TWI865064B - Minimum risk decision system and method for changing lane and non-transitory computer readable media - Google Patents

Abstract

The present disclosure provides a minimum risk decision system for changing lane which is disposed at a vehicle. The minimum risk decision system includes at least one processor. The processor includes an external risk region calculating module, a system failure judging system, a self-driving starting condition confirming module and a decision module. The external risk region calculating module calculates an external risk region. The system failure judging system judges whether the system fails and whether a lateral module of a backup system is normal. The self-driving starting condition confirming module confirms whether a self-driving starting condition is satisfied. A decision conduct unit conducts a minimum risk decision as any one of an external risk, a system failure risk or a self-driving starting condition unsatisfied risk happens. Therefore, the self-driving satisfies the rule and the driving safety is increased.

Description

Lane change minimum risk decision system and method and non-transient computer-readable medium

The present invention relates to a minimum risk decision system and method and a non-transitory computer-readable medium, and more particularly to a lane change minimum risk decision system and method and a non-transitory computer-readable medium.

In recent years, the autonomous driving of vehicles has developed rapidly, and the Society of Autonomous Vehicle Engineers (SAE International) and the National Highway Traffic Safety Administration (NHTSA) of the United States have divided the degree of autonomous driving into five levels according to the design of different driving assistance and automation levels. Among them, level 3 autonomous driving means that the autonomous driving system can have the ability to recognize the environment, but the driver can intervene appropriately.

In order to improve the safety of autonomous driving, the United Nations World Forum for Harmonization of Vehicle Regulation issued UN R157 as the standard for level 3 autonomous vehicles. Its purpose is to have a protection mechanism for driving transfer and minimum risk when an emergency occurs (such as system failure) or the autonomous vehicle does not meet the start-up standards. In current technology, there is a lack of minimum risk decision-making during lane change. Therefore, how to make the lane change of autonomous vehicles meet the regulations of UN R157 has become the goal of relevant industry players.

To solve the above problems, the present invention provides a lane change minimum risk decision system and method and a non-transient computer-readable medium. Through the system architecture and method steps, a self-driving car can perform a minimum risk decision during a lane change process.

According to an embodiment of the present invention, a lane change minimum risk decision system is provided, which is arranged in a vehicle, and the lane change minimum risk decision system includes at least one processor. The at least one processor includes an external risk interval calculation module, a system failure judgment module, an automatic driving start condition confirmation module and a decision module. The external risk interval calculation module calculates the collision time between a current position of the vehicle and a front object at any time point among a plurality of time points after the vehicle enters a lane change decision, and uses the collision time to calculate a lateral acceleration of the vehicle when changing lanes from the current position to a target lane, and uses the larger of the lateral acceleration and a maximum normal lateral acceleration and a lateral distance of the vehicle when changing lanes to the target lane to calculate an external risk interval. The system failure judgment module judges whether a system of the vehicle fails at any time point, and whether a lateral module of a backup system of the vehicle is valid. The self-driving start condition confirmation module confirms whether a self-driving start condition is met based on whether a driving status of the vehicle is available at any time point. The decision module includes a judgment unit, a lateral component calculation unit, a feasible space calculation unit and a decision making unit. The judgment unit judges whether the vehicle generates an external risk, a system failure risk, and a risk of not satisfying the self-driving start condition, wherein, when the system is in a normal state, if a relative longitudinal distance between the vehicle and the front object at a certain point in time is less than or equal to the external risk interval, it is determined that the external risk is generated; if the system is failed at a certain point in time but the lateral module of the backup system is valid, it is determined that the system failure risk is generated; when the system is normal and the relative longitudinal distance at a certain point in time is greater than the external risk interval, but the self-driving start condition is not satisfied, it is determined that the risk of not satisfying the self-driving start condition is generated. When any one of the external risk, the system failure risk, and the risk that the self-driving start condition is not satisfied occurs, the lateral component calculation unit calculates a lane change time for the vehicle to change lanes to the target lane at the aforementioned time point, thereby calculating an emergency lateral acceleration of the vehicle, and calculates a lateral component of the vehicle using the emergency lateral acceleration or a normal lateral acceleration. When any one of the external risk, the system failure risk, and the risk that the self-driving start condition is not satisfied occurs, the feasible space calculation unit calculates a feasible space in the target lane at the aforementioned time point using the lateral distance, the lateral component, and a rear object speed of a rear object of the target lane. The decision making unit makes a minimum risk decision when any one of the external risk, the system failure risk and the risk of not satisfying the self-driving start condition occurs, wherein when the external risk or the system failure risk occurs, the vehicle changes lanes with the lateral component to enter a feasible space, or decelerates to a standstill; and when the risk of not satisfying the self-driving start condition occurs, the vehicle changes lanes with the lateral component to enter a feasible space, or moves along an original trajectory of the lane change decision for at least a period of time to another time point, and if the self-driving start condition continues to be not satisfied, the target lane, the lateral component of the lane change and the feasible space at the aforementioned other time point are reconfirmed.

In this way, by calculating the external risk interval and detecting whether the system fails and whether the conditions for automatic start are met, it is possible to determine whether external risks, system failure risks, and risks of not meeting the conditions for automatic start are generated, and calculate the feasible space to make the minimum risk decision, so that lane changes can meet the provisions of UN R157 and improve safety.

According to the lane change minimum risk decision system of the aforementioned implementation method, the external risk interval calculation module can calculate , , and , _Tc is the collision time, _Dr is the relative longitudinal distance between the vehicle and the front object, _vh is the vehicle's ego velocity, _vt is the front object velocity of the front object, _ay is the vehicle's lateral acceleration calculated based on the collision time, _ayrgmax is the maximum normal lateral acceleration, _aymax is the larger of the lateral acceleration and the maximum normal lateral acceleration, _Dy is the lateral distance, _vx is a longitudinal component of the ego velocity, and _Df is the external risk interval.

According to the lane change minimum risk decision system of the above-mentioned implementation method, when the judgment unit determines that an external risk or a system failure risk occurs at a certain point in time, the lateral component calculation unit can calculate and , T _Lc is the lane change time, L is the length of the vehicle, t _{ISO limit} is the time during which the vehicle cannot cross the lane as stipulated by the law, and a _y ^E is the emergency lateral acceleration.

According to the lane change minimum risk decision system of the aforementioned implementation method, if there is no obstacle in the feasible space at a certain point in time and the emergency lateral acceleration is less than or equal to the emergency lateral acceleration upper limit specified by the regulations, the lane can be changed to enter the feasible space, otherwise, the vehicle will decelerate to a stop.

According to the lane change minimum risk decision system of the aforementioned implementation method, when the judgment unit determines that the self-driving start condition does not meet the risk at a certain point in time, the lateral component calculation unit can calculate the lateral component using the conventional lateral acceleration.

According to the lane change minimum risk decision system of the aforementioned implementation method, if there is an obstacle in the feasible space, the vehicle can wait for the obstacle to leave before entering the feasible space with normal lateral acceleration.

According to the lane change minimum risk decision system of the aforementioned implementation method, the vehicle can change lanes to the right one by one to a feasible space on the shoulder of the road and slow down to a stop.

According to the lane change minimum risk decision system of the aforementioned implementation method, when the judgment unit determines that the self-driving start condition does not meet the risk at a certain point in time, the decision making unit can issue a warning within the aforementioned period of time.

According to another embodiment of the present invention, a lane change minimum risk decision method is provided, which includes an external risk interval calculation step, a system failure judgment step, an automatic driving start condition judgment step, a risk judgment step, a lateral component calculation step, a feasible space calculation step and a minimum risk decision execution step. In the external risk interval calculation step, an external risk interval calculation module of at least one processor calculates a collision time between a current position of a vehicle and a front object at any time point in a plurality of time points after the vehicle enters a lane change decision, and calculates a lateral acceleration of the vehicle when changing lanes from the current position to a target lane based on the collision time, and calculates an external risk interval based on the larger of the lateral acceleration and a maximum conventional lateral acceleration and a lateral distance of the vehicle when changing lanes to the target lane. In the system failure judgment step, a system failure judgment module of the at least one processor judges whether a system of the vehicle fails at any time point, and whether a lateral module of a backup system of the vehicle is effective. In the self-driving start condition determination step, an self-driving start condition confirmation module of the at least one processor is enabled to confirm whether an self-driving start condition is satisfied based on whether a driving state of the vehicle at any time point is available. In the risk judgment step, a judgment unit of the at least one processor is made to judge whether the vehicle generates an external risk, a system failure risk, and a risk of not satisfying an automatic start condition, wherein, when the system is in a normal state, if a relative longitudinal distance between the vehicle and the front object at a certain point in time is less than or equal to the external risk interval, it is determined that an external risk is generated; if the system is failed at a certain point in time but the lateral module of the backup system is valid, it is determined that a system failure risk is generated; if the system is normal and the relative longitudinal distance at a certain point in time is greater than the external risk interval, but the automatic start condition is not satisfied, it is determined that a risk of not satisfying the automatic start condition is generated. In the lateral component calculation step, a lateral component calculation unit of the at least one processor calculates a lane change time for the vehicle to change lanes to the target lane at the aforementioned time point when any one of the external risks, the system failure risk, and the risk that the self-driving start condition is not satisfied occurs, so as to calculate an emergency lateral acceleration of the vehicle, and calculate a lateral component of the vehicle using the emergency lateral acceleration or a normal lateral acceleration. In the feasible space calculation step, a feasible space calculation module of the at least one processor is used to calculate a feasible space at the aforementioned time point using the lateral distance, lateral component, and a rear object speed of a rear object of the target lane when any one of the external risks, system failure risks, and the risk of not satisfying the self-driving start condition occurs. In the minimum risk decision making step, a decision making unit of the at least one processor is used to make a minimum risk decision when any one of the external risks, system failure risks, and the risk of not satisfying the self-driving start condition occurs. When external risks or system failure risks occur, the vehicle changes lanes with a lateral component to enter a feasible space, or decelerates to a standstill; and when the risk of unsatisfied autonomous driving start conditions occurs, the vehicle changes lanes with a lateral component to enter a feasible space, or moves along an original trajectory of the lane change decision for at least a period of time to another time point. If the autonomous driving start conditions continue to be unsatisfied, the target lane, the lateral component of the lane change, and the feasible space at the aforementioned other time point are reconfirmed.

According to the lane change minimum risk decision method of the aforementioned implementation method, the external risk interval calculation module can calculate , , and , _Tc is the collision time, _Dr is the relative longitudinal distance between the vehicle and the front object, _vh is the vehicle's ego velocity, _vt is the front object velocity of the front object, _ay is the vehicle's lateral acceleration calculated based on the collision time, _ayrgmax is the maximum normal lateral acceleration, _aymax is the larger of the lateral acceleration and the maximum normal lateral acceleration, _Dy is the lateral distance, _vx is a longitudinal component of the ego velocity, and _Df is the external risk interval.

According to the lane change minimum risk decision method of the above-mentioned implementation method, when the judgment unit determines that an external risk or a system failure risk occurs at a certain point in time, the lateral component calculation unit can calculate and , T _Lc is the lane change time, L is the length of the vehicle, t _{ISO limit} is the time during which the vehicle cannot cross the lane as stipulated by the law, and a _y ^E is the emergency lateral acceleration.

According to the lane change minimum risk decision method of the aforementioned implementation method, if the feasible space at the aforementioned time point is free of obstacles and the emergency lateral acceleration is less than or equal to the emergency lateral acceleration upper limit specified by the regulations, the lane can be changed to enter the feasible space, otherwise, the vehicle will decelerate to a stop.

According to the lane change minimum risk decision method of the aforementioned implementation method, when the judgment unit determines that the self-driving start condition does not meet the risk at a certain point in time, the lateral component calculation unit can calculate the lateral component using the conventional lateral acceleration.

According to the lane change minimum risk decision method of the aforementioned implementation method, if there is an obstacle in the feasible space, the vehicle can wait for the obstacle to leave before entering the feasible space with a normal lateral acceleration.

According to the lane change minimum risk decision method of the aforementioned implementation method, the vehicle can change lanes to the right one by one to a feasible space on the shoulder of the road and slow down to a stop.

According to the lane change minimum risk decision method of the aforementioned implementation method, when the judgment unit determines that the self-driving start condition does not meet the risk at a certain point in time, the decision making unit may issue a warning within at least a period of time mentioned above.

According to another embodiment of the present invention, a non-transitory computer-readable medium is provided, which stores a computer program to enable at least one processor to execute the following. Calculate the collision time between a current position of a vehicle and a front object at any time point among a plurality of time points after the vehicle enters a lane change decision, and calculate a lateral acceleration of the vehicle when changing lanes from the current position to a target lane based on the collision time, and calculate an external risk interval based on the larger of the lateral acceleration and a maximum conventional lateral acceleration and a lateral distance of the vehicle when changing lanes to the target lane. Determine whether a system of the vehicle fails at any time point, and whether a lateral module of a backup system of the vehicle is effective. Whether an autonomous driving start condition is satisfied is determined based on whether a driving status of the vehicle is available at any point in time. Determine whether the vehicle generates an external risk, a system failure risk, and a risk of not satisfying the self-driving start condition, wherein, when the system is in a normal state, if a relative longitudinal distance between the vehicle and the front object at a certain point in time is less than or equal to the external risk interval, it is determined that the external risk is generated; if the system is failed at a certain point in time but the lateral module of the backup system is valid, it is determined that the system failure risk is generated; when the system is normal and the relative longitudinal distance at a certain point in time is greater than the external risk interval, but the self-driving start condition is not satisfied, it is determined that the risk of not satisfying the self-driving start condition is generated. When any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs, a lane change time of the vehicle changing lanes to the target lane at the aforementioned time point is calculated to calculate an emergency lateral acceleration of the vehicle, and a lateral component of the vehicle is calculated using the emergency lateral acceleration or a normal lateral acceleration. When any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs, a feasible space at the aforementioned time point is calculated using the lateral distance, the lateral component and a rear object speed of a rear object of the target lane. When any one of the external risk, the system failure risk and the risk of not satisfying the self-driving start condition occurs, a minimum risk decision is made. When the external risk or the system failure risk occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or decelerates to a standstill; and when the risk of not satisfying the self-driving start condition occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or moves along an original trajectory of the lane change decision for at least a period of time to another time point. If the self-driving start condition continues to be not satisfied, the target lane, the lateral component of the lane change and the feasible space at the aforementioned other time point are reconfirmed.

The following will describe the embodiments of the present invention with reference to the drawings. For the purpose of clarity, many practical details will be described together in the following description. However, the reader should understand that these practical details should not be used to limit the present invention. That is to say, in some embodiments of the present invention, these practical details are not necessary. In addition, in order to simplify the drawings, some commonly used structures and components will be shown in the drawings in a simple schematic manner; and repeated components may be represented by the same number or similar number.

In addition, the terms "first", "second", "third", etc. in this article are only used to describe different elements or components, and do not limit the elements/components themselves. Therefore, the first element/component can also be renamed as the second element/component. Moreover, the combination of elements/components/mechanisms/modules in this article is not a generally known, conventional or familiar combination in this field. Whether the elements/components/mechanisms/modules themselves are known cannot be used to determine whether their combination relationship is easy to be completed by ordinary knowledge in the technical field.

Please refer to FIG. 1 and FIG. 2, wherein FIG. 1 shows a system block diagram of a lane change minimum risk decision system 1000 according to an embodiment of the present invention, and FIG. 2 shows a schematic diagram of the lane change minimum risk decision system 1000 of the embodiment of FIG. 1 being installed on a vehicle CH and applied on a road. The lane change minimum risk decision system 1000 is set on the vehicle CH, and the lane change minimum risk decision system 1000 includes at least one processor 1100. The aforementioned at least one processor 1100 includes an external risk interval calculation module 1110, a system failure judgment module 1120, an automatic driving start condition confirmation module 1130 and a decision module 1140.

The external risk interval calculation module 1110 calculates a collision time between a current position of the vehicle CH and a front object CT at any time point in a plurality of time points after the vehicle CH enters a lane change decision, and calculates a lateral acceleration of the vehicle CH from the current position to a target lane using the collision time, and calculates an external risk interval _DF using the larger of the lateral acceleration and a maximum normal lateral acceleration and a lateral distance Dy of the vehicle CH from the lane change to the target _lane . The system failure judgment module 1120 judges whether a system of the vehicle CH fails at any time point, and whether a lateral module of a backup system of the vehicle CH is effective. The self-driving start condition confirmation module 1130 confirms whether a self-driving start condition is satisfied based on whether a driving state of the vehicle CH is available at any time point.

The decision module 1140 includes a judgment unit 1141, a lateral component calculation unit 1142, a feasible space calculation unit 1143 and a decision execution unit 1144. The judging unit 1141 judges whether the vehicle CH generates an external risk, a system failure risk, and a risk that the self-driving start condition is not satisfied. In the case where the system is normal, if a relative longitudinal distance D _r between the vehicle CH and the front object CT at the aforementioned time point is less than or equal to the external risk interval D _F (i.e., D _r ≤ _{D F} ), it is determined that the external risk is generated; if the system is failed at a time point but the lateral module of the backup system is valid, it is determined that the system failure risk is generated; if the system is normal and the relative longitudinal distance D _r at the aforementioned time point is greater than the external risk interval D _F (i.e., D _r > D _F ), but the automatic start conditions are not met, it is determined that the risk of automatic start conditions not being met occurs.

When any one of the external risk, the system failure risk and the risk of the self-driving start condition not being satisfied occurs, the lateral component calculation unit 1142 calculates a lane change time of the vehicle CH to change lanes to a target lane at a certain point in time, thereby calculating an emergency lateral acceleration of the vehicle CH, and calculates a lateral component v _y of the vehicle CH using the emergency lateral acceleration or a normal lateral acceleration.

When any one of the external risk, the system failure risk and the risk that the autonomous driving start condition is not satisfied occurs, the feasible space calculation unit 1143 calculates a feasible space S1 at a time point using the lateral distance D _y , the lateral component v _y and a rear object speed v _r of a rear object CR in the target lane.

The decision making unit 1144 makes a minimum risk decision when any one of the external risk, the system failure risk and the risk of not satisfying the self-driving start condition occurs, wherein when the external risk or the system failure risk occurs, the vehicle CH changes lanes with the lateral component v _y to enter the feasible space S1, or decelerates to a standstill; and when the risk of not satisfying the self-driving start condition occurs, the vehicle CH changes lanes with the lateral component v _y to enter the feasible space S1, or moves at least for a period of time to another time point according to an original trajectory of the lane change decision. If the self-driving start condition continues to be not satisfied, the target lane at the aforementioned other time point and the lateral component v y of the lane change are re-confirmed. _y and feasible space S1.

Thus, by calculating the external risk interval _DF and detecting whether the system fails and whether the automatic start condition is satisfied, it is possible to judge whether external risk, system failure risk and risk of not satisfying the automatic start condition are generated, and calculate the feasible space S1 to make a minimum risk decision, so that the lane change meets the provisions of UN R157 and improves safety. The details of the lane change minimum risk decision system 1000 will be described in detail later.

At least one processor 1100 of the lane change minimum risk decision system 1000 is installed on the vehicle CH. Specifically, the number of processors 1100 can be two, one processor 1100 includes a control program of the system (main system), and the other processor 1100 includes a control program of the backup system. The processor 1100 can be, for example, a central processing unit (CPU), a digital signal processor (DSP), a microprocessor (MPU), a microcontroller (MCU), etc.; the processor 1100 can be programmed to achieve specific functions. In this embodiment, each processor 1100 is programmed to be divided into an external risk interval calculation module 1110, a system failure judgment module 1120, an automatic driving start condition confirmation module 1130 and a decision module 1140, and different program parts in the decision module 1140 can be divided into a judgment unit 1141, a lateral component calculation unit 1142, a feasible space calculation unit 1143 and a decision making unit 1144. In addition to the above functions, the processor 1100 can also be programmed to achieve general control and image processing functions of the vehicle CH, but is not limited thereto. In one embodiment, the main system may also include multiple mechanisms on the vehicle, which can be controlled by the control program of the main system to achieve the operation of the vehicle, such as the brake mechanism and the steering mechanism, etc. The control program of the backup system can also control these mechanisms, or there are two sets of the same mechanism, one set is controlled by the main system and the other set is controlled by the backup system.

The lane change minimum risk decision system 1000 may further include a driving state detection device 1200 and a sensing module 1300. The sensing module 1300 may include a plurality of cameras for detecting the external environment of the vehicle CH, and can shoot the current environment through the camera, so as to calculate the relative longitudinal distance D _r between the vehicle CH and the front object CT, a front object speed v _t of the front object CT, the current position of the vehicle CH on the lane L1 and the lane line, etc. The sensing module 1300 may also include a radar to detect obstacles near the vehicle CH, but is not limited to the above. The driving state detection device 1200 can be used to detect the driver's eyes, posture, seat belt and other states, and the self-driving start condition confirmation module 1130 determines whether the self-driving start condition is met. In this embodiment, the self-driving start condition can be, for example, the PALS state machine of ISO21202, UNR79 or the specification of 47-2 of Taiwan VACC.

The external risk interval calculation module 1110 can calculate T _c =D _r /v _t -v _h , , and , where _Tc is the collision time, _Dr is the relative longitudinal distance between the vehicle CH and the front object CT, _vh is the vehicle velocity of the vehicle CH, _vt is the front object velocity of the front object CT, _ay is the lateral acceleration of the vehicle CH calculated at the collision time, _ayrgmax is the maximum normal lateral acceleration, _aymax is the larger of the lateral acceleration and the maximum normal lateral acceleration, _Dy is the lateral distance, _vx is a longitudinal component of the vehicle velocity, and _Df is the external risk interval. Please note that in general, the position that the vehicle wants to reach is the center line of the target lane, so the lateral distance is the distance between the center of the vehicle and the center line of the target lane.

As shown in FIG. 1 and FIG. 2, when the vehicle CH is in the state of automatic driving, the system can decide whether to change lanes according to the road conditions of the current lane (e.g., lane L1) and the adjacent lane (e.g., lane L2) where the vehicle CH is located. After the system makes a lane change decision, the vehicle CH starts to change lanes automatically. During the lane change process, the external risk interval calculation module 1110 continuously detects the vehicle speed v _h of the vehicle CH, the relative longitudinal distance D _r between the vehicle CH and the front object CT, and the front object speed v _t of the front object CT, and calculates To obtain the collision time (i.e. T _c ). Afterwards, the collision time can be used to calculate the lateral acceleration of vehicle CH, and then compared with the maximum normal lateral acceleration specified by the law. , we can obtain the larger of the lateral acceleration and the maximum normal lateral acceleration, and use the concept of time-to-space conversion to calculate To obtain the external risk interval D _F , and when any object including the front object CT is within the external risk interval D _F , if the vehicle CH continues to move in the way originally planned by the system (for example, at the originally planned speed), there may be a risk of collision, so it is necessary to change the route to improve driving safety. Based on this, the lateral component v _y and the feasible space S1 can be calculated to plan a new trajectory.

When the judgment unit 1141 determines that an external risk or a system failure risk occurs at a certain point in time, the transverse component calculation unit 1142 can calculate and , where T _Lc is the lane change time, L is the length of vehicle CH, t _{ISO limit} is a legal time limit for lane crossing, and a _y ^E is the emergency lateral acceleration. After determining the feasible space S1, vehicle CH can prepare to start lateral displacement to the feasible space S1. Since vehicle CH will also move forward longitudinally when moving laterally, the lane change time must consider the difference between the relative longitudinal distance D _r between vehicle CH and the object in front CT and the length L of vehicle CH, as well as the relationship between the vehicle speed v _h and the object in front speed v _t . It is also necessary to consider entering the feasible space S1 within half of the collision time, and the legal time limit for lane crossing stipulated in the laws of various countries must be considered. Therefore, it can be calculated based on the time limit. The largest of these is taken as the lane change time, and then used as the basis for calculating the emergency lateral acceleration. Afterwards, the lateral component v _y is calculated using the emergency lateral acceleration. Since the vehicle CH has a high probability of being at risk at this time, it must move laterally at a faster speed to avoid a collision. In addition, the lateral component calculation unit 1142 can further calculate the steering wheel angle corresponding to this lateral component v _y .

Furthermore, when the determination unit 1141 determines that the self-driving start condition does not meet the risk at a certain point in time, the lateral component calculation unit 1142 also calculates , but the lateral component v _y can be calculated using the normal lateral acceleration. In this case, since the self-driving start condition is not met, it means that only driving is unavailable, while the system is normal and there is no external risk, the vehicle CH can change lanes using the normal lateral acceleration.

When planning the feasible space S1, the feasible space calculation unit 1143 refers to the rear object speed v _r of the rear object CR to calculate the forward longitudinal extension distance D _NLF of the vehicle CH (especially the front of the vehicle CH) and the rear longitudinal extension distance D _NLR of the vehicle CH (especially the rear of the vehicle CH) in the feasible space S1. Based on this, D _NLF = (D y /v y ) can be calculated according to the length of the vehicle CH, the lateral component v _y and longitudinal component v _x of the vehicle CH, the rear object speed v _r , and the lateral distance D _y between the vehicle _CH and the target _lane . v _x is used to obtain the forward longitudinal extension distance D _NLF , and D _NLR =(D _y /v _y )．max(v _x ,v _r ) is used to obtain the rearward longitudinal extension distance D _NLR . In one embodiment, when the feasible space is located on a lane or a shoulder, the width of the feasible space can be the width of the lane or the width of the shoulder. When the shoulder is not included, an outer boundary of the feasible space (the boundary farther from the vehicle) is an edge, and a width of the feasible space is a car width, or is fixed to 3.5 meters, and at this time, the center line of the target lane is half a car width inward from the edge, or 1.75 meters inward from the edge, but it is not limited thereto. In the present invention, the target lane may refer to a nearby lane, or a nearby space available for vehicles to travel, and is not limited to a lane with lane line markings.

As shown in FIG. 2 , in a hypothetical situation, the vehicle CH starts to change from lane L1 to lane L2. When the front object CT suddenly decelerates or brakes at a certain point in time, the current position of the vehicle CH is still on lane L1, so that the relative longitudinal distance D _r at this point in time is less than or equal to the external risk interval D _F , then the judgment unit 1141 determines that an external risk occurs, and can calculate the emergency lateral acceleration at this point in time to calculate the lateral component _vy , and calculate the feasible space S1 on the lane L2 adjacent to the lane L1. If the decision making unit 1144 determines that the feasible space S1 on the lane L2 at this time point has no obstacles and the emergency lateral acceleration is less than or equal to the upper limit of the emergency lateral acceleration specified by the regulations, the steering mechanism can be notified to make the vehicle CH change lanes and enter the feasible space S1 on the lane L2. Otherwise, the braking mechanism is notified to make the vehicle CH decelerate to a stop. Specifically, when the emergency lateral acceleration is less than or equal to the upper limit of the emergency lateral acceleration specified by the regulations (for example, 4 meters/second squared), it means that this emergency lateral acceleration is safe and permitted by the regulations, so the vehicle can enter the feasible space S1 on the lane L2 according to the lateral component v _y calculated by the emergency lateral acceleration and the steering wheel angle. When the emergency lateral acceleration is greater than the upper limit of the emergency lateral acceleration specified by the regulations, it means that this emergency lateral acceleration is not safe, so the lane change is not performed and the vehicle decelerates to a stop directly. In addition, when there is an obstacle in the feasible space S1 on the lane L2, it means that there is a possibility of collision when entering this feasible space S1, so the lane change cannot be performed, but the vehicle decelerates to a stop directly. Please note that since the lateral control of the system is effective, before decelerating to a stop, the body of vehicle CH will first return to the straight position and then decelerate to a stop. At this time, because there is a possibility of collision, the deceleration can be set to 4 meters/second squared.

Furthermore, vehicle CH can change lanes to the right one by one to the feasible space S1 on the shoulder L3, and decelerate to a stop. Specifically, when vehicle CH has moved from lane L1 to lane L2, that is, in the feasible space S1 on lane L2, it can move rightward to shoulder L3, that is, the target lane at this point is shoulder L3. At this time, the feasible space S1 on shoulder L3 and the emergency lateral acceleration will be calculated again, and then vehicle CH will change lanes to shoulder L3. As shown in Figure 2, after vehicle CH changes lanes to the feasible space S1 on shoulder L3, it can decelerate to a stop. Since shoulder L3 is relatively safe, the deceleration at this time can be less than 4 meters/second squared, for example, 1 meter/second squared to 3 meters/second squared.

In another hypothetical situation, vehicle CH starts to change from lane L1 to lane L2, the system fails but the lateral module of the backup system is effective. At this time, the judgment unit 1141 determines that the risk of system failure occurs, and can calculate the emergency lateral acceleration at this point in time to calculate the lateral component v _y and the feasible space S1 on lane L2 adjacent to lane L1. If the feasible space S1 at this point in time has no obstacles and the emergency lateral acceleration is less than or equal to the upper limit of the emergency lateral acceleration specified by the regulations, the vehicle can change lanes and enter the feasible space S1 on lane L2. Otherwise, the vehicle decelerates to a stop. Please note that the minimum risk decision when the system failure risk occurs is similar to the minimum risk decision when the external risk occurs. However, when the system failure risk occurs, the processor 1100 including the backup system control program may perform the relevant calculations.

In another hypothetical situation, the vehicle CH starts to change from lane L1 to lane L2. At a certain point in time, the driver becomes unconscious for some reason. The driving state detection device 1200 detects the driver's eyes and sends a driving abnormality notification to the self-driving start condition confirmation module 1130. Subsequently, the self-driving start condition confirmation module 1130 confirms that the self-driving start condition is not satisfied. If the system is normal at this time and no external risk occurs, the judgment unit 1141 will determine that the risk of self-driving start condition not being satisfied occurs. In a minimum risk decision, the feasible space S1 on lane L2 at this point in time can be calculated. If there is no obstacle in the feasible space S1 on lane L2 at this time, the lateral component v _y calculated by the conventional lateral acceleration can be used to change lanes and enter the feasible space S1; if there is an obstacle in the feasible space S1 on lane L2 at this time, wait until the obstacle is removed, and then change lanes to enter the feasible space S1 on lane L2 using the lateral component v _y calculated by the conventional lateral acceleration. Please note that the system is normal at this time, so vehicle CH can first return to the center and wait for the obstacle to leave before entering the feasible space S1 on lane L2. Furthermore, after entering the feasible space S1 on the lane L2, the feasible space S1 on the shoulder L3 is calculated, and the lane change is performed using the lateral component v _y calculated using the conventional lateral acceleration. That is, the vehicle CH can change lanes to the right successively to the feasible space S1 on the shoulder L3, and finally decelerate to stop.

In another minimum risk decision, when the judgment unit 1141 determines that the risk of the self-driving start condition not being met is generated, it first moves along an original trajectory of the lane change decision for at least a period of time to another aforementioned time point, and then confirms whether the self-driving start condition continues to be unsatisfied. Specifically, the original trajectory refers to the steering wheel angle, speed, acceleration, etc. when changing lanes from lane L1 to lane L2 in the lane change decision. Because the system is normal and there is no external risk, the decision making unit 1144 can issue a warning within the aforementioned at least a period of time, for example, a warning of 10 seconds. Within 10 seconds, the movement of the original trajectory may have been completed, that is, the vehicle CH has changed lanes from lane L1 to lane L2. If the self-driving start condition is not met at this time, it means that the driver may not be able to respond and should stop at the roadside to avoid danger. Therefore, the current position of the vehicle CH at another time point is lane L2, and the target lane is shoulder L3. The feasible space S1 on the shoulder L3 can be calculated, and the vehicle can be decelerated to stop. The deceleration rate can be less than 3 meters/second squared. Please note that although the embodiment of Figure 2 is based on two lanes L1 and L2 plus a shoulder L3 as an example, it should be known that in other embodiments with multiple lanes, the vehicle moves to the right one by one to the feasible space on the shoulder or the edge of the road.

Please refer to FIG. 3, wherein FIG. 3 shows a block flow chart of a lane change minimum risk decision method 2000 according to another embodiment of the present invention. The lane change minimum risk decision method 2000 includes an external risk interval calculation step 2100, a system failure judgment step 2200, an automatic driving start condition judgment step 2300, a risk judgment step 2400, a lateral component calculation step 2500, a feasible space calculation step 2600 and a minimum risk decision execution step 2700. The following will explain the details of the lane change minimum risk decision method 2000 with reference to the lane change minimum risk decision system 1000 in Figures 1 and 2.

In the external risk interval calculation step 2100, an external risk interval calculation module 1110 of at least one processor 1100 calculates a collision time between a current position of the vehicle CH and a front object CT at any time point in a plurality of time points after the vehicle CH enters a lane change decision, and calculates a lateral acceleration of the vehicle CH when changing lanes from the current position to a target lane based on the collision time, and calculates an external risk interval _DF based on the larger one of the lateral acceleration and a maximum normal lateral acceleration and a lateral distance _Dy when the vehicle CH changes lanes to the target lane.

In the system failure judgment step 2200, a system failure judgment module 1120 of the aforementioned at least one processor 1100 is used to judge whether a system of the vehicle CH fails at any time point, and whether a lateral module of a backup system of the vehicle CH is valid.

In the automatic start condition determination step 2300, an automatic start condition confirmation module 1130 of the at least one processor 1100 is enabled to confirm whether an automatic start condition is satisfied based on whether a driving state of the vehicle CH at any time point is available.

In the risk determination step 2400, a determination unit 1141 of the at least one processor 1100 determines whether the vehicle CH generates an external risk, a system failure risk, and a risk that the self-driving start condition is not satisfied. In the case where the system is normal, if a relative longitudinal distance D _r between the vehicle CH and the front object CT at a certain point in time is less than or equal to the external risk interval D _F (i.e., D _r ≤ _{D F} ), it is determined that the external risk is generated; if the system is failed at a certain point in time but the lateral module of the backup system is valid, it is determined that the system failure risk is generated; if the system is normal and the relative longitudinal distance D _r at a certain point in time is greater than the external risk interval D _F (i.e., D _r > D _{F )} , it is determined that the self-driving start condition is not satisfied. ), but the automatic start conditions are not met, it is determined that the risk of automatic start conditions not being met occurs.

In the lateral component calculation step 2500, a lateral component calculation unit 1142 of the at least one processor 1100 calculates a lane change time for the vehicle CH to change lanes to the target space at the aforementioned time point when any one of the external risk, the system failure risk, and the risk that the self-driving start condition is not satisfied occurs, so as to calculate an emergency lateral acceleration of the vehicle CH, and calculate a lateral component v _y of the vehicle CH using the emergency lateral acceleration or a normal lateral acceleration.

In the feasible space calculation step 2600, a feasible space calculation unit 1143 of the at least one processor 1100 calculates a feasible space S1 at the aforementioned time point using the lateral distance D _y , the lateral component v _y and a rear object speed v _r of a rear object CR in the target lane when any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs.

In the minimum risk decision making step 2700, a decision making unit 1144 of the at least one processor 1100 makes a minimum risk decision when any one of the external risk, the system failure risk and the risk of the self-driving start condition not being satisfied occurs. When an external risk or a system failure risk occurs, the vehicle CH changes lanes with the lateral component v _y and enters the feasible space S1, or decelerates to a standstill; and when a risk that the self-driving start condition is not met occurs, the vehicle CH changes lanes with the lateral component v _y and enters the feasible space S1, or moves along an original trajectory of the lane change decision for at least a period of time to another time point. If the self-driving start condition continues to be unsatisfied, the target lane, the lateral component v _y for lane change, and the feasible space S1 at the aforementioned other time point are reconfirmed.

Please refer to FIG. 4, wherein FIG. 4 shows a first detailed flow chart of a lane change minimum risk decision method 2000 of the embodiment of FIG. 3. Initially, in step S01, the system (i.e., the main system of the self-driving car) operates normally, and in step S02, it can be confirmed whether the driver intervenes, that is, whether the driver intervenes to control the vehicle CH by itself. If so, it enters step S03 to return the driving right to the driver; if not, it enters step S04 to confirm whether to enter the lane change decision, that is, whether the system decides to change lanes according to the current road conditions. If not, it returns to step S01 and the system continues to operate; if so, it enters step S05 to determine whether the system fails.

If the system failure judgment module 1120 confirms that the system has failed in step S05, then the process proceeds to step S06 to further confirm whether the backup system has failed. If so, then the process proceeds to step S09, and the vehicle CH decelerates to a stop; if not, then the process proceeds to step S07, and the system failure judgment module 1120 confirms whether the lateral module of the backup system has failed. If the lateral module has failed, then the process proceeds to step S09 to decelerate the vehicle CH to a stop; if the lateral module is valid, then the process proceeds to step S08, and the judgment unit 1141 judges that the vehicle CH has a system failure risk.

If the system failure judgment module 1120 confirms in step S05 that the system is normal and has not failed, the process proceeds to step S10, where the external risk interval calculation module 1110 calculates the external risk interval _DF , and the calculation method of the external risk interval _DF is the same as above and will not be repeated. Thereafter, the process proceeds to step S11, where the judgment unit 1141 compares whether the relative longitudinal distance D _r between the vehicle CH and the front object CT is less than or equal to the external risk interval _DF (i.e., whether D _r ≤ _DF is satisfied). If yes, the process proceeds to step S12, where the judging unit 1141 judges whether the vehicle CH generates an external risk; if no, the process proceeds to step S13, where the self-driving start condition confirmation module 1130 judges whether driving is available. If driving is available, the process returns to step S01, and the system continues to operate; if driving is not available, the process proceeds to step S14, where the judging unit 1141 judges whether the vehicle CH generates a risk of not satisfying the self-driving start condition.

Please refer to FIG. 5 together with FIG. 1 and FIG. 2, wherein FIG. 5 shows a second detailed flow chart of a lane change minimum risk decision method 2000 of the embodiment of FIG. 3. If it is confirmed in step S08 that a system failure risk is generated, or if it is confirmed in step S12 that an external risk is generated, then step S15 may be entered, and the lateral component calculation unit 1142 calculates the emergency lateral acceleration and the lateral component _vy , and then step S16 may be entered, and the feasible space calculation unit 1143 calculates the feasible space S1, for example, when the vehicle CH is located on the lane L1, the feasible space S1 on the lane L2 is calculated.

When a system failure risk or external risk occurs, if there is no obstacle in the feasible space S1 and the emergency lateral acceleration is less than or equal to the upper limit of the emergency lateral acceleration specified by the regulations, the vehicle can change lanes and enter the feasible space S1. Otherwise, the vehicle will decelerate to a stop. Therefore, in step S17, it can be confirmed whether there is an obstacle in the feasible space S1 on the lane L2. If so, the vehicle will proceed to step S21 and decelerate to a stop in the lane L1. Otherwise, the vehicle will proceed to step S18 to further confirm whether the calculated emergency lateral acceleration meets the regulations, that is, whether the emergency lateral acceleration is less than or equal to the upper limit of the emergency lateral acceleration specified by the regulations. If yes, then the process proceeds to step S19 to enter the feasible space S1 on lane L2 according to the calculated steering wheel angle and lateral component v _y . Otherwise, the process proceeds to step S21 to decelerate to a stop in lane L1. After entering the feasible space S1 on lane L2, it can be determined whether the feasible space S1 is located on the shoulder (e.g., shoulder L3) or near the curb. In step S20, if it is determined that lane L2 is not the shoulder L3 or the adjacent curb, then return to steps S15 and S16, calculate the lateral component v _y and the next feasible space S1 again, for example, the feasible space S1 on the shoulder L3, and then perform steps S17, S18, S19, and S20 in sequence. At this time, it is confirmed that the location of the feasible space S1 is the shoulder L3, so enter step S21 and decelerate to stop.

Please refer to FIG. 6 together with FIG. 1 and FIG. 2, wherein FIG. 6 shows a third detailed flow chart of a lane change minimum risk decision method 2000 of the embodiment of FIG. 3. After confirming in step S14 that the self-driving start condition does not meet the risk, the process proceeds to step S22, where the lateral component calculation unit 1142 calculates the lateral component _vy using conventional lateral acceleration, and then proceeds to step S23, where the feasible space calculation unit 1143 calculates the feasible space S1, for example, the feasible space S1 on lane L2. Afterwards, the vehicle enters step S24 to check whether there are obstacles in the feasible space S1 on the lane L2. If there are obstacles, the vehicle enters step S26 to wait in the lane L1. At this time, the vehicle can return to the normal position and continue to wait. The vehicle can return to steps S22 and S23 to update the feasible space S1 on the lane L2 again with the current position and relevant data of the vehicle CH, and enter step S24 to check whether there are obstacles in the feasible space S1 on the lane L2. If there are no obstacles, the vehicle can enter step S25 to change lanes and enter the feasible space S1 on the lane L2. Steps S27 and S28 are similar to steps S20 and S21, and will not be repeated.

Please refer to FIG. 7, and refer to FIG. 1 and FIG. 2 in conjunction, wherein FIG. 7 shows a fourth detailed flow chart of a lane change minimum risk decision method 2000 of the embodiment of FIG. 3. In another scheme, after confirming in step S14 that the self-driving start condition does not meet the risk, step S29 can be entered to first warn and maintain the original track, that is, maintain the original track of changing lanes from lane L1 to lane L2. After that, step S30 is entered to confirm whether the driver is available. If so, it means that the driver may have fallen asleep, and has awakened after the warning, and step S32 can be entered to restart the driver system. If not, the process proceeds to step S31, where the lateral component calculation unit 1142 calculates the lateral component v _y using conventional lateral acceleration, and then proceeds to step S33 to calculate the feasible space S1 on the shoulder L3. Steps S33, S34, S35, S36, S37, and S38 are similar to steps S23 to S28 and will not be described in detail.

Another embodiment of the present invention is a non-transitory computer-readable medium storing a computer program for at least one processor to execute the following. Calculate the collision time between a current position of a vehicle and a front object at any time point in a plurality of time points after the vehicle enters a lane change decision, and calculate a lateral acceleration of the vehicle when changing lanes from the current position to a target lane using the collision time, and calculate an external risk interval using the larger of the lateral acceleration and a maximum conventional lateral acceleration and a lateral distance of the vehicle when changing lanes to the target lane. Determine whether a system of the vehicle fails at any time point, and whether a lateral module of a backup system of the vehicle is effective. Whether an autonomous driving start condition is satisfied is determined based on whether a driving status of the vehicle is available at any point in time. Determine whether the vehicle generates an external risk, a system failure risk, and a risk of not satisfying the self-driving start condition, wherein, when the system is in a normal state, if a relative longitudinal distance between the vehicle and the front object at a certain point in time is less than or equal to the external risk interval, it is determined that the external risk is generated; if the system is failed at a certain point in time but the lateral module of the backup system is valid, it is determined that the system failure risk is generated; when the system is normal and the relative longitudinal distance at a certain point in time is greater than the external risk interval, but the self-driving start condition is not satisfied, it is determined that the risk of not satisfying the self-driving start condition is generated. When any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs, a lane change time of the vehicle changing lanes to the target lane at the aforementioned time point is calculated to calculate an emergency lateral acceleration of the vehicle, and a lateral component of the vehicle is calculated using the emergency lateral acceleration or a normal lateral acceleration. When any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs, a feasible space at the aforementioned time point is calculated using the lateral distance, the lateral component and a rear object speed of a rear object of the target lane. When any one of the external risk, the system failure risk and the risk of not satisfying the self-driving start condition occurs, a minimum risk decision is made. When the external risk or the system failure risk occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or decelerates to a standstill; and when the risk of not satisfying the self-driving start condition occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or moves along an original trajectory of the lane change decision for at least a period of time to another time point. If the self-driving start condition continues to be not satisfied, the target lane, the lateral component of the lane change and the feasible space at the aforementioned other time point are reconfirmed.

Non-transitory computer-readable media can be any data storage hardware unit that can store data and can be subsequently read by a computer device, such as a memory device. Non-transitory computer-readable media can be hard drives, network attached storage devices (NAS), read-only memory (ROM), random-access memory (RAM), compact discs-read-only memory (CD-ROM), recordable compact discs (CD-R), rewritable compact discs (CD-RW), magnetic tapes, and other optical or non-optical storage hardware units. In this way, computer programs stored on non-transitory computer-readable media can be read and executed.

Although the present invention has been disclosed as above by way of embodiments, it is not intended to limit the present invention. Anyone skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the scope defined in the attached patent application.

1000: Lane change minimum risk decision system 1100: Processor 1110: External risk interval calculation module 1120: System failure judgment module 1130: Self-driving start condition confirmation module 1140: Decision module 1141: Judgment unit 1142: Lateral component calculation unit 1143: Feasible space calculation unit 1144: Decision making unit 1200: Driving state detection device 13 00: Sensing module 2000: Lane change minimum risk decision method 2100: External risk interval calculation step 2200: System failure judgment step 2300: Self-driving start condition judgment step 2400: Risk judgment step 2500: Transverse component calculation step 2600: Feasible space calculation step 2700: Minimum risk decision step CH: Vehicle CT: Front object CR: Rear object D _F : External risk interval D _NLF : Forward longitudinal extension distance D _NLR : Rear longitudinal extension distance D _r : Relative longitudinal distance D _y : lateral distance L1, L2: lane L3: shoulder S1: feasible space S01, S02, S03, S04, S05, S06, S07, S08, S09, S10, S11, S12, S13, S14, S15, S16, S17, S18, S19, S20, S21, S22, S23, S24, S25, S26, S27, S28, S29, S30, S31, S32, S33, S34, S35, S36, S37, S38: step v _h : vehicle speed v _r : rear object speed v _t : front object speed v _x : longitudinal component v _y : lateral component

FIG. 1 shows a system block diagram of a lane change minimum risk decision system according to an embodiment of the present invention; FIG. 2 shows a schematic diagram of the lane change minimum risk decision system of the embodiment of FIG. 1 installed on a vehicle and applied on a road; FIG. 3 shows a block flow chart of a lane change minimum risk decision method according to another embodiment of the present invention; FIG. 4 shows a first detailed flow chart of a lane change minimum risk decision method of the embodiment of FIG. 3; FIG. 5 shows a second detailed flow chart of a lane change minimum risk decision method of the embodiment of FIG. 3; FIG. 6 shows a third detailed flow chart of a lane change minimum risk decision method of the embodiment of FIG. 3; and FIG. 7 shows a fourth detailed flow chart of a lane change minimum risk decision method of the embodiment of FIG. 3.

1000: Lane change minimum risk decision system

1100: Processor

1110: External risk interval calculation module

1120: System failure judgment module

1130: Automatic start condition confirmation module

1140: Decision-making module

1141: Judgment unit

1142: Transverse component calculation unit

1143: Feasible space calculation unit

1144: Decision making unit

1200: Driving status detection device

1300:Sensor module

Claims (17)

A lane change minimum risk decision system is provided in a vehicle. The lane change minimum risk decision system comprises: at least one processor, comprising: an external risk interval calculation module, which calculates the collision time between a current position of the vehicle and a front object at any of a plurality of time points after the vehicle enters a lane change decision, and calculates a lateral acceleration of the vehicle when changing lanes from the current position to a target lane based on the collision time, and calculates an external risk interval based on the larger of the lateral acceleration and a maximum conventional lateral acceleration and a lateral distance of the vehicle when changing lanes to the target lane; a system failure judgment module, which judges whether a system of the vehicle fails at any of the time points, and a lateral distance of a backup system of the vehicle. a self-driving start condition confirmation module, based on whether a driving state of the vehicle at any point in time is available, to confirm whether a self-driving start condition is met; and a decision module, including: a judgment unit, judging whether the vehicle generates an external risk, a system failure risk, and a risk of not meeting the self-driving start condition, wherein, in the When the system is in a normal state, if the relative longitudinal distance between the vehicle and the object in front at a certain point in time is less than or equal to the external risk interval, then the external risk is determined to have occurred; if the system is in failure at a certain point in time but the lateral module of the backup system is effective, then the system failure risk is determined to have occurred; when the system is normal and the relative longitudinal distance at a certain point in time is greater than In the external risk interval, but the self-driving start condition is not satisfied, it is determined that the risk of the self-driving start condition not being satisfied is generated; a lateral component calculation unit calculates a lane change time for the vehicle to change lanes to the target lane at the time point when any one of the external risk, the system failure risk and the risk of the self-driving start condition not being satisfied is generated, so as to calculate the vehicle an emergency lateral acceleration of the vehicle, and using the emergency lateral acceleration or a normal lateral acceleration to calculate a lateral component of the vehicle; a feasible space calculation unit, when any one of the external risk, the system failure risk, and the risk that the self-driving start condition is not satisfied occurs, using the lateral distance, the lateral component, and a rear object speed of a rear object of the target lane to calculate a feasible space calculation unit; a degree, calculating a feasible space at the target lane at the time point; and a decision making unit, when any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs, making a minimum risk decision, wherein: when the external risk or the system failure risk occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or decelerate to a standstill; and when the risk of the self-driving start condition not being met occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or moves according to an original trajectory of the lane change decision for at least a period of time to another time point. If the self-driving start condition continues to be unsatisfied, re-confirm the target lane, the lateral component of the lane change and the feasible space at the other time point. The lane change minimum risk decision system as claimed in claim 1, wherein the external risk interval calculation module calculates T _c =D _r /v _t -v _h , a _y =2D _y /(T _C ² ), a _ymax =max(a _y ,a _yrgmax ) and

, T _c is the collision time, D _r is the relative longitudinal distance between the vehicle and the front object, v _h is the vehicle's own vehicle speed, v _t is a front object speed of the front object, a _y is the lateral acceleration of the vehicle calculated at the collision time, a _yrgmax is the maximum normal lateral acceleration, a _ymax is the larger of the lateral acceleration and the maximum normal lateral acceleration, D _y is the lateral distance, v _x is a longitudinal component of the vehicle's own vehicle speed, and D _F is the external risk interval. The lane change minimum risk decision system as described in claim 2, wherein when the judgment unit determines that the external risk or the system failure risk occurs at the time point, the lateral component calculation unit calculates T _Lc = max[(D _r -L)/(v _t -v _h ),T _c /2,t _{ISO limit} ] and

, T _Lc is the lane change time, L is the length of the vehicle, t _{ISO limit is} the time during which lane crossing is prohibited according to regulations, and a _y ^E is the emergency lateral acceleration. The lane change minimum risk decision system as described in claim 3, wherein if the feasible space at that point in time has no obstacles and the emergency lateral acceleration is less than or equal to the emergency lateral acceleration upper limit specified by the regulations, then the lane is changed to enter the feasible space, otherwise, the vehicle is decelerated to a stop. The lane change minimum risk decision system as described in claim 2, wherein when the judgment unit determines that the self-driving start condition does not meet the risk at the time point, the lateral component calculation unit calculates the lateral component using the conventional lateral acceleration. The lane change minimum risk decision system as described in claim 5, wherein if there is an obstacle in the feasible space, wait for the obstacle to leave before entering the feasible space with the normal lateral acceleration. A lane change minimum risk decision system as described in claim 4 or 6, wherein the vehicle changes lanes to the right one by one to the feasible space on the shoulder of the road and decelerates to a stop. The lane change minimum risk decision system as described in claim 1, wherein when the judgment unit determines that the self-driving start condition does not meet the risk at the time point, the decision making unit issues a warning within the at least one period of time. A lane change minimum risk decision method includes: an external risk interval calculation step, which enables an external risk interval calculation module of at least one processor to calculate a collision time between a current position of a vehicle and a front object at any of a plurality of time points after a vehicle enters a lane change decision, and calculates a lateral acceleration of the vehicle when changing lanes from the current position to a target lane based on the collision time, An external risk interval is calculated based on the larger of the lateral acceleration and a maximum conventional lateral acceleration and a lateral distance of the vehicle from changing lanes to the target lane; a system failure determination step is performed to enable a system failure determination module of the at least one processor to determine whether a system of the vehicle fails at any point in time and whether a lateral module of a backup system of the vehicle is effective; an automatic driving start A condition determination step, in which an automatic start condition confirmation module of the at least one processor determines whether an automatic start condition is satisfied based on whether a driving state of the vehicle at any point in time is available; a risk determination step, in which a determination unit of the at least one processor determines whether the vehicle generates an external risk, a system failure risk, and a risk of not satisfying the automatic start condition, wherein When the system is in a normal state, if the relative longitudinal distance between the vehicle and the front object at a certain point in time is less than or equal to the external risk interval, it is determined that the external risk occurs; if the system is in failure at a certain point in time but the lateral module of the backup system is effective, it is determined that the system failure risk occurs; when the system is normal and the relative longitudinal distance at a certain point in time is greater than the external risk interval, but If the self-driving start condition is not satisfied, it is determined that the self-driving start condition is not satisfied risk; a lateral component calculation step, so that a lateral component calculation unit of the at least one processor calculates a lane change time of the vehicle changing lanes to the target lane at the time point when any one of the external risk, the system failure risk and the self-driving start condition is not satisfied risk occurs, so as to calculate the vehicle an emergency lateral acceleration, and using the emergency lateral acceleration or a normal lateral acceleration to calculate a lateral component of the vehicle; a feasible space calculation step, causing a feasible space calculation module of the at least one processor to calculate a feasible space calculation module based on the lateral distance, the lateral component and a rear object of the target lane when any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs. a speed of a rear object, calculate a feasible space at that time point; and a minimum risk decision making step, so that a decision making unit of the at least one processor makes a minimum risk decision when any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs, wherein: when the external risk or the system failure risk occurs, the vehicle is made to move with the lateral component Change lanes to enter the feasible space, or decelerate to a standstill; and when the risk of the self-driving start condition not being met occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or moves according to an original trajectory of the lane change decision for at least a period of time to another time point. If the self-driving start condition continues to be unsatisfied, re-confirm the target lane, the lateral component of the lane change and the feasible space at the other time point. The lane change minimum risk decision method as claimed in claim 9, wherein the external risk interval calculation module calculates T _c =D _r /v _t -v _h , a _y =2D _y /(T _C ² ), a _ymax =max(a _y ,a _yrgmax ) and

, T _c is the collision time, D _r is the relative longitudinal distance between the vehicle and the front object, v _h is the vehicle's own vehicle speed, v _t is a front object speed of the front object, a _y is the lateral acceleration of the vehicle calculated at the collision time, a _yrgmax is the maximum normal lateral acceleration, a _ymax is the larger of the lateral acceleration and the maximum normal lateral acceleration, D _y is the lateral distance, v _x is a longitudinal component of the vehicle's own vehicle speed, and D _F is the external risk interval. The lane change minimum risk decision method as described in claim 10, wherein, when the judgment unit determines that the external risk or the system failure risk occurs at a time point, the lateral component calculation unit calculates T _Lc =max[(D _r -L)/(v _t -v _h ),T _c /2,t _{ISO limit} ] and a _y ^E =D _y /(T _Lc ² ), T _Lc is the lane change time, L is the length of the vehicle, t _{ISO limit} is a regulatory time limit for not crossing lanes, and a _y ^E is the emergency lateral acceleration. The lane change minimum risk decision method as described in claim 11, wherein if the feasible space at that point in time has no obstacles and the emergency lateral acceleration is less than or equal to the emergency lateral acceleration upper limit specified by the regulations, then the lane is changed to enter the feasible space, otherwise, the vehicle is decelerated to a stop. The lane change minimum risk decision method as described in claim 10, wherein when the judgment unit determines that the self-driving start condition does not meet the risk at the time point, the lateral component calculation unit calculates the lateral component using the conventional lateral acceleration. The lane change minimum risk decision method as described in claim 13, wherein if there is an obstacle in the feasible space, wait for the obstacle to leave before entering the feasible space with the normal lateral acceleration. The lane change minimum risk decision method as described in claim 12 or 14, wherein the vehicle changes lanes to the right one by one to the feasible space on the shoulder of the road and decelerates to a stop. The lane change minimum risk decision method as described in claim 9, wherein when the judgment unit determines that the self-driving start condition does not meet the risk at the time point, the decision making unit issues a warning within the at least one period of time. A non-transitory computer-readable medium stores a computer program that enables at least one processor to execute the following: calculate a collision time between a current position of a vehicle and a front object at any of a plurality of time points after a vehicle enters a lane-changing decision, and use the collision time to calculate a lateral acceleration of the vehicle when changing lanes from the current position to a target lane, and use the larger of the lateral acceleration and a maximum conventional lateral acceleration and a lateral distance of the vehicle when changing lanes to the target lane to calculate an external risk interval; determine whether a system of the vehicle fails at any of the time points, and whether a lateral module of a backup system of the vehicle is effective; based on the vehicle's Whether the state of a driver at any point in time is available to confirm whether an automatic start condition is met; determine whether the vehicle generates an external risk, a system failure risk, and a risk of not meeting the automatic start condition, wherein, when the system is in a normal state, if a relative longitudinal distance between the vehicle and the front object at that point in time is less than or equal to the external risk interval, then it is determined that the external risk is generated; if the system is failed at that point in time but the lateral module of the backup system is valid, then it is determined that the system failure risk is generated; when the system is normal and the relative longitudinal distance at that point in time is greater than the external risk interval, but the automatic start condition is If the condition is not satisfied, it is determined that the risk of the self-driving start condition not being satisfied occurs; when any one of the external risk, the system failure risk and the risk of the self-driving start condition not being satisfied occurs, a lane change time of the vehicle changing lanes to the target lane at that time point is calculated, so as to calculate an emergency lateral acceleration of the vehicle, and the The method comprises: calculating a lateral component of the vehicle using an emergency lateral acceleration or a normal lateral acceleration; and calculating a possible lateral component of the vehicle using the lateral distance, the lateral component and a rear object speed of a rear object of the target lane when any one of the external risk, the system failure risk and the risk that the self-driving start condition is not satisfied occurs. and when any one of the external risk, the system failure risk, and the risk that the self-driving start condition is not satisfied occurs, a minimum risk decision is made, wherein: when the external risk or the system failure risk occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or decelerates to a standstill; and when the self-driving start condition is When the risk of the lane change condition not being met occurs, the vehicle changes lanes with the lateral component to enter the feasible space, or moves according to an original track of the lane change decision for at least a period of time to another time point. If the self-driving start condition continues to be unsatisfied, the target lane, the lateral component of the lane change and the feasible space at the other time point are reconfirmed.

Images