系统50的架构示意图。如图5所示,智能驾驶系统50可以包括感知装置51和智能驾驶装置52。其中,感知装置51和智能驾驶装置52之间通信连接。
在本申请实施例中,感知装置51用于检测第一障碍物,感知装置51还用于感知第一障碍物的信息和车辆的行驶信息,感知装置51还用于将第一障碍物的信息和车辆的行驶信息发送至智能驾驶装置52。
可选的,感知装置51包括车辆传感器,车辆传感器通常设置于车辆内部。车辆传感器可以用于感知车辆周围第一障碍物的信息以及车辆的行驶信息。其中,车辆传感器可以是加速度传感器、陀螺仪传感器、轮速传感器、气压传感器、超声波传感器、摄像头传感器、定位传感器、雷达传感器等。
可以理解,由于车辆传感器设置于车辆中,因此,当感知装置51是车辆传感器时,感知装置51随车辆移动,则感知装置51和智能驾驶装置52之间可以通过无线通信网络进行通信。
应理解,以上对感知装置51的描述仅为示例性说明。
在本申请实施例中,智能驾驶装置52用于执行本申请提供的智能驾驶方法。具体的,智能驾驶装置52可以用于根据障碍物的信息和车辆的行驶信息确定当前风险等级,该风险等级用于表示当前障碍物和车辆发生碰撞的可能性大小;智能驾驶装置52还可以用于获取与当前风险等级匹配的巡航速度,该巡航速度为在当前风险等级下,车辆避免与障碍物发生碰撞的最低行驶速度;智能驾驶装置52还用于控制车辆以该巡航速度行驶。
可选的,智能驾驶装置52可以是智能驾驶计算平台,智能驾驶计算平台是实现智能驾驶、决策、规划、控制等功能的计算平台,是整个车辆的核心部件。该智能驾驶计算平台与车辆中各部件交互,获取各部件的实时数据,控制各部件工作。
可选的,智能驾驶装置52可以是服务器,服务器可以是Linux服务器、Windows服务器或其他可提供多设备同时接入的服务器设备,亦可以是多地域、多机房、多服务器所组成的服务器集群等。作为示例,该智能驾驶装置52可以是智能交通系统的服务器,例如物理服务或云服务器,本申请实施例对此不作限定。
可以理解的是,上述智能驾驶系统50中感知装置51和智能驾驶装置52是各自独立的部件,各部件之间交互完成智能驾驶。在实际应用中,智能驾驶系统50也可以仅包括智能驾驶装置52,其他部件由车辆中的其他系统控制,各系统交互完成智能驾驶。
应理解,以上对智能驾驶系统的描述仅为示例性说明。上述智能驾驶系统中的各个模块是按照功能逻辑进行划分,实际还可能为其他划分方式。此外,上述模块可以为其他名称。此外,各个模块均可以由硬件来实现,也可以由软件来实现,或者由硬件和软件的结合来实现,具体某个模块究竟以硬件、软件、或者软硬件结合的方式来实现,取决于技术方案的特定应用和设计约束条件。不同模块可以由不同的硬件实现,多个模块也可以由同一个硬件来实现。
可以理解的是,本申请描述的系统架构及业务场景是为了更加清楚的说明本申请的技术方案,并不构成对于本申请提供的技术方案的唯一限定,本领域普通技术人员可知,随着系统架构的演变和新业务场景的出现,本申请提供的技术方案对于类似的技术问题,同样适用。
示例性的,图6示出了本申请实施例提供的一种智能驾驶方法的流程示意图,该方法的执行主体可以诸如上文所述的可移动智能设备,或者也可以为可移动智能设备中的处理器,本申请实施例以执行主体为可移动智能设备,该可移动智能设备为车辆为例。如图6所示的,该方法包括以下步骤:
S601、车辆检测到第一障碍物,该第一障碍物为动态障碍物。
其中,车辆处于自动驾驶模式或智能驾驶模式,车辆处于巡航状态行驶。
可以理解的是,自动驾驶模式或智能驾驶模式用于通过车载传感系统(也可以描述为感知系统)感知道路环境,并根据感知所获得的道路、车辆信息和障碍物信息,控制车辆的转向和速度,从而使车辆能够安全、可靠地在道路上行驶并到达预定地点。
根据美国汽车工程师协会(society of automotive engineers,SAE)发布的标准,将自动驾驶分为多个等级,即L0至L5。具体的,L0:无自动化,该级别表示车辆完全由人类驾驶员全权操作,没有任何自动化系统的支持;L1:驾驶员协助:该级别表示车辆对方向盘和加减速其中一项操作提供驾驶支援,其他由驾驶员操作;L2:部分自动化:该级别表示车辆对方向盘和加减速中多项操作提供驾驶支援,其他由驾驶员操作(驾驶员需要保持对道路状况的观察和准备随时接管控制);L3:有条件自动化:该级别表示车辆可以完成大部分驾驶操作,在特定条件下实现自主驾驶,驾驶员需要保持注意力以备不时之需;L4:高度自动化:该级别表示车辆可以在大多数情况下完成所有驾驶操作,独立驾驶,驾驶员只在极少数情况下需要介入,如极端天气或复杂交通情况;L5:完全自动化:该级别表示车辆完全可以在任何道路和所有条件下自主驾驶,无需驾驶员介入。
在本申请实施例中,自动驾驶模式适用于L2-L4级。车辆可以辅助驾驶员,控制车辆加减速,避让动态障碍物。或者,车辆通过传感器、计算机视觉、人工智能等技术,完成大部分驾驶操作,避让障碍物。或者,车辆感知车周环境,独立完成所有驾驶操作,避让障碍物。
巡航状态是指车辆启动巡航控制模式后,车辆保持预定的速度向前行驶,无需驾驶员踩油门或刹车来维持车速。在该状态下,车辆会自动调整油门和刹车,以保持稳定的速度。
在一些实施例中,第一障碍物可以是车辆周围与车辆可能发生碰撞风险的动态障碍物。动态障碍物为保持运动状态的障碍物,如行驶的车辆、跑步的行人、行走的行人、奔跑的动物等。第一障碍物包括但不限于移动的物体、车辆、行人等。
在一些实施例中,车辆通过各类传感器检测到车辆周围的障碍物。具体的,车辆可以通过车辆上搭载的摄像头捕捉车辆周围的图像信息,通过图像处理和计算机视觉技术,识别道路上的障碍物。车辆还可以通过雷达系统发射无线电波并接收其反射波来探测周围环境中的物体,从而检测到障碍物。车辆还可以通过激光雷达发射激光束并根据激光束反射时间确定障碍物的位置和形状。车辆还可以通过超声波传感器检测障碍物。
在一些实施例中,车辆在检测到车辆周围的障碍物时,也会获取该障碍物的信息(如障碍物的位置、速度、方向、形状等信息)。车辆根据各障碍物的信息分析各障碍物与车辆是否存在碰撞的风险,以及障碍物是否为动态障碍物,将存在碰撞风险的动态障碍物确定为第一障碍物。
在一些实施例中,车辆会实时检测车辆的自车状态信息(如行驶信息等)。
在一些实施例中,车辆通过各类传感器获取车辆周围的障碍物的信息(如障碍物的位置、速度、方向、加速度、形状、大小等信息),以及车辆的行驶信息(车辆的位置、速度、加速度、方向等信息)。根据障碍物的信息和车辆的行驶信息预测障碍物未来的行为和车辆未来的行为,判断是否存在碰撞的风险。车辆根据障碍物的信息确定该障碍物是否为动态障碍物。
示例性的,根据障碍物的信息(如行人)确定为动态障碍物,车辆根据车辆的行驶信息预测车辆的未来运动轨迹,车辆根据障碍物的信息预测障碍物的未来运动轨迹,若车辆的未来运动轨迹和障碍物的未来运动轨迹于同一时间点相交,则确定该动态障碍物存在碰撞的风险,将该动态障碍物确定为第一障碍物。
例如,基于上文图1的示例,车辆11获取车辆12的信息,车辆12以一定的速度向左行驶,车辆11的未来运动轨迹与车辆12的未来运动轨迹存在于同一时间点相交的可能性,则确定车辆12为第一障碍物。
又如,如图7所示,车辆左转,车辆根据自车状态确定的未来运动轨迹为图7中的实线所示,车辆检测到左前方有行人行走,距离车辆越来越近,根据检测到的行人信息确定到行人的未来运动轨迹为图7中的虚线所示。行人的未来运动轨迹和车辆的未来运行轨迹相交,则车辆将确定行人为第一障碍物。
在另一些实施例中,车辆通过各类传感器获取车辆周围的障碍物的信息(如障碍物的位置、速度、方向、加速度、形状、大小等信息),以及车辆的行驶信息(车辆的位置、速度、加速度、方向等信息)。车辆根据各类信息判断当前场景是否为动障交互场景,若是,则将该场景中的动态障碍物确定为第一障碍物。
其中,动障交互场景可以理解为,在自动驾驶或智能驾驶场景中,车辆与动态障碍物交互博弈避障的场景。
具体的,车辆检测周围障碍物,先判断是否存在动态障碍物;若存在,则确定该障碍物是否存在与车辆碰撞的风险,若存在,则确定当前场景为动障交互场景,将该场景中的障碍物确定为第一障碍物。
例如,车辆向前直行,车辆检测到车辆左前方有行人向右跑步,行人的未来运动轨迹与车辆的未来运动轨迹存在相交的可能性,车辆与行人的距离不断缩减,则确定该场景为动障交互场景,将行人确定为第一障碍物。车辆与行人的交互情况包括以下几种:车辆检测到行人,降低行驶速度让行;或者,行人看到车辆,减速让行;或者,行人看到车辆,但两者距离较远,行人加速通过等。
在另一些实施例中,车辆还会实时检测车辆的周围环境信息(如道路信息、天气信息、信号灯信息等)。车辆根据周围环境信息、障碍物的信息和车辆的行驶信息筛选车辆周围的障碍物,确定第一障碍物。
可以理解的是,上述实施例以只有一个动态障碍物为例进行说明,在实际应用中,车辆周围可能存在多个动态障碍物,当存在多个动态障碍物时,将距离车辆最近,且存在碰撞风险的障碍物确定为第一障碍物。或者,车辆可以将每个动态障碍物都作为第一障碍物,综合多个第一障碍物信息进行速度规划。
可以理解的是,本申请实施例中对于车辆周围的障碍物进行筛选,将与车辆存在碰撞风险的障碍物确定为第一障碍物。如此,车辆可以更准确的识别潜在的碰撞威胁,以便后续评估该障碍物的风险等级,根据风险等级高效调整车辆的行驶速度,灵活控制车辆,减少不必要的制动或避让操作,提高车辆通行率和行驶效率,以实现智能驾驶,提高驾乘体验。
S602、车辆根据该第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级;该第一风险等级用于表示在该第一信息和该第一行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小。
其中,第一信息包括但不限于:第一障碍物当前的位置(如第一障碍物相对于车辆的坐标信息)、第一障碍物当前的距离(如第一障碍物与车辆之间的距离,以便评估第一障碍物和车辆的空间关系),第一障碍物当前的速度(如通过连续检测第一障碍物的位置变化,确定第一障碍物的运动速度)、第一障碍物当前的加速度、第一障碍物当前的方向、第一障碍物的形状、第一障碍物的大小、第一障碍物的属性(如车辆、行人、建筑物等)、第一障碍物当前的状态(如静止、移动、加速、减速等)等。
其中,第一行驶信息包括但不限于:车辆当前的位置(如车辆的坐标信息)、车辆当前的速度、车辆当前的加速度、车辆当前的方向等。
在一些实施例中,车辆根据第一障碍物的第一信息和车辆的第一行驶信息预测风险信息,再根据该风险信息评估横纵向风险,确定第一风险等级。
其中,风险信息包括但不限于:第一碰撞点、碰撞时间、当前车辆与第一碰撞点之间的距离,当前第一障碍物与第一碰撞点之间的距离等。示例性的,第一碰撞点为根据第一信息预测的第一障碍物的未来运动轨迹和根据第一行驶信息预测的车辆的未来运动轨迹的相交点。风险信息还可以包括障碍物的信息、车辆的信息和表示障碍物和车辆之间相对关系的信息,如障碍物的速度、车辆的速度,障碍物和车辆速度的速度差、障碍物与车辆之间的距离等。
在一些实施例中,车辆根据预定义的算法或模型处理第一障碍物的第一信息和车辆的第一行驶信息,确定风险信息。车辆再根据风险信息评估横向方向的方向(也可以描述为横向风险)和纵向方向的风险(也可以描述为纵向风险),根据横向风险和纵向风险,确定风险等级。
其中,横向风险为车辆左右方向的风险,包括躲避障碍物或与障碍物碰撞的可能性。纵向风险为车辆前后方向的风险,包括减速刹车或加速前进以避免碰撞的可能性。
其中,预定义的算法包括但不限于运动预测算法、轨迹生成算法、碰撞检测算法等。示例性的,车辆通过轨迹生成算法处理第一障碍物的第一信息确定第一障碍物的未来运动轨迹,车辆通过轨迹生成算法处理车辆的第一行驶信息确定车辆的未来运动轨迹,车辆通过碰撞检测算法处理第一障碍物的未来运动轨迹和车辆的未来运动轨迹,分析这两个轨迹的交叉情况,确定风险信息。
可以理解的是,预定义的算法还可以包括用于预测障碍物行为(如第一障碍物的未来运动轨迹)的机器学习算法,以及用于融合各传感器检测的数据,准确估计第一障碍物和车辆之间距离和速度信息的贝叶斯滤波算法等。
其中,预定义的模型可以是训练得到的机器学习模型或深度学习模型(神经网络模型)等。例如,该神经网络模型可以是车辆中预配置的神经网络模型,也可以是从开源平台或云服务获取的神经网络模型。该神经网络模型可以是卷积神经网络(convolutional neural network,CNN)等。
例如,预定义的模型为用于评估碰撞风险的CNN模型,CNN模型的训练过程包括:采集大量真实驾驶场景下的数据作为输入数据,输入数据包括但不限于车辆的行驶信息,障碍物的信息等。对采集到的输入数据进行特征提取等预处理操作,基于历史数据或经验,对预处理后的数据进行标注得到输入数据对应的标签(例如,为每组输入数据(障碍物信息和车辆信息)标注相应的风险信息等。选择合适的机器学习算法或深度学习模型来构建CNN模型。将标注后的数据集输入到CNN模型中,基于前向传播算法训练得到预测结果。根据预测结果和标签计算损失函数,并启动反向传播算法根据该损失函数调整CNN模型的模型参数。调整模型参数后的CNN模型重复执行上述的前向传播、计算损失、反向传播和参数调整的过程,直至CNN模型收敛或达到设定的迭代次数,结束训练,输出CNN模型。训练输出的CNN模型可应用于实际的预测、识别等任务中。
在实际应用中,车辆通过CNN模型确定风险信息。示例性的,将第一障碍物的第一信息和车辆的第一行驶信息输入CNN模型,由CNN模型输入风险信息。
可以理解的是,本申请实施例中的模型可以是基于大量的真实驾驶数据训练得到的,真实驾驶数据可以理解成驾驶员应对障碍物时的相关数据。以便车辆做出的决策更符合人类驾驶员的决策,提升车辆的智能性。
在一些实施例中,车辆根据第一障碍物的第一信息和车辆的第一行驶信息进行预测,得到预测信息,再根据预测信息进行短时推演和碰撞检测确定风险信息。
其中,预测信息可以理解成未来障碍物的行为、未来车辆的行为。预测信息包括但不限于障碍物的未来速度、未来位置、未来状态、未来加速度和未来运动轨迹,车辆的未来速度、未来位置、未来状态、未来加速度和未来运动轨迹等。
具体的,第一障碍物为动态障碍物,车辆根据第一障碍物的第一信息和预定义的算法或模型确定第一障碍物的预测信息,如第一障碍物的未来运动轨迹、第一障碍物的未来速度等。车辆根据车辆的第一行驶信息和预定义的算法或模型确定车辆的预测信息,如预测车辆的未来运动轨迹、车辆的未来速度等。车辆根据第一障碍物的未来运动轨迹和车辆的未来运动轨迹,确定第一碰撞点。根据第一碰撞点、第一信息和第一行驶信息确定碰撞时间、当前车辆与第一碰撞点之间的距离,当前第一障碍物与第一碰撞点之间的距离等风险信息。
其中,预定义的算法包括但不限于运动预测算法、轨迹生成算法。预定义的模型包括训练得到的机器学习模型等。
在得到预测信息之后,车辆根据预测信息进行短时推演,模拟未来一小段时间内车辆和障碍物的相对运动情况。车辆在短时推演的基础上进行碰撞检测,比较车辆和障碍物的速度、位置等信息,确定风险信息。
其中,碰撞检测包括但不限于几何碰撞检测、时间到达分析等。
在另一些实施例中,在车辆确定风险信息时,还可以参考车辆的周围环境信息进行判断。例如,根据道路曲率、实时车况、天气情况等适应性调整风险信息。
在车辆确定风险信息后,确定风险等级。在一些实施例中,车辆基于纵向风险评估算法或模型、横向风险评估算法或模型处理风险信息,确定纵向风险和横向风险。车辆综合分析纵向风险和横向风险,考虑不同风险对驾驶的综合影响,确定风险等级。
在另一些实施例中,车辆中预配置有风险信息和风险等级的映射表。该映射表中包括不同维度的风险信息,以及其对应的风险等级。车辆根据该映射表确定当前风险信息对应的风险等级。例如,通过对历史数据、模拟数据进行大数据分析和评估,确定不同维度的风险信息(如距离、速度、轨迹交叉等)与风险等级之间的关系,制定风险信息和风险等级的映射表,将映射表集成到车辆中。可以理解的是,映射表可以跟随技术发展和实际使用情况而不断更新,改进,以适应更多的交通环境。
示例性的,基于图1的示例,车辆11执行本申请的智能驾驶方法,车辆11行驶过程中,检测到车辆12,确定自身与车辆12有碰撞风险,将车辆12确定为第一障碍物。车辆11获取车辆12当前的速度、位置、方向、加速度等信息,车辆11还获取自身当前的位置、速度、加速度、方向等行驶信息。车辆11根据各类信息确定自身的未来运动轨迹、车辆12的未来运动轨迹、碰撞点、车辆11到碰撞点的时间大于车辆12到碰撞点的时间、车辆11到碰撞点的距离大于车辆12到碰撞点的距离、车辆11的速度小于车辆12的速度,车辆11和车辆12之间的距离较远等风险信息。基于上述信息可以确定,自车到碰撞点的时间小于车辆12到碰撞点的时间,两者碰撞的风险不高,所以确定风险等级为中风险。
示例性的,基于上述图7的示例,车辆执行本申请的智能驾驶方法,车辆行驶过程中检测到行人信息,并确定该行人与自车有碰撞风险。车辆根据自车信息和行人信息,确定行人速度小于车辆速度,行人到碰撞点的距离小于自车到碰撞点的距离。行人到碰撞点的时间小于自车到碰撞点的时间,但两者较为接近,且行人还在行走。基于上述信息可以确定若行人和车辆都以当前速度前进,两者碰撞的风险较大,所以风险等级为高风险。
可选的,在确定风险等级之后,车辆还可以根据预设规则确定风险等级对应的控制策略。其中,预设规则包括风险等级与控制策略的对应关系,控制策略包括让行策略或抢行策略。
应理解,车辆根据第一障碍物的信息和车辆的信息预测碰撞风险,根据碰撞风险确定相应的控制策略,基于控制策略和碰撞风险规划车辆速度,高效调整车辆的行驶速度、控制车辆采取更合适的方式避免碰撞,提升车辆的通行率,减少不必要的刹停,提升驾乘体验。
本申请中,车辆综合考虑障碍物信息和车辆信息,可以准确确定障碍物与车辆发生碰撞的风险等级,以便车辆根据风险等级做出更合理的决策,控制车辆的行驶速度,提高车辆的通行效率,提升驾乘体验以及车辆安全性。
S603、车辆获取与该第一风险等级匹配的第一巡航速度;该第一巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度。
在一些实施例中,车辆中预配置有风险等级和巡航速度的映射表。车辆可以根据该对应关系确定与第一风险等级匹配的第一巡航速度。
在另一些实施例中,车辆根据第一风险等级、第一信息和第一行驶信息,通过预定义的规则或模型计算出第一风险等级对应的第一巡航速度。
示例性的,预定义的规则可以是经逻辑推理、分析历史数据或经验数据确定的关于风险等级与巡航速度的一系列规则。预定义的模型可以是通过学习训练风险等级和巡航速度的历史数据等得到的机器学习模型或深度学习模型等。
可以理解的是,第一巡航速度应为一个有效数值。例如,第一巡航速度大于或等于0。
示例性的,基于上述步骤S602中图1及图7的示例,图1中车辆11与车辆12的碰撞风险不高,风险等级为中风险,根据预配置的映射表,确定车辆11的第一巡航速度为速度1。图7中车辆与行人的碰撞风险较大,风险等级为高风险,根据预配置的映射表,确定车辆的第一巡航速度为速度2。速度1应大于速度2,风险等级越高,表明车辆与障碍物碰撞的可能性越大,所以要尽快降低车辆的速度,让车辆以一个较低的速度巡航行驶。
S604、车辆控制该车辆以第一巡航速度行驶。
第一巡航速度可以理解为车辆根据当前风险等级规划的,避免车辆未来与动态障碍物碰撞的目标行驶速度,因此,车辆需要将当前的行驶速度调整为第一巡航速度。
控制该车辆以第一巡航速度行驶应理解为控制车辆的行驶速度从当前速度调整至第一巡航速度,再控制车辆以第一巡航速度稳定行驶的这一控制过程,包含速度调整过程以及速度维持过程。
在一些实施例中,车辆根据车辆的当前速度和第一巡航速度生成速度规划曲线,以便车辆根据该速度规划曲线将车辆的行驶速度调整至第一巡航速度。从而,使得车辆可以以第一巡航速度行驶。
具体的,在速度规划模型中引入第一巡航速度作为约束条件,确保速度规划曲线中车辆的行驶速度不会低于设定的第一巡航速度。在速度规划模型中将速度规划视为一个最优化问题,在满足各类约束条件的前提下,求解最佳的速度曲线。速度规划模型生成连接当前速度和第一巡航速度的平滑的速度规划曲线,使车辆速度从当前速度平稳的过渡到第一巡航速度。车辆根据速度规划曲线上的信息调整车辆的速度,逐步将其调整为第一巡航速度。
示例性的,如图8所示,基于上述步骤S603中的示例,图1中车辆11的速度规划曲线为速度规划曲线1。速度规划曲线1规划的速度为:车辆11从当前速度V1平滑下降至速度1,随后保持速度1继续行驶。图7中车辆的速度规划曲线为速度规划曲线2。速度规划曲线2规划的速度为:车辆从当前速度V2平滑下降至速度2,随后保持速度2继续行驶。图1中车辆11的风险等级为中风险,图7中车辆的风险等级为高风险,速度1大于速度2。图1中车辆11相较于图7中的车辆无需很快下降到速度1,所以车辆降速时间大于图7中车辆降速时间。
可以理解的是,上述示例中车辆的行驶速度平滑的调整至第一巡航速度,可以避免车辆行驶速度骤变引起车辆失控、舒适性下降等问题,保证车辆的安全性和驾乘体验。
在一些实施例中,在控制该车辆以第一巡航速度行驶之前,车辆还需判断能否采用第一巡航速度进行速度规划。
可以理解的是,车辆在巡航时需要与前方物体保持一定的安全距离,以便车辆操控。因此,使用巡航避障这一策略之前,还需要判断当前车辆距离碰撞点是否有充足的安全距离。
可选的,车辆确定当前车辆位置与第一碰撞点之间的距离是否大于第一阈值;若大于,则车辆采用第一巡航速度规划车辆的行驶速度,否则,不采用第一巡航速度规划车辆的行驶速度。
其中,第一阈值为该车辆与该第一障碍物之间的最小安全距离。第一阈值是基于动力学特性和安全考虑,预先计算的车辆与第一障碍物之间的最小安全距离。
可以理解的是,基于上述示例,如图1中的车辆11和车辆12,车辆11与车辆12朝碰撞点13行驶。车辆12可能会先通过碰撞点,但若车辆12经过碰撞点时,车辆12突然熄火。此时,车辆11与碰撞点13之间的距离小于第一阈值,距离过短,无法为车辆11提供充足的反应时间,以便车辆11应对该突发情况。即便车辆11紧急刹车,也可能发生碰撞危险。所以,在车辆进入巡航之前,需要确保车辆周围已留有足够的空间以便提供充足的反应时间,及时应对各种交通情况。
可选的,车辆判断车辆的未来运动轨迹上是否存在第二障碍物,该第二障碍物为静态障碍物,且当前车辆位置与第二障碍物之间的距离大于当前车辆位置与第一碰撞点之间的距离;若存在,则确定当前车辆位置与第二障碍物之间的距离是否大于第二阈值;若大于,则车辆采用第一巡航速度规划车辆的行驶速度;否则,不采用第一巡航速度规划车辆的行驶速度。
其中,第二阈值为该车辆与该第二障碍物之间的最小安全距离,第一阈值是基于动力学特性和安全考虑,预先计算的车辆与第二障碍物之间的最小安全距离。
如图9所示,车辆91向前行驶,车辆92向左行驶,车辆91执行本申请的智能驾驶方法,车辆91检测到车辆92为动态障碍物,根据车辆91的信息和车辆92的信息,确定风险等级1,并根据该风险等级1确定车辆91的第一巡航速度为速度3。但车辆91还检测到预测的车辆91的未来运动轨迹上,在碰撞点93之后还存在一个石头94,且石头94到车辆91的距离小于第二阈值。此时,车辆91不采用速度3规划车辆的行驶速度,主动刹停。可以理解的是,若车辆以速度3规划的规划曲线降速向前行驶,车辆92通过障碍物时,车辆在95处,虽然车辆还未到碰撞点,但此时车辆距离石头94很近,即便车辆主动刹停,也可能发生碰撞危险。所以不能采用巡航避障策略,应主动刹停,为车辆保留足够的空间以便提供充足的反应时间。
可以理解的是,上述示例中为了给巡航状态的车辆提供充足的反应时间,保障车辆安全,所以在进入巡航状态之前,判定周围环境能否使用该巡航状态。
可选的,车辆确定当前车辆位置与第一碰撞点之间的距离是否大于第一阈值;若大于,则继续判断车辆的未来运动轨迹上是否存在第二障碍物,第二障碍物为静态障碍物,且当前车辆位置与第二障碍物之间的距离大于当前车辆位置与第一碰撞点之间的距离;若存在,则确定当前车辆位置与第二障碍物之间的距离是否大于第二阈值;若大于,则车辆采用第一巡航速度规划车辆的行驶速度;否则,不采用第一巡航速度规划车辆的行驶速度。
可以理解的是,动态障碍物是可移动的,静态障碍物不可以移动,车辆在躲避障碍物的同时也应保证自身与障碍物留有足够的安全距离。第一阈值小于第二阈值。所以,在使用第一巡航速度规划车辆行驶速度之前,需要判定下当前车辆与第一碰撞点、车辆与静态障碍物之间是否留有足够的空间以便通过巡航避让障碍物。若空间不足,则不采用巡航的方式避让障碍物,可以采用紧急刹停的方式,为车辆保留更多的运动空间。例如,若车辆与第一障碍物都不减速,车辆与第一障碍物距离第一碰撞点越来越近,当车辆距离第一碰撞点的距离小于或等于第一阈值时,车辆与第一障碍物碰撞的风险非常高,即使车辆紧急刹停,但车辆无法在碰撞点前刹停,无法避免碰撞。所以,车辆应确保自身相对于碰撞点有足够的安全距离,以便车辆在碰撞点前刹停,保障一定的运动空间。
本申请中,基于安全距离确定车辆能否采用第一巡航速度规划车辆的行驶速度,可以更好的保障车辆安全。
可以理解的是,在确定车辆可以根据第一巡航速度规划车辆的行驶速度之后,车辆根据第一巡航速度和当前速度规划的速度规划曲线改变自车的行驶速度,车辆改变行驶速度的同时,还会实时采集第一障碍物的信息和车辆的信息,根据实时采集的信息确定其对应的风险等级,根据该风险等级适应性调整巡航速度。
可以理解的是,车辆在确定风险等级对应的巡航速度后,先根据该巡航速度进行速度规划,再确定能否采用该巡航速度(也即根据该巡航速度推演是否留有足够的安全距离,以便车辆及时响应),若可以采用,则根据第一巡航速度和当前速度确定的速度规划曲线调整车辆的行驶速度。
在一些实施例中,在车辆采用第一巡航速度和当前速度确定的速度规划曲线调整车辆的行驶速度的过程中,若车辆实时采集的第一障碍物信息和车辆信息对应的风险等级一直未改变,则不改变巡航速度,在车辆行驶速度调整为第一巡航速度后,车辆以第一巡航速度行驶。
在另一些实施例中,在车辆采用第一巡航速度和当前速度确定的速度规划曲线调整车辆的行驶速度的过程中,若车辆实时采集的第一障碍物信息和车辆信息对应的风险等级改变,则车辆根据改变后的风险等级确定其匹配的巡航速度,基于该巡航速度重新规划车辆的行驶速度,以便车辆根据重新规划后的巡航速度行驶。
示例性的,在车辆根据第一巡航速度规划车辆的行驶速度之后,车辆实时检测第一障碍物信息和车辆的信息,若根据该第一障碍物的第二信息和车辆的第二行驶信息,确定第二风险等级;该第二风险等级高于该第一风险等级,该第二风险等级用于表示在该第二信息和该第二行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;则获取与该第二风险等级匹配的第二巡航速度;该第二巡航速度低于该第一巡航速度,该第二巡航速度为在该第二风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制车辆以第二巡航速度行驶。
示例性的,在车辆根据第一巡航速度规划车辆的行驶速度之后,车辆实时检测第一障碍物信息和车辆的信息,若根据该第一障碍物的第三信息和车辆的第三行驶信息,确定第三风险等级;该第三风险等级低于该第一风险等级,该第三风险等级用于表示在该第三信息和该第三行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;则获取与该第三风险等级匹配的第三巡航速度;该第三巡航速度高于该第一巡航速度,该第三巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;控制车辆以第三巡航速度行驶。
上文的第一信息可以理解为第一时刻采集的障碍物信息,第二信息/第三信息可以理解为第N时刻采集的障碍物信息。N为整数,且N大于1。第一行驶信息可以理解为第一时刻采集的车辆行驶信息,第二行驶信息/第三行驶信息可以理解为第N时刻采集的车辆行驶信息。第二风险等级为根据第二信息和第二行驶信息确定的第一障碍物和车辆的碰撞风险,第三风险等级为根据第三信息和第三行驶信息确定的第一障碍物和车辆的碰撞风险。第二风险等级、第二风险等级均不同于第一风险等级。
第二风险等级高于第一风险等级可以理解成第N时刻第一障碍物和车辆之间的碰撞风险相较于第一时刻的碰撞风险增高了(如第一障碍物不减速,第一障碍物抢行),所以,车辆应降低巡航速度,避让第一障碍物,让第一障碍物先通过碰撞点。因此,第N时刻对应的第二巡航速度低于第一时刻的第一巡航速度,车辆采用新确定的第二巡航速度进行速度规划,以便车辆根据重新确定的第二巡航速度行驶。
第三风险等级高于第一风险等级可以理解成第N时刻第一障碍物和车辆之间的碰撞风险相较于第一时刻的碰撞风险降低了(如第一障碍物减速,第一障碍物让行),所以,车辆应可以提高巡航速度,先通过碰撞点。因此,第N时刻对应的第三巡航速度低于第一时刻的第一巡航速度,车辆采用新确定的第三巡航速度进行速度规划,以便车辆根据重新确定的第三巡航速度行驶。
示例性的,基于上述图1的示例,在根据速度1规划车辆11速度后,车辆继续检测车辆11和车辆12的信息,若检测车辆12减速,车辆12的速度小于车辆11的速度,车辆12到碰撞点的时间与车辆11到碰撞点的时间相差不大,则确定两者碰撞风险增高,此时的风险等级为高风险,车辆11应降低巡航速度,如速度4,基于速度4规划车辆速度,以便车辆根据速度4行驶。
示例性的,基于上述图7的示例,在车辆根据速度2规划速度后,车辆继续检测车辆和行人的信息,若检测到行人停止或者后退,则可以确定车辆与行人碰撞风险降低,此时风险等级为低风险,车辆可以提高巡航速度,如速度5,基于速度5规划车辆速度,以便车辆根据速度5行驶。
可以理解的是,控制车辆以第二巡航速度行驶、控制车辆以第三巡航速度行驶的具体实现方式参见上文所述,此处不再赘述。
可以理解的是,车辆根据第二巡航速度规划该车辆的行驶速度,以及根据该第三巡航速度规划该车辆的行驶速度的前提条件应为:第N时刻车辆位置与第一碰撞点之间的距离大于第一阈值;或者,第N时刻车辆位置与第一碰撞点之间的距离大于第一阈值,且车辆未来运动轨迹上的存在静态的第二障碍物,第N时刻车辆位置与第二障碍物之间的距离大于第N时刻车辆位置与第一碰撞点之间的距离,第N时刻车辆位置与第二障碍物之间的距离也大于第二阈值。
可以理解的是,上述实施例中,都以车辆根据巡航速度成功规划车辆行驶速度为例进行说明。而实际应用中还存在主动退出巡航的策略。下面,详细介绍下车辆主动退出巡航的具体实现方式。
在另一些实施例中,在车辆根据该第一巡航速度规划该车辆的行驶速度之后,车辆实时检测障碍物和车辆的信息,根据该第一障碍物的第四信息和车辆的第四行驶信息,确定第四风险等级;该第四风险等级高于该第一风险等级,该第四风险等级用于表示在该第四信息和该第四行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第四风险等级匹配的第四巡航速度;该第四巡航速度低于该第一巡航速度,该第四巡航速度为在该第四风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;确定当前该车辆位置与第一碰撞点之间的距离小于或等于第一阈值;该第一碰撞点为根据该第一信息和该第一行驶信息预测的障碍物未来运动轨迹和车辆未来运动轨迹的相交点,该第一阈值为该车辆与该第一障碍物之间的最小安全距离;将该车辆的行驶速度归零。
示例性的,基于上述示例图1的示例,在根据速度1规划车辆11速度后,车辆继续检测车辆11和车辆12的信息,若检测车辆12减速,车辆12的速度小于车辆11的速度,车辆12到碰撞点的时间与车辆11到碰撞点的时间相差不大,则可以两者碰撞风险增高,此时的风险等级为高风险,车辆11应降低巡航速度,如速度4,基于速度4规划车辆速度。但检测到当前车辆位置到碰撞点之间的距离小于第一阈值,车辆无法及时响应意外情况,所以车辆应退出巡航策略,将行驶速度逐步降为零。
在另一些实施例中,若在该车辆的未来运动轨迹上还存在第二障碍物,第二障碍物为静态障碍物,当前车辆位置与第二障碍物之间的距离大于当前车辆位置与第一碰撞点之间的距离,在根据第一巡航速度规划该车辆的行驶速度之后,车辆实时检测障碍物和车辆的信息,根据该第一障碍物的第五信息和车辆的第五行驶信息,确定第五风险等级;该第五风险等级高于该第一风险等级,该第五风险等级用于表示在该第五信息和该第五行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小;获取与该第五风险等级匹配的第五巡航速度;该第五巡航速度低于该第一巡航速度,该第五巡航速度为在该第五风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度;确定当前该车辆位置与该第二障碍物之间的距离小于或等于第二阈值;该第二阈值为该车辆与该第二障碍物之间的最小安全距离;将该车辆的行驶速度归零。
可以理解的是,将该车辆的行驶速度归零是一个逐步变化过程,车辆从当前车速减速,直至变为零。如此,可以保证用户的舒适性,保障车辆的稳定性。
基于上文第二信息、第三信息、第二行驶信息、第三行驶信息的描述,第四信息/第五信息可以理解为第N时刻采集的障碍物信息,第四行驶信息/第五行驶信息可以理解为第N时刻采集的车辆行驶信息。第四风险等级为根据第四信息和第四行驶信息确定的第一障碍物和车辆的碰撞风险,第五风险等级为根据第五信息和第五行驶信息确定的第一障碍物和车辆的碰撞风险。
第四风险等级高于第一风险等级,车辆本应降低巡航速度,避让第一障碍物,让第一障碍物先通过碰撞点。但车辆还检测到第N时刻车辆位置与第一碰撞点之间的距离小于或等于第一阈值,也即,第N时刻车辆位置与第一碰撞点之间的距离小于或等于最小安全距离,继续采用巡航策略无法保障安全,所以应平滑退出巡航策略,从车辆第N时刻的速度平稳过渡到0,以便车辆在碰撞点前停刹,留有足够的空间。
第五风险等级低于第一风险等级,车辆本应提高巡航速度,先通过障碍点。但车辆检测到车辆的未来运动轨迹上的还存在静态的第二障碍物,第N时刻车辆位置与第二障碍物之间的距离大于第N时刻车辆位置与第一碰撞点之间的距离,第N时刻车辆位置与第二障碍物之间的距离小于或等于第二阈值。这种情况下,第N时刻车辆位置与第二障碍物之间小于或等于最小安全距离,即便采用巡航策略车辆优先通过碰撞点,再减速刹停,但车辆与第二障碍物之间的距离太近,限制车辆的行驶,无法保障安全,所以应平滑退出巡航策略,从车辆第N时刻的速度平稳过渡到0,以便车辆在碰撞点前停刹,留有足够的空间,以便车辆操控。
可以理解的是,在车辆采用巡航策略后,车辆仍向前行驶,车辆与碰撞点之间的距离越来越近,车辆与障碍物之间的距离也越来越近,当交通情况不满足巡航状态应保持的安全距离时,主动退出巡航状态,保障车辆安全行驶,以便车辆及时应对各种交通情况。
可以理解的是,本申请的应用场景为自动驾驶场景,车辆自主感知周围环境,做出相应的决策,车辆做出的决策应接近人类驾驶员的行为和决策。
本申请中,在车辆根据当前风险等级对应的巡航速度避让障碍物时,根据实时的距离信息判断是否满足安全距离,当实时距离小于安全距离时,平稳退出巡航,保障车辆的安全性,以及用户的舒适性。
可以理解的是,现有技术中,车辆面对动态障碍物时都主动刹停,避让动态障碍物。这样,车辆每遇到一个动态障碍物就会刹停,导致车辆行驶效率、和通行效率低,车辆不智能,影响用户体验。
本申请的技术方案中,车辆面对动态障碍物时,采集动态障碍物和车辆的实时信息,确定动态障碍物和车辆之间的碰撞风险,基于碰撞风险,确定车辆的最低巡航速度,基于该最低巡航速度规划车辆的行驶速度,让车辆以最低巡航速度向碰撞点行驶。车辆实时检测动态障碍物的行为,进行博弈。若动态障碍物不避让,则车辆可以以一个较低的速度行驶,在距离碰撞点较近的位置刹停,等动态障碍物通过碰撞点后再行驶。或者,动态障碍物不避让,则车辆可以以一个更低的速度行驶,动态障碍物通过碰撞点时,车辆与碰撞点还有一定距离,动态障碍物通过碰撞点后,车辆可加速行驶通过碰撞点。或者,若动态障碍物避让,则可以根据风险等级逐步提升速度,优先通过碰撞点。
示例性的,如图10所示,车辆1001检测到车辆1002将横穿自己的未来行驶路径,两车的碰撞点为1003,现有技术中车辆1001主动刹停,刹停点为1004,本申请的技术方案中车辆1001采用巡航策略,以一个较低的速度巡航行驶,最终的刹停点为1005,距离碰撞点更近。
可以理解的是,本申请的技术方案在车辆面对动态障碍物时,根据实时的碰撞风险,降低或提高车辆的巡航速度,基于该巡航速度与动态障碍物博弈,灵活控制车辆,避免不必要的刹停或制动操作,使得车辆输出类人的速度规划策略,增加车辆的博弈能力和类人性,提升车辆的通行效率和行驶效率,提高驾乘体验。
可选的,第一障碍物也可以是静态障碍物,如不能移动的物体,建筑物等。车辆检测到第一障碍物之后,可以进行风险评估,根据最低巡航速度缓慢接近障碍物直至刹停。或者,车辆检测到第一障碍物后,缓慢退出巡航状态。增加了车辆巡航行驶时,对静态障碍物的处理方式。
本申请中,通过上述方式,车辆根据车辆与障碍物之间的风险确定匹配的巡航速度,基于巡航速度规划车辆行驶速度,控制车辆以巡航速度行驶。在面对动态障碍物时,车辆不直接刹停,而是以一个较低的速度缓慢行驶,与动态障碍物博弈,根据实时风险灵活控制车辆,优化了智能驾驶策略,提高了车辆在自动驾驶巡航状态下避让动态障碍物的灵活性,在保障车辆安全的同时,提高车辆的通行效率和抢行能力,提升驾乘体验,以便车辆更好的适应复杂的交通环境。
上述主要从方法的角度对本申请实施例提供的方案进行了介绍。为了实现上述功能,其包含了执行各个功能相应的硬件结构和/或软件模块。本领域技术人员应该很容易意识到,结合本文中所公开的实施例描述的各示例的单元及算法步骤,本申请能够以硬件或硬件和计算机软件的结合形式来实现。某个功能究竟以硬件还是计算机软件驱动硬件的方式来执行,取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本申请的范围。
如图11所示,为本申请实施例提供的另一种智能驾驶装置的结构示意图。该智能驾驶装置1100包括检测模块1101和处理模块1102。该智能驾驶装置1100用于执行上述的智能驾驶方法,例如用于执行图6-图10所示的智能驾驶方法。当然,智能驾驶装置1100还可以包括其他模块,或者智能驾驶装置1100可以包括更少的模块。本申请实施例对智能驾驶装置的具体形态以及实现方式不作具体限定。
检测模块1101,用于检测到第一障碍物;该第一障碍物为动态障碍物。
处理模块1102,用于根据该第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级;该第一风险等级用于表示在该第一信息和该第一行驶信息下,该第一障碍物和该车辆发生碰撞的可能性大小。
处理模块1102,还用于获取与该第一风险等级匹配的第一巡航速度,该第一巡航速度为在该第一风险等级下,该车辆避免与该第一障碍物发生碰撞的最低行驶速度。
处理模块1102,还用于控制车辆以第一巡航速度行驶。
智能驾驶装置1100中的各个模块的操作和/或功能分别为了实现上述方法实施例中所述的智能驾驶方法的相应流程,上述方法实施例涉及的各步骤的所有相关内容均可以援引到对应功能单元的功能描述,为了简洁,在此不再赘述。
可选的,图11所示的智能驾驶装置1100还可以包括存储模块(图11中未示出),该存储模块中存储有程序或指令。当检测模块1101和处理模块1102执行该程序或指令时,使得图11所示的智能驾驶装置1100可以执行上述方法实施例中所述的智能驾驶方法。
图11所示的智能驾驶装置1100的技术效果可以参考上述方法实施例中所述的智能驾驶方法的技术效果,此处不再赘述。
本申请实施例还提供一种芯片系统,如图12所示,该芯片系统1200包括至少一个处理器1201和至少一个接口电路1202。作为示例,当该芯片系统1200包括一个处理器和一个接口电路时,则该一个处理器可以是图12中实线框所示的处理器1201(或者是虚线框所示的处理器1201),该一个接口电路可以是图12中实线框所示的接口电路1202(或者是虚线框所示的接口电路1202)。当该芯片系统1200包括两个处理器和两个接口电路时,则该两个处理器包括图12中实线框所示的处理器1201和虚线框所示的处理器1201,该两个接口电路包括图12中实线框所示的接口电路1202和虚线框所示的接口电路1202。对此不作限定。
处理器1201和接口电路1202可通过线路互联。例如,接口电路1202可用于接收信号。又例如,接口电路1202可用于向其它装置(例如处理器1201)发送信号。示例性的,接口电路1202可读取存储器中存储的指令,并将该指令发送给处理器1201。当所述指令被处理器1201执行时,可使得智能驾驶装置执行上述实施例中的各个步骤。当然,该芯片系统还可以包含其他分立器件,本申请实施例对此不作具体限定。
示例性地,该芯片系统可以是现场可编程门阵列(field programmable gate array,FPGA),可以是专用集成芯片(application-specific integrated circuit,ASIC),系统级芯片(system on a chip,SoC),中央处理器(central processing unit,CPU),网络处理器(network processor,NP),数字信号处理器(digital signal processor,DSP),微控制器单元(microcontroller unit,MCU),可编程逻辑器件(programmable logic device,PLD)或其他集成芯片。
应理解,上述方法实施例中的各步骤可以通过处理器中的硬件的集成逻辑电路或者软件形式的指令完成。结合本申请实施例所公开的方法步骤可以直接体现为硬件处理器执行完成,或者用处理器中的硬件及软件模块组合执行完成。
本申请实施例还提供一种存储一个或多个计算机程序的计算机可读存储介质,所述一个或多个计算机程序包括指令,所述指令当被计算机执行时使计算机执行上述实施例中智能驾驶方法的相应流程。
在一些实施例中,所公开的方法可以实施为以机器可读格式被编码在计算机可读存储介质上的或者被编码在其它非瞬时性介质或者制品上的计算机程序指令。
本申请实施例还提供一种计算机程序产品,当该计算机程序产品在计算机上运行时,使得计算机执行上述相关步骤,以实现上述实施例中智能驾驶方法。
其中,本申请实施例提供的装置、计算机可读存储介质、计算机程序产品或芯片均用于执行上文所提供的对应的方法。因此,其所能达到的有益效果可参考上文所提供的对应的方法中的有益效果,此处不再赘述。
以上所述,仅为本申请的具体实施方式,但本申请的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本申请揭露的技术范围内,可轻易想到变化或替换,都应涵盖在本申请的保护范围之内。因此,本申请的保护范围应以所述权利要求的保护范围为准。
Claims (14)
- 一种智能驾驶方法,其特征在于,包括:检测到第一障碍物;所述第一障碍物为动态障碍物;根据所述第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级;所述第一风险等级用于表示在所述第一信息和所述第一行驶信息下,所述第一障碍物和所述车辆发生碰撞的可能性大小;获取与所述第一风险等级匹配的第一巡航速度,所述第一巡航速度为在所述第一风险等级下,所述车辆避免与所述第一障碍物发生碰撞的最低行驶速度;控制所述车辆以所述第一巡航速度行驶。
- 根据权利要求1所述的方法,其特征在于,所述方法还包括:根据所述第一障碍物的第二信息和车辆的第二行驶信息,确定第二风险等级;所述第二风险等级高于所述第一风险等级,所述第二风险等级用于表示在所述第二信息和所述第二行驶信息下,所述第一障碍物和所述车辆发生碰撞的可能性大小;获取与所述第二风险等级匹配的第二巡航速度;所述第二巡航速度低于所述第一巡航速度,所述第二巡航速度为在所述第二风险等级下,所述车辆避免与所述第一障碍物发生碰撞的最低行驶速度;控制所述车辆以所述第二巡航速度行驶。
- 根据权利要求1所述的方法,其特征在于,所述方法还包括:根据所述第一障碍物的第三信息和车辆的第三行驶信息,确定第三风险等级;所述第三风险等级低于所述第一风险等级,所述第三风险等级用于表示在所述第三信息和所述第三行驶信息下,所述第一障碍物和所述车辆发生碰撞的可能性大小;获取与所述第三风险等级匹配的第三巡航速度;所述第三巡航速度高于所述第一巡航速度,所述第三巡航速度为在所述第一风险等级下,所述车辆避免与所述第一障碍物发生碰撞的最低行驶速度;控制所述车辆以所述第三巡航速度行驶。
- 根据权利要求1至3中任一项所述的方法,其特征在于,所述根据所述第一障碍物的第一信息和车辆的第一行驶信息,确定第一风险等级,包括:根据所述第一障碍物的第一信息和车辆的第一行驶信息预测风险信息;所述风险信息包括以下一项或多项:第一碰撞点,碰撞时间,当前所述车辆位置与所述第一碰撞点之间的距离,当前所述第一障碍物与所述第一碰撞点之间的距离;所述第一碰撞点为根据所述第一信息和所述第一行驶信息预测的所述第一障碍物的未来运动轨迹和所述车辆的未来运动轨迹的相交点;根据所述风险信息评估横纵向风险,确定所述第一风险等级。
- 根据权利要求1至4中任一项所述的方法,其特征在于,所述控制所述车辆以所述第一巡航速度行驶,包括:根据所述车辆的当前速度和所述第一巡航速度生成速度规划曲线;根据所述速度规划曲线将所述车辆的行驶速度调整至所述第一巡航速度;控制所述车辆以所述第一巡航速度行驶。
- 根据权利要求1至5中任一项所述的方法,其特征在于,所述控制所述车辆以所述第一巡航速度行驶,包括:若当前所述车辆位置与所述第一碰撞点之间的距离大于第一阈值,控制所述车辆以所述第一巡航速度行驶;所述第一碰撞点为根据所述第一信息和所述第一行驶信息预测的第一障碍物的未来运动轨迹和车辆的未来运动轨迹的相交点,所述第一阈值为所述车辆与所述第一障碍物之间的最小安全距离。
- 根据权利要求6所述的方法,其特征在于,在所述车辆的未来运动轨迹上存在第二障碍物,所述第二障碍物为静态障碍物,当前所述车辆位置与所述第二障碍物之间的距离大于当前所述车辆位置与所述第一碰撞点之间的距离,所述控制所述车辆以所述第一巡航速度行驶,包括:若当前所述车辆位置与所述第二障碍物之间的距离大于第二阈值,控制所述车辆以所述第一巡航速度行驶;所述第二阈值为所述车辆与所述第二障碍物之间的最小安全距离。
- 根据权利要求1至7中任一项所述的方法,其特征在于,所述方法还包括:根据所述第一障碍物的第四信息和车辆的第四行驶信息,确定第四风险等级;所述第四风险等级高于所述第一风险等级,所述第四风险等级用于表示在所述第四信息和所述第四行驶信息下,所述第一障碍物和所述车辆发生碰撞的可能性大小;获取与所述第四风险等级匹配的第四巡航速度;所述第四巡航速度低于所述第一巡航速度,所述第四巡航速度为在所述第四风险等级下,所述车辆避免与所述第一障碍物发生碰撞的最低行驶速度;若当前所述车辆位置与第一碰撞点之间的距离小于或等于第一阈值,将所述车辆的行驶速度归零;所述第一碰撞点为根据所述第一信息和所述第一行驶信息预测的所述第一障碍物的未来运动轨迹和所述车辆的未来运动轨迹的相交点,所述第一阈值为所述车辆与所述第一障碍物之间的最小安全距离。
- 根据权利要求1至8中任一项所述的方法,其特征在于,在所述车辆的未来运动轨迹上存在第二障碍物,所述第二障碍物为静态障碍物,当前所述车辆位置与所述第二障碍物之间的距离大于当前所述车辆位置与所述第一碰撞点之间的距离,所述方法还包括:根据所述第一障碍物的第五信息和车辆的第五行驶信息,确定第五风险等级;所述第五风险等级高于所述第一风险等级,所述第五风险等级用于表示在所述第五信息和所述第五行驶信息下,所述第一障碍物和所述车辆发生碰撞的可能性大小;获取与所述第五风险等级匹配的第五巡航速度;所述第五巡航速度低于所述第一巡航速度,所述第五巡航速度为在所述第五风险等级下,所述车辆避免与所述第一障碍物发生碰撞的最低行驶速度;若当前所述车辆位置与所述第二障碍物之间的距离小于或等于第二阈值,将所述车辆的行驶速度归零;所述第二阈值为所述车辆与所述第二障碍物之间的最小安全距离。
- 一种智能驾驶装置,其特征在于,包括:处理器和存储器,所述存储器与所述处理器耦合,所述存储器用于存储计算机可读指令,当所述处理器从所述存储器中读取所述计算机可读指令,使得所述智能驾驶装置执行如权利要求1-9中任意一项所述的方法。
- 一种车辆,其特征在于,所述车辆包括如权利要求10所述的智能驾驶装置。
- 一种芯片系统,其特征在于,包括至少一个处理器和至少一个接口电路,所述至少一个接口电路用于执行收发功能,所述至少一个处理器用于执行如权利要求1-9中任意一项所述的方法。
- 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质包括计算机程序,当所述计算机程序被处理器执行时实现如权利要求1-9中任意一项所述的方法。
- 一种计算机程序产品,其特征在于,所述计算机程序产品包括计算机程序,当所述计算机程序被处理器执行时实现如权利要求1-9中任意一项所述的方法。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410366424.3 | 2024-03-27 | ||
| CN202410366424.3A CN120756514A (zh) | 2024-03-27 | 2024-03-27 | 智能驾驶方法、装置及车辆 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2025200580A1 true WO2025200580A1 (zh) | 2025-10-02 |
Family
ID=97217998
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2024/138104 Pending WO2025200580A1 (zh) | 2024-03-27 | 2024-12-10 | 智能驾驶方法、装置及车辆 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN120756514A (zh) |
| WO (1) | WO2025200580A1 (zh) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN121375854A (zh) * | 2025-12-24 | 2026-01-23 | 成都通广网联科技有限公司 | 一种基于行为预测模型的远程接管风险预警系统及方法 |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109774714A (zh) * | 2018-12-29 | 2019-05-21 | 百度在线网络技术(北京)有限公司 | 用于自动驾驶车辆的控制方法及装置 |
| KR20210083462A (ko) * | 2019-12-26 | 2021-07-07 | 주식회사 만도 | 운전자 보조 시스템, 그를 가지는 차량 및 그 제어 방법 |
| CN114248794A (zh) * | 2020-09-23 | 2022-03-29 | 华为技术有限公司 | 车辆的控制方法、装置及车辆 |
| CN115273512A (zh) * | 2022-07-22 | 2022-11-01 | 重庆长安汽车股份有限公司 | 自动驾驶车辆的防碰撞辅助方法、装置、设备及介质 |
| CN117141472A (zh) * | 2023-10-26 | 2023-12-01 | 集美大学 | 一种车辆行驶主动避障控制系统及方法 |
| CN118722601A (zh) * | 2024-08-07 | 2024-10-01 | 重庆长安汽车股份有限公司 | 一种自动泊车避撞方法、装置、介质及车辆 |
-
2024
- 2024-03-27 CN CN202410366424.3A patent/CN120756514A/zh active Pending
- 2024-12-10 WO PCT/CN2024/138104 patent/WO2025200580A1/zh active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109774714A (zh) * | 2018-12-29 | 2019-05-21 | 百度在线网络技术(北京)有限公司 | 用于自动驾驶车辆的控制方法及装置 |
| KR20210083462A (ko) * | 2019-12-26 | 2021-07-07 | 주식회사 만도 | 운전자 보조 시스템, 그를 가지는 차량 및 그 제어 방법 |
| CN114248794A (zh) * | 2020-09-23 | 2022-03-29 | 华为技术有限公司 | 车辆的控制方法、装置及车辆 |
| CN115273512A (zh) * | 2022-07-22 | 2022-11-01 | 重庆长安汽车股份有限公司 | 自动驾驶车辆的防碰撞辅助方法、装置、设备及介质 |
| CN117141472A (zh) * | 2023-10-26 | 2023-12-01 | 集美大学 | 一种车辆行驶主动避障控制系统及方法 |
| CN118722601A (zh) * | 2024-08-07 | 2024-10-01 | 重庆长安汽车股份有限公司 | 一种自动泊车避撞方法、装置、介质及车辆 |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN121375854A (zh) * | 2025-12-24 | 2026-01-23 | 成都通广网联科技有限公司 | 一种基于行为预测模型的远程接管风险预警系统及方法 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN120756514A (zh) | 2025-10-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113460042B (zh) | 车辆驾驶行为的识别方法以及识别装置 | |
| CN113160547B (zh) | 一种自动驾驶方法及相关设备 | |
| CN113460081B (zh) | 车辆控制装置、车辆控制方法及存储介质 | |
| CN110362077B (zh) | 无人驾驶车辆紧急避险决策系统、方法及介质 | |
| CN112230642B (zh) | 道路可行驶区域推理方法及装置 | |
| US11801864B1 (en) | Cost-based action determination | |
| CN113525413B (zh) | 车辆控制装置、车辆控制方法及存储介质 | |
| CN113460079B (zh) | 车辆控制装置、车辆控制方法及存储介质 | |
| CN110379193A (zh) | 自动驾驶车辆的行为规划方法及行为规划装置 | |
| US12539884B2 (en) | Implementing slowdown manoeuvres in autonomous vehicles | |
| WO2022016901A1 (zh) | 一种规划车辆行驶路线的方法以及智能汽车 | |
| US12049236B2 (en) | Complementary control system detecting imminent collision of autonomous vehicle in fallback monitoring region | |
| CN114248794A (zh) | 车辆的控制方法、装置及车辆 | |
| CN114261404B (zh) | 一种自动驾驶方法及相关装置 | |
| US12409865B2 (en) | Autonomous driving method and apparatus | |
| CN113631452A (zh) | 一种变道区域获取方法以及装置 | |
| CN113963535B (zh) | 行驶决策确定方法、装置、电子设备存储介质 | |
| CN113460083B (zh) | 车辆控制装置、车辆控制方法及存储介质 | |
| US20240092350A1 (en) | Vehicle safety system | |
| US12552410B2 (en) | Path generation based on predicted actions | |
| US20230053243A1 (en) | Hybrid Performance Critic for Planning Module's Parameter Tuning in Autonomous Driving Vehicles | |
| JP2021144276A (ja) | 移動体制御装置、移動体制御方法、及びプログラム | |
| KR20250044689A (ko) | 예측된 액션들에 기초한 경로 생성 | |
| WO2025200580A1 (zh) | 智能驾驶方法、装置及车辆 | |
| CN114932901A (zh) | 一种自适应速度规划方法、装置和域控制器 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 24931515 Country of ref document: EP Kind code of ref document: A1 |