A method for optimizing intersection delay time and guiding speed of intelligent connected vehicles

Abstract

The present invention belongs to the field of intelligent transportation, and specifically relates to an intersection delay time optimization and speed guidance method for intelligent networked vehicles. The present invention includes: constructing a Petri net model; numbering vehicles that are about to arrive at the intersection area and generating an intersection vehicle passage sequence according to the passage process of each vehicle passing through each road right point on its path, and decoding it into a transition sequence; detecting the current transition sequence, and repairing it if a deadlock phenomenon exists; obtaining the input information of the vehicle, optimizing the Delaytime, and finally adjusting the speed. Provide the optimal driving speed and passing time recommendations for each intelligent networked vehicle. The vehicle can pass through the intersection at the optimal speed while ensuring safety and avoiding deadlock, effectively reducing waiting time and alleviating traffic congestion.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent transportation, and in particular relates to an intersection delay time optimization and speed guidance method for an intelligent networked vehicle. Background Art

[0002] Intelligent connected vehicle intersections, as core nodes in the urban road traffic network, are directly related to the smoothness and safety of the entire traffic system through their efficient operation. These intersections are not only key hubs connecting different sections of the road, but also often become areas prone to traffic accidents due to high-density traffic flow. Traditionally, intersection management relies on signal control methods such as timing control, induction control, and adaptive control. However, in the face of increasingly complex traffic environments and growing traffic demands, these traditional methods are often difficult to adapt in real time, resulting in unnecessary extension of vehicle waiting time, increased traffic congestion, and even deadlock in extreme cases.

[0003] With the rapid development of perception technology, communication technology, positioning technology and assisted driving technology, intelligent networked vehicles have emerged, bringing changes to intersection management. In the intelligent networked environment, intersection control has achieved a qualitative leap and has become more intelligent, flexible and efficient. Through the design of signal-free intersections, the efficient information exchange between the main control center and the sub-control center is used to achieve real-time scheduling and precise control of intelligent networked vehicles. Through advanced perception technology, the system can obtain key data such as traffic information, vehicle location and speed in all directions of the intersection in real time, and conduct comprehensive analysis combined with intelligent algorithms to provide each intelligent networked vehicle with the optimal driving speed and passing time recommendations. In this way, the vehicle can pass through the intersection at the optimal speed while ensuring safety and avoiding deadlock, effectively reducing waiting time and alleviating traffic congestion. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a method for optimizing the intersection delay time and speed guidance of an intelligent connected vehicle, which can effectively reduce the waiting time of vehicles at intersections and avoid unnecessary traffic delays, thereby alleviating traffic congestion and improving traffic efficiency, thereby maximizing the utilization of intersection road rights and resources and providing a set of reliable speed regulation strategies for intersection vehicle traffic under the intelligent transportation system.

[0005] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solution: a method for optimizing intersection delay time and guiding speed of an intelligent networked vehicle, comprising the following steps:

[0006] Step 1) (Modeling): Construct a Petri net model of the unsignalized intersection system (P, T, F, M) based on the spatial resource allocation within the intersection and the dynamic occupation of these resources when vehicles continuously cross the intersection.0 ), where P is a finite set of places; T is a finite set of transitions; F = (P × T) ∪ (T × P) is a directed arc set; M 0 Initial logo set.

[0007] Step 2): Obtain vehicle information in the unsignalized intersection area, including the number of vehicles, speed, and lane number.

[0008] Step 3) (Encoding and decoding): Since the encoding reflects the vehicle's passage sequence, the position and speed of the particles, the vehicle's passage sequence is encoded according to the characteristics of the intelligent connected vehicle system and the particle swarm algorithm. The specific steps are as follows:

[0009] Step 3-1): The vehicle number is two integers AB, A represents the vehicle driving path, and B represents the vehicle driving priority.

[0010] Step 3-2): Sort the vehicles from small to large, and then expand the sequence according to the number of transitions that each vehicle needs to pass, and the expanded vehicle sequence is Car_sequence.

[0011] Step 3-3): Decode the expanded vehicle traffic sequence Car_sequence into a transition sequence Tran.

[0012] Step 4) (deadlock detection and repair): The transition sequence obtained by decoding in step 3 is not necessarily a feasible pass sequence, and deadlock detection and repair are required. Using the deadlock avoidance algorithm in the prior art, the transition sequence of each particle is detected and repaired to ensure that each transition in the transition sequence can be triggered in sequence. In the process of repairing the particle, the position and speed of the particle remain unchanged.

[0013] Step 5) (Optimize Delaytime): Obtain the repaired particles through step 4, calculate the delay time Delaytime of all particles in each generation, adjust the particle sequence through the particle swarm algorithm, optimize the Delaytime, and finally obtain the best particle sequence and optimal Delaytime in all generations. The specific steps are as follows:

[0014] Step 5-1): Calculate the start_times and end_times list of each particle in each generation of particles, store the start time start_time of each vehicle occupying each resource in start_times, and store the end time end_time in end_times.

[0015] Step 5-2): Determine whether the end_time of the current resource occupied by the vehicle is equal to the start_time of the next resource occupied. If they are the same, then Delaytime = 0, otherwise Delaytime = Delaytime_pre + start_time - end_time. This process continues until all vehicles have finished the determination.

[0016] Step 5-3): Output the optimal particle sequence and its corresponding BestDelaytime in this generation of particle swarm.

[0017] Step 5-4): Use the particle swarm of each generation as the input of the particle swarm algorithm, optimize the current particle swarm through the particle swarm algorithm, and obtain the optimized particle swarm.

[0018] Step 5-5): Obtain particles after each generation of optimization. The optimization content includes the vehicle sequence, the transition sequence corresponding to the vehicle sequence, and the optimal Delaytime value.

[0019] Step 5-6): The particles obtained after optimization are subjected to step 5-1, step 5-2, step 5-3, and step 5-4 again until the last round of iteration is completed, and the optimal particle and its corresponding BestDelaytime after all rounds of iteration are output.

[0020] Step 6) (Adjust speed): After completing the iteration in step 5 and determining the optimal particle and its corresponding optimal delay time (BestD elaytime), the speed model needs to be optimized in order to be closer to the dynamic characteristics of the real world. The current speed setting is limited to two states: uniform speed and static speed, which is more ideal in practical applications. Therefore, it is planned to introduce three speed modes: deceleration, uniform speed, and acceleration, to construct a motion process that is more in line with the laws of physics. The specific steps are as follows:

[0021] Step 6-1): Get the discontinuous vehicle sequence in the optimal particle and store it in the nocontinuoucar list.

[0022] Step 6-2): Find the start time t of discontinuous vehicles occupying the current resource 1 , the end time is t 2 , and the start time t of occupying the next resource 3 , then the delay time (Delay) is t 3 -t 2 .

[0023] Step 6-3): Under the condition that the total time and total distance remain unchanged, the speed is adjusted using parabola fitting, and the adjusted speed value is output.

[0024] Step 6-4): Divide the total time total_time into 100 parts, calculate the speed once at each time point, store the calculated speed in the adjusted_speeds list, and draw the final speed adjustment image.

[0025] As a preferred technical solution of the present invention, step 5-3 outputs the optimal particle sequence and its corresponding BestDelaytime in this generation of particle swarm, and calculates the fitness value Adapt of the particle swarm at the same time. The inertia weight adaptive adjustment function is used for the calculation of Adapt. The specific formula is as follows:

[0026]

[0027] Where w(t) is the inertia weight value in the Gen-th iteration; maxGen is the maximum number of iterations; p(t) is the fitness update rate of the population individuals. update(Gen) is the number of fitness improvements of the population individuals in the Gen-th iteration; Adapt[i] is the fitness value after the Gen-th update, Adapt[i-1] is the fitness value of the Gen-th update, and r i is a random number between [0,1], Adapt best is the optimal fitness value, ND is a random integer between [1,2], Adapt mean is the mean fitness of the particle swarm.

[0028] As a preferred technical solution of the present invention, step 5-4 optimizes the current particle swarm through the particle swarm algorithm to obtain the optimized particle swarm. The specific adaptive formula in the optimization process is as follows:

[0029]

[0030] where f avg is the average fitness value of the particle swarm, f max is the maximum fitness value in the particle swarm, f i is the current particle fitness value, p c1 is 0.8, p c2 is 0.5.

[0031] As a preferred technical solution of the present invention, in step 6-1, the discontinuous vehicle sequence in the optimal particle is stored in the nocontinuoucar list. The specific steps are as follows:

[0032] Calculate whether the end_time of each vehicle occupying the current resource in the optimal particle is equal to the start_time of the next resource to be occupied. If they are the same, the vehicles are running continuously at this time. Otherwise, the vehicles are not running continuously, and the discontinuous vehicle sequence is stored in the nocontinuoucar list.

[0033] As a preferred technical solution of the present invention, in step 6-3, under the condition that the total time and total distance remain unchanged, the speed adjustment is fitted by a parabola, and the specific speed adjustment formula is as follows:

[0034]

[0035] For t ≤ mid_time, if the currently calculated speed < c, then set the current speed speed to the last speed value greater than zero within the time range. For mid_time < t ≤ total_time, if the currently calculated speed > 0, then set the current speed speed = c. The speed1 in the formula refers to the last calculated speed when t ≤ mid_time.

[0036] For the method for optimizing intersection delay time and guiding speed of an intelligent connected vehicle described in the present invention, compared with the prior art by adopting the above technical solutions, the following technical effects are achieved:

[0037] (1) By introducing an inertia weight adaptive adjustment function to calculate the Adapt value, the present invention significantly improves the accuracy and efficiency of fitness calculation when solving specific research topics. The core of this method lies in its ability to dynamically adjust the inertia weight according to the particle behavior and environmental feedback during the search process, so as to more accurately guide the particles to move towards the optimal solution region.

[0038] (2) By introducing an improved adaptive formula, the present invention can quickly obtain the optimal or sub-optimal scheduling strategy, thereby improving the traffic efficiency of intersections.

[0039] (3) The present invention finds the discontinuous sequences of vehicles in the optimal particles, and visualizes the discontinuous vehicles at the same time, revealing the internal logic of the driving process of the delayed vehicles, which facilitates the subsequent verification work.

[0040] (4) By applying the vehicle speed regulation mechanism, the present invention realizes the instant identification and response to potential unreasonable situations during the vehicle driving process. When it is found that the vehicle driving speed does not meet the preset standard or the actual situation, the vehicle speed is adjusted to ensure that the driving of each vehicle can meet the requirements of road safety and traffic fluency. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a flow chart of a method for optimizing intersection delay time and guiding speed of an intelligent connected vehicle according to the present invention.

[0042] Figure 2 It is a Petri net model of a two-way four-lane intersection according to the present invention.

[0043] Figure 3 It is a Gantt chart corresponding to the delayed vehicle scheduling sequence in the specific example of the present invention.

[0044] Figure 4 This is a delayed vehicle speed adjustment diagram in a specific example of the present invention. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present application, some symbols in the embodiments of the present application are first explained to facilitate understanding by those skilled in the art.

[0046] like Figure 1 As shown, a method for optimizing intersection delay time and guiding speed of an intelligent connected vehicle comprises the following steps:

[0047] Step 1) (Modeling): Divide the right-of-way points and resource points at the unsignalized intersection and establish a Petri net model, such as Figure 2 shown.

[0048] Step 2) (Vehicle Coding and Decoding): Number the vehicles that are about to arrive at the intersection area and generate the intersection vehicle traffic sequence according to the order in which each vehicle passes through each right-of-way point on its path. Specifically, there are 16 vehicles in the eight lanes of the two-way four-lane intersection area. The 16 vehicles are numbered according to the rules and shuffled to obtain the vehicle sequence L 0 :(12,13,41,21,33,16,25,22,32,14,34,23,31,15,11,24), expand the total number of places occupied by each vehicle, and 0 The vehicle sequence L obtained after expansion 1 :(12,12,12,12,12,13,13,13,13,13,41,41,41,41,41,21,21,21,21,21,33,33,33,33,33,16,16,16,16,16,25,25,25,25,25,22,22,22,22,32,32,32,32,32,14,14,14,14,14,34,34,34,34,34,23,23,23,23,31,31,31,31,31,15,15,15,15,11,11,11,11,11,24,24,24,24,24). By L 1Get the vehicle pass sequence Pass = (11,11,11,11,11,12,12,12,12,12,41,41,41,41,41,21,21,21,21,31,31,31,31,31,13,13,13,13,22,22,22,22,23,23,23,23,23,32 ,32,32,32,32,14,14,14,14,14,33,33,33,33,33,24,24,24,24,24,34,34,34,34,34,15,15,15,15,15,16,16,16,16,25,25,25,25,25),decode the Pass sequence into a transition sequence T = (t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 41 ,t 42 ,t 43 ,t 44 ,t 45 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13 ,t14 ,t 15 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ).

[0049] Step 3): Get the total number of vehicles Car_total = 16, Carroadnum[i] is the total number of vehicles on the i-th road, Carroadnum[1] = 6, Carroadnum[2] = 5, Carroadnum[3] = 4, Carroadnum[4] = 1, Carroadspeed ij is the speed of the jth car on the ith road at a constant speed, Carroadspeed 11 =15, Carroadspeed 12 =10, Carroadspeed 13 =13, Carroadspeed 14 =8, Carroadspeed 15 =15, Carroadspeed 16 =17, Carroadspeed 21 =5, Carroadspeed 22 =12, Carroadspeed 23 =10, Carroadspeed 24 =15, Carroadspeed 25 =6, Carroadspeed 31=7, Carroadspeed 32 =14, Carroadspeed 33 =15, Carroadspeed 34 =13, Carroadspeed 41 =10, Carroadspeed 42 =11, Carroadspeed 43 =8, Carroadspeed 44 =15, Carroadspeed 45 =13, Carroadspeed 51 =5, Carroadspeed 52 =9, Carroadspeed 53 =6, Carroadspeed 61 =4, Carroadspeed 71 =6, Carroadspeed 72 =7, Carroadspeed 73 =14, Carroadspeed 74 =11, Carroadspeed 81 =10.

[0050] Step 4) (Detect and repair deadlock): By iteratively analyzing the transitions and resources in the Petri net, the deadlock of the model is avoided to ensure that the system is safe and free of deadlock. The specific process is as follows: Transition T = (t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 41 ,t 42 ,t 43 ,t 44 ,t 45 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ), initial identifier M 0 =(6,0,0,0,0,0,0,0,0,0,0,0,5,0,0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1), at the initial identifier M 0Under this condition, the enabling conditions of transitions are checked one by one and whether a deadlock will occur after initiation. If a deadlock will occur after initiation, the enabling transition after T is selected and placed in front of the transition, and T is updated. Otherwise, T is triggered and the current flag is updated. The specific process is as follows: Starting from i=1, determine transition T[1]=t 11 , find the transition t 11 The library r that will be reduced when triggered 10 , at this time M 0 [r 10 ]=1, then the transition t 11 Enable, and trigger t 11 Will not cause deadlock, trigger t 11 , and update the current identifier to M 1 =(5,1,0,0,0,0,0,0,0,0,0,5,0,0,0,0,0,4,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1), i=i+1. After testing, there is no deadlock in this example. If a deadlock occurs, such as: T=(t 11 ,t 41 ,t 12 ,t 71 ,t 11 ,t 31 ,t 72 ,t 73 ,t 71 ,t 72 ,t 74 ,t 42 ,t 13 ,t 12 ,t 43 ,t 41 ,t 32 ,t 75 ,t 73 ,t 74 ,t 14 ,t 33 ,t 71 ,t 72 ,t 11 ,t 42 ,t 41 ,t 73 ,t 71 ,t 72 ,t 71 ,t 15 ,t 13 ,t 76 ,t 14 ,t12 ,t 72 ,t 44 ,t 13 ,t 45 ,t 43 ,t 77 ,t 15 ,t 75 ,t 61 ,t 76 ,t 74 ,t 73 ,t 73 ,t 14 ,t 44 ,t 45 ,t 77 ,t 62 ,t 75 ,t 76 ,t 74 ,t 15 ,t 77 ,t 74 ,t 75 ,t 75 ,t 42 ,t 76 ,t 77 ,t 63 ,t 34 ,t 43 ,t 35 ,t 44 ,t 45 ,t 64 ,t 76 ,t 65 ,t 77 ,t 67 ,t 67 ), initial identification M 0 =(3,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,5,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1), when i=30, T[i]=t 71 When a deadlock occurs, the deadlock can be repaired by using the deadlock avoidance algorithm in the prior art to obtain a set of deadlock-free transition sequences T = (t 11 ,t 41 ,t 12 ,t 71 ,t 11 ,t 31 ,t 72 ,t73 ,t 71 ,t 72 ,t 74 ,t 42 ,t 13 ,t 12 ,t 43 ,t 41 ,t 32 ,t 75 ,t 73 ,t 74 ,t 14 ,t 33 ,t 71 ,t 72 ,t 11 ,t 42 ,t 41 ,t 73 ,t 71 ,t 72 ,t 15 ,t 13 ,t 14 ,t 12 ,t 71 ,t 44 ,t 45 ,t 43 ,t 76 ,t 77 ,t 15 ,t 75 ,t 76 ,t 74 ,t 73 ,t 72 ,t 13 ,t 14 ,t 44 ,t 45,t77 ,t 61 ,t 75 ,t 76 ,t 74 ,t 15 ,t 77 ,t 73 ,t 75 ,t 76 ,t 77 ,t 74 ,t 42 ,t 62 ,t 34 ,t 35 ,t 43 ,t 44 ,t 75 ,t 76 ,t 45 ,t 63 ,t 64 ,t 65 ,t 77,t 66 ,t 67 ), thus obtaining the deadlock-free vehicle passage sequence Pass=(11,41,11,71,12,31,71,71,72,72,71,41,11,12,41,42,31,71,72,72,11,31,73,73,13,42,43,73,74,74,11,12,12,13,75, 41,41,42,71,71,12,72,72,73,74,75,13,13,42,42,72,61,73,73,74,13,73,75,74,74,75,43,61,31,31,43,43,75,75,43,61,61,61,75,61,61).

[0051] Step 5) (Calculate Delaytime): The specific steps are as follows:

[0052] Step 5-1): Calculate the number of particles N = [[12,12,12,12,12,31,31,31,31,31,15,15,15,15,25,25,25,25,25,33,33,33,33,33,14,14,14,14,14,21,21,21,21,21,23,23,23, 23,23,16,16,16,16,16,32,32,32,32,32,34,34,34,34,34,41,41,41,41,4 1,24,24,24,24,24,13,13,13,13,13,22,22,22,22,22,11,11,11,11,11],[t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 31 ,t 32 ,t 33 ,t 34 ,t 35 ,t 41 ,t 42 ,t 43 ,t 44 ,t 45 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15 ,t 21 ,t 22 ,t 23 ,t 24 ,t 25 ,t 11 ,t 12 ,t 13 ,t 14 ,t 15],0] with start_times = [[0,0.5333,1.3333,1.8666,2.4],[0.5333,1.3333,2.1333,2.9333,3.7333],[1.3333,2.1333,2.9333,3.7333,4.3487],[2.1333,3.1333],[0,1.6,3.2,4.8000,6.4],[2.7666,5.0],[0,1.3333,1.8333,3.5666,4.5],[5.8,6.9428,7.3714]] and end_times es=[[0.5333,1.0666,1.8666,2.4,2.6666],[1.3333,2.1333,2.9333,3.7333,4.1333],[1.9487,2.7487,3.5487,4.3487,4.6564],[3.1333,4.1333],[1.6,3.2,4.8000,6.4,7.2],[3.5666,5.8],[1.3333,1.8333,2.7666,4.5,5.0],[6.9428,7.3714,8.1714]], all time units are s. The specific steps are as follows:

[0053] Step 5-2): After obtaining the deadlock-free particle N, calculate the start time start_time and end time end_time of each vehicle occupying each resource, store the start_time in the start_times list, and store the end_time in the end_times list.

[0054] Step 5-3): Based on the obtained start_times and end_times, determine whether the end_time of the current resource is equal to the start_time of the next occupied resource. If they are the same, then Delaytime = 0, otherwise Delaytime = Delaytime_pre + start_time - end_time. This process continues until all vehicles have finished judging.

[0055] Step 5-4): Output BestDelaytime of particle N = 13.335229476405944s.

[0056] Step 6) (Adjust speed): The specific steps are as follows

[0057] Step 6-1): Obtain the vehicle sequence with discontinuous vehicles in the particles and store it in the nocontinuoucar list. nocontinuoucar = [11, 11, 11, 11, 14, 14, 14, 14, 15, 15, 15, 15, 16, 16, 16, 16, 23, 23, 23, 23, 24, 24, 24, 24], and draw the discontinuous vehicles and the resources they occupy, as Figure 3 shown.

[0058] Step 6-2): Find that the start time for the discontinuous vehicle to occupy the current resource is t 1 , the end time is t 2 , and the start time t 3 for occupying the next resource. Then the delay time (Delay) is t 3 - t 2 .

[0059] Step 6-3): Under the condition that the total time and total distance remain unchanged, use parabolic fitting to adjust the speed. The speed adjustment formula is as follows:

[0060]

[0061] For t ≤ mid_time, if the currently calculated speed < c, then set the current speed speed to the last speed value greater than zero within the time range. For mid_time < t ≤ total_time, if the currently calculated speed > 0, then set the current speed speed = c. speed1 in the formula refers to the speed finally calculated when t ≤ mid_time.

[0062] Step 6-4): Divide the total time total_time into 100 parts, calculate the speed once at each time point, and store the calculated speed in the adjusted_speeds list, adjusted_speeds = [10.0, 9.8716, 9.735, 9.5903, 9.4374, 9.2764, 9.1072, 8.9298, 8.7442, 8.5505, 8.3487, 8.1387, 7.9205, 7.6941, 7.4 596,7.2169,6.9661,6.7071,6.4399,6.1646,5.8811,5.5895,5.2896,4.9817,4.6655,4.3412,4.0088,3.6682,3.3194,2.9624,2.5973,2.224,1.8426,1.453,1.0552,0.6493,0.2352,0.2352,0.2352,0.2352,0.2 352,0.2352,0.2352,0.2352,0.2352,0.2352,0.2352,0.2352,0.2352,0.2984,0.4309,0.5715,0.7203,0.8773,1.0424,1.2157,1.3972,1.5868,1.7846,1.9905,2.2046,2.4269,2.6573,2.8959,3.1427,3.3976,3.6607,3 .9319,4.2113,4.4989,4.7946,5.0985,5.4106,5.7308,6.0592,6.3957,6.7404,7.0933,7.4543,7.8235,8.2009,8.5864,8.9801,9.3819,9.7919,10 ... Figure 4 .

[0063] Through practical verification, the present invention can provide optimal driving speed and passing time suggestions for each intelligent networked vehicle. In this way, the vehicle can pass through the intersection at the optimal speed while ensuring safety and avoiding deadlock, effectively reducing waiting time and alleviating traffic congestion.

[0064] The specific implementation scheme described above further describes in detail the purpose, technical scheme and beneficial effects of the present invention. It should be understood that the above is only a specific implementation scheme of the present invention and is not intended to limit the scope of the present invention. Any equivalent changes and modifications made by any technician in the field without departing from the concept and principle of the present invention should fall within the scope of protection of the present invention.

Claims

1. A method for optimizing intersection delay time and guiding speed of an intelligent connected vehicle, characterized in that: The following steps are involved: S1. Modeling: In view of the spatial resource allocation inside the intersection and the dynamic occupation of these resources by vehicles in the process of continuous crossing of the intersection, a Petri net model (P, T, F, M0) of the unsignalized intersection system is constructed, where P is a finite place set; T is a finite transition set; F = (P × T) ∪ (T × P) is a directed arc set; M0 is an initial identification set; S2, obtaining vehicle information in the unsignalized intersection area, including the number of vehicles, speed, and lane number; S3, encoding and decoding: Since the encoding reflects the vehicle's passage sequence, the position and speed of the particles, the vehicle's passage sequence is encoded according to the characteristics of the intelligent connected vehicle system and the particle swarm algorithm. The specific steps are as follows: S3-1: The vehicle number is two integers AB, A represents the vehicle's driving path, and B represents the vehicle's driving priority; S3-2: Sort the vehicles from small to large, and then expand the sequence according to the number of transitions that each vehicle needs to pass, and the vehicle sequence after expansion is Car_sequence; S3-3: Decode the expanded vehicle traffic sequence Car_sequence into a transition sequence Tran; S4, deadlock detection and repair: The transition sequence obtained by decoding in S3 may not be a feasible pass sequence, so deadlock detection and repair are performed on it; S5, optimize Delaytime: obtain the repaired particles through S4, calculate the delay time Delaytime of all particles in each generation, adjust the particle sequence through the particle swarm algorithm, optimize the Delaytime, and finally obtain the best particle sequence and optimal Delaytime in all generations. The specific steps are as follows: S5-1, calculate the start_times and end_times list of each particle in each generation of particles, store the start time start_time of each vehicle occupying each resource in start_times, and store the end time end_time in end_times; S5-2, determine whether the end_time of the current resource occupied by the vehicle is equal to the start_time of the next resource occupied. If they are the same, then Delaytime = 0, otherwise Delaytime = Delaytime_pre + start_time - end_time; until all vehicles are judged to be finished; S5-3, output the optimal particle sequence and its corresponding BestDelaytime in this generation of particle swarm; S5-4, taking the particle swarm of each generation as the input of the particle swarm algorithm, optimizing the current particle swarm through the particle swarm algorithm, and obtaining the optimized particle swarm; S5-5, obtaining particles after each generation of optimization, the optimization content includes the vehicle sequence, the transition sequence corresponding to the vehicle sequence, and the optimal Delaytime value; S5-6. Re-perform S5-1, S5-2, S5-4, and S5-5 on the particles obtained after optimization until the last round of iteration is completed, and output the optimal particle and its corresponding BestDelaytime after all rounds of iteration are completed; S6. Adjust the speed: After completing the iteration through S5 and determining the optimal particle and its corresponding best delay time BestDelaytime, in order to better approximate the dynamic characteristics of the real world, it is necessary to optimize the speed model. It is planned to introduce three speed modes: deceleration, uniform speed, and acceleration to construct a more physically consistent motion process. The specific steps are as follows: S6-1. Obtain the discontinuous vehicle sequence in the optimal particle and store it in the nocontinuoucar list; S6-2. Find the start time t1, end time t2 when the discontinuous vehicle occupies the current resource, and the start time t3 when it occupies the next resource. Then the delay time Delay is t3 - t2; S6-3. Under the condition that the total time and total distance remain unchanged, use parabolic fitting to adjust the speed and output the adjusted speed value; S6-4. Divide the total time total_time into 100 parts, calculate the speed at each time point, store the calculated speeds in the adjusted_speeds list, and draw the final speed adjustment graph; In S5-3, output the optimal particle sequence in this generation of particle swarm and its corresponding BestDelaytime. At the same time, calculate the fitness value Adapt of the particle swarm. The calculation of Adapt uses an inertia weight self-adaptive adjustment function. The specific formula is as follows: Where w(t) is the value of the inertia weight in the Gen-th iteration; maxGen is the maximum number of iterations; p(t) is the update rate of the fitness of the population individuals; update(Gen) is the number of fitness improvements of the population individuals in the Gen-th iteration; Adapt[i] is the fitness value after the Gen-th update, Adapt[i-1] is the fitness value after the Gen-th update, i is a random number between [0,1], Adapt best is the optimal fitness value, ND is a random integer between [1,2], Adapt mean is the mean fitness of the particle swarm.

2. The method for optimizing intersection delay time and guiding speed of an intelligent networked vehicle according to claim 1, characterized in that: In S5-4, optimize the current particle swarm through the particle swarm algorithm to obtain the optimized particle swarm. The specific self-adaptive formula in the optimization process is as follows: where f avg is the average fitness value of the particle swarm, f max is the maximum fitness value in the particle swarm, f i is the current particle fitness value, p c1 is 0.8, p c2 is 0.

5.

3. The method for optimizing intersection delay time and guiding speed of an intelligent networked vehicle according to claim 2, characterized in that: The specific steps of S6-1 are as follows: Calculate whether the end_time when each vehicle in the optimal particle occupies the current resource is equal to the start_time when it occupies the next resource. If they are the same, the vehicle is traveling continuously at this time; otherwise, the vehicle is discontinuous. Store the discontinuous vehicle sequence in the nocontinuoucar list.

4. The method for optimizing intersection delay time and guiding speed of an intelligent networked vehicle according to claim 3, characterized in that: In S6-3, under the condition that the total time and total distance remain unchanged, use parabolic fitting to adjust the speed. The specific speed adjustment formula is as follows: For t ≤ mid_time, if the currently calculated speed < c, then set the current speed speed to the last speed value greater than zero within the time range; for mid_time < t ≤ total_time, if the currently calculated speed > 0, then set the current speed speed = c; speed1 in the formula refers to the speed finally calculated when t ≤ mid_time.

Images