A speed compensation based collision avoidance method and system for a steering process

Abstract

The application discloses a kind of based on speed compensation's steering process collision avoidance method and system, the method includes: in the steering process of ego car, the environment information and vehicle information obtained by collection are filtered and compensated processing;The environment information and vehicle information after filtering and compensation processing are judged and scene screening, output system state information and potential collision risk scene information;According to system state information and potential collision risk scene information, risk judgment is carried out, and risk assessment result is generated and whether trigger emergency brake action is judged.The application solves the problem that target speed is inaccurate in the scene of large steering, reduces the probability of false triggering of emergency brake.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of vehicle control, and particularly relates to a steering process collision avoidance method and system based on speed compensation. BACKGROUND

[0002] Generally, when a dangerous obstacle such as a main moving object (for example, a movable device such as a vehicle) or an external object (for example, a moving person such as a pedestrian or a fixed device such as a precast paver, a guardrail, and a curbstone) is detected in a front region of a vehicle during vehicle travel, an autonomous emergency braking (AEB) system can use detected dangerous obstacle detection information to determine that a front obstacle is a control object and warn a driver of a high possibility of a collision occurring when a possibility of a collision with the dangerous obstacle is high. If necessary, the AEB system performs autonomous emergency braking to prevent the vehicle from colliding with any obstacle. The AEB application field has recently increased and been standardized, so that many developers and companies are conducting intensive research on an advanced evolved AEB system that can expand the use and utility in a manner in which the advanced AEB system can operate in more cases than a conventional AEB system and guarantee higher safety and stability in a manner in which the advanced AEB system can prevent the vehicle from colliding with any obstacle.

[0003] Specifically, a conventional AEB system can use fusion of a front sensor designed to detect whether there is a surrounding vehicle in a front longitudinal region of a host vehicle to identify whether there is a surrounding vehicle, and can determine movement of the identified surrounding vehicle (i.e., a front vehicle), so that the conventional AEB system can centrally control the detected front vehicle. However, as the AEB system has expanded to more applications, recently, an advanced safety system that can warn a vehicle driver of a close vehicle moving in a horizontal direction at an intersection or a pedestrian running in a horizontal direction at an intersection or can perform autonomous emergency braking has been developed and established by law, thereby obtaining vehicle safety and stability and greater user convenience.

[0004] At present, the related JA (intersection assistant) system needs the support of corner radar in product development and later maintenance, and is generally equipped in vehicles with L3 level advanced driving assistant function, and is generally arranged in four groups around the vehicle; for example, a certain international supplier, the corner radar product supporting the advanced driving assistant system (ADAS, Advanced Driving Assistant System) has a development cost of about 400-600 yuan per unit, and a cost and R&D sharing of about 2000 yuan. Although the popularization rate of ADAS system in intelligent networking is increasing, but according to the authoritative data of China Automobile Consumers Association, the sales of vehicles equipped with four corner radars is less than 0.5%, and the sales proportion is expected to be less than 5% after 2020. But at present, due to the requirement of CNCAP, the products equipped with forward millimeter wave radar and camera are increasingly popular, and the cost has become the mainstream, and the equipment rate is also increasing. For example, a vehicle model with monthly sales of more than 30,000 of a certain mainstream OEM, all models above level two are equipped with forward millimeter wave radar, and the ADAS equipment coverage is high.

[0005] The existing intersection assistant is to judge the oncoming vehicle on the curve by corner radar, and the disadvantage is that corner radar must be supported, and the equipment rate of corner radar is not high and the development cost is high, which leads to the inability to popularize in large quantities. At present, the AEB (Autonomous Emergency Braking, Autonomous Emergency Braking) autonomous emergency braking system related to the front millimeter wave radar considers the front target in product development and testing, and the field of view is narrow, but due to its own limitations, it cannot realize the intersection assistant function. SUMMARY

[0006] The purpose of the present application is to provide a steering process collision avoidance method and system based on speed compensation, which solves the problem of inaccurate target speed in large steering scenarios and reduces the probability of false triggering of emergency braking.

[0007] To solve the above technical problems, the technical scheme of the present application is: a steering process collision avoidance method based on speed compensation, comprising:

[0008] In the steering process of the vehicle, the collected environmental information and vehicle information are filtered and compensated;

[0009] The filtered and compensated environmental information and vehicle information are subjected to state judgment and scene screening, and system state information and potential collision risk scene information are output;

[0010] According to the system state information and potential collision risk scene information, risk judgment is carried out, and risk assessment results are generated and it is judged whether to trigger emergency braking action.

[0011] The specific steps for compensating the vehicle information are as follows:

[0012] A vehicle coordinate system x-y is established, a camera arranged on the ego vehicle is taken as an origin of the vehicle coordinate system, a forward direction of the ego vehicle is taken as a positive direction of an x-axis of the vehicle coordinate system, and a direction perpendicular to the x-axis and pointing to a left side of the ego vehicle is taken as a positive direction of a y-axis of the vehicle coordinate system;

[0013] A target object is set as a target object in the vehicle coordinate system into which the ego vehicle enters in the steering process;

[0014] A horizontal / longitudinal speed of the target object after rotation compensation in the vehicle coordinate system is equal to a horizontal / longitudinal relative speed output by a sensor of the ego vehicle plus a rotation-compensated horizontal / longitudinal speed, and is expressed as:

[0015]

[0016] In the formula, Vx represents a longitudinal relative speed of the target object output by the camera; Vy represents a horizontal relative speed of the target object output by the camera; Vx rel represents a rotation-compensated longitudinal relative speed; Vy rel represents a rotation-compensated horizontal relative speed; Vx deal represents a target object longitudinal speed after rotation compensation; and Vy deal represents a target object horizontal speed after rotation compensation.

[0017] The ego vehicle rotates, and the target object has an angular velocity-W opposite to a direction relative to the ego vehicle, and a speed of the target object relative to the ego vehicle caused by rotation of the ego vehicle is expressed as:

[0018] V rel =-W×R(2)

[0019] In the formula, V rel represents a rotation-compensated target object speed; W represents a yaw angular velocity of the ego vehicle; and R represents a turning radius of the target vehicle.

[0020] A geometric relationship is obtained as follows:

[0021]

[0022] In the formula, θ represents an included angle between the speed Vrel and the x-axis of the vehicle coordinate system; r represents the turning radius of the ego vehicle; lon represents a longitudinal distance of the target object from the ego vehicle; and lat represents a horizontal distance of the target object from the ego vehicle.

[0023] A longitudinal and horizontal speed of the target vehicle relative to the ego vehicle caused by rotation is expressed as:

[0024]

[0025] The formula (2), (3), (4) is brought into the formula (1), and the lateral / longitudinal velocity of the target object after rotation compensation in the vehicle coordinate system is obtained.

[0026]

[0027] In the formula, V is the speed of the ego vehicle relative to the earth.

[0028] The method for judging the risk according to the system state information and the potential collision risk scene information is specifically: the nonlinear steering motion of the ego vehicle is simplified as a linear problem decoupled in lateral and longitudinal directions, a lateral-longitudinal linear collision formula is established, data characteristics of the steering motion of the ego vehicle are introduced, a minimum threshold value LatTTC_Min of lateral collision time and a minimum threshold value LonTTC_Min of longitudinal collision time are set to filter the false triggering caused by the mutation of data, the maximum threshold value LatTTC_Max of lateral collision time, the minimum threshold value LatTTC_Min of lateral collision time and the maximum threshold value LonTTC_Max of longitudinal collision time, the minimum threshold value LonTTC_Min of longitudinal collision time are determined according to the real vehicle data characteristics in the system state information, and when LatTTC_Min≤LatTTC≤LatTTC_Max and / or LonTTC_Min≤LonTTC≤LonTTC_Max, it indicates that there is a collision risk; wherein, LatTTC is the lateral collision time of the ego vehicle, and LonTTC is the longitudinal collision time of the ego vehicle.

[0029] The data characteristics of the steering motion of the ego vehicle are introduced into the lateral-longitudinal linear collision formula, and the specific representation is:

[0030] It is assumed that the ego vehicle moves at a constant speed in a straight line at the current vehicle speed, and if a collision occurs with the target object, the longitudinal collision time should be equal to the lateral collision time, and then:

[0031]

[0032] In the formula, LonTTC represents the calculated value of the longitudinal collision time, and LatTTC represents the calculated value of the lateral collision time.

[0033] The rotation-compensated steering process of the ego vehicle is introduced, and the formula (5) after rotation compensation is brought into the formula (6) to obtain:

[0034]

[0035] In the calculation of the collision time, Vy≈0 can be ignored, it is assumed that the ego vehicle turns left, then Lon>0, lat>0, W>0, Vx deal <0, Vy deal <0, so formula (7) is obtained:

[0036]

[0037] Let The relationship between the longitudinal collision time and the transverse collision time is constructed, and formula (8) can be simplified as:

[0038]

[0039] When k is much smaller than -1, c is much larger than 1, and LocTTC=LatTTC, formula (9) has two unequal positive solutions T1, T2, that is, there is an interval LatTTC∈(0,T1), so that LonTTC>LatTTC, and it is concluded that in the range close to zero, the longitudinal collision time is not equal to the transverse collision time, and the longitudinal collision time is larger.

[0040] The environmental information at least includes lane line information, target object information and sign information, and the vehicle information at least includes vehicle chassis information, vehicle body information, vehicle speed, yaw angular velocity and steering wheel rotation angle.

[0041] The application also provides a steering process collision avoidance system based on speed compensation, comprising a calibration observation management module, a perception information processing module, a state machine management module, a scene screening module, a risk judgment module and a brake output module, wherein,

[0042] The calibration observation management module is used for sending vehicle configuration information to other modules, and the vehicle configuration information at least includes vehicle model configuration words, vehicle wheelbase and camera installation position.

[0043] The perception information processing module is used for filtering and compensating the environmental information and the vehicle information, and outputting the filtered and compensated vehicle information to the state machine management module and the brake output module, and outputting the filtered and compensated environmental information to the scene screening module and the brake output module.

[0044] The state machine management module is used for judging the state of the filtered and compensated vehicle information, and outputting system state information to the risk judgment module and the brake output module.

[0045] The scene screening module is used for screening the filtered and compensated environmental information, and outputting potential collision risk scene information to the risk judgment module.

[0046] The risk judgment module is used for judging the risk according to the system state information and the potential collision risk scene information, generating a risk evaluation result and sending it to the brake output module.

[0047] The brake output module is used for generating corresponding brake instructions according to the risk evaluation result, and braking according to the brake instructions.

[0048] The specific steps for compensating the vehicle information are as follows:

[0049] A vehicle coordinate system x-y is established, a camera arranged on the ego vehicle is taken as an origin of the vehicle coordinate system, a forward direction of the ego vehicle is taken as a positive direction of an x-axis of the vehicle coordinate system, and a direction perpendicular to the x-axis and pointing to a left side of the ego vehicle is taken as a positive direction of a y-axis of the vehicle coordinate system;

[0050] A target object is set as a target object in the vehicle coordinate system into which the ego vehicle enters in the steering process;

[0051] A horizontal / longitudinal speed of the target object after rotation compensation in the vehicle coordinate system is equal to a horizontal / longitudinal relative speed output by a sensor of the ego vehicle plus the rotation-compensated horizontal / longitudinal speed, and is expressed as:

[0052]

[0053] In the formula, Vx represents a longitudinal relative speed of the target object output by the camera; Vy represents a horizontal relative speed of the target object output by the camera; Vx rel is the rotation-compensated longitudinal relative speed; Vy rel is the rotation-compensated horizontal relative speed; Vx deal is the target object longitudinal speed after rotation compensation; Vy deal is the target object horizontal speed after rotation compensation;

[0054] The ego vehicle rotates, and the target object has an angular velocity-W opposite to a direction relative to the ego vehicle, and a speed of the target object relative to the ego vehicle caused by the rotation of the ego vehicle is expressed as:

[0055] V rel =-WxR(2)

[0056] In the formula, Vrel is the rotation-compensated target object speed; W is a yaw angular velocity of the ego vehicle; and R is a turning radius of the target vehicle;

[0057] According to a geometric relationship, the following formula is obtained:

[0058]

[0059] In the formula, θ is an included angle between the speed Vrel and the x-axis of the vehicle coordinate system; r is the turning radius of the ego vehicle; lon represents a longitudinal distance of the target object from the ego vehicle; and lat represents a horizontal distance of the target object from the ego vehicle;

[0060] The longitudinal and horizontal speeds of the target vehicle relative to the ego vehicle caused by the rotation are expressed as:

[0061]

[0062] The formula (2), (3), (4) is brought into the formula (1), and the lateral / longitudinal velocity of the target object after rotation compensation in the vehicle coordinate system is obtained.

[0063]

[0064] In the formula, V is the speed of the ego vehicle relative to the earth.

[0065] The method for judging the risk according to the system state information and the potential collision risk scene information is specifically: the nonlinear steering motion of the ego vehicle is simplified as a linear problem decoupled in lateral and longitudinal directions, a lateral-longitudinal linear collision formula is established, data characteristics of the steering motion of the ego vehicle are introduced, a minimum threshold value LatTTC_Min of lateral collision time and a minimum threshold value LonTTC_Min of longitudinal collision time are set to filter the false triggering caused by the mutation of data, the maximum threshold value LatTTC_Max of lateral collision time, the minimum threshold value LatTTC_Min of lateral collision time and the maximum threshold value LonTTC_Max of longitudinal collision time, the minimum threshold value LonTTC_Min of longitudinal collision time are determined according to the real vehicle data characteristics in the system state information, and when LatTTC_Min≤LatTTC≤LatTTC_Max and / or LonTTC_Min≤LonTTC≤LonTTC_Max, it indicates that there is a collision risk; wherein, LatTTC is the lateral collision time of the ego vehicle, and LonTTC is the longitudinal collision time of the ego vehicle.

[0066] The data characteristics of the steering motion of the ego vehicle are introduced into the lateral-longitudinal linear collision formula, and the specific representation is:

[0067] It is assumed that the ego vehicle moves at a constant speed in a straight line at the current vehicle speed, and if a collision occurs with the target object, the longitudinal collision time should be equal to the lateral collision time, and then:

[0068]

[0069] In the formula, LonTTC represents the calculated value of the longitudinal collision time, and LatTTC represents the calculated value of the lateral collision time.

[0070] The rotation-compensated steering process of the ego vehicle is introduced, and formula (5) after rotation compensation is brought into formula (6) to obtain:

[0071]

[0072] In the calculation of the collision time, Vy≈0 can be ignored, it is assumed that the ego vehicle turns left, then lon>0, lat>0, W>0, Vx dea l<0, Vy deal <0, so formula (7) is obtained:

[0073]

[0074] Let The formula (8) can be simplified as:

[0075]

[0076] When k is far less than -1, c is far greater than 1, and LonTTC=LatTTC, the formula (9) has two unequal positive solutions T1, T2, that is, there is an interval LattTTC∈(0,T1), so that LonTTC>LatTTC, and it is concluded that in the range close to zero, the longitudinal collision time and the lateral collision time are not equal, and the longitudinal collision time is greater.

[0077] The setting position of the camera is the center of the vehicle head.

[0078] Compared with the prior art, the present application has the beneficial effects that:

[0079] The present application solves the problem of inaccurate target object speed in the scene of large steering by rotating compensation of the relative speed of the target object detected by the vehicle sensor relative to the ego vehicle through the perception information processing module; the present application decouples the lateral and longitudinal risks through the risk judgment module, selects appropriate maximum and minimum thresholds of TTC, simplifies the algorithm, reduces the number of examples, and reduces the probability of false triggering of emergency braking. BRIEF DESCRIPTION OF DRAWINGS

[0080] Figure 1 The flowchart of the embodiment of the present application is shown in the figure;

[0081] Figure 2 The functional architecture diagram of the embodiment of the present application is shown in the figure;

[0082] Figure 3 The schematic diagram of rotating compensation in the embodiment of the present application is shown in the figure;

[0083] Figure 4 The principle schematic diagram of risk judgment in the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical scheme and advantages of the present application clearer and more apparent, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0085] The technical scheme of the present application is:

[0086] Embodiment 1:

[0087] A speed compensation-based collision avoidance method in a steering process, as shown, comprising: Figure 1

[0088] S1. Filtering and compensating the collected environmental information and vehicle information during the steering process of the ego vehicle;

[0089] S2. Judging the state and screening the scene of the filtered and compensated environmental information and vehicle information, outputting system state information and potential collision risk scene information;

[0090] S3. Judging the risk according to the system state information and potential collision risk scene information, generating a risk assessment result and judging whether to trigger an emergency braking action.

[0091] The specific steps for compensating the vehicle information are:

[0092] Establishing a vehicle coordinate system x-y, taking the camera mounted on the ego vehicle as the origin of the vehicle coordinate system, the forward direction of the ego vehicle as the positive direction of the x-axis of the vehicle coordinate system, and the direction perpendicular to the x-axis and pointing to the left side of the ego vehicle as the positive direction of the y-axis of the vehicle coordinate system;

[0093] Setting the other vehicle entering the vehicle coordinate system during the steering process of the ego vehicle as the target object;

[0094] The horizontal / longitudinal speed of the target object after rotation compensation in the vehicle coordinate system is equal to the horizontal / longitudinal relative speed output by the sensor of the ego vehicle plus the rotation-compensated horizontal / longitudinal speed, which is expressed as:

[0095]

[0096] In the formula, Vx represents the longitudinal relative speed of the target object output by the camera; Vy represents the horizontal relative speed of the target object output by the camera; Vx rel is the rotation-compensated longitudinal relative speed; Vy rel is the rotation-compensated horizontal relative speed; Vx deal is the longitudinal speed of the target object after rotation compensation; Vy deal is the horizontal speed of the target object after rotation compensation;

[0097] When the ego vehicle rotates, the target object has an angular velocity-W opposite to the direction relative to the ego vehicle, and the speed of the target object relative to the ego vehicle caused by the rotation of the ego vehicle is expressed as:

[0098] rel = -W x R (2)

[0099] In the formula, V rel is the rotation-compensated speed of the target object; W is the yaw angular velocity of the ego vehicle; R is the turning radius of the target vehicle;

[0100] ​​We obtain this from geometric relations:

[0101]

[0102] In the formula, θ is the angle between the velocity Vrel and the x-axis of the vehicle coordinate system; r is the turning radius of the vehicle; lon represents the longitudinal distance of the target object from the vehicle; lat represents the lateral distance of the target object from the vehicle.

[0103] The longitudinal and lateral velocities of the target vehicle relative to its own vehicle due to rotation are expressed as follows:

[0104]

[0105] Substituting equations (2), (3), and (4) into equation (1), the lateral / longitudinal velocities of the target object in the vehicle coordinate system after rotation compensation are expressed as follows:

[0106]

[0107] In the formula, V is the speed of the vehicle relative to the ground.

[0108] like Figure 4 As shown, the method for risk assessment based on system status information and potential collision risk scenario information is as follows: The nonlinear steering motion of the vehicle is simplified into a linear problem decoupled laterally and longitudinally. A linear collision formula is established, and data characteristics of the vehicle's steering motion are introduced. Minimum thresholds for lateral collision time (LatTTC_Min) and longitudinal collision time (LonTTC_Min) are set to filter for false triggering caused by data abrupt changes. Combined with real-vehicle data characteristics in the system status information, the maximum thresholds for lateral collision time (LatTTC_max), minimum thresholds for lateral collision time (LatTTC_Min), maximum thresholds for longitudinal collision time (LonTTC_Max), and minimum thresholds for longitudinal collision time (LonTTC_Min) are determined. A collision risk is indicated when LatTTC_Min ≤ LatTTC ≤ LatTTC_Max and / or LonTTC_Min ≤ LonTTC ≤ LonTTC_Max. Here, LatTTC represents the lateral collision time of the vehicle, and LonTTC represents the longitudinal collision time.

[0109] Introducing data characteristics of the vehicle's steering motion into the lateral and longitudinal linear collision formula, its specific representation is as follows:

[0110] Assuming the vehicle is moving at a constant linear speed at its current speed, if it collides with the target object, the longitudinal collision time should equal the lateral collision time. Therefore:

[0111]

[0112] In the formula, LonTTC represents a longitudinal collision time calculation value; and LatTTC represents a lateral collision time calculation value.

[0113] The self-vehicle steering process after the rotation compensation is introduced, and the formula (5) after the rotation compensation is brought into the formula (6) to obtain:

[0114]

[0115] In the calculation of the collision time, Vy can be ignored, assuming that the self-vehicle turns left, then lon>0, lat>0, W>0, Vx deal <0, Vy deal <0, and thus the formula (7) is obtained:

[0116]

[0117] Let The relationship between the longitudinal collision time and the lateral collision time is constructed, and thus the formula (8) can be simplified as:

[0118]

[0119] When k is much smaller than -1 and c is much larger than 1, LonTTC=latTTC, the formula (9) has two unequal positive solutions T1 and T2, that is, there is an interval LatTTC∈(0,T1) such that LonTTC>LatTTC, and thus it is concluded that in the range close to zero, the longitudinal collision time is not equal to the lateral collision time, and the longitudinal collision time is larger.

[0120] The environmental information at least includes lane line information, target object information and sign information; and the vehicle information at least includes self-vehicle chassis information, vehicle body information, vehicle speed, yaw angular velocity and steering wheel turning angle.

[0121] The above-mentioned embodiments can be stored in a computer readable storage medium, and the storage medium stores executable instructions, which are executed by a processor to enable the processor to implement the method described in the above-mentioned embodiments.

[0122] Embodiment 2:

[0123] A steering process collision avoidance system based on speed compensation, as shown in Figure 2 The system includes a calibration observation management module, a perception information processing module, a state machine management module, a scene screening module, a risk judgment module and a braking output module; wherein,

[0124] The calibration observation management module is configured to send vehicle configuration information to other modules, the vehicle configuration information at least including a vehicle model configuration word, an axle distance of the ego vehicle, and a camera installation position; the module is configured to define variables that need to be set as adjustable as calibration quantities, and to define internal variables that need to be observed in real time as observation quantities, and is applicable to various places in a software module to realize real-time adjustment and observation of internal variables, and the module is an important way for real vehicle debugging.

[0125] The perception information processing module is configured to perform filtering and compensation processing on the environment information and the vehicle information, and output the vehicle information after the filtering and compensation processing to the state machine management module and the braking output module, and output the environment information after the filtering and compensation processing to the scene screening module and the braking output module; the sensor data is processed through filtering, denoising, calibration and other post-processing to ensure the accuracy and stability of the perception information, including secondary modules such as data filtering and rotation compensation. Specifically, the rotation compensation module: mainly compensates the transverse and longitudinal relative speeds of the target object detected by the sensor when the ego vehicle turns. Many sensor suppliers have errors in detecting the transverse and longitudinal speeds of the target object when the ego vehicle turns, and do not consider the influence of the inertial coordinate system on the non-inertial coordinate system when the ego vehicle turns, so the relative speed of the target object is compensated by the speed compensation amount caused by the rotation of the ego vehicle, as shown in the following formula. Figure 3

[0126] The state machine management module is configured to perform state judgment on the vehicle information after the filtering and compensation processing, and output system state information to the risk judgment module and the braking output module; through the ego vehicle state related signals including the vehicle speed, the yaw rate, the steering wheel angle and the like provided by the ego vehicle sensors, the classification judgment is performed according to the states that will cause the function to enter a fault, enter a prohibition or enter a normal enablement, and the extensible states such as the suppression state, the initialization state and the like are reserved. The module outputs the system state signal state and the function enable signal enable as the state driving signals for subsequent braking judgment.

[0127] The scene screening module is configured to perform scene screening on the environment information after the filtering and compensation processing, and output potential collision risk scene information to the risk judgment module; analyze the road vehicle information and the like output by the perception information processing module, judge whether the current is near a crossroad, determine to enter the scene screening stage, and screen out potential collision risk scenes according to the vehicle position, the speed and the road obstacle structure relationship, including secondary modules such as crossroad screening and target object screening.

[0128] ​The risk judgment module is configured to perform risk judgment based on the system state information and the potential collision risk scene information, generate a risk assessment result, and send the risk assessment result to the brake output module; based on the screened scene, a pre-set rule or a machine learning algorithm is used to evaluate the collision risk, and then based on the risk assessment result, it is judged whether an emergency braking action needs to be triggered, including a two-level module of lateral risk judgment, longitudinal risk judgment, and comprehensive risk judgment.

[0129] The brake output module is configured to generate a corresponding brake instruction based on the risk assessment result, and perform braking according to the brake instruction. Based on the result of the risk judgment, a corresponding brake instruction is generated, including the intensity of braking and the duration of braking, and the braking process is adjusted according to real-time sensing information to ensure the smoothness and safety of the braking operation, including a two-level module of deceleration setting and state jump.

[0130] The specific steps of compensating the vehicle information are as follows:

[0131] A vehicle coordinate system x-y is established, a camera arranged on the ego vehicle is taken as an origin of the vehicle coordinate system, a forward direction of the ego vehicle is taken as a positive direction of an x-axis of the vehicle coordinate system, and a direction perpendicular to the x-axis and pointing to a left side of the ego vehicle is taken as a positive direction of a y-axis of the vehicle coordinate system;

[0132] A target object is set as a target object in the vehicle coordinate system in a steering process of the ego vehicle;

[0133] A horizontal / longitudinal velocity of the target object after rotation compensation in the vehicle coordinate system is equal to a horizontal / longitudinal relative velocity output by a sensor of the ego vehicle plus a rotation-compensated horizontal / longitudinal velocity, and is expressed as:

[0134]

[0135] In the formula, Vx represents a longitudinal relative velocity of the target object output by the camera, and has a unit of meters per second m / s; Vy represents a horizontal relative velocity of the target object output by the camera, and has a unit of meters per second m / s; Vx rel is a rotation-compensated longitudinal relative velocity, and has a unit of meters per second m / s; Vy rel is a rotation-compensated horizontal relative velocity, and has a unit of meters per second m / s; Vxdeal is a compensated longitudinal velocity of the target object, and has a unit of meters per second m / s; Vydeal is a compensated horizontal velocity of the target object, and has a unit of meters per second m / s.

[0136] The ego vehicle rotates, and the target object has an angular velocity-W opposite to a direction relative to the ego vehicle, and a velocity of the target object relative to the ego vehicle caused by the rotation of the ego vehicle is expressed as:

[0137] V rel =-WxR(2)

[0138] In the formula, V relVrel is the target vehicle speed in the vehicle coordinate system, in unit of meter per second m / s; W is the yaw rate of the ego vehicle, in unit of radian per second rad / s; R is the target vehicle turning radius, in unit of meter m;

[0139] According to the geometric relationship, the following is obtained:

[0140]

[0141] In the formula, theta is the angle between the speed Vrel and the lateral axis of the ego vehicle, in unit of radian rag; r is the turning radius of the ego vehicle, in unit of meter m; lon represents the longitudinal distance between the target vehicle and the ego vehicle, in unit of meter m; lat represents the lateral distance between the target vehicle and the ego vehicle, in unit of meter m;

[0142] The longitudinal and lateral speeds of the target vehicle relative to the ego vehicle caused by rotation are represented as:

[0143]

[0144] The lateral and longitudinal speeds of the target vehicle in the vehicle coordinate system after rotation compensation are represented as:

[0145]

[0146] In the formula, V is the speed of the ego vehicle relative to the earth, in unit of meter per second m / s.

[0147] The method for judging the risk according to the system state information and the potential collision risk scene information is specifically as follows: the nonlinear steering motion of the ego vehicle is simplified as a linear problem of lateral and longitudinal decoupling, a lateral and longitudinal linear collision formula is established, data characteristics of the steering motion of the ego vehicle are introduced, a minimum threshold value LatTTC_Min of lateral collision time and a minimum threshold value LonTTC_Min of longitudinal collision time are set to filter the false triggering caused by sudden changes of data, the maximum threshold value LatTTC_Max of lateral collision time, the minimum threshold value LatTTC_Min of lateral collision time and the maximum threshold value LonTTC_Max of longitudinal collision time, the minimum threshold value LonTTC_Min of longitudinal collision time are determined according to the real vehicle data characteristics in the system state information, and when LatTTC_Min≤LatTTC≤LonTTC_Max and / or lonTTC_Min≤LonTTC≤LonTTC_Max, it indicates that there is a collision risk; wherein, latTTC is the lateral collision time of the ego vehicle, and LonTTC is the longitudinal collision time of the ego vehicle.

[0148] The data characteristics of the steering motion of the ego vehicle are introduced into the lateral and longitudinal linear collision formula, and the specific representation is as follows:

[0149] Assuming that the ego vehicle is moving at a constant speed, if a collision occurs with the target object, the longitudinal collision time should be equal to the lateral collision time, that is,

[0150]

[0151] In the formula, LonTTC represents the longitudinal collision time calculation value; LatTTC represents the lateral collision time calculation value;

[0152] The ego vehicle steering process after rotation compensation is introduced, and formula (5) after rotation compensation is brought into formula (6) to obtain:

[0153]

[0154] In the calculation of the collision time, Vy≈0 can be ignored, assuming that the ego vehicle turns left, then lon>0, lat>0, W>0, Vx deal <0, Vy deal <0, so formula (7) is obtained:

[0155]

[0156] Let The relationship between the longitudinal collision time and the lateral collision time is constructed, and formula (8) can be simplified as:

[0157]

[0158] When k is much smaller than -1 and c is much larger than 1, LonTTC=LatTTC, formula (9) has two unequal positive solutions T1, T2, that is, there is an interval LatTTC∈(0, T1) such that LonTTC>LatTTC, so the conclusion is drawn that in the range close to zero, the longitudinal collision time is not equal to the lateral collision time, and the longitudinal collision time is larger.

[0159] The horizontal and longitudinal collision time trigger thresholds are preliminarily positioned through simulation, and the horizontal and longitudinal collision time trigger thresholds are finally determined in combination with the characteristics of the real vehicle data.

[0160] In the embodiment, the setting position of the camera is the center of the vehicle head, as shown in Figure 3 .

[0161] Figure 3 Symbol explanation in the formula:

[0162] a) x-y represents the vehicle coordinate system, taking the camera as the origin, the vehicle longitudinal axis is the positive direction of the x-axis, and the vertical x-axis points to the left side of the vehicle as the positive direction of the y-axis;

[0163] b) lon represents the longitudinal distance of the target object in the vehicle coordinate system; lat represents the lateral distance of the target object in the vehicle coordinate system;

[0164] c) V is the speed of the ego vehicle relative to the ground, W is the yaw rate of the ego vehicle, and r is the turning radius of the ego vehicle;

[0165] d) Vx is the longitudinal relative speed of the target object in the vehicle coordinate system output by the sensor, and Vy is the lateral relative speed of the target object in the vehicle coordinate system output by the sensor;

[0166] e) R is the turning radius of the target vehicle relative to the ego vehicle;

[0167] f) Vrel is the rotation-compensated speed, Vxrel is the rotation-compensated longitudinal speed, Vyrel is the rotation-compensated lateral speed, and θ is the angle between the speed Vrel and the lateral axis of the ego vehicle;

[0168] g) Vxdeal is the compensated longitudinal speed of the target object in the vehicle coordinate system, and Vydeal is the compensated lateral speed of the target object in the vehicle coordinate system.

[0169] Embodiment 3:

[0170] The application further provides an electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete mutual communication through the communication bus; the memory stores a computer program, and when the program is executed by the processor, the processor executes the steps of the steering process collision avoidance method based on speed compensation.

[0171] Embodiment 4:

[0172] The application further provides a computer readable storage medium, which stores executable instructions, and the instructions make the processor implement the steering process collision avoidance method based on speed compensation when executed by the processor.

[0173] Those skilled in the art should understand that the embodiments of the application can be provided as a method, a system or a computer program product. Therefore, the application can be in the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the application can be in the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0174] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing the steps listed in the flowchart

[0175] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing the steps listed in the flowchart

[0176] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 means for functionally implementing the steps listed in the flowchart

[0177] Those skilled in the art will readily understand that the above described embodiments are merely intended to be illustrative and should not be taken as limiting the application. Any modification, equivalent replacement or improvement made without departing from the spirit and principle of the application shall fall within the scope of the application.

Claims

1. A speed compensation based collision avoidance method for a steering process, characterized in that, The application relates to a vehicle state risk assessment method and system. During the steering of the ego vehicle, the collected environmental information and vehicle information are filtered and compensated; The environmental information at least includes lane line information, target object information and sign information; the vehicle information at least includes ego vehicle chassis information, vehicle body information, vehicle speed, yaw angular velocity and steering wheel rotation angle; the specific steps for compensating the vehicle information are as follows: A vehicle coordinate system x-y is established, a camera arranged on the ego vehicle is used as the origin of the vehicle coordinate system, the forward direction of the ego vehicle is used as the positive direction of the x-axis of the vehicle coordinate system, and the direction perpendicular to the x-axis and pointing to the left side of the ego vehicle is used as the positive direction of the y-axis of the vehicle coordinate system; A target object is set as a target object in the vehicle coordinate system during the steering of the ego vehicle; The rotation-compensated lateral / longitudinal speed of the target object in the vehicle coordinate system is equal to the lateral / longitudinal relative speed output by the sensor of the ego vehicle plus the rotation-compensated lateral / longitudinal speed, and is expressed as: In the formula, Vx represents the longitudinal relative speed of the target output by the camera; Vy represents the transverse relative speed of the target output by the camera; Vx rel is the longitudinal relative speed compensated for rotation; Vy rel is the transverse relative speed compensated for rotation; Vx deal is the longitudinal speed of the target after rotation compensation; Vy deal is the transverse speed of the target after rotation compensation; The filtered and compensated environmental information and vehicle information are subjected to state judgment and scene screening, and system state information and potential collision risk scene information are output; Risk judgment is performed according to the system state information and the potential collision risk scene information, a risk assessment result is generated, and it is judged whether an emergency braking action is triggered.

2. The speed compensation based collision avoidance method for a turning process according to claim 1, wherein, In the step of compensating the vehicle information: The ego vehicle rotates, and the target object has an angular velocity-W opposite to the ego vehicle, so the speed of the target object relative to the ego vehicle caused by the rotation of the ego vehicle is expressed as: V rel = -W x R (2) In the formula, V rel is the target vehicle speed for rotation compensation; W is the yaw angular velocity of the ego vehicle; and R is the target vehicle turning radius. According to the geometric relationship, the following formula is obtained: In the formula, theta is the included angle between the speed Vrel and the x-axis of the vehicle coordinate system; r is the turning radius of the ego vehicle; lon represents the longitudinal distance of the target object from the ego vehicle; and lat represents the lateral distance of the target object from the ego vehicle. The longitudinal and lateral speeds of the target vehicle relative to the ego vehicle caused by the rotation are expressed as: The formula (2), (3) and (4) are brought into the formula (1), and the rotation-compensated lateral / longitudinal speed of the target object in the vehicle coordinate system is expressed as: In the formula, V is the speed of the ego vehicle relative to the earth.

3. The speed compensation based collision avoidance method for a turning process according to claim 2, wherein, The method for performing risk judgment according to the system state information and the potential collision risk scene information is as follows: the nonlinear steering motion of the ego vehicle is simplified as a linear problem of lateral and longitudinal decoupling, a lateral and longitudinal linear collision formula is established, the data characteristics of the steering motion of the ego vehicle are introduced, the minimum threshold values of the lateral and longitudinal collision times LatTTC_Min and LonTTC_Min are set to filter the false triggering caused by the mutation of data, the maximum threshold values of the lateral and longitudinal collision times LatTTC_Max and LonTTC_Max are determined in combination with the real vehicle data characteristics in the system state information, and when LatTTC_Min<=LatTTC<=LatTTC_Max and / or LonTTC_Min<=LonTTC<=LonTTC_Max, it is indicated that there is a collision risk; wherein, LatTTC is the lateral collision time of the ego vehicle, and LonTTC is the longitudinal collision time of the ego vehicle.

4. The speed compensation based collision avoidance method for a turning process according to claim 3, wherein, The data characteristics of the steering motion of the ego vehicle are introduced into the lateral and longitudinal linear collision formula, and the specific expression is as follows: Assuming that the ego vehicle is moving at a constant speed, if a collision occurs with the target object, the longitudinal collision time should be equal to the lateral collision time, that is, In the formula, LonTTC represents the longitudinal collision time calculation value; LatTTC represents the lateral collision time calculation value; The formula (5) after the rotation compensation is brought into the formula (6) to obtain: In the calculation of the collision time, Vy≈0 can be ignored, assuming that the ego vehicle turns left, then lon>0, lat>0, W>0, Vx deal <0, Vy deal <0, so from equation (7): Let Constructing the relationship between longitudinal collision time and transverse collision time, formula (8) can be simplified as: When k is much smaller than -1, c is much larger than 1, and LonTTC=LatTTC, the formula (9) has two unequal positive solutions T1 and T2, that is, there is an interval LatTTC∈(0,T1) such that LonTTC>LatTTC, and it is concluded that in the range close to zero, the longitudinal collision time is not equal to the lateral collision time, and the longitudinal collision time is greater.

5. A speed compensation based collision avoidance system for a steering process, characterized in that The calibration observation management module, the perception information processing module, the state machine management module, the scene screening module, the risk judgment module and the brake output module are included. The calibration observation management module is configured to send vehicle configuration information to other modules, wherein the vehicle configuration information at least includes vehicle model configuration word, ego vehicle wheelbase and camera installation position. The perception information processing module is configured to filter and compensate the environmental information and the vehicle information, and output the filtered and compensated vehicle information to the state machine management module and the brake output module, and output the filtered and compensated environmental information to the scene screening module and the brake output module. The environmental information at least includes lane line information, target object information and sign information; the vehicle information at least includes ego vehicle chassis information, vehicle body information, vehicle speed, yaw rate and steering wheel angle; and the specific steps for compensating the vehicle information are as follows: A vehicle coordinate system x-y is established, wherein a camera arranged on the ego vehicle is taken as the origin of the vehicle coordinate system, the forward direction of the ego vehicle is taken as the positive direction of the x-axis of the vehicle coordinate system, and the direction perpendicular to the x-axis and pointing to the left side of the ego vehicle is taken as the positive direction of the y-axis of the vehicle coordinate system; A target object is set as a he vehicle entering the vehicle coordinate system in the steering process of the ego vehicle; The lateral / longitudinal speed of the target object after rotation compensation in the vehicle coordinate system is equal to the lateral / longitudinal relative speed output by the sensor of the ego vehicle plus the rotation-compensated lateral / longitudinal speed, and is expressed as: The state machine management module is configured to judge the state of the filtered and compensated vehicle information, and output system state information to the risk judgment module and the brake output module. In the formula, Vx represents the longitudinal relative speed of the target output by the camera; Vy represents the transverse relative speed of the target output by the camera; Vx rel is the longitudinal relative speed compensated for rotation; Vy rel is the transverse relative speed compensated for rotation; Vx deal is the longitudinal speed of the target after rotation compensation; Vy deal is the transverse speed of the target after rotation compensation; The scene screening module is configured to screen the filtered and compensated environmental information, and output potential collision risk scene information to the risk judgment module. The risk judgment module is configured to judge the risk according to the system state information and the potential collision risk scene information, generate a risk evaluation result and send the risk evaluation result to the brake output module. The brake output module is configured to generate corresponding brake instructions according to the risk evaluation result, and brake according to the brake instructions. The specific steps for compensating the vehicle information are as follows:

6. The speed compensation based collision avoidance system for a turning maneuver according to claim 5, characterized in that, A vehicle coordinate system x-y is established, wherein a camera arranged on the ego vehicle is taken as the origin of the vehicle coordinate system, the forward direction of the ego vehicle is taken as the positive direction of the x-axis of the vehicle coordinate system, and the direction perpendicular to the x-axis and pointing to the left side of the ego vehicle is taken as the positive direction of the y-axis of the vehicle coordinate system; ​ The ego vehicle is set as a target object in the vehicle coordinate system during the steering process; The horizontal / longitudinal speed of the target object in the vehicle coordinate system after rotation compensation is equal to the horizontal / longitudinal relative speed output by the sensor of the ego vehicle plus the rotation-compensated horizontal / longitudinal speed, and is expressed as: In the formula, Vx represents the longitudinal relative speed of the target output by the camera; Vy represents the transverse relative speed of the target output by the camera; Vx rel is the longitudinal relative speed compensated for rotation; Vy rel is the transverse relative speed compensated for rotation; Vx deal is the longitudinal speed of the target after rotation compensation; Vy deal is the transverse speed of the target after rotation compensation; The ego vehicle rotates, and the target object has an angular velocity opposite to the ego vehicle, that is, -W, so the speed of the target object relative to the ego vehicle caused by the rotation of the ego vehicle is expressed as: V rel = -W x R (2) In the formula, V rel is the target vehicle speed for rotation compensation; W is the yaw rate of the ego vehicle; and R is the target vehicle turning radius. From the geometric relationship, we have: In the formula, θ is the angle between the speed Vrel and the x-axis of the vehicle coordinate system; r is the turning radius of the ego vehicle; lon represents the longitudinal distance of the target object from the ego vehicle; and lat represents the lateral distance of the target object from the ego vehicle. The longitudinal and lateral speeds of the target vehicle relative to the ego vehicle caused by the rotation are expressed as: The horizontal / longitudinal speed of the target object in the vehicle coordinate system after rotation compensation is obtained by substituting formulas (2), (3), and (4) into formula (1), and is expressed as: In the formula, V is the speed of the ego vehicle relative to the earth.

7. The speed compensation based collision avoidance system for a turning maneuver according to claim 6, characterized in that, The method for judging the risk according to the system state information and the potential collision risk scenario information is as follows: the nonlinear steering motion of the ego vehicle is simplified as a linear problem of horizontal and longitudinal decoupling, a horizontal and longitudinal linear collision formula is established, data characteristics of the steering motion of the ego vehicle are introduced, a minimum threshold value LatTTC_Min of the horizontal collision time and a minimum threshold value LonTTC_Min of the longitudinal collision time are set to filter the false triggering caused by the mutation of the data, the maximum threshold value LatTTC_Max of the horizontal collision time, the minimum threshold value LatTTC_Min of the horizontal collision time, the maximum threshold value LonTTC_Max of the longitudinal collision time, and the minimum threshold value LonTTC_Min of the longitudinal collision time are determined according to the real vehicle data characteristics in the system state information, and it is indicated that there is a collision risk when LatTTC_Min≤LatTTC≤LatTTC_Max and / or LonTTC_Min≤LonTTC≤LonTTC_Max, wherein LatTTC is the horizontal collision time of the ego vehicle, and LonTTC is the longitudinal collision time of the ego vehicle.

8. The speed compensation based collision avoidance system for a turning maneuver according to claim 7, characterized in that, The data characteristics of the steering motion of the ego vehicle are introduced into the horizontal and longitudinal linear collision formula, and are specifically expressed as: It is assumed that the ego vehicle moves at a constant speed in a straight line at the current vehicle speed, and if a collision occurs with the target object, the longitudinal collision time should be equal to the horizontal collision time, and then we have: In the formula, LonTTC represents the calculated value of the longitudinal collision time, and LatTTC represents the calculated value of the horizontal collision time. The rotation-compensated steering process of the ego vehicle is introduced, and formula (5) after rotation compensation is substituted into formula (6) to obtain: In the calculation of the collision time, Vy≈0 can be ignored, assuming that the ego vehicle turns left, then lon>0, lat>0, W>0, Vx deal <0, Vy deal <0, so from equation (7): Let Constructing the relationship between longitudinal collision time and transverse collision time, formula (8) can be simplified as: When k is much smaller than -1 and c is much larger than 1, LonTTC=LatTTC, formula (9) has two unequal positive solutions T1 and T2, that is, there is an interval LatTTC∈(0,T1) such that LonTTC>LatTTC, so it is concluded that in the range close to zero, the longitudinal collision time and the horizontal collision time are not equal, and the longitudinal collision time is larger.

9. The speed compensation based collision avoidance system for a turning maneuver according to claim 5, wherein, The setting position of the camera is the center of the vehicle head of the ego vehicle.

Images