A real-time risk monitoring method for automatic driving vehicle safe lane changing

Abstract

This invention relates to a real-time risk monitoring method for safe lane changing in autonomous vehicles, belonging to the field of intelligent connected vehicle technology. Addressing the problems of coupling between lane changing safety monitoring modules and decision-making systems, insufficient intervention during lane changing, and reliance on high-precision maps in existing technologies, this invention proposes an independent three-layer safety architecture (lane changing function layer, lane changing function monitoring layer, and basic safety monitoring layer). Through lightweight maps or mapless positioning systems, it integrates real-time weather, road conditions, and traffic congestion information to dynamically calculate lane changing safety boundaries, assess lane changing risks in real time, and trigger intervention. The technical solution includes core modules such as road congestion estimation, moving target behavior prediction, maximum lane changing time estimation, safe distance calculation, lane changing safety assessment, and safety boundary control. The advantages of this invention are improved lane changing safety independence, reduced reliance on high-precision maps, and the ability to achieve real-time risk monitoring and emergency intervention in all weather conditions, effectively reducing the risk of accidental lane changes and traffic accidents.

Description

Technical Field

[0001] This invention belongs to the field of intelligent connected vehicle technology, specifically relating to a real-time risk monitoring method for safe lane changing in autonomous vehicles. Background Technology

[0002] Current mainstream rule-based lane-changing decision-making systems typically rely on perception sensors and high-definition maps to detect the poses of surrounding traffic participants during lane changes, predicting their behavior and deciding on a lane change based on whether the distance between the vehicles in front and behind in the target lane meets safety requirements. However, existing autonomous driving lane-changing decision-making systems, primarily based on rules and high-definition maps, suffer from the following drawbacks: The safety assessment module is coupled with the decision-making system: if the lane-switching function module fails, the safety assessment capability is lost.

[0003] Insufficient intervention during lane changing: It only affects decision-making and lacks dynamic safety control during the execution phase.

[0004] Poor environmental adaptability: The impact of weather and road conditions on perception and control is not fully considered, which can easily lead to wrong lane changes or accidents.

[0005] Highly dependent on maps: When roads are under construction or maps are not updated, drivers may mistakenly switch to lanes that are impassable. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a real-time risk monitoring method for safe lane changing of autonomous vehicles. Through an independent safety monitoring architecture and a real-time risk assessment mechanism, the method improves the safety of lane changing, reduces the dependence on high-precision maps, and adapts to complex weather and road environments.

[0007] This invention provides a real-time risk monitoring method for safe lane changing of autonomous vehicles, comprising the following steps: Construct a lane-changing safety monitoring system independent of the lane-changing planning and control module. The system includes a lane-changing function layer, a lane-changing function monitoring layer, and a basic safety monitoring layer. By using lightweight maps or mapless positioning systems, real-time weather data, road condition information, and road congestion information are integrated to generate a dynamic distribution relationship between vehicles and traffic participants in a local space. Based on the dynamic distribution relationship, the maximum time, safe time distance and lane-changing safety boundary of the lane-changing direction are calculated in real time. Based on the lane-changing safety boundary, the behavioral parameters of the lane-changing planning and control module are dynamically constrained, including minimum and maximum vehicle speeds, lane-changing time windows, and lateral and longitudinal acceleration ranges.

[0008] As a further technical solution of the present invention, the lane-changing function monitoring layer includes: The road congestion estimation module obtains the vehicle's position and the speed and trajectory of surrounding traffic participants through sensing sensors, and estimates the current road congestion level by combining historical data. The moving target behavior prediction module predicts short-term behavior based on the speed, trajectory, and road congestion of traffic participants; The maximum lane change time estimation module combines weather and road condition detection results to correct lane change performance parameters and calculate the maximum lane change time. The safe distance calculation module generates real-time lateral and longitudinal safe distances between the vehicle and other road users; The lane change safety assessment module assesses lane change risks based on the maximum travel time and safe time interval, and triggers intervention when the risk exceeds the limit. The lane-changing safety boundary control module dynamically generates and publishes constraint parameters for lane-changing planning and control.

[0009] Furthermore, the basic security monitoring layer includes: An independent safety task monitoring module is used to detect the operating status of the lane change safety monitoring system and trigger an emergency stop command when a system failure is detected. The hardware resource isolation module allocates independent computing and storage resources to the lane change safety monitoring system and the safety task monitoring system to avoid common-cause failures.

[0010] Furthermore, intervention mechanisms include: During the lane-changing decision-making phase, if the environmental risks exceed the threshold, lane changing will be prohibited or aggressive lane changing will be suppressed. If the actual response exceeds the safety boundary during the lane change execution phase, the lane change will be forcibly withdrawn and the vehicle will be restored to the original lane. In the event of a system failure, an emergency shutdown command is triggered at the basic safety monitoring layer.

[0011] Furthermore, the moving target behavior prediction module supports the prediction of behaviors including: cutting in, cutting out, emergency braking, overtaking, yielding, and driving close to the edge.

[0012] Furthermore, weather and road condition monitoring includes: road surface roughness, potholes, curb condition, and the impact of extreme weather on handling.

[0013] Furthermore, the lane-changing safety boundary control module ensures that the lane-changing process is within safety constraints by dynamically adjusting the lane-changing time window and acceleration range.

[0014] Furthermore, the security task monitoring module determines whether the system has failed through heartbeat detection or redundancy checks.

[0015] The advantages of this invention lie in its ability to dynamically calculate lane-changing safety boundaries by integrating weather, road, and traffic data in real time through an independent three-layer security architecture and a map-free perception system. Throughout the lane-changing decision-making and execution process, safety distance assessment and risk threshold judgment trigger suppression or withdrawal interventions, significantly improving lane-changing safety in complex scenarios and reducing reliance on high-precision maps. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the three-layer security architecture of the present invention; Figure 2 This is a flowchart of the lane-changing function monitoring strategy of the present invention. Detailed Implementation

[0017] This embodiment provides a real-time risk monitoring method for safe lane changing of autonomous vehicles according to the present invention, including the following steps: Construct a lane-changing safety monitoring system independent of the lane-changing planning and control module. The system includes a lane-changing function layer, a lane-changing function monitoring layer, and a basic safety monitoring layer. By using lightweight maps or mapless positioning systems, real-time weather data, road condition information, and road congestion information are integrated to generate a dynamic distribution relationship between vehicles and traffic participants in a local space. Based on the dynamic distribution relationship, the maximum time, safe time distance and lane-changing safety boundary of the lane-changing direction are calculated in real time. Based on the lane-changing safety boundary, the behavioral parameters of the lane-changing planning and control module are dynamically constrained, including minimum and maximum vehicle speeds, lane-changing time windows, and lateral and longitudinal acceleration ranges.

[0018] like Figure 1 As shown, the lane change safety monitoring system has a three-layer security architecture. It is mainly divided into the lane change function layer, the lane change function monitoring layer, and the basic safety monitoring layer. The lane change function layer is the monitored object, subject to real-time monitoring and intervention by an independent lane change function monitoring module.

[0019] Lane change function monitoring layer If the current vehicle or environment does not meet the requirements or the risk of changing lanes is too high, the lane change function monitoring system will directly intervene to suppress the lane change decision function, such as suppressing aggressive lane changes or prohibiting lane changes.

[0020] The lane change function monitoring system publishes information such as lane change safety boundaries to the lane change planning and control system in real time to ensure that the lane change process is controlled within the safety boundaries. If environmental or vehicle control state changes beyond system expectations and the vehicle's response exceeds the lane-changing safety boundary, it will directly intervene in the original vehicle control system and urgently implement safety boundary over-limit suppression. Basic security monitoring layer Although the lane change safety monitoring system is an independent system, there is still a possibility that the monitoring function may fail due to software or hardware failure. Therefore, a more basic safety monitoring layer is designed, which uses an independent safety task monitoring module to detect whether the lane change safety monitoring system is operating normally. If an abnormal situation such as no response or deadlock of the lane change safety monitoring system is detected, it will trigger intervention to execute an emergency stop of the original vehicle control system, minimizing the safety risks after the lane change safety monitoring system fails.

[0021] Independent resource allocation is set up for the computing and storage resources required by the lane change safety monitoring system and the safety task monitoring system to achieve hardware resource isolation, avoid common cause failures, and enhance the operational safety and reliability of the safety function modules.

[0022] like Figure 2 As shown, the lane-changing function monitoring system mainly includes the design of six core components: road congestion estimation, moving target behavior prediction, maximum lane-changing time estimation, safe distance calculation, lane-changing safety assessment, and lane-changing safety boundary control. The detailed design strategy is as follows: Road congestion estimates By using perception sensors such as visual cameras, lidar, and millimeter-wave radar to calculate the vehicle's position and pose, as well as the passable space where the vehicle is located, the detection of targets and speeds of other traffic participants, and the detection and inference of lane lines within a limited range, the system can fuse and output the real-time distribution relationship between the vehicle and other traffic participants in a local space. Combined with historical data, the current road congestion situation can be estimated.

[0023] Predicting the behavior of moving targets Based on the speed and trajectory of the moving target, and considering the current road congestion, the short-term behavior of the moving target can be predicted relatively accurately, such as whether there will be expected behaviors such as cut-in, cut-out, lane change, sudden braking, overtaking, yielding, or driving close to the edge.

[0024] Maximum lane change time estimate By detecting the current weather conditions, as well as the condition of the road surface, such as roughness, potholes, and curbs, the basic performance of the vehicle's lane-changing ability is further corrected, and the maximum time for lane-changing in each lane is calculated.

[0025] Safety time interval calculation Based on the predicted behavior of the moving target, the vehicle's pose, and the relative position of the vehicle with traffic participants, the vehicle's current time distance in both the horizontal and vertical directions is calculated in real time.

[0026] Lane change safety assessment Based on the estimated maximum lane change time and safe distance, if the vehicle has a serious malfunction, lane change is suppressed or withdrawn; if the vehicle is in normal condition, the lane change function layer is intervened to trigger risk lane change withdrawal if the risk of front and rear collisions in the target lane is too high during the actual lane change process.

[0027] Lane change safety boundary control Based on the estimated maximum time for lane changing and the safe distance between the vehicle and other moving targets, the minimum and maximum permissible speeds, minimum and maximum lane changing time windows, and minimum and maximum lateral and longitudinal acceleration ranges for each lane can be calculated to determine the lane changing safety boundaries. These boundaries are then sent to the lane changing planning and control system to dynamically constrain the planning and control behavior of lane changing. If the planning and control system exceeds the lateral or longitudinal dynamic constraint safety boundaries during actual execution, the lane changing safety monitoring system will urgently intervene in the original vehicle control system to suppress the over-limit behavior and minimize the lane changing safety risks.

[0028] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the specific embodiments described above. The specific embodiments and descriptions in the specification are merely for further illustrating the principles of the invention. Various changes and modifications can be made to the present invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the claims and their equivalents.

Claims

1. A real-time risk monitoring method for safe lane changing of autonomous vehicles, characterized in that, Includes the following steps: A lane-changing safety monitoring system, independent of the lane-changing planning and control module, is constructed. This system includes a lane-changing function layer, a lane-changing function monitoring layer, and a basic safety monitoring layer. Using a lightweight map or mapless positioning system, real-time weather data, road condition information, and traffic congestion data are integrated to generate a dynamic distribution relationship between vehicles and other road users within a local space. Based on this dynamic distribution relationship, the maximum travel time, safe time distance, and lane-changing safety boundary for each lane-changing direction are calculated in real time. According to the lane-changing safety boundary, the behavioral parameters of the lane-changing planning and control module are dynamically constrained, including minimum and maximum vehicle speeds, lane-changing time windows, and lateral and longitudinal acceleration ranges. During lane-changing, if the vehicle's state or environmental conditions exceed the lane-changing safety boundary, an intervention mechanism is triggered, including suppressing lane-changing decisions, emergency lane-changing reversal, or forced stopping. The lane-changing function monitoring layer includes: a road congestion estimation module, which acquires the vehicle's position and the speed and trajectory of surrounding road users through sensing sensors and estimates the current road congestion level by combining historical data; and a moving target behavior prediction module, which predicts short-term behavior based on the speed, trajectory, and road congestion conditions of road users. The system includes: a maximum lane-changing time estimation module, which combines weather and road condition detection results to correct lane-changing performance parameters and calculate the maximum lane-changing time; a safe distance calculation module, which generates real-time lateral and longitudinal safe distances between the vehicle and other road users; a lane-changing safety assessment module, which assesses lane-changing risks based on the maximum time and safe distance, and triggers intervention when risks exceed limits; and a lane-changing safety boundary control module, which dynamically generates and publishes constraint parameters for lane-changing planning and control. The basic safety monitoring layer includes: an independent safety task monitoring module, used to detect the operating status of the lane-changing safety monitoring system and trigger an emergency stop command when system failure is detected; and a hardware resource isolation module, which allocates independent computing and storage resources to the lane-changing safety monitoring system and the safety task monitoring system to avoid common-cause failures.

2. The real-time risk monitoring method for safe lane changing of autonomous vehicles according to claim 1, characterized in that, The intervention mechanisms include: during the lane-changing decision-making phase, if the environmental risk exceeds the threshold, prohibiting lane changing or suppressing aggressive lane changing; during the lane-changing execution phase, if the actual response exceeds the safety boundary, forcibly withdrawing the lane change and restoring driving in the original lane; and triggering an emergency stop command in the basic safety monitoring layer when the system fails.

3. The real-time risk monitoring method for safe lane changing of autonomous vehicles according to claim 2, characterized in that, The predicted behaviors supported by the moving target behavior prediction module include: cutting in, cutting out, sudden braking, overtaking, yielding, and driving close to the edge.

4. The real-time risk monitoring method for safe lane changing of an autonomous vehicle according to claim 2, characterized in that, The weather and road condition detection includes: road surface roughness, potholes, curb condition, and the impact of extreme weather on handling.

5. The real-time risk monitoring method for safe lane changing of autonomous vehicles according to claim 1, characterized in that, The lane-changing safety boundary control module ensures that the lane-changing process is within safety constraints by dynamically adjusting the lane-changing time window and acceleration range.

6. A real-time risk monitoring method for safe lane changing of autonomous vehicles according to claim 3, characterized in that, The security task monitoring module determines whether the system has failed through heartbeat detection or redundancy check.

Images