Speed optimization method and device for zero-waiting passing intersection of unmanned vehicle
By constructing a conflict function between autonomous vehicles to optimize speed planning and generating a target speed planning curve, the problems of low efficiency and high energy consumption caused by traditional traffic light systems are solved, and zero-wait passage and safe passage at intersections are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 罗光富
- Filing Date
- 2025-12-29
- Publication Date
- 2026-04-17
AI Technical Summary
Traditional traffic light control systems result in low average vehicle speeds, high energy consumption, severe pollution, and numerous accidents at intersections, especially in environments with driverless vehicles.
By constructing a conflict function among autonomous vehicles at intersections, vehicle speed planning is optimized, and a target speed planning curve is generated to achieve zero-wait passage.
While meeting safety constraints, it significantly improves the traffic efficiency of intersections, enables driverless vehicles to pass through with zero waiting time, and reduces energy consumption and pollution.
Smart Images
Figure CN121884583A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of intelligent traffic control technology, and in particular to a speed optimization method and device for unmanned vehicles to pass through intersections with zero waiting time. Background Technology
[0002] Traffic light systems are currently the mainstream solution for controlling traffic flow at intersections in urban areas. By alternately opening red / green lights for different directions, this system allows traffic to pass through intersections in batches within a limited time period, thereby reducing the risk of vehicle conflicts and ensuring traffic safety. However, extensive practice has shown that while traffic lights sacrifice safety through "batch release," they also significantly reduce the average speed of vehicles, making intersections a key bottleneck for traffic congestion during peak hours and holidays. For example, at a typical intersection approximately 100 meters long, with a red light waiting time of about 1 minute and a single vehicle taking about 10 seconds to pass through, the average speed of vehicles passing through the intersection is only about 5.1 kilometers per hour (km / h), equivalent to 1 / 12 to 1 / 6 of the common urban road speed (approximately 30-60 km / h). When adjacent intersections are densely distributed, vehicles that slow down or stop at one intersection often encounter another red light at the next, creating an inefficient "stop-go-stop" cycle that severely weakens the overall traffic capacity of the road. Meanwhile, frequent stops and starts also lead to significant energy losses and exacerbate air pollution and noise emissions. Complex intersections are also high-risk areas for traffic accidents. With the continuous increase in the number of motor vehicles and the rapid development of autonomous driving technology, the limitations of traditional traffic light control systems in terms of traffic efficiency and environmental friendliness are becoming increasingly prominent. Summary of the Invention
[0003] This application provides a speed optimization method and apparatus for unmanned vehicles to pass through intersections with zero waiting time, which can enable unmanned vehicles to pass through intersections with zero waiting time while meeting safety constraints.
[0004] The technical solution of this application embodiment is as follows: This application provides a speed optimization method for autonomous vehicles passing through intersections with zero waiting time, applied to a dispatch controller at the intersection. The method includes: determining a speed planning function for each autonomous vehicle managed by the dispatch controller within a target time period to be optimized; wherein the speed planning function includes driving planning variables describing the autonomous vehicle's passage through the intersection; obtaining driving information for each autonomous vehicle at an initial time within the target time period to be optimized; wherein the driving information includes at least an initial position vector, an initial driving speed, and a driving route; and for each sampling time, optimizing the speed planning function based on a preset safety distance. The initial position vector, velocity planning function, and driving route of each autonomous vehicle are used to construct a sub-objective optimization function. This sub-objective optimization function includes the sum of conflict functions between any two autonomous vehicles waiting to pass through the intersection at the sampling time. The function value of the conflict function is used to measure the safety status of the corresponding vehicle at the sampling time. Based on the sub-objective optimization functions at each sampling time, a target optimization function is constructed for the target time period to be optimized. Under preset constraints, the target optimization function is minimized to obtain the target value of the driving planning variable in the velocity planning function corresponding to each autonomous vehicle.
[0005] This application embodiment also provides a speed optimization device for autonomous vehicles passing through intersections with zero waiting time, applied to the dispatch controller of the intersection; the device includes: an information acquisition unit, used to acquire the driving information corresponding to each autonomous vehicle at the initial time within the target time period to be optimized; wherein, the driving information includes at least an initial position vector, an initial driving speed, and a driving route; a speed optimization unit, used to determine the speed planning function corresponding to each autonomous vehicle managed by the dispatch controller within the target time period to be optimized; wherein, the speed planning function includes driving planning variables describing the autonomous vehicle passing through the intersection; for each sampling time, based on a preset safety distance, the initial position vector of each autonomous vehicle, the speed planning function, and the driving route, the device optimizes the speed planning function. The route is constructed by establishing a sub-objective optimization function. This sub-objective optimization function includes the sum of conflict functions between any two autonomous vehicles at each sampling time, where the function value of the conflict function measures the safety status of the corresponding vehicle at the sampling time. Based on the sub-objective optimization function corresponding to each sampling time, an objective optimization function is constructed for the target time period to be optimized. Under preset constraints, the objective optimization function is minimized to obtain the target value of the driving planning variable in the speed planning function corresponding to each autonomous vehicle. For each autonomous vehicle, the target value is substituted into the speed planning function to obtain a target speed planning curve. The target speed planning curve is used to optimize the driving speed of the autonomous vehicle passing through the intersection.
[0006] This application also provides a scheduling controller, including: a memory for storing computer-executable instructions or computer programs; and a processor for executing the computer-executable instructions or computer programs stored in the memory to implement the speed optimization method provided in this application.
[0007] This application also provides a computer program product, which includes computer-executable instructions or a computer program. When the computer-executable instructions or the computer program are executed by a processor, they implement the speed optimization method provided in this application.
[0008] The embodiments of this application have the following beneficial effects: by constructing a target optimization function composed of conflict functions between each pair of autonomous vehicles in the intersection, and minimizing the target optimization function under constraints, the target values of each vehicle's driving planning variables are obtained, and then a target speed planning curve is generated to optimize vehicle passage. This can achieve safe, zero-waiting, and high-speed passage of vehicles in various types of intersection scenarios, thereby significantly improving the passage efficiency of intersections. Attached Figure Description
[0009] Figure 1 This is a schematic diagram of the structure of an unmanned vehicle dispatching system at intersections provided in an embodiment of this application; Figure 2 This is a schematic diagram of the workflow of an unmanned vehicle dispatching system at intersections provided in an embodiment of this application. Figure 3 A flowchart illustrating the speed optimization method for autonomous vehicles at intersections provided in this application embodiment. Figure 1 ; Figure 4 A flowchart illustrating the speed optimization method for autonomous vehicles at intersections provided in this application embodiment. Figure 2 ; Figure 5 A flowchart illustrating the speed optimization method for autonomous vehicles at intersections provided in this application embodiment. Figure 3 ; Figures 6 to 18 The results are for verifying the speed optimization method for autonomous vehicles passing through intersections provided in this application; Figure 19 A structural block diagram of the speed optimization device for unmanned vehicles at intersections provided in the embodiments of this application; Figure 20 This is a structural block diagram of the scheduling controller provided in an embodiment of this application.
[0010] It should be noted that the terms "first" and "second" mentioned above are only used to distinguish different technical solutions or technical features, and are not used to limit their importance, priority, or superiority / inferiority. Detailed Implementation
[0011] To make the objectives, technical solutions, and beneficial effects of this application clearer, the application will be further described below in conjunction with the accompanying drawings. It should be understood that the embodiments described below are only for explaining this application and do not constitute a limitation on the scope of protection of this application. Other embodiments obtained by those skilled in the art based on the disclosure of this application without creative effort are all within the scope of protection of this application.
[0012] Figure 1This diagram illustrates the structure of an intersection dispatching system according to an embodiment of this application. The dispatching system 100 may include a dispatch controller 101 and an onboard communicator 102. The dispatch controller 101 is located at the intersection and may include an information acquisition unit 1011 and a speed optimization unit 1012 capable of communicating with each other. Specifically, the information acquisition unit 1011 is used to collect real-time driving information of each unmanned vehicle within the monitoring range of the intersection. The driving information includes at least position vectors, driving speeds, and driving routes. An optional acquisition method is that the monitored unmanned vehicles actively report their driving information to the dispatch controller 101, and the reported information is verified and supplemented by the detection device included in the information acquisition unit 1011 to improve the accuracy and completeness of the collected information. The speed optimization unit 1012 is used to optimize the driving speed of each unmanned vehicle within the intersection range according to a preset unmanned vehicle intersection speed optimization method and generate corresponding speed planning information. The vehicle-mounted communicator 102 is installed on each unmanned vehicle and is used to report the vehicle's driving information to the dispatch controller 101 and receive the speed planning function or speed planning curve issued by the dispatch controller 101.
[0013] Figure 2 The above describes the overall workflow of the intersection dispatching system. Specifically, the workflow may include the following steps: S1: When an autonomous vehicle enters the preset intersection dispatching range, the dispatch controller 101 establishes a communication connection with the onboard communicator 102 of the autonomous vehicle through the information acquisition unit 1011, and obtains the driving information of all autonomous vehicles within the dispatching range; S2: The dispatch controller 101 calls the speed optimization unit 1012 to predict and generate the target speed planning curve for each autonomous vehicle in the time period to be optimized after the current time (i.e., the initial time) based on the driving information at the current time (i.e., the initial time); S3: The target speed planning curve is sent out. S4: After a preset time interval is reached, the dispatch controller 101 updates the driving information of the autonomous vehicles within the dispatch range through the information acquisition unit 1011 and the vehicle communicator 102. The update includes adding newly entered vehicles and removing vehicles that have left the dispatch range. Then, the process returns to step S2 to form rolling optimization and continuous dispatch. S5-S6: When an autonomous vehicle leaves the dispatch range of the intersection, the dispatch controller 101 terminates the communication connection with the autonomous vehicle and stops dispatching the vehicle.
[0014] The speed optimization method used by the intersection autonomous vehicle speed optimization unit 1012 provided in this application embodiment is as follows: Figure 3 As shown, it may include steps S301 to S305.
[0015] S301: Determine the speed planning function for each autonomous vehicle managed by the scheduling controller within the target time period to be optimized; wherein the speed planning function includes driving planning variables for describing the autonomous vehicle passing through the intersection.
[0016] It should be noted that the intersection can be any type of intersection, such as a three-way intersection, a four-way intersection, or a five-way intersection. The target time period to be optimized is the time window required for the autonomous vehicle to pass through the intersection. In practical applications, this time window can be estimated based on a preset minimum driving speed and the intersection length. For example, if the minimum driving speed is 20 km / h and the intersection length is 100 meters (m), then the maximum time for the vehicle to pass through the intersection is approximately 18 seconds (s). That is, within approximately 18 seconds after the current moment (initial moment), vehicles within the scheduling range can complete the passage through the intersection. In this case, the target time period to be optimized can be set as an 18-second time interval after the current moment. The speed planning function can be understood as a time-varying driving speed function planned for the autonomous vehicle within the target time period.
[0017] In some embodiments, the speed planning function includes at least one sub-speed planning function; each sub-speed planning function includes a first constant speed driving stage, a variable speed driving stage, and a second constant speed driving stage; the driving planning variables include a first time variable, an acceleration variable, and a second time variable. Based on this, determining the speed planning function corresponding to each autonomous vehicle managed by the scheduling controller within the target time period to be optimized may include: for each sub-speed planning function of each autonomous vehicle, determining a first speed expression corresponding to the first constant speed driving stage based on the initial driving speed; or, determining the first speed expression based on the driving speed corresponding to the second constant speed driving stage of the previous sub-speed planning function; wherein, the first constant speed driving stage is located between the initial time and the first time variable corresponding to the sub-speed planning function, or located between the second time variable corresponding to the previous sub-speed planning function and the first time variable corresponding to the sub-speed planning function; based on the first speed expression, the first time variable, and the... The acceleration variable and the second time-moment variable are used to determine the second speed expression corresponding to the variable-speed driving phase; wherein, the variable-speed driving phase is located between the first time-moment variable corresponding to the sub-speed planning function and the second time-moment variable corresponding to the sub-speed planning function; based on the driving speed of the autonomous vehicle after the variable-speed driving phase ends and the second time-moment variable, the third speed expression corresponding to the second constant-speed driving phase is determined; wherein, the second constant-speed driving phase is located after the second time-moment variable corresponding to the sub-speed planning function; based on the first speed expression, the second speed expression, and the third speed expression, the sub-speed planning function is determined; based on each sub-speed planning function, the speed planning function corresponding to each autonomous vehicle is determined.
[0018] It should be noted that the speed planning functions for all autonomous vehicles within an intersection are identical in form, differing only in their specific parameters. Therefore, the following explanation uses only the speed planning function and parameters of any one autonomous vehicle as an example; other autonomous vehicles can be understood by referring to this example.
[0019] The above sub-speed planning function can be in the form of a three-segment piecewise function, with each sub-speed planning function corresponding to one speed adjustment. When the traffic flow scenario is complex, a single speed adjustment may not be sufficient to maintain the distance between all autonomous vehicles at the intersection within the preset safe range during the target time period. Therefore, multiple speed adjustments can be performed. Based on this, the sub-speed planning function corresponding to the first speed adjustment can be expressed as formula (1).
[0020] (1).
[0021] in, For variables at the first moment; For the second time step; t =0 indicates the initial time of the target time period to be optimized; Let be the acceleration variable, where the value of the acceleration variable can be positive or negative.
[0022] Formula (1) above means: In t < t i1 At that time, driverless vehicles i The initial driving speed is entered into the dispatch controller. The vehicle is traveling at a constant speed; this is the first stage of constant speed travel. The first expression for this stage is: .exist At that time, driverless vehicles i driving speed Based on acceleration Acceleration or deceleration occurs during this phase, which is the gear shifting phase. The second expression at this time is... .exist At that time, driverless vehicles i The speed after gear shift The vehicle is traveling at a constant speed; this is the second stage of constant speed travel. The third expression at this point is: When using the above velocity planning method, subsequent optimization only requires adjustments to the variables. , , By solving this problem, the autonomous vehicle can be obtained. i The target speed optimization curve within the target time period will be further explained later.
[0023] For the reasons mentioned above, if multiple speed adjustments are required, the sub-speed planning function corresponding to the second speed adjustment can be expressed as formula (2).
[0024] (2).
[0025] Among them, the newly introduced variables , , The physical meaning of is consistent with the corresponding parameter in formula (1), and will not be repeated here. The formula for adjusting the speed more times is similar to that in formula (2).
[0026] It should be noted that variables , , and variables , , These are the driving planning variables included in the aforementioned sub-speed planning function.
[0027] In other embodiments, the sub-velocity planning function may also take other functional forms, such as the S-shaped (Sigmoid) function shown in formula (3).
[0028] (3).
[0029] Among them, variables , , The physical meaning is the same as described above, and will not be repeated here; the coefficients in the formula can be adjusted according to the requirements.
[0030] S302: Obtain the driving information of each unmanned vehicle at the initial time within the target time period to be optimized; wherein, the driving information includes at least the initial position vector, the initial driving speed, and the driving route.
[0031] Before performing optimization, it is necessary to obtain the driving information of each autonomous vehicle within the intersection area at the initial time of the target time period to be optimized. The driving information can be provided by the autonomous vehicle and reported to the dispatch controller 101 through the vehicle communication device 102; at the same time, the information acquisition unit 1011 of the dispatch controller 101 can also independently obtain the vehicle's position vector and driving speed to verify and supplement the reported information, so as to improve the integrity and reliability of the information.
[0032] S303: For each sampling time, a sub-objective optimization function is constructed based on a preset safety distance, the initial position vector of each autonomous vehicle, the speed planning function, and the driving route; wherein, the sub-objective optimization function includes the sum of the conflict functions between any two autonomous vehicles waiting to pass through the intersection at the sampling time, and the function value of the conflict function is used to measure the safety status of the corresponding vehicle at the sampling time.
[0033] It should be noted that the speed optimization method in this application aims to enable autonomous vehicles to pass through intersections with zero waiting time while ensuring safety. Therefore, it is necessary to uniformly optimize all autonomous vehicles within the intersection area at all sampling times. In practical applications, various functions within the target time period to be optimized can remain consistent; therefore, the following only considers arbitrary sampling times. T The example given is 0. The number of sampling times within the target time period to be optimized can be set according to factors such as control accuracy and computing resources.
[0034] It should be understood that the steps for constructing the conflict function and sub-objective optimization function between any two autonomous vehicles within the intersection area can be the same. For ease of explanation, the following will only use any first and second autonomous vehicle as examples to illustrate the construction process of their conflict function and sub-objective optimization function.
[0035] In some embodiments, such as Figure 4 As shown, S303 may specifically include the following steps 401 to 405.
[0036] S401: For any first unmanned vehicle and second unmanned vehicle in the intersection, determine the distance expression of the first unmanned vehicle and the second unmanned vehicle at the sampling time based on the first initial position vector, the first velocity planning function and the first driving route corresponding to the first unmanned vehicle, and the second initial position vector, the second velocity planning function and the second driving route corresponding to the second unmanned vehicle.
[0037] In some embodiments, step S401 may further include: determining a first position vector of the first autonomous vehicle at the sampling time based on a first initial position vector, a first velocity planning function, and a first driving route corresponding to the first autonomous vehicle; wherein the first initial position vector is the position vector of the first autonomous vehicle at the initial time of the target time period to be optimized; determining a second position vector of the second autonomous vehicle at the sampling time based on a second initial position vector, a second velocity planning function, and a second driving route corresponding to the second autonomous vehicle; wherein the second initial position vector is the position vector of the second autonomous vehicle at the initial time of the target time period to be optimized; and determining a distance expression based on the first position vector and the second position vector; wherein the distance expression is the absolute value of the difference between the first position vector and the second position vector.
[0038] It should be noted that the first position vector of the first autonomous vehicle and the second position vector of the second autonomous vehicle are calculated in the same way. Let the first autonomous vehicle be denoted as... i The vector function corresponding to its driving route is denoted as The position vector is denoted as Then the first unmanned vehicle at the sampling time T 0's first position vector It can be calculated using formula (4).
[0039] (4).
[0040] in, For the first driverless vehiclei The first initial position vector; For the first driverless vehicle i The driving speed determined based on the first speed planning function; This represents the time period from the initial moment of the target time period to the sampling moment. T The cumulative driving distance is 0.
[0041] Similarly, the second driverless vehicle (denoted as...) can be obtained. j At the sampling time T 0's second position vector As shown in formula (5).
[0042] (5).
[0043] in, For the second driverless vehicle j The vector function corresponding to the driving route; For the second driverless vehicle j The second initial position vector; For the second driverless vehicle j The driving speed is determined based on the second speed planning function.
[0044] Then, based on the first position vector and the second position vector, the first unmanned vehicle is determined. i With the second driverless vehicle j At sampling time T Distance of 0 As shown in formula (6).
[0045] (6).
[0046] S402: Determine the first driverless vehicle based on the first driving route and the second driving route. i With the second driverless vehicle j The first overlapping sub-function between the two vehicles; wherein, the first overlapping sub-function is used to characterize whether the driving routes of the two vehicles intersect or overlap. It can be expressed as formula (7).
[0047] (7).
[0048] When the first driverless vehicle i First driving route s i With the second driverless vehicle j Second driving route s jWhen there is intersection or overlap, the value of the first overlapping sub-function is 1; otherwise, the value of the first overlapping sub-function is 0.
[0049] S403: Based on the first driving direction corresponding to the first unmanned vehicle and the second driving direction corresponding to the second unmanned vehicle, determine the first unmanned vehicle. i With the second driverless vehicle j The second overlapping sub-function; wherein, the second overlapping sub-function is used to characterize whether the driving directions of the two vehicles intersect. The second overlapping sub-function It can be expressed as formula (8).
[0050] (8).
[0051] in, For the first driverless vehicle i At any moment t The driving azimuth angle; For the second driverless vehicle j At any moment t The driving azimuth angle. When the two vehicles are not traveling in completely opposite directions, that is... When the condition is met, the value of the second overlapping subfunction is 1; otherwise, the value of the second overlapping subfunction is 0.
[0052] S404: Based on the preset safety distance, the distance expression, the first overlapping sub-function, and the second overlapping sub-function, determine the conflict function corresponding to the first unmanned vehicle and the second unmanned vehicle at the sampling time.
[0053] In some embodiments, such as Figure 5 As shown, S404 may include the following steps.
[0054] S501: Determine the safety sub-function based on the distance expression and the preset safety distance.
[0055] S501 may include: taking the distance expression ( ) and the preset safety distance (denoted as d The difference between 0 and 0 is defined as the first safety margin ( Alternatively, the quotient of the distance expression and the preset safety distance can be defined as the second safety margin. ), and construct the security sub-function based on the first security margin or the second security margin.
[0056] In some embodiments, the safety subfunction may take the form of a power function constructed based on a first safety margin. For example, it may be: .
[0057] In other embodiments, the safety subfunction may take the form of an exponential function constructed based on the first safety margin. For example, it may be: Each coefficient can be adjusted according to actual control requirements.
[0058] In other embodiments, the safety subfunction may also take the form of a Logistic function constructed based on the first safety margin. For example, it may be: The correlation coefficient can also be adjusted according to actual needs.
[0059] In other embodiments, the safety subfunction may take the form of a Gaussian function constructed based on the second safety margin. For example, it may be: The correlation coefficient can be adjusted according to actual needs.
[0060] S502: The product of the security sub-function, the first overlapping sub-function, and the second overlapping sub-function is determined as the conflict function.
[0061] Based on the above safety sub-functions, various conflict function forms can be obtained, as shown in formulas (9)-(12).
[0062] (9).
[0063] (10).
[0064] (11).
[0065] (12).
[0066] in, For the first driverless vehicle i With the second driverless vehicle j Between t The conflict function at each time step. In practical applications, the appropriate conflict function form can be selected based on factors such as control accuracy and computational cost.
[0067] S405: Regarding the sampling time T The conflict functions of all vehicles at time 0 are summed to obtain the sub-objective optimization function corresponding to that sampling time.
[0068] Specifically, the sampling time T The sub-objective optimization function for 0 can be expressed as: .
[0069] S304: Based on the sub-objective optimization function corresponding to each sampling time, construct the target optimization function within the target time period to be optimized.
[0070] After obtaining the sub-objective optimization function at each sampling time, the sub-objective optimization functions at all sampling times within the target time period to be optimized are accumulated to obtain the target optimization function, which can be expressed as formula (13) for example.
[0071] (13).
[0072] In practical applications, to ensure that driving speeds meet regulatory requirements, improve passenger comfort, and achieve zero-wait passage through intersections, the driving speeds of each autonomous vehicle within the intersection, as well as the acceleration and time variables in the driving planning variables, are limited to a reasonable range. These limitations are the preset constraints. Specifically, autonomous vehicles... i The corresponding preset constraints can be represented by formulas (14)-(16).
[0073] (14).
[0074] (15).
[0075] (16).
[0076] in, This is the minimum passing speed, used to ensure vehicles can pass through intersections with zero waiting time and maintain high traffic efficiency. Other preset constraints for autonomous vehicles can be set in the same manner as described above, and will not be repeated here. For example, 20 km / h is acceptable. A speed of 60 km / h is acceptable. -2.2 m / s is acceptable. 2 , 2.2 m / s is acceptable. 2 .
[0077] S305: Under the condition of satisfying the preset constraints, the objective optimization function is minimized to obtain the target value of the driving planning variable in the speed planning function corresponding to each autonomous vehicle; for each autonomous vehicle, the target value is substituted into the speed planning function to obtain the target speed planning curve; wherein, the target speed planning curve is used to optimize the driving speed of the autonomous vehicle through the intersection.
[0078] Specifically, firstly, under the premise of satisfying the preset constraints, the objective optimization function is minimized, i.e. min (17).
[0079] This yields the target values of the driving planning variables for each autonomous vehicle when the objective optimization function reaches its minimum. The purpose of minimization is to ensure that the autonomous vehicle maintains the safe distance throughout its journey. The minimization algorithm may include, but is not limited to, interior point methods, projected gradient methods, augmented Lagrange multiplication, simulated annealing, and genetic algorithms.
[0080] Then, the target value of the driving planning variable is substituted into the corresponding speed planning function to obtain the target speed planning curve for the corresponding autonomous vehicle. The dispatch controller 101 generates control commands based on the target speed planning curve and sends them to the corresponding autonomous vehicle to optimize its driving speed, thereby enabling each autonomous vehicle to pass through the intersection safely, with zero waiting time and high efficiency.
[0081] To facilitate understanding of this application, the following will be combined with Figures 6 to 18 The verification results shown illustrate the effectiveness of the speed optimization method for unmanned vehicles passing through intersections (hereinafter referred to as the "intelligent scheduling system") proposed in this application.
[0082] Figure 6 The paper presents a comparison of the time required for traffic flow to completely cross an intersection under different intersection types and vehicle numbers, using a traditional traffic light control system and the intelligent dispatching system proposed in this application. Here, "all vehicles have passed the 60-meter boundary outside the center point of the intersection" is defined as the traffic flow completely crossing the intersection. The two values in parentheses in the figure represent the passage time under the traffic light control system and the intelligent dispatching system, respectively. Three independent tests were conducted for each intersection type and vehicle number combination; the initial driving information (including position vector, speed, and route) of the vehicles in each test group was generated randomly. In the traffic light control system, the green light duration is 45 seconds, and the autonomous vehicle must travel at a speed of 2.2 m / s after the green light turns on. 2 The acceleration continues to build up to 50 km / h, while at red lights it accelerates at -2.2 m / s². 2 The acceleration decelerates to a stop. An average analysis of the three independent test results for each combination shows that, compared to traditional traffic light control systems, the intelligent dispatching system of this application can increase traffic flow speed by more than four times.
[0083] Figures 7 to 10 This is presented as a group to illustrate the traffic flow distribution and sub-objective optimization function of 18 vehicles traveling at constant speeds along their respective routes in a three-way intersection scenario. It also shows the optimized traffic flow distribution, sub-objective optimization function, and target speed planning curve using the intelligent scheduling system described in this application, under the same initial conditions. Figure 7This image shows the distribution of traffic flow at several moments while vehicles travel at a constant speed along a predetermined route. Black arrows indicate vehicles that are too close together and are in serious conflict. Figure 8 The image shows the optimized traffic flow distribution based on the intelligent dispatch system described in this application; it can be seen that vehicles that were originally in serious conflict have been moved to a safe distance. Figure 9 The sub-objective optimization function at each sampling time: Under the condition of uniform speed travel, the sub-objective optimization function at some sampling times showed extremely large values; however, after optimization by the intelligent scheduling system of this application, the sub-objective optimization function was reduced to zero at all sampling times. Figure 10 This refers to the target speed planning curve for each vehicle after optimization by the intelligent scheduling system described in this application. (Summary) Figures 8 to 10 It is evident that the intelligent dispatching system of this application can enable all vehicles to pass through the three-way intersection at high speed while ensuring a safe distance.
[0084] Figures 11 to 14 This is presented as a group to illustrate the traffic flow distribution and sub-objective optimization function of 16 vehicles traveling at constant speeds along their own routes in a four-way intersection scenario. It also shows the optimized traffic flow distribution, sub-objective optimization function, and target speed planning curve using the intelligent scheduling system proposed in this application, under the same initial conditions. Figure 11 This image shows the distribution of traffic flow at several moments while vehicles travel at a constant speed along a predetermined route. Black arrows indicate vehicles that are too close together and are in serious conflict. Figure 12 The image shows the traffic flow distribution optimized by the intelligent scheduling method described in this application; it can be seen that vehicles that were originally in serious conflict have been moved to a safe distance. Figure 13 The sub-objective optimization function is defined for each sampling time. Under the condition of uniform speed travel, the sub-objective optimization function exhibits extremely large values at some sampling times. However, after optimization by the intelligent scheduling system of this application, the sub-objective optimization function is reduced to zero at all sampling times. Figure 14 This is the target speed planning curve for each vehicle after intelligent scheduling optimization according to this application. (Summary) Figures 12 to 14 It is evident that the intelligent dispatching system of this application can enable all vehicles to pass through the four-way intersection at high speed while ensuring a safe distance.
[0085] Figures 15 to 18 This is presented as a group to demonstrate the traffic flow distribution and sub-objective optimization function of 30 vehicles traveling at constant speeds along their own routes in a five-way intersection scenario. It also shows the optimized traffic flow distribution, sub-objective optimization function, and target speed planning curve using the intelligent scheduling system proposed in this application, under the same initial conditions. Figure 15 This image shows the distribution of traffic flow at several moments while vehicles travel at a constant speed along a predetermined route. Black arrows indicate vehicles that are too close together and are in serious conflict. Figure 16The image shows the traffic flow distribution optimized by the intelligent scheduling method described in this application; it can be seen that vehicles that were originally in serious conflict have been moved to a safe distance. Figure 17 The sub-objective optimization function is defined for each sampling time. Under the condition of uniform speed travel, the sub-objective optimization function exhibits extremely large values at some sampling times. However, after optimization by the intelligent scheduling system of this application, the sub-objective optimization function is reduced to zero at all sampling times. Figure 18 This refers to the target speed planning curve for each vehicle after optimization by the intelligent scheduling system described in this application. (Summary) Figure 16 and Figure 18 It is evident that the intelligent dispatching system of this application can enable all vehicles to pass through the five-way intersection at high speed while ensuring a safe distance.
[0086] like Figure 19 As shown in the figure, this application embodiment also provides a speed optimization device 190 for unmanned vehicles at intersections, which may include an information acquisition unit 1011 and a speed optimization unit 1012.
[0087] The information acquisition unit 1011 is used to acquire the driving information of each unmanned vehicle at the initial moment within the target time period to be optimized; wherein, the driving information includes at least the initial position vector, the initial driving speed, and the driving route.
[0088] The speed optimization unit 1012 is used to determine the speed planning function corresponding to each autonomous vehicle managed by the scheduling controller within the target time period to be optimized; wherein the speed planning function includes driving planning variables describing the autonomous vehicle passing through the intersection; for each sampling time, a sub-objective optimization function is constructed based on a preset safety distance, the initial position vector of each autonomous vehicle, the speed planning function, and the driving route; wherein the sub-objective optimization function includes the sum of conflict functions between any two autonomous vehicles at the sampling time, and the function value of the conflict function is used to measure the safety status of the corresponding vehicle at the sampling time; based on the sub-objective optimization function corresponding to each sampling time, a target optimization function is constructed within the target time period to be optimized; under the condition of satisfying preset constraints, the target optimization function is minimized to obtain the target value of the driving planning variable in the speed planning function corresponding to each autonomous vehicle; for each autonomous vehicle, the target value is substituted into the speed planning function to obtain a target speed planning curve; wherein the target speed planning curve is used to optimize the driving speed of the autonomous vehicle passing through the intersection.
[0089] In some embodiments, the driving planning variables include a first time-of-flight variable, an acceleration variable, and a second time-of-flight variable; the speed planning function includes at least one sub-speed planning function; each sub-speed planning function includes a first constant-speed driving stage, a variable-speed driving stage, and a second constant-speed driving stage; the speed optimization unit 1012 may be specifically used to: for each sub-speed planning function of each autonomous vehicle, determine a first speed expression corresponding to the first constant-speed driving stage based on the initial driving speed; or, determine the first speed expression based on the driving speed obtained from the second constant-speed driving stage of the previous sub-speed planning function; wherein, the first constant-speed driving stage is: between the initial time and the first time-of-flight variable corresponding to the sub-speed planning function, or between the second time-of-flight variable corresponding to the previous sub-speed planning function and the sub-speed planning function. The first time-of-flight variables are defined as follows: A second speed expression is determined based on the first speed expression, the first time-of-flight variable, the acceleration variable, and the second time-of-flight variable; wherein the variable-of-flight phase is located between the first time-of-flight variable corresponding to the sub-speed planning function and the second time-of-flight variable; a third speed expression is determined based on the driving speed of the autonomous vehicle after the variable-of-flight phase and the second time-of-flight variable; wherein the second uniform speed driving phase is located after the second time-of-flight variable corresponding to the sub-speed planning function; the sub-speed planning function is determined based on the first speed expression, the second speed expression, and the third speed expression; and the speed planning function corresponding to each autonomous vehicle is determined based on each sub-speed planning function.
[0090] In some embodiments, the speed optimization unit 1012 is specifically configured to: for any first autonomous vehicle and a second autonomous vehicle at the intersection, based on the first initial position vector, the first speed planning function, and the first driving route corresponding to the first autonomous vehicle, and the second initial position vector, the second speed planning function, and the second driving route corresponding to the second autonomous vehicle, determine the distance expression between the first autonomous vehicle and the second autonomous vehicle at the sampling time; determine a first overlap sub-function between the first autonomous vehicle and the second autonomous vehicle according to the first driving route and the second driving route, wherein the first overlap sub-function is used to characterize whether the driving routes of the two vehicles intersect or overlap; determine a second overlap sub-function between the first autonomous vehicle and the second autonomous vehicle according to the first driving direction corresponding to the first autonomous vehicle and the second driving direction corresponding to the second autonomous vehicle, wherein the second overlap sub-function is used to characterize whether the driving directions of the two vehicles intersect; further determine the conflict function corresponding to the first autonomous vehicle and the second autonomous vehicle at the sampling time based on the preset safety distance, the distance expression, the first overlap sub-function, and the second overlap sub-function; and sum the conflict functions between any two autonomous vehicles to obtain the sub-objective optimization function at the sampling time.
[0091] In some embodiments, the speed optimization unit 1012 is further configured to: determine a first position vector of the first unmanned vehicle at the sampling time based on a first initial position vector, a first speed planning function, and a first driving route corresponding to the first unmanned vehicle; wherein the first initial position vector is the position vector of the first unmanned vehicle at the initial time of the target time period to be optimized; determine a second position vector of the second unmanned vehicle at the sampling time based on a second initial position vector, a second speed planning function, and a second driving route corresponding to the second unmanned vehicle; wherein the second initial position vector is the position vector of the second unmanned vehicle at the initial time of the target time period to be optimized; and determine a distance expression based on the first position vector and the second position vector; wherein the distance expression is the absolute value of the difference between the first position vector and the second position vector.
[0092] In some embodiments, the speed optimization unit 1012 is further configured to: determine a safety sub-function based on the distance expression and the preset safety distance; and determine the product of the safety sub-function, the first overlapping sub-function, and the second overlapping sub-function as the conflict function.
[0093] In some embodiments, the speed optimization unit 1012 is further configured to: determine the difference between the distance expression and the preset safety distance as a first safety margin, or determine the quotient between the distance expression and the preset safety distance as a second safety margin; and determine the safety sub-function based on the first safety margin or the second safety margin.
[0094] In some embodiments, the speed optimization unit 1012 is further configured to: determine the power function constructed based on the first safety margin as the safety sub-function; or, determine the exponential function constructed based on the first safety margin as the safety sub-function; or, determine the Logistic function constructed based on the first safety margin as the safety sub-function; or, determine the Gaussian function constructed based on the second safety margin as the safety sub-function.
[0095] It should be noted that the autonomous driving speed optimization device provided in this application embodiment is used to implement the aforementioned speed optimization method. Since its implementation steps have been described in detail above in conjunction with the method embodiments, they will not be repeated here.
[0096] Based on the same inventive concept, such as Figure 20 As shown in the illustration, this application embodiment also provides a scheduling controller 101, which may include a memory 2001 and a processor 2002. The memory 2001 stores computer-executable instructions or computer programs; the processor 2002, when executing the computer-executable instructions or computer programs stored in the memory 2001, implements the steps of the speed optimization method described in this application embodiment. In some embodiments, the processor 2002 may integrate or implement the aforementioned speed optimization unit 1012. The processor 2002 may be, but is not limited to, a central processing unit (CPU); the memory 2001 may be a read-only memory (ROM), random access memory (RAM), flash memory, etc.
[0097] Similarly, embodiments of this application also provide a computer-readable storage medium storing computer-executable instructions or computer programs thereon, which, when executed by a processor, implement the steps of the speed optimization method described in the embodiments of this application.
[0098] This application also provides a computer program product, which includes computer-executable instructions or a computer program. When the computer-executable instructions or the computer program are executed by a processor, they are used to implement the speed optimization method described in this application.
[0099] It should be noted that the above embodiments of the device, storage medium, and scheduling controller are based on the same inventive concept as the foregoing method embodiments, and their implementation principles and beneficial effects correspond accordingly. For technical details not described in detail in the embodiments of the device, storage medium, and scheduling controller, please refer to the relevant descriptions in the foregoing method embodiments for understanding.
[0100] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, and improvements made within the spirit and scope of this application should be included within the scope of protection of this application.
Claims
1. A speed optimization method for unmanned vehicles passing through intersections with zero waiting time, characterized in that, The method, applied to the dispatch controller at the intersection, includes: Determine the speed planning function for each autonomous vehicle managed by the scheduling controller within the target time period to be optimized; wherein the speed planning function includes driving planning variables for describing the autonomous vehicle passing through the intersection; Obtain the driving information of each autonomous vehicle at the initial time within the target time period to be optimized; wherein, the driving information includes at least the initial position vector, the initial driving speed, and the driving route; For each sampling time, a sub-objective optimization function is constructed based on a preset safety distance, the initial position vector of each autonomous vehicle, the speed planning function, and the driving route; wherein, the sub-objective optimization function includes the sum of the conflict functions between any two autonomous vehicles waiting to pass through the intersection at the sampling time, and the function value of the conflict function is used to measure the safety status of the corresponding vehicle at the sampling time; Based on the sub-objective optimization function at each sampling time, a target optimization function is constructed within the target time period to be optimized; Under the condition of satisfying the preset constraints, the objective optimization function is minimized to obtain the target value of the driving planning variable in the speed planning function corresponding to each autonomous vehicle. For each autonomous vehicle, the target value is substituted into the speed planning function to obtain the target speed planning curve; wherein, the target speed planning curve is used to optimize the driving speed of the autonomous vehicle through the intersection.
2. The method according to claim 1, characterized in that, The driving planning variables include a first-time variable, an acceleration variable, and a second-time variable; the speed planning function includes at least one sub-speed planning function; each sub-speed planning function includes a first constant-speed driving phase, a variable-speed driving phase, and a second constant-speed driving phase. The step of determining the speed planning function for each autonomous vehicle managed by the scheduling controller within the target time period to be optimized includes: For each sub-vehicle speed planning function of each autonomous vehicle, a first speed expression corresponding to the first uniform speed driving stage is determined based on the initial driving speed; or, the first speed expression is determined based on the driving speed obtained in the second uniform speed driving stage of the previous sub-vehicle speed planning function; wherein, the first uniform speed driving stage is: between the initial time and the first time variable corresponding to the sub-vehicle speed planning function, or between the second time variable corresponding to the previous sub-vehicle speed planning function and the first time variable corresponding to the sub-vehicle speed planning function; Based on the first speed expression, the first time-of-flight variable, the acceleration variable, and the second time-of-flight variable, a second speed expression corresponding to the variable-speed driving stage is determined; wherein, the variable-speed driving stage is located between the first time-of-flight variable corresponding to the sub-speed planning function and the second time-of-flight variable corresponding to the sub-speed planning function; Based on the driving speed of the autonomous vehicle after the variable speed driving phase and the second time-moment variable, the third speed expression corresponding to the second constant speed driving phase is determined; wherein, the second constant speed driving phase is located after the second time-moment variable corresponding to the sub-speed planning function; The sub-velocity planning function is determined based on the first velocity expression, the second velocity expression, and the third velocity expression; Based on each sub-velocity planning function, the velocity planning function corresponding to each autonomous vehicle is determined.
3. The method according to claim 1, characterized in that, The sub-objective optimization function, constructed based on a preset safety distance, the initial position vector of each autonomous vehicle, the velocity planning function, and the driving route, includes: For any first unmanned vehicle and second unmanned vehicle in the intersection, the distance expression of the first unmanned vehicle and the second unmanned vehicle at the sampling time is determined according to the first initial position vector, the first velocity planning function and the first driving route corresponding to the first unmanned vehicle, and the second initial position vector, the second velocity planning function and the second driving route corresponding to the second unmanned vehicle. Based on the first driving route and the second driving route, a first overlapping sub-function is determined between the first autonomous vehicle and the second autonomous vehicle; wherein, the first overlapping sub-function is used to characterize whether the driving routes between the first autonomous vehicle and the second autonomous vehicle intersect or overlap; Based on the first driving direction corresponding to the first autonomous vehicle and the second driving direction corresponding to the second autonomous vehicle, a second overlapping sub-function is determined between the first autonomous vehicle and the second autonomous vehicle; wherein, the second overlapping sub-function is used to characterize whether the driving directions of the first autonomous vehicle and the second autonomous vehicle intersect; Based on the preset safety distance, the distance expression, the first overlapping sub-function, and the second overlapping sub-function, determine the conflict function corresponding to the first unmanned vehicle and the second unmanned vehicle at the sampling time; The sub-objective optimization function corresponding to the sampling time is obtained by summing the conflict functions between any two autonomous vehicles at the sampling time.
4. The method according to claim 3, characterized in that, The step of determining the distance expression between the first autonomous vehicle and the second autonomous vehicle at the sampling time based on the first initial position vector, the first velocity planning function, and the first driving route corresponding to the first autonomous vehicle, and the second initial position vector, the second velocity planning function, and the second driving route corresponding to the second autonomous vehicle, includes: Based on the first initial position vector, the first velocity planning function, and the first driving route corresponding to the first unmanned vehicle, the first position vector of the first unmanned vehicle at the sampling time is determined; wherein, the first initial position vector is the position vector corresponding to the first unmanned vehicle at the initial time of the target time period to be optimized; Based on the second initial position vector, the second velocity planning function, and the second driving route corresponding to the second unmanned vehicle, the second position vector of the second unmanned vehicle at the sampling time is determined; wherein, the second initial position vector is the position vector corresponding to the second unmanned vehicle at the initial time of the target time period to be optimized; The distance expression is determined based on the first position vector and the second position vector; wherein the distance expression is the absolute value of the difference between the first position vector and the second position vector.
5. The method according to claim 3, characterized in that, The step of determining the conflict function between the first autonomous vehicle and the second autonomous vehicle at the sampling time based on the preset safety distance, the distance expression, the first overlapping sub-function, and the second overlapping sub-function includes: Based on the distance expression and the preset safety distance, determine the safety sub-function; The product of the security sub-function, the first overlapping sub-function, and the second overlapping sub-function is determined as the conflict function.
6. The method according to claim 5, characterized in that, The method for determining the safety sub-function based on the distance expression and the preset safety distance includes: The difference between the distance expression and the preset safety distance is determined as the first safety margin, or the quotient between the distance expression and the preset safety distance is determined as the second safety margin. The safety sub-function is determined based on either the first safety margin or the second safety margin.
7. The method according to claim 6, characterized in that, Determining the safety sub-function based on the first safety margin or the second safety margin includes: The power function constructed based on the first safety margin is determined as the safety sub-function; or... The exponential function constructed based on the first safety margin is determined as the safety sub-function; or, The Logistic function constructed based on the first safety margin is determined as the safety sub-function; or... The Gaussian function constructed based on the second safety margin is determined as the safety sub-function.
8. A speed optimization device for driverless vehicles to pass through intersections with zero waiting time, characterized in that, The device is applied to the dispatch controller of the intersection, and the device includes: An information acquisition unit is used to acquire the driving information of each of the autonomous vehicles at the initial moment within the target time period to be optimized; wherein the driving information includes at least the initial position vector, the initial driving speed, and the driving route; A speed optimization unit is used to determine the speed planning function corresponding to each autonomous vehicle managed by the scheduling controller within the target time period to be optimized. The speed planning function includes driving planning variables describing the autonomous vehicle's passage through the intersection. For each sampling time, a sub-objective optimization function is constructed based on a preset safety distance, the initial position vector of each autonomous vehicle, the speed planning function, and the driving route. The sub-objective optimization function includes the sum of conflict functions between any two autonomous vehicles at the sampling time, where the function value of the conflict function measures the safety status of the corresponding vehicle at the sampling time. Based on the sub-objective optimization functions corresponding to each sampling time, a target optimization function is constructed within the target time period to be optimized. Under preset constraints, the target optimization function is minimized to obtain the target value of the driving planning variable in the speed planning function corresponding to each autonomous vehicle. For each autonomous vehicle, the target value is substituted into the speed planning function to obtain a target speed planning curve. The target speed planning curve is used to optimize the driving speed of the autonomous vehicle passing through the intersection.
9. A scheduling controller, characterized in that, include: Memory is used to store executable instructions or computer programs. A processor for executing computer-executable instructions or computer programs stored in the memory to implement the speed optimization method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions or computer programs for implementing the speed optimization method according to any one of claims 1 to 7 when executed by a processor.