Collision avoidance method during secondary interaction of automatic driving watering cart and roadside static driving-out vehicle

Abstract

The invention belongs to the field of traffic safety, and particularly relates to a collision avoidance method during secondary interaction between an automatic driving watering cart and a roadside static driving-out vehicle, which comprises the following steps: firstly, after the watering cart and a front roadside vehicle stop for the first time, turning on a right steering lamp of the watering cart, and turning on a left steering lamp of the watering cart; and then the watering cart interacts with the front roadside vehicle for the second time. According to the method, it can be ensured that the automatic driving watering cart dynamically fuses transverse and longitudinal risks in a complex secondary interaction scene, safe collision avoidance is achieved, and the problem that a conflict time window is extremely easy to form due to the influence of multiple factors is solved.

Description

Technical Field

[0001] The present invention belongs to the field of traffic safety, and particularly relates to a collision avoidance method for an autonomous sprinkler truck during the secondary interaction with a static vehicle exiting from the roadside. Background Art

[0002] During the urban road cleaning operation, a low-speed autonomous sprinkler truck needs to frequently drive along the roadside to perform the sprinkling operation. Its special operation mode causes the vehicle to often run close to the road edge. When it detects that a stationary vehicle in the right front starts and invades its driving path, based on the advantages of real-time perception algorithms, the sprinkler truck can quickly identify the motion trajectory of the target vehicle within a short distance and trigger the emergency braking system to achieve deceleration and stop. At this time, the driver of the leading vehicle observes the approaching autonomous sprinkler truck through the rearview system and will also generate a stress braking behavior, forming a two-way parking situation in the first interaction.

[0003] The potential risks in this scenario mainly exist in the secondary interaction stage: According to the right-of-way priority given by traffic rules, the autonomous driving system usually executes the restart driving instruction after the first stop; based on experience judgment, human drivers are likely to misinterpret the active avoidance behavior of the sprinkler truck as a courtesy signal to give way and then start the vehicle synchronously. When both vehicles perform acceleration operations within a limited space, affected by multiple factors such as the detection blind area of sensors, system response delay, and human reaction time, it is extremely easy to form a conflict time window, resulting in the autonomous driving system being unable to avoid collisions in time.

[0004] The contradiction between operation efficiency and safety under this special working condition has become the core bottleneck restricting the commercial implementation of autonomous sprinkler trucks. How to construct a decision-making model that takes into account both right-of-way rules and humanized interaction, and develop a cooperative control algorithm with predictive negotiation capabilities, is the key technical challenge that the current industry urgently needs to break through.

[0005] In view of this, the present invention is specifically proposed. Summary of the Invention

[0006] The purpose of the present invention is to overcome the above-mentioned shortcomings of the prior art and propose a collision avoidance method for an autonomous sprinkler truck during the secondary interaction with a static vehicle exiting from the roadside, so as to achieve the safe operation of the autonomous sprinkler truck.

[0007] To achieve the above purpose, the present invention adopts the following technical solutions:

[0008] The present invention provides a collision avoidance method for an autonomous sprinkler truck during the secondary interaction with a static vehicle exiting from the roadside, including first, after the first interaction and parking of the sprinkler truck and the vehicle on the roadside in front, turning on the right turn signal of the sprinkler truck, and then the sprinkler truck and the vehicle on the roadside in front perform a secondary interaction. The specific process of the secondary interaction is as follows:

[0009] Step 1: When the sprinkler truck restarts, it flashes the high beam once and turns on the left turn signal. At the same time, it obtains the lateral position, lateral speed, and longitudinal speed of the vehicle on the roadside ahead, the width of the lane where the sprinkler truck is traveling, and measures the longitudinal distance and relative speed between the sprinkler truck and the vehicle on the roadside ahead in real time.

[0010] Step 2: The sprinkler truck starts with a set initial acceleration, calculates safety critical indicators based on the data obtained in Step 1, analyzes the risk level through the safety critical indicators, and obtains the risk level result during the secondary interaction.

[0011] Step 3: Based on the risk level result in Step 2, the speed and braking behavior of the sprinkler truck corresponding to each risk level are controlled respectively. During this process, a fault tolerance mechanism is also set up to ensure that the sprinkler truck avoids collisions when performing secondary interaction with the vehicle on the roadside ahead.

[0012] Further, in Step 1, the lateral position, lateral speed, and longitudinal speed of the vehicle on the roadside ahead are detected by a lidar or camera on the sprinkler truck.

[0013] Further, in Step 1, the longitudinal distance and relative speed between the sprinkler truck and the vehicle on the roadside ahead are measured in real time by a millimeter-wave radar on the sprinkler truck.

[0014] Further, in Step 1, the width of the lane where the sprinkler truck is traveling is obtained through a high-precision map on the sprinkler truck.

[0015] Further, in Step 2, the set initial acceleration is a value less than a init The calculation formula of a init is as follows;

[0016]

[0017] where: d0 is the longitudinal initial distance between the two vehicles at the start of the secondary interaction; b max is the maximum braking deceleration of the braking system of the sprinkler truck; τ is the response time of the power system of the sprinkler truck.

[0018] Further, in Step 2, the safety critical indicator includes a dynamic safety distance, and the calculation formula of the dynamic safety distance is as follows:

[0019] d safe2 =1.2·d safe(t) (2)

[0020] where, d safe(t) is the standard dynamic safety distance considering both lateral and longitudinal risks, and its calculation formula is as follows:

[0021]

[0022] Among them, v A is the speed of the sprinkler truck; T LC is the lateral intrusion time of the vehicle on the roadside ahead, and its calculation formula is as follows:

[0023]

[0024] Among them, W lane is the width of the driving lane of the sprinkler truck, y o is the lateral distance that the vehicle on the roadside ahead has intruded into the driving lane of the sprinkler truck, and v y is the lateral speed of the vehicle on the roadside ahead.

[0025] Furthermore, in step 2, the risk level results during the secondary interaction include low risk, medium risk, and high risk;

[0026] When d (t) > d safe2 and T LC > 2τ, it is determined that the secondary interaction between the sprinkler truck and the vehicle on the roadside ahead is a low risk, where d(t) is the longitudinal distance between the sprinkler truck and the vehicle on the roadside ahead measured in real time;

[0027] When d (t) ≤ d safe2 or T LC ≤ 2τ, it is determined that the secondary interaction between the sprinkler truck and the vehicle on the roadside ahead is a medium risk;

[0028] When T LC ≤ τ or d (t) < d safe2 it is determined that the secondary interaction between the sprinkler truck and the vehicle on the roadside ahead is a high risk.

[0029] Furthermore, the automatic driving control of the sprinkler truck adopts the intelligent driver IDM model:

[0030]

[0031] Among them, s0 is the static safety distance: when the sprinkler truck and the vehicle on the roadside ahead are completely stationary, the safety distance that needs to be maintained between the two vehicles (since the sprinkler truck has a long body and requires a larger buffer during operation, 3.5 m can be taken. In the case of sensor (various radars) failure mode, 5 m is forcibly taken to improve the fault tolerance ability), Δv = v B (t) - v A (t) is the relative speed of the two vehicles, v B is the longitudinal speed of the vehicle on the roadside ahead, T is the minimum time headway: the time required for the sprinkler truck to travel uniformly to the position of the vehicle on the roadside ahead currently, a max$a_{max}$ is the maximum acceleration of the sprinkler truck, $b$ is the comfortable deceleration of the sprinkler truck, and $v_0$ is the optimal working speed of the sprinkler truck;

[0032] In step 3, the specific process of controlling the speed and braking behavior of the sprinkler truck corresponding to each risk level is as follows:

[0033] When the secondary interaction between the sprinkler truck and the vehicle on the roadside ahead is of low risk, the sprinkler truck follows the vehicle normally and limits the acceleration to:

[0034]

[0035] where $K$ is the smoothing coefficient, with a value of 0.5, to control the smoothness of the acceleration change;

[0036] When the secondary interaction between the sprinkler truck and the vehicle on the roadside ahead is of medium risk, the sprinkler truck actively decelerates and compensates for lateral intrusion, and the acceleration is limited to:

[0037]

[0038] where $\lambda$ is the lateral compensation coefficient, with a value of 0.3, to control the sensitivity of lateral intrusion;

[0039] When the secondary interaction between the sprinkler truck and the vehicle on the roadside ahead is of high risk, the sprinkler truck brakes in stages to avoid collision, and the acceleration is limited to:

[0040]

[0041] Further, in step 3, the fault tolerance mechanism includes that when the camera or radar signal of the sprinkler truck is lost, the redundant sensor of the sprinkler truck is turned on, and the speed of the sprinkler truck $v$ A $\leq 5m / s$. If the redundant sensor signal is also lost, then $d(t)$ is calculated based on the last obtained data, and the vehicle is decelerated to a stop within 8 - 10s, and at the same time, the hazard warning light of the sprinkler truck is turned on;

[0042] When the communication system of the sprinkler truck has a delay, the control system of the sprinkler truck issues an acceleration or deceleration command 0.2 - 0.5s in advance to offset the response delay of the power system.

[0043] Further, the redundant sensor is selected as a LiDAR or an ultrasonic sensor.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] A collision avoidance method for an autonomous sprinkler truck during secondary interaction with a static vehicle exiting the roadside. Based on various data collected by the autonomous sprinkler truck, safety key indicators are calculated, and risk level analysis is performed through the safety key indicators to obtain three risk levels. Subsequently, corresponding driving speeds and braking conditions are obtained for these three risk levels respectively. At the same time, a fault tolerance mechanism is also set up. The method of the present invention can ensure that the autonomous sprinkler truck dynamically integrates lateral and longitudinal risks in complex secondary interaction scenarios, realizes safe collision avoidance, and solves the problem of easily forming a conflict time window due to multiple factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings herein are incorporated into and constitute a part of this specification, and together with the specification are used to explain the principles of the present invention.

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a flowchart of the collision avoidance method of the present invention;

[0049] Figure 2 It is a schematic diagram of the geometric relationship between the sprinkler truck and the vehicle on the roadside in front in the embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] Here, the exemplary embodiments will be described in detail. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present invention. On the contrary, they are only examples consistent with some aspects of the present invention detailed in the appended claims.

[0051] Please refer to Figures 1-2 , the embodiment of the present invention provides a collision avoidance method for an autonomous sprinkler truck during secondary interaction with a static vehicle exiting the roadside, including the following steps:

[0052] Step 1: After the sprinkler truck stops during the first interaction with the vehicle on the roadside in front, turn on the right turn signal of the sprinkler truck;

[0053] Step 2: After the first stop, restart for secondary interaction. When the secondary interaction decision is to restart, the sprinkler truck flashes the high beam once and turns on the left turn signal. At the same time, obtain the lateral position, lateral speed, and longitudinal speed of the vehicle on the roadside in front through the lidar on the sprinkler truck, obtain the width of the driving lane of the sprinkler truck through the high-precision map, and measure the longitudinal distance and relative speed between the sprinkler truck and the vehicle on the roadside in front in real time through the millimeter-wave radar on the sprinkler truck;

[0054] Step 3: The sprinkler starts with a set initial acceleration and calculates safety critical indicators based on the data obtained in Step 2 (the lateral position, lateral speed, longitudinal speed of the vehicle on the roadside ahead, the width of the lane in which the sprinkler is traveling, the longitudinal distance and relative speed between the sprinkler and the vehicle on the roadside ahead), and conducts risk analysis through the safety critical indicators to obtain the risk level result during the secondary interaction;

[0055] Step 4: Based on the risk level result in Step 3, the speed and braking behavior of the sprinkler corresponding to each risk level are controlled respectively. During this process, a fault tolerance mechanism is also set up to ensure that the sprinkler avoids collision when making a secondary interaction with the vehicle on the roadside ahead.

[0056] Among them, in Step 2, the set initial acceleration is less than the value of a init which is restricted by the response time of the power system of the sprinkler, and its calculation formula is as follows; init where d0 is the longitudinal initial distance between the two vehicles at the start of the secondary interaction; b

[0057]

[0058] is the maximum braking deceleration of the braking system of the sprinkler; τ is the response time of the power system of the sprinkler. max

[0059] In Step 2, the safety critical indicators include the dynamic safety distance, and the calculation formula of the dynamic safety distance is as follows:

[0060] d safe2 = 1.2·d safe(t) (2)

[0061] where d safe(t) is the standard dynamic safety distance considering both lateral and longitudinal risks, and its calculation formula is as follows:

[0062]

[0063] where v A is the speed of the sprinkler; T LC is the lateral intrusion time of the vehicle on the roadside ahead, and its calculation formula is as follows:

[0064]

[0065] where W lane is the width of the lane in which the sprinkler is traveling, y o is the lateral distance that the vehicle on the roadside ahead has intruded into the lane of the sprinkler, and v y is the lateral speed of the vehicle on the roadside ahead.

[0066] ​It should be noted that when the design is in the secondary interaction stage, 20% is added to the standard dynamic safety distance to ensure safety.

[0067] In step 2, the risk level results of the second interaction include low risk, medium risk and high risk;

[0068] When (t) >d safe2 And T LC When >2τ, the safety distance is sufficient and the lateral intrusion time is sufficient, and the secondary interaction between the sprinkler truck and the roadside vehicle in front is judged to be low risk, where d(t) is the longitudinal distance between the sprinkler truck and the roadside vehicle in front measured in real time;

[0069] When (t) ≤d safe2 or T LC When ≤2τ, the safety distance is close to the threshold or the lateral intrusion time is short, and the secondary interaction between the sprinkler truck and the roadside vehicle ahead is judged to be medium risk;

[0070] When T LC ≤τ or d (t) <d safe2 When the safety distance is insufficient or the lateral intrusion time is extremely short, it is judged that the sprinkler truck is at high risk when interacting with the roadside vehicle in front for the second time.

[0071] The automatic driving control of the sprinkler truck adopts the intelligent driver IDM model:

[0072]

[0073] in, s0 is the static safety distance: the safety distance that needs to be maintained between the sprinkler truck and the roadside vehicle in front when the two vehicles are completely stationary (in this embodiment, 3.5m is taken, but in the case of sensor (various radars) failure, 5m is taken), Δv = v B (t)-v A (t) is the relative speed of the two vehicles, v B is the longitudinal speed of the roadside vehicle ahead, T is the minimum time distance: the time required for the sprinkler truck to travel at a constant speed to the current position of the roadside vehicle ahead, a max is the maximum acceleration of the sprinkler truck, b is the comfortable deceleration of the sprinkler truck, v0 is the optimal working speed of the sprinkler truck, and in this embodiment, the value is 20 km / h (i.e., the ideal working speed of the sprinkler truck during spraying operation. If there is no vehicle ahead, the sprinkler truck accelerates to v0 and maintains it; if there is a vehicle ahead, the speed is adjusted according to the IDM model, but the ultimate goal is still to return to v0);

[0074] It should be noted that: according to the braking performance of the sprinkler truck (such as b max)Adjust τ and T in formula (1) and formula (5). Specifically, vehicle performance directly affects the acceleration, deceleration, and braking capabilities of the vehicle, which in turn affect the choice of time headway. The specific impacts include:

[0075] Braking performance: A vehicle with good braking performance can stop within a shorter distance, so a shorter safety time headway can be selected; conversely, a longer safety time headway is chosen.

[0076] Response time of the powertrain: A vehicle with a short response time of the powertrain can adjust its speed faster to adapt to the changes of the vehicle in front, so a shorter safety time headway can be selected; conversely, a longer safety time headway is chosen.

[0077] Furthermore, in step 3, the specific process of controlling the speed and braking behavior of the sprinkler corresponding to each risk level is as follows:

[0078] When the secondary interaction between the sprinkler and the vehicle on the roadside ahead is of low risk, it means that the intrusion time of the vehicle on the roadside ahead is long, and the sprinkler has sufficient time to react and follow the vehicle normally (control in the following stage: use the intelligent driver model (IDM) considering the longitudinal-lateral combination), and limit the acceleration to:

[0079]

[0080] where K is the smoothing coefficient, with a value of 0.5;

[0081] When the secondary interaction between the sprinkler and the vehicle on the roadside ahead is of medium risk, the sprinkler actively decelerates and compensates for the lateral intrusion, and the acceleration is limited to:

[0082]

[0083] where λ is the lateral compensation coefficient, with a value of 0.3;

[0084] When the secondary interaction between the sprinkler and the vehicle on the roadside ahead is of high risk, the sprinkler brakes in stages to avoid collision, and the acceleration is limited to:

[0085]

[0086] Furthermore, in step 3, the fault tolerance mechanism includes that when the camera or radar signal of the sprinkler is lost, the redundant sensor of the sprinkler is turned on, and the speed v of the sprinkler is limited A ≤ 5 m / s. If the redundant sensor signal is also lost, then d(t) is calculated based on the last acquired data, and the vehicle is decelerated to a stop within 8 - 10 s, and at the same time, the hazard warning lights of the sprinkler are turned on;

[0087] When the communication system of the sprinkler truck has a delay, the control system of the sprinkler truck issues an acceleration or deceleration command 0.2 - 0.5 s in advance, with 0.3 s being optimal, to offset the response delay of the power system.

[0088] In this embodiment, the redundant sensor is an ultrasonic sensor.

[0089] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention.

[0090] It should be understood that the present invention is not limited to the content already described above, and various modifications and changes can be made without departing from its scope. The scope of the present invention is only limited by the appended claims.

Claims

1. A collision avoidance method for a secondary interaction between an autonomous driving water sprinkler truck and a static vehicle exiting from the roadside, characterized in that: The process includes firstly turning on the right turn signal of the watering truck after the watering truck and the roadside vehicle in front interact with each other for the first time and then the watering truck and the roadside vehicle in front interact for the second time. The specific process of the second interaction is as follows: Step 1: When the water truck restarts, it flashes its high beam and turns on its left turn signal, and at the same time obtains the lateral position, lateral speed, and longitudinal speed of the roadside vehicle ahead, the width of the lane where the water truck is traveling, and measures the longitudinal distance and relative speed between the water truck and the roadside vehicle ahead in real time; Step 2: The watering truck starts at a set initial acceleration, and calculates key safety indicators according to the data obtained in step 1, analyzes the risk level through the key safety indicators, and obtains the risk level result during the secondary interaction; Step 3: Based on the risk level results of step 2, the speed and braking behavior of the sprinkler truck corresponding to each risk level are controlled respectively. In this process, a fault tolerance mechanism is also set up to ensure that the sprinkler truck avoids collision when it interacts with the roadside vehicle in front for the second time.

2. The collision avoidance method according to claim 1, characterized in that: In step 1, the lateral position, lateral speed and longitudinal speed of the vehicle on the roadside ahead are detected by the laser radar or camera on the sprinkler truck.

3. The collision avoidance method according to claim 1, characterized in that: In step 1, the millimeter wave radar on the sprinkler truck measures the longitudinal distance and relative speed between the sprinkler truck and the roadside vehicle ahead in real time.

4. The collision avoidance method according to claim 1, characterized in that: In step 1, the width of the lane where the sprinkler truck is traveling is obtained through the high-precision map on the sprinkler truck.

5. The collision avoidance method according to claim 1, characterized in that: In step 2, the initial acceleration is set to be less than a init The value of a init The calculation formula is as follows; Where d0 is the initial longitudinal distance between the two vehicles at the beginning of the secondary interaction; b max is the maximum braking deceleration of the sprinkler truck’s braking system; τ is the response time of the sprinkler truck’s power system.

6. The collision avoidance method according to claim 1, characterized in that: In step 2, the key safety indicator includes a dynamic safety distance, and the calculation formula of the dynamic safety distance is as follows: d safe2 =1.2·d safe(t) (2) Among them, d safe(t) In order to integrate the standard dynamic safety distance considering horizontal and vertical risks, the calculation formula is as follows: Among them, v A is the speed of the sprinkler truck; T LC is the lateral intrusion time of the roadside vehicle ahead, and its calculation formula is as follows: Among them, W lane is the width of the lane where the sprinkler truck drives, y o is the lateral distance that the roadside vehicle ahead has invaded the sprinkler lane, v y is the lateral speed of the vehicle on the roadside ahead.

7. The collision avoidance method according to claim 6, characterized in that: In step 2, the risk level results of the second interaction include low risk, medium risk and high risk; When (t) >d safe2 And T LC When t > 2τ, the secondary interaction between the sprinkler truck and the roadside vehicle ahead is judged to be low risk, where d(t) is the longitudinal distance between the sprinkler truck and the roadside vehicle ahead measured in real time; When (t) ≤d safe2 or T LC When ≤2τ, the secondary interaction between the sprinkler truck and the roadside vehicle ahead is judged as medium risk; When T LC ≤τ or d (t) <d safe2 When the sprinkler truck interacts with the roadside vehicle ahead for the second time, it is judged to be high risk.

8. The collision avoidance method according to claim 7, characterized in that: The automatic driving control of the sprinkler truck adopts the intelligent driver IDM model: in, s0 is the static safety distance: the safety distance that needs to be maintained between the sprinkler truck and the roadside vehicle in front when the two vehicles are completely stationary, Δv = v B (t)-v A (t) is the relative speed of the two vehicles, v B is the longitudinal speed of the roadside vehicle ahead, T is the minimum time distance: the time required for the sprinkler truck to travel at a constant speed to the current position of the roadside vehicle ahead, a max is the maximum acceleration of the sprinkler truck, b is the comfortable deceleration of the sprinkler truck, and v0 is the optimal working speed of the sprinkler truck; In step 3, the specific process of controlling the speed and braking behavior of the sprinkler truck corresponding to each risk level is as follows: When the water truck interacts with the roadside vehicle ahead for the second time and the risk is low, the water truck follows the vehicle normally and limits the acceleration to: Where K is the smoothing coefficient; When the water truck interacts with the roadside vehicle ahead for the second time at a medium risk, the water truck actively slows down and compensates for the lateral intrusion, and the acceleration limit is: Where, λ is the lateral compensation coefficient; When the water truck interacts with the roadside vehicle ahead for the second time and the risk is high, the water truck applies graded braking to avoid collision and limits the acceleration to:

9. The collision avoidance method according to claim 1, characterized in that: In step 3, the fault tolerance mechanism includes turning on the redundant sensors of the sprinkler truck and limiting the speed of the sprinkler truck to v when the camera or radar signal of the sprinkler truck is lost. A ≤5m / s. If the redundant sensor signal is also lost, d(t) is calculated based on the last acquired data, and the vehicle is decelerated to a stop within 8 to 10 seconds, while turning on the hazard warning lights of the sprinkler truck. When the communication system of the sprinkler truck is delayed, the control system of the sprinkler truck issues an acceleration or deceleration command 0.2 to 0.5 seconds in advance to offset the response delay of the power system.

10. The collision avoidance method according to claim 9, characterized in that: The redundant sensor is a LiDAR or ultrasonic sensor.

Images