Online control method for single-point intersection signal based on spatiotemporal trajectory data of vehicles on the entrance lane

Abstract

The present invention provides a method for online signal control at a single-point intersection based on the spatiotemporal trajectory data of vehicles on the entrance lane. The method comprises the following steps: preprocessing; obtaining vehicle data at time T and updating the accumulated delay of each vehicle; obtaining traffic light information and determining whether time T is in the scheme update phase; dynamically segmenting each vehicle flow and analyzing the spatiotemporal distribution characteristics of the traffic flow in each phase at time T; establishing an online signal control optimization model for a single-point intersection; solving the model based on a genetic algorithm to obtain the optimal signal timing scheme after time T+1; and controlling the traffic lights according to the optimal scheme. Application of this method can achieve a comprehensive analysis of the real-time traffic conditions at the intersection, meet drivers' needs for signal light countdowns, and simultaneously match signal control decisions with traffic demand, thereby ensuring fair distribution of right-of-way and improving the utilization rate of spatiotemporal resources at the intersection.

Description

Technical Field

[0001] The present invention relates to the field of road traffic control, and in particular to an online control method for single-point intersection signals based on spatiotemporal trajectory data of vehicles on an inlet road. Background Art

[0002] At present, the research results in the field of single-point signal light control can be mainly divided into three categories: timing control, induction control and adaptive control.

[0003] Timed control uses a fixed signal control scheme, meaning that control parameters such as cycle duration and green-to-signal ratio remain constant. Because it cannot respond to randomly changing traffic demands, timed control is only suitable for situations with simple traffic conditions. However, with increasingly complex traffic flows and increasing congestion, the application scope of timed control is shrinking.

[0004] While inductive control can respond to real-time traffic demands, it can only detect the arrival of vehicles and cannot predict future arrivals. If sporadic vehicles continue to arrive at the target single-point intersection during the green light phase within the maximum extension time, the green light will be extended even if there are many vehicles in the red light queue, resulting in unfair right-of-way allocation. Furthermore, inductive control is a stimulus-response control model, generally suitable for signalized intersections with low traffic volume but generally ineffective for signalized intersections with high traffic volume.

[0005] While adaptive control can make optimal control decisions with optimization goals like minimizing average delay and idle time, its decision-making is often based on short-term traffic flow forecasts based on real-time cross-sectional traffic data. This places high demands on data accuracy, model building, and algorithm design, resulting in poor practical application. Furthermore, cross-sectional traffic data often fails to reflect the comprehensive traffic conditions at a single target intersection, making it difficult for signal timing plans derived from this data to align with actual traffic conditions.

[0006] At the same time, compared with timing control, although inductive control and adaptive control can better respond to randomly changing traffic demands, there are currently few studies that combine these two methods with traffic light countdown, and it is impossible to provide a countdown solution that is consistent with the actual situation. Summary of the Invention

[0007] An embodiment of the present invention provides an online signal control method for a single-point intersection based on spatiotemporal trajectory data of vehicles on an entrance lane, so as to effectively improve the utilization rate of spatiotemporal resources of a target single-point intersection.

[0008] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions.

[0009] A single-point intersection signal online control method based on spatiotemporal trajectory data of vehicles on an entrance lane comprises:

[0010] Step S0: Determine the vehicle data acquisition frequency, obtain the spatial morphological information, spatial division scheme, signal timing basic scheme and traffic characteristic parameters of the target single-point intersection, and determine the constraint conditions of each phase of the target single-point intersection;

[0011] Step S1: Obtain vehicle data at the target single-point intersection at time T and update the cumulative delay of each vehicle at the entrance of the target single-point intersection;

[0012] Step S2: Obtain traffic light information for the target single-point intersection and determine whether time T is in the plan update phase. If so, execute steps S3 to S6. Otherwise, calculate the optimal plan after time T+1 based on the optimal timing plan generated by the genetic algorithm at the last moment of the previous plan update phase, and execute step S6.

[0013] Step S3, dynamically segmenting each traffic flow and analyzing the spatiotemporal distribution characteristics of traffic flows in each phase of the target single-point intersection at time T;

[0014] Step S4, establishing an online signal control optimization model for the target single-point intersection based on the spatiotemporal distribution characteristics of traffic flow in each phase of the target single-point intersection and the constraints of each phase;

[0015] Step S5, solving the target single-point intersection signal online control optimization model based on a genetic algorithm to obtain the optimal signal timing plan after time T+1;

[0016] Step S6: Control the traffic lights at the target single-point intersection according to the optimal signal timing plan after time T+1.

[0017] Preferably, the step S0, determining the vehicle data acquisition frequency, acquiring the spatial morphological information, spatial division scheme, signal timing basic scheme, and traffic characteristic parameters of the target single-point intersection, and determining the constraint conditions of each phase of the target single-point intersection, includes:

[0018] S0-1) Determine the vehicle data acquisition frequency Δt, and the calculation formula of Δt is as follows: Where t1 represents the data collection time required to obtain vehicle data at a certain moment, and t2 represents the calculation time required for the single-point intersection signal online control method;

[0019] S0-2) obtaining spatial morphological information of the target single-point intersection, the spatial morphological information including the number and type of lanes at each entrance, and the positions of each lane and stop line;

[0020] S0-3) determining the space occupied by a standard vehicle, and dividing each lane in the entrance area of the target single-point intersection into a number of cells of equal length using the space occupied by the standard vehicle as a unit length;

[0021] S0-4) Determine the basic signal timing plan for the target single-point intersection, including the green light countdown time requirements, phase division, and green light interval time required for phase transition;

[0022] S0-5) Determine the constraints for each phase of the target single-point intersection. These constraints include: a minimum green light time constraint, which is determined by the shortest time required for pedestrians to cross the street in that phase; a maximum red light time constraint, which is determined by the maximum allowable waiting time for pedestrians before crossing the street; a maximum vehicle delay constraint, which is determined based on the traffic volume and importance of the phase; and a maximum queue length constraint, which is determined based on the road conditions and the maximum queue length allowed for each vehicle flow in each phase.

[0023] S0-6) Determine the traffic characteristic parameters of the target single-point intersection: including the saturation flow rate of each phase at the stop line section.

[0024] Preferably, the step S1, obtaining vehicle data of the target single-point intersection at time T and updating the accumulated delay of each vehicle located at the entrance of the target single-point intersection, includes:

[0025] S1-1) Deploy a phased array radar at each entrance to the target single-point intersection, establish an entrance vehicle detection system, and adjust the stop line and lane line at each entrance to be parallel to the X-axis and Y-axis of the corresponding phased array radar coordinate system; obtain vehicle data detected by the entrance vehicle detection system at the target single-point intersection at time T, the vehicle data including the ID assigned to each vehicle by the entrance vehicle detection system, as well as the speed and position of the vehicle;

[0026] S1-2) updating the cumulative delay of each vehicle in the entrance lane based on the vehicle speed data.

[0027] Preferably, step S2, obtaining traffic light information of the target single-point intersection, determining whether time T belongs to the solution update stage, and if so, executing steps S3 to S6; otherwise, calculating the optimal solution after time T+1 based on the optimal timing solution generated by the genetic algorithm at the last moment of the previous solution update stage, and executing step S6, includes:

[0028] S2-1) Obtaining traffic light information at the target single-point intersection at time T, the traffic light information including the light color corresponding to each phase, the duration of the light color, and whether a green light or red light countdown is being displayed;

[0029] S2-2) The solution update phase refers to the period after the green light of a phase turns on and before the green light countdown is displayed. Determine whether time T belongs to the solution update phase. If so, execute step S2-4; otherwise, execute step S2-3;

[0030] S2-3) When time T does not belong to the plan update phase, the optimal plan after time T+1 is calculated based on the optimal timing plan generated by the genetic algorithm at the last moment of the previous plan update phase, and step S6 is executed;

[0031] S2-4) When time T belongs to the plan update stage, the processing of steps S3 to S6 is continued to complete the plan update and traffic light control.

[0032] Preferably, the step S2-1) acquires traffic light information of the target single-point intersection at time T, the traffic light information including the light color corresponding to each phase, the duration of the light color, and whether a green light or red light countdown is being displayed, including:

[0033] t c represents the required duration of the green light countdown, Δt represents the frequency of data acquisition, and the judgment results of the target single-point intersection traffic status and whether the plan is updated corresponding to the different traffic light information at time T are divided into the following three cases:

[0034] (1) A phase is in green light status, but the signal light has not yet displayed the green light countdown: the traffic flow corresponding to this phase has the right of way. This moment belongs to the scheme update stage, and the subsequent signal timing needs to be updated based on the vehicle data at this moment;

[0035] (2) A phase is in green light state, and the traffic light is showing the green light countdown: the traffic flow corresponding to this phase has the right of way. This moment does not belong to the scheme update stage. When the green light countdown ends, the traffic state of the target single-point intersection enters situation (3);

[0036] (3) No green light phase: Traffic in all phases has no right of way. This moment does not belong to the plan update stage, and the subsequent signal timing plan has been determined.

[0037] The step S3 dynamically segments each traffic flow and analyzes the spatiotemporal distribution characteristics of traffic flows in each phase of the target single-point intersection at time T, including:

[0038] S3-1) Determine vehicle type: Based on the spatiotemporal trajectory data of traffic flow at time T and before, a vehicle speed equal to 0 is used as a sign that the vehicle has arrived at the target single-point intersection. Vehicles with a speed equal to 0 at time T or before are determined to be arriving vehicles, and the remaining vehicles are considered upstream vehicles.

[0039] S3-2) Set up the traffic flow matrix: Based on the position of the vehicles at time T, follow the one-to-one principle and map all vehicles to specific cells in the entrance lane. Based on the cell matching results, for each traffic flow, set the position matrix M of the arrival traffic flow in the entrance lane. ap and the delay matrix M ad , where the upstream traffic flow is set at the position matrix M of the entrance lane up and the velocity matrix M uv The size of each of the four matrices is m rows and n columns, where m = the number of lanes corresponding to the traffic flow, n = the length of the lane corresponding to the traffic flow in the study area / the length of the cell, and the first column represents the cell next to the stop line; the position matrix, speed matrix, and delay matrix are all initialized to zero matrices. If there is a vehicle in a cell, the element in the position matrix corresponding to the cell is changed to 1, the element in the speed matrix is changed to the speed of the vehicle, and the element in the delay matrix is changed to the accumulated delay of the vehicle;

[0040] S3-3) Count vehicle arrivals: The distance between the last arriving vehicle in each traffic flow and the stop line is recorded as the queue length L at time T. 0 , the number of arriving vehicles in the traffic flow is recorded as The sum of the cumulative delays of the arriving vehicles in this traffic flow is recorded as L 0 、 and The calculation formula is as follows, where L pcu Indicates the cell length:

[0041] L 0 =max(J a )×L pcu (1)

[0042] J a (i, j) = P a (i, j) × j (2)

[0043]

[0044]

[0045] S3-4) Divide the upstream fleet: Determine the fleet division threshold L f , when the distance between two moving vehicles is less than L f, the two vehicles are classified into the same fleet; the cells corresponding to the elements in the same column of the upstream traffic flow position matrix are merged into a unit area, the number of vehicles in each unit area is calculated, and the unit area is marked as "no car" or "car" according to the number of vehicles; if the states of two adjacent unit areas are different, the dividing line between the two unit areas is used as the alternative dividing line for traffic flow segmentation; if the area between the two adjacent alternative dividing lines is "no car state" and the distance between the two is less than L f Delete these two alternative dividing lines;

[0046] S3-5) Calculate the fleet parameters: Consider each fleet as a uniform entity, and use s for the xth fleet in the upstream traffic flow. x，1 Indicates the distance between the front of the convoy and the stop line, s x，2 represents the distance between the rear end of the convoy and the stop line, q x 、v x 、k x 、N x represents the flow, speed, density, and number of vehicles of the fleet, j x，1 、j x，2 Indicates the matrix column subscript corresponding to the cell where the first and last vehicles of the team are located, L pcu represents the space occupied by a standard vehicle, and the calculation formula for the relevant parameters of the fleet is as follows:

[0047] s x，1 =L pc u ×(j x，1 ―1) (5)

[0048] s x，2 =L pc u ×j x，2 (6)

[0049]

[0050]

[0051]

[0052] Preferably, the step S4, based on the spatiotemporal distribution characteristics of traffic flow in each phase of the target single-point intersection and the constraints of each phase, establishes an online signal control optimization model for the target single-point intersection, including:

[0053] S4-1) Select the decision variables and target variables of the target single-point intersection signal online control optimization model, and take the n after time T. k time periods, numbered with k, and t k represents the duration of time period k, represents the status of the signal light in phase i during time period k, Indicates green light; Indicates yellow or red light, with t k and As the decision variable; D represents the number of all import vehicles detected at time T in n k The sum of the cumulative delays in each period, with D as the target variable;

[0054] S4-2) Determine the expression of the objective function:

[0055]

[0056] Among them, K represents the time period set, P represents the phase set, represents the added delay of phase i in time period k, The calculation of is related to the decision variable values and the results of the spatiotemporal characteristics analysis of the traffic flow in phase i in S3;

[0057] S4-3) Determine the constraints:

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067]

[0068]

[0069]

[0070] 1≤k≤n k ,k∈Z (24)

[0071] t k ∈Z + (25)

[0072]

[0073] Among them: (12) (13) In order to meet the basic conditions of the signal timing plan; (14) indicates that the status of the signal lights of each phase in period 1 should be consistent with time T, represents the signal light status of phase i at time T; (15) indicates that if there is a phase in the green light state in time period k, then t k It must be at least longer than the green light countdown time requirement, t c represents the duration requirement of the green light countdown, Δt represents the frequency of data acquisition; (16)(17) In order to meet the minimum green light time constraint, G 0 Represents the green light phase at time T, Indicates the duration of the green light in this phase. represents the minimum green light time of phase i; (18)(19) In order to meet the maximum red light time constraint, R 0 Represents the red light phase set at time T, (i∈R 0 ) represents the duration of the red light phase up to time T. represents the maximum red light time of phase i, and M represents the maximum value; (20) In order to meet the green light interval time constraint, I is a matrix, and I(i, j) represents the green light interval time required for the green light phase to switch from i to j; (21) In order to ensure that all the vehicles entering the lane detected at time T can enter the lane n k Pass the target single-point intersection in the time period, represents the number of vehicles entering the lane detected at time T in phase i, represents the number of vehicles passing through the target single-point intersection in phase i during time period k; (22) In order to satisfy the maximum average vehicle delay constraint in each phase, represents the accumulated delay of phase i up to time T, represents the cumulative delay of phase i in time period k; (23) In order to meet the maximum queue length constraint of the corresponding traffic flow in each phase, represents the maximum queue length of vehicle flow m corresponding to phase i in time period k, represents the maximum queue length constraint of vehicle flow m corresponding to phase i; (24)(25)(26) are the values of the variables.

[0074] Preferably, the step S5, solving the target single-point intersection signal online control optimization model based on a genetic algorithm to obtain the optimal signal timing plan after time T+1, includes:

[0075] S5-1) Coding: Set two chromosomes for each individual in the genetic algorithm. Chromosome I represents the phase sequence, and chromosome II represents the corresponding duration. The length of both chromosomes is n. p ×n c, where n p is the number of phases, and is 1, 2,, n p As the phase number, n c is the preset number of cycles; the encoding method of chromosome I is: 1, 2,, n p ×n c Randomly shuffle the order and move the number of the current green light phase to the first position; the encoding method of chromosome II is to randomly generate n p ×n c A number greater than the countdown duration requirement and less than 60s;

[0076] S5-2) Decoding: In chromosome Ⅰ, i, i+n p 、i+2×n p ,,i+(n c ―1)×n p Indicates the phase numbered i, according to the 1, 2, n p ×n c The order of the phases obtaining the right of way after time T is determined by the sequence of chromosomes II and I. Chromosome II corresponds to chromosome I and represents the green light time allocated to the phases of the corresponding order. If there are multiple adjacent numbers representing the same phase in chromosome I, the position of the first digit is recorded as the reserved position, the positions of the following digits are recorded as the deleted positions, and the corresponding digits are deleted. At the same time, the digits corresponding to the deleted positions in chromosome II are added to the reserved positions and deleted as the green light time allocated to the phase.

[0077] S5-3) Calculate fitness: According to the decoded timing plan, calculate the objective function value D of the signal control optimization model under the plan; check whether each constraint condition is satisfied. If there is a constraint condition j that is not satisfied, calculate the corresponding penalty value e j ; The calculation formula of fitness is as follows:

[0078]

[0079] S5-4) Determine the termination condition: set the number of iterations as the termination condition;

[0080] S5-5) Selection: Eliminate individuals whose fitness ranks in the bottom third of the population;

[0081] S5-6) Crossover: Select individuals ranked in the top third of fitness and pair them up for crossover. Chromosomes I and II cross simultaneously, and each crossover produces two new individuals.

[0082] S5-7) Mutation: The top 30 individuals in fitness are retained, and the remaining individuals mutate according to a certain probability. There are two types of mutation: one is to randomly swap the phase order of two phases, and the other is to randomly change the green light duration of one phase between 50% and 150%. When an individual needs to mutate, a certain number of offspring are generated according to the two mutation types, and the offspring with the highest fitness is selected as the mutation result of the individual;

[0083] S5-8) Determine the optimal signal timing scheme after time T+1, the optimal signal timing scheme includes the remaining green light time t of the current green light phase at time T+1 g_after(T+1) , the next green light phase P g_next The selection of , and the initial value of the green light time t set for the next green light phase g_next After the iteration ends, select the individual with the largest fitness, decode it, and take the first and second digits in chromosome II as t g_after(T+1) and t g_next , take the phase corresponding to the second digit in chromosome I as P g_next ;.

[0084] Preferably, the step S6 of controlling the traffic lights at the target single-point intersection according to the optimal signal timing plan after time T+1 includes:

[0085] S6-1) In the optimal signal timing scheme finally obtained, if t 0 >t c +Δt, determine that the countdown is not displayed at time T+1, and after obtaining the vehicle data at time T+1, repeat steps S1-S6;

[0086] S6-2) In the final optimal signal timing solution, if t 0 =t c +Δt, it is determined that the countdown starts at time T+1, and the traffic light is controlled according to the scheme until the green light of the next green light phase is turned on.

[0087] It can be seen from the technical solutions provided by the above-mentioned embodiments of the present invention that the method of the present invention can realize a comprehensive analysis of the real-time traffic conditions of the target single-point intersection, meet the driver's demand for traffic light countdown, and at the same time match the signal control decision with the traffic demand, thereby ensuring the fair distribution of the right of way and improving the utilization rate of the time and space resources of the target single-point intersection.

[0088] Additional aspects and advantages of the present invention will be set forth in part in the following description, will become apparent from the following description, or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0090] Figure 1 A processing flow chart of a single-point intersection signal online control method based on the spatiotemporal trajectory data of vehicles on the entrance lane provided by an embodiment of the present invention;

[0091] Figure 2 A signal timing phase division diagram provided by an embodiment of the present invention;

[0092] Figure 3 A genetic algorithm flow chart provided in an embodiment of the present invention;

[0093] Figure 4 A schematic diagram of a vehicle dynamic segmentation method provided by an embodiment of the present invention;

[0094] Figure 5 A schematic diagram of the calculation of vehicle passage, queuing, and delay in an optimization model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0095] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limiting the present invention.

[0096] It will be understood by those skilled in the art that, unless expressly stated otherwise, the singular forms "a", "an", "said" and "the" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the description of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, "connected" or "coupled" as used herein may include wireless connections or couplings. The term "and / or" used herein includes any unit and all combinations of one or more associated listed items.

[0097] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art in the art to which the present invention pertains. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and, unless defined as such herein, will not be interpreted in an idealized or overly formal sense.

[0098] To facilitate understanding of the embodiments of the present invention, several specific embodiments will be further explained below with reference to the accompanying drawings, and each embodiment does not constitute a limitation on the embodiments of the present invention.

[0099] The present invention provides an online signal control method for single-point intersections based on the spatiotemporal trajectory data of vehicles on the approach lanes, which addresses at least some of the aforementioned technical issues. Based on the spatiotemporal trajectory data of vehicles on the approach lanes, this method generates a signal light timing plan based on the vehicle queue delays, through-traffic volume, and speed at each approach lane, and provides a countdown for when the green light will appear.

[0100] The processing flow of a single-point intersection signal online control method based on the spatiotemporal trajectory data of the entrance lane provided by the embodiment of the present invention is as follows: Figure 1 As shown, the following processing steps are included:

[0101] Step S0, pre-processing: determining the vehicle data acquisition frequency, obtaining the spatial morphological information, spatial division scheme, signal timing basic scheme and traffic characteristic parameters of the target single-point intersection, and determining the constraint conditions of each phase of the target single-point intersection;

[0102] Step S1: Obtain vehicle data at the target single-point intersection at time T and update the cumulative delay of each vehicle at the entrance of the target single-point intersection;

[0103] Step S2: Obtain traffic light information for the target single-point intersection and determine whether time T is in the plan update phase. If so, execute steps S3 to S6. Otherwise, calculate the optimal plan after time T+1 based on the optimal timing plan generated by the genetic algorithm at the last moment of the previous plan update phase, and execute step S6.

[0104] Step S3, dynamically segmenting each traffic flow and analyzing the spatiotemporal distribution characteristics of traffic flows in each phase of the target single-point intersection at time T;

[0105] Step S4, establishing an online signal control optimization model for the target single-point intersection based on the spatiotemporal distribution characteristics of traffic flow in each phase of the target single-point intersection and the constraints of each phase;

[0106] Step S5, solving the target single-point intersection signal online control optimization model based on a genetic algorithm to obtain the optimal signal timing plan after time T+1;

[0107] Step S6: Control the traffic lights at the target single-point intersection according to the optimal signal timing plan after time T+1.

[0108] Furthermore, in step S0:

[0109] S0-1) Determine the frequency of vehicle data acquisition: This refers to determining the time difference Δt between time T and time T-1. The calculation formula for Δt is as follows: Where t1 represents the data collection time required to obtain the vehicle data at a certain moment, and t2 represents the calculation time required by the algorithm proposed in the present invention;

[0110] S0-2) Acquire spatial morphological information of the target single-point intersection, including the number and type of lanes at each entrance, and the position of each lane and stop line in the coordinate system of the entrance vehicle detection system; deploy a phased array radar at each entrance of the target single-point intersection, establish an entrance vehicle detection system, and adjust the stop line and lane line at each entrance to be parallel to the X-axis and Y-axis of the corresponding phased array radar coordinate system; S0-3) Complete a spatial division plan for the target single-point intersection area: determine the space occupied by a standard vehicle (equal to the standard vehicle body length + the minimum allowable distance to the preceding vehicle); and divide each lane into a number of equal-length cells, using the standard vehicle space as the unit length;

[0111] S0-4) Determine the basic signal timing plan for the target single-point intersection: including the green light countdown time requirements, phase division, and the green light interval required for phase transition;

[0112] S0-5) Determine the constraints for each phase of the target single-point intersection. These constraints include: a minimum green light time constraint, which is determined by the shortest time required for pedestrians to cross the street in that phase; a maximum red light time constraint, which is determined by the maximum allowable waiting time for pedestrians before crossing the street; a maximum vehicle delay constraint, which is determined based on the traffic volume and importance of the phase and the maximum allowable vehicle delay per vehicle in that phase; and a maximum queue length constraint, which is determined based on road conditions and the maximum allowable queue length per vehicle flow in each phase to prevent queue overflow.

[0113] S0-6) Determine the traffic characteristic parameters of the target single-point intersection: including the saturation flow rate of each phase at the stop line section, etc.

[0114] Furthermore, in step S1:

[0115] S1-1) obtaining vehicle data detected by the inbound lane vehicle detection system at the target single-point intersection at time T, the vehicle data including an ID assigned to each vehicle by the inbound lane vehicle detection system, and the speed and position of the vehicle;

[0116] S1-2) updating the cumulative delay of each vehicle in the entrance lane based on the vehicle speed data.

[0117] Furthermore, in step S2:

[0118] S2-1) Figure 2 A signal timing phase division diagram provided by an embodiment of the present invention. Obtain traffic light information for a target single-point intersection at time T, including the light color corresponding to each phase, the duration of the light color, and whether a green light (red light) countdown is being displayed;

[0119] t c represents the required duration of the green light countdown, and Δt represents the frequency of data acquisition. The traffic status of the target single-point intersection corresponding to the traffic light information at time T can be divided into the following three situations:

[0120] (1) A phase is in green light state, and the traffic flow corresponding to the phase has the right of way, but the traffic light has not yet shown the green light countdown (see Figure 2 The T1 moment in the scheme update phase): This situation indicates that the signal timing scheme after this moment has not yet been determined, and the subsequent signal timing needs to be updated according to the vehicle data at this moment; the optimal scheme is obtained through steps S3-S5, including the remaining green light time of the green light phase after time T, the selection of the next green light phase, and the tentative green light time for the next green light phase; when the remaining green light time > t c +Δt, it means the remaining green light time at time T1+1>t c , that is, the green light countdown does not need to be displayed at time T1+1, which belongs to case (1); when the remaining green light time = t c +Δt, which means the remaining green light time at time T1+1 = t c , that is, the green light countdown needs to start at time T1+1, and the next green light phase has been selected, and situation (2) begins from time T1+1.

[0121] (2) A phase is in green light state, the traffic flow corresponding to the phase has the right of way, and the traffic light is showing the green light countdown (see Figure 2 The T2 moment in the figure belongs to the green light countdown stage): This situation indicates that the signal timing plan after this moment has been determined, that is, the next green light phase and its on-time are clear; when the green light countdown ends, the traffic status of the target single-point intersection enters situation (3).

[0122] (3) In the phase without green light, all traffic flows in the phase have no right of way (see Figure 2 The T3 moment in the green light interval period is the same as (2). This situation indicates that the signal timing plan after this moment has been determined. It should be noted that the tentative green light time for the next green light phase must be greater than t c Therefore, the countdown will not be displayed for the next green light phase, that is, the turn-on time of each green light phase belongs to case (1).

[0123] S2-2) The plan update phase refers to the period after the green light of a phase turns on and before the green light countdown is displayed. During the plan update phase, the signal timing plan after time T needs to be updated based on the actual traffic conditions reflected by vehicle data;

[0124] S2-3) When time T does not belong to the plan update phase, the optimal plan after time T+1 is calculated based on the optimal timing plan generated by the genetic algorithm at the last moment of the previous plan update phase, and steps S3-S5 are skipped to directly enter step S6;

[0125] S2-4) When time T belongs to the plan update stage, the processing of steps S3 to S6 is continued to complete the plan update and traffic light control.

[0126] Furthermore, in step S3:

[0127] S3-1) Determine vehicle type: Based on the spatiotemporal trajectory data of traffic flow at time T and before, a vehicle speed equal to 0 is used as a sign that the vehicle has arrived at the target single-point intersection. Vehicles with a speed equal to 0 at time T or before are determined to be arriving vehicles, and the remaining vehicles are considered upstream vehicles.

[0128] S3-2) Set up the traffic flow matrix: Based on the position of the vehicles at time T, follow the one-to-one principle and map all vehicles to specific cells in the entrance lane. Based on the cell matching results, for each traffic flow, set the position matrix M of the arrival traffic flow in the entrance lane. ap and the delay matrix M ad , where the upstream traffic flow is set at the position matrix M of the entrance lane up and the velocity matrix M uv The size of each of the four matrices is m rows and n columns, where m = the number of lanes corresponding to the traffic flow, n = the length of the lane corresponding to the traffic flow in the study area / the length of the cell, and the first column represents the cell next to the stop line; the position matrix, speed matrix, and delay matrix are all initialized to zero matrices. If there is a vehicle in a cell, the element in the position matrix corresponding to the cell is changed to 1, the element in the speed matrix is changed to the speed of the vehicle, and the element in the delay matrix is changed to the accumulated delay of the vehicle;

[0129] S3-3) Count vehicle arrivals: The distance between the last arriving vehicle in each traffic flow and the stop line is recorded as the queue length L at time T. 0 , the number of arriving vehicles in the traffic flow is recorded as The sum of the cumulative delays of the arriving vehicles in this traffic flow is recorded as L 0 、 and The calculation formula is as follows, where L pcu Indicates the cell length:

[0130] L 0 =max(J a )×L pcu (1)

[0131] J a (i, j) = P a (i, j) × j (2)

[0132]

[0133]

[0134] S3-4) Divide the upstream fleet: Determine the fleet division threshold L f , when the distance between two moving vehicles is less than L f , the two vehicles are classified into the same fleet; the cells corresponding to the elements in the same column of the upstream traffic flow position matrix are merged into a unit area, the number of vehicles in each unit area is calculated, and the unit area is marked as "no car" or "car" according to the number of vehicles; if the states of two adjacent unit areas are different, the dividing line between the two unit areas is used as the alternative dividing line for traffic flow segmentation; if the area between the two adjacent alternative dividing lines is "no car state" and the distance between the two is less than L f Delete these two alternative dividing lines;

[0135] S3-5) Calculate the fleet parameters: Consider each fleet as a uniform entity, and use s for the xth fleet in the upstream traffic flow. x，1 Indicates the distance between the front of the convoy and the stop line, s x，2 represents the distance between the rear end of the convoy and the stop line, q x 、v x 、k x 、N x represents the flow, speed, density, and number of vehicles of the fleet, j x，1 、j x，2 Indicates the matrix column subscript corresponding to the cell where the first and last vehicles of the team are located, L pcurepresents the space occupied by a standard vehicle, and the calculation formula for the relevant parameters of the fleet is as follows:

[0136] s x，1 =L pc u ×(j x，1 ―1) (5)

[0137] s x，2 =L pc u ×j x，2 (6)

[0138]

[0139]

[0140]

[0141] q x =k x ×v x (10).

[0142] Furthermore, in step S4:

[0143] S4-1) Select the decision variables and target variables of the target single-point intersection signal online control optimization model: Take n after time T k time periods, numbered with k, and t k represents the duration of time period k, represents the status of the signal light in phase i during time period k ( Indicates green light; Indicates yellow or red light) and t k and As the decision variable; D represents the number of all import vehicles detected at time T in n k The sum of the cumulative delays in each period, with D as the target variable;

[0144] S4-2) Determine the expression of the objective function:

[0145]

[0146] Among them, K represents the time period set, P represents the phase set, represents the added delay of phase i in time period k, The calculation of is related to the decision variable values and the results of the spatiotemporal characteristics analysis of the traffic flow in phase i in S3;

[0147] S4-3) Determine the constraints:

[0148]

[0149]

[0150]

[0151]

[0152]

[0153]

[0154]

[0155]

[0156]

[0157]

[0158]

[0159]

[0160] 1≤k≤n k ,k∈Z (24)

[0161] t k ∈Z + (25)

[0162]

[0163] Among them: (12) (13) In order to meet the basic conditions of the signal timing plan; (14) indicates that the status of the signal lights of each phase in period 1 should be consistent with time T, represents the signal light status of phase i at time T; (15) indicates that if there is a phase in the green light state in time period k, then t k It must be at least longer than the green light countdown time requirement, t c represents the duration requirement of the green light countdown, Δt represents the frequency of data acquisition; (16)(17) In order to meet the minimum green light time constraint, G 0 Represents the green light phase at time T, Indicates the duration of the green light in this phase. represents the minimum green light time of phase i; (18)(19) In order to meet the maximum red light time constraint, R 0 Represents the red light phase set at time T, Indicates the duration of the red light phase up to time T. represents the maximum red light time of phase i, and M represents the maximum value; (20) In order to meet the green light interval time constraint, I is a matrix, and I(i, j) represents the green light interval time required for the green light phase to switch from i to j; (21) In order to ensure that all the vehicles entering the lane detected at time T can enter the lane n k Pass the target single-point intersection in the time period, represents the number of vehicles entering the lane detected at time T in phase i, represents the number of vehicles passing through the target single-point intersection in phase i during time period k; (22) In order to satisfy the maximum average vehicle delay constraint in each phase, represents the accumulated delay of phase i up to time T, represents the cumulative delay of phase i in time period k; (23) In order to meet the maximum queue length constraint of the corresponding traffic flow in each phase, represents the maximum queue length of vehicle flow m corresponding to phase i in time period k, represents the maximum queue length constraint of vehicle flow m corresponding to phase i; (24)(25)(26) are the values of the variables.

[0164] Furthermore, in step S5: the signal online control optimization model proposed by the present invention is solved by using a genetic algorithm. A flowchart of a genetic algorithm provided by an embodiment of the present invention is as follows: Figure 3 As shown in the figure, the specific processing process includes:

[0165] S5-1) Coding: Set two chromosomes for each individual, chromosome I represents the phase sequence, chromosome II represents the corresponding duration, and the length of both chromosomes is n p ×n c , where n p is the number of phases, and is 1, 2,, n p As the phase number, n c is the preset number of cycles; the encoding method of chromosome I is to convert 1, 2,, n p ×n c Randomly shuffle the order and move the number of the current green light phase to the first position; the encoding method of chromosome II is to randomly generate n p ×n c A number that is greater than the countdown duration requirement and less than 60s (note that 60s is just an example).

[0166] S5-2) Decoding: with n p =4, n c=2, the countdown duration is required to be 10s, chromosome I (2-1-6-3-7-4-5-8) and chromosome II (15-20-30-25-10-23-40-35) are used as examples to illustrate the decoding steps; in chromosome I, 1 and 5 correspond to phase 1, 2 and 6 correspond to phase 2, 3 and 7 correspond to phase 3, and 4 and 8 correspond to phase 4. Therefore, the order of the green light phases after time T represented by chromosome I is 2-1-2-3-3-4-1-4; in chromosome II, each number represents the duration of the corresponding green light phase in chromosome I; merge the same phases. When the same phase appears adjacently, there is no need to consider the green light in the middle. Interval time, so 2-1-2-3-3-4-1-4 is further changed to 2-1-2-3-4-1-4, and correspondingly, chromosome II is further decoded to 15-20-30-35-23-40-35; then, 0 is inserted between each two phases to indicate the green light interval stage, and the green light interval time is inserted in the middle of the corresponding green light duration, and finally the completely decoded timing scheme is obtained (2-0-1-0-2-0-3-0-4-0-1-0-4, 15-3-20-3-30-3-35-3-23-3-40-3-35, taking the green light interval time of 3s as an example).

[0167] S5-3) Calculate fitness: According to the decoded timing plan, calculate the objective function value D of the signal control optimization model under the plan; at the same time, check whether each constraint condition is satisfied. If there is a constraint condition j that is not satisfied, calculate the corresponding penalty value e j ; The calculation formula of fitness is as follows:

[0168]

[0169] S5-4) Determine the termination condition: set the number of iterations as the termination condition.

[0170] S5-5) Selection: Eliminate individuals whose fitness ranks in the bottom third of the population.

[0171] S5-6) Crossover: Select individuals ranked in the top one-third of fitness and pair them up for crossover. Chromosomes I and II cross simultaneously, and two new individuals are obtained each time they cross.

[0172] S5-7) Mutation: The top 30 individuals in terms of fitness are retained, and the remaining individuals mutate according to a certain probability. There are two types of mutation: one is to randomly swap the phase sequence of two phases, and the other is to randomly change the green light duration of one phase between 50% and 150%. When an individual needs to mutate, a certain number of offspring are generated according to the two mutation types, and the offspring with the highest fitness are selected as the result of the individual's mutation.

[0173] Furthermore, in step S6:

[0174] S6-1) In the optimal signal timing scheme finally obtained, if t 0 >t c +Δt, determine that the countdown is not displayed at time T+1, and after obtaining the vehicle data at time T+1, repeat steps S1-S6;

[0175] S6-2) In the final optimal signal timing solution, if t 0 =t c +Δt, it is determined that the countdown starts at time T+1, and the traffic light is controlled according to the scheme until the green light of the next green light phase is turned on.

[0176] A schematic diagram of a vehicle dynamic segmentation method provided by an embodiment of the present invention is shown in FIG. Figure 4 Taking a straight traffic flow at a certain entrance (without widening section) as an example, the process of dynamically segmenting each traffic flow in step S3 includes:

[0177] (1) Determine vehicle type: Based on the spatiotemporal trajectory data of traffic flow at time T and before, a vehicle speed equal to 0 is used as a sign that the vehicle has arrived at the target single-point intersection. Vehicles with a speed equal to 0 at time T or before are determined to be arriving vehicles, and the remaining vehicles are considered upstream vehicles.

[0178] (2) Set up the traffic flow matrix: According to the position of the vehicle at time T, follow the one-to-one principle and map all vehicles to specific cells in the entrance lane. Based on the cell matching results, set the position matrix M of the entrance lane for the arriving traffic flow. ap and the delay matrix M ad , is the position matrix M of the upstream traffic flow set at the entrance lane up and the velocity matrix M uv The matrix size is m rows and n columns, where m = the number of through lanes, n = the length of through lanes in the study area / cell length, and the first column represents the cell next to the stop line. The position matrix and speed matrix are both initialized to 0 matrices. If there is a vehicle in a cell, the element in the position matrix corresponding to the cell is changed to 1, the element in the speed matrix is changed to the speed of the vehicle, and the element in the delay matrix is changed to the accumulated delay of the vehicle.

[0179] (3) Count vehicle arrivals: The distance between the last arriving vehicle in the traffic flow and the stop line is recorded as the queue length L at time T 0 , the number of arriving vehicles in the traffic flow is recorded as The sum of the cumulative delays of the arriving vehicles in this traffic flow is recorded as L 0 、 and The calculation formula is as follows, where L pcu Indicates the cell length:

[0180] L 0 =max(J a )×L pcu (28)

[0181] J a (i, j) = P a (i, j)×j (29)

[0182]

[0183]

[0184] (4) Divide the upstream fleet: merge the cells corresponding to the elements in the same column of the upstream traffic flow position matrix into a unit area, calculate the number of vehicles in each unit area, and mark the unit area as "no vehicle" or "with vehicle" according to the number of vehicles; if the states of two adjacent unit areas are different, the dividing line between the two unit areas is used as the alternative dividing line for traffic flow segmentation; if the area between the two adjacent alternative dividing lines is "no vehicle state" and the distance between the two is less than L f Delete these two alternative dividing lines (assuming that when the distance between the two moving vehicles is less than L f When , two vehicles can be put into the same team); Figure 2 In, L f The value of is 3×L pcu , so the upstream traffic is finally divided into two fleets. It should be noted that L f =3×L pcu This is just an example. There is not only one value that can be used. It should be adjusted according to the actual situation.

[0185] (5) Calculation of platoon parameters: The calculation is based on the following assumptions: each platoon is considered as a homogeneous entity, that is, the speeds of all vehicles in it are exactly the same and they are evenly distributed in space; for the xth platoon in the upstream traffic flow, s is used. x，1 Indicates the distance between the front of the convoy and the stop line, s x，2 represents the distance between the rear end of the convoy and the stop line, q x 、v x 、k x 、N x represents the flow, speed, density, and number of vehicles of the fleet, j x，1 、j x，2 The matrix column subscripts corresponding to the cells where the first and last vehicles of the fleet are located are represented. The calculation formulas for the fleet-related parameters are as follows:

[0186] sx，1 =L pc u× (j x，1 ―1) (32)

[0187] s x，2 =L pc u ×j x，2 (33)

[0188]

[0189]

[0190]

[0191] q x =k x ×v x (37)

[0192] In the above target single-point intersection signal online control optimization model The calculation method is as follows:

[0193]

[0194] It is closely related to the model objective function and constraints.

[0195] The calculation formula is as follows:

[0196]

[0197] It is equal to the sum of the cumulative delays of all arriving vehicles in phase i, where the delay of each vehicle is equal to the time difference between the moment it is judged to be an arriving vehicle and time T.

[0198]

[0199]

[0200] A schematic diagram of calculation of vehicle passing, queuing and delay in an optimization model provided by an embodiment of the present invention is shown as follows: Figure 5 Take a certain import through traffic flow as an example to illustrate The calculation method is as follows:

[0201] (1) Vehicle arrival prediction: Based on the dynamic segmentation results of traffic flow, the vehicle arrival situation in the future is predicted; x represents the arrival flow rate of the xth upstream convoy when it arrives at the target single-point intersection, t x，1It represents the time difference between the first vehicle of the xth upstream convoy and the last vehicle of the x-1th upstream convoy arriving at the target single-point intersection, t x，2 k represents the time required for the xth upstream team to arrive from the first car to the last car. j represents the blocking density of the traffic flow; x , t x，1 , t x，2 The calculation formula is as follows,

[0202] λ1=q1 (41)

[0203]

[0204]

[0205]

[0206]

[0207]

[0208] (2) Vehicle passing, queuing, and delay analysis: Assume that during the green light period, when there is a queue, the traffic flow passes through the target single-point intersection at a saturation flow rate, and when there is no queue, the traffic flow passes through the target single-point intersection at an arrival flow rate; based on the vehicle arrival prediction results, a cumulative number of arriving vehicles-time graph is drawn, and combined with the decision variable (t k and ) values to draw the cumulative number of vehicles passing through - time diagram, and then calculate the corresponding value of each decision variable group

[0209] In summary, the online signal control method for a single-point intersection based on the spatiotemporal trajectory data of vehicles on the entrance lane proposed in an embodiment of the present invention meets the driver's demand for a traffic light countdown. At the same time, it can make signal control decisions that match the traffic demand based on the real-time traffic conditions of the target single-point intersection, thereby ensuring the fair distribution of the right of way and improving the utilization rate of the spatiotemporal resources of the target single-point intersection.

[0210] By acquiring the spatiotemporal trajectory data of vehicles at the target single-point intersection through the entrance vehicle detection system, a comprehensive traffic condition analysis of the target single-point intersection is realized, providing a guarantee for generating signal timing that matches traffic demand.

[0211] By setting a countdown duration requirement, the green light time for each phase is divided into two stages: the solution update stage and the countdown stage. The former supports real-time response of signal timing to changes in traffic demand, while the latter satisfies drivers' reliance on signal light countdown.

[0212] The dynamic segmentation of traffic flow simplifies the spatiotemporal feature analysis process and ensures timely feedback of real-time data.

[0213] Based on the principle of minimizing total delay, an online signal control optimization model for the target single-point intersection is constructed under the constraints of minimum green light time, maximum red light time, maximum queue length, and maximum average vehicle delay. This model is conducive to improving the service level of the target single-point intersection.

[0214] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of an embodiment, and the modules or processes in the accompanying drawings are not necessarily required to implement the present invention.

[0215] From the above description of the embodiments, it can be seen that those skilled in the art can clearly understand that the present invention can be implemented by means of software plus the necessary general-purpose hardware platform. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments of the present invention or certain parts of the embodiments.

[0216] Each embodiment in this specification is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the device or system embodiments, since they are basically similar to the method embodiments, the description is relatively simple. For the relevant parts, refer to the partial description of the method embodiments. The device and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. A person of ordinary skill in the art can understand and implement it without making any creative efforts.

[0217] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A single-point intersection signal online control method based on the spatiotemporal trajectory data of vehicles on the entrance lane, characterized in that: include: Step S0: Determine the vehicle data acquisition frequency, obtain the spatial morphological information, spatial division scheme, signal timing basic scheme and traffic characteristic parameters of the target single-point intersection, and determine the constraint conditions of each phase of the target single-point intersection; Step S1: Obtain vehicle data at the target single-point intersection at time T and update the cumulative delay of each vehicle at the entrance of the target single-point intersection; Step S2: Obtain traffic light information for the target single-point intersection and determine whether time T is in the plan update phase. The plan update phase is the period between the time when the green light of a phase turns on and the time when the green light countdown is displayed. If so, execute steps S3 to S6. Otherwise, calculate the optimal plan after time T+1 based on the optimal timing plan generated by the genetic algorithm at the last moment of the previous plan update phase, and execute step S6. Step S3, dynamically segmenting each traffic flow and analyzing the spatiotemporal distribution characteristics of traffic flows in each phase of the target single-point intersection at time T; Step S4, establishing an online signal control optimization model for the target single-point intersection based on the spatiotemporal distribution characteristics of traffic flow in each phase of the target single-point intersection and the constraints of each phase; Step S5, solving the target single-point intersection signal online control optimization model based on a genetic algorithm to obtain the optimal signal timing plan after time T+1; Step S6, controlling the traffic lights at the target single-point intersection according to the optimal signal timing plan after time T+1, includes: S6-1) In the optimal signal timing scheme finally obtained, if t 0 >t c +Δt, where t 0 Represents the remaining green light time of the current green light phase, t c represents the required duration of the green light countdown, Δt represents the frequency of data acquisition, determines that the countdown is not displayed at time T+1, and after acquiring the vehicle data at time T+1, repeats steps S1-S6; S6-2) In the final optimal signal timing solution, if t 0 =t c +Δt, it is determined that the countdown starts at time T+1, and the traffic light is controlled according to the scheme until the green light of the next green light phase is turned on.

2. The method according to claim 1, characterized in that The step S0, determining the vehicle data acquisition frequency, obtaining the spatial morphological information, spatial division scheme, signal timing basic scheme and traffic characteristic parameters of the target single-point intersection, and determining the constraint conditions of each phase of the target single-point intersection, includes: S0-1) Determine the vehicle data acquisition frequency Δt, and the calculation formula of Δt is as follows: Where t1 represents the data collection time required to obtain vehicle data at a certain moment, and t2 represents the calculation time required for the single-point intersection signal online control method; S0-2) obtaining spatial morphological information of the target single-point intersection, the spatial morphological information including the number and type of lanes at each entrance, and the positions of each lane and stop line; S0-3) determining the space occupied by a standard vehicle, and dividing each lane in the entrance area of the target single-point intersection into a number of cells of equal length using the space occupied by the standard vehicle as a unit length; S0-4) Determine the basic signal timing plan for the target single-point intersection, including the green light countdown time requirements, phase division, and green light interval time required for phase transition; S0-5) Determine the constraints for each phase of the target single-point intersection. These constraints include: a minimum green light time constraint, which is determined by the shortest time required for pedestrians to cross the street in that phase; a maximum red light time constraint, which is determined by the maximum allowable waiting time for pedestrians before crossing the street; a maximum vehicle delay constraint, which is determined based on the traffic volume and importance of the phase; and a maximum queue length constraint, which is determined based on the road conditions and the maximum queue length allowed for each vehicle flow in each phase. S0-6) Determine the traffic characteristic parameters of the target single-point intersection: including the saturation flow rate of each phase at the stop line section.

3. The method according to claim 2, characterized in that The step S1, obtaining vehicle data of the target single-point intersection at time T and updating the accumulated delay of each vehicle at the entrance of the target single-point intersection, includes: S1-1) Deploy a phased array radar at each entrance to the target single-point intersection, establish an entrance vehicle detection system, and adjust the stop line and lane line at each entrance to be parallel to the X-axis and Y-axis of the corresponding phased array radar coordinate system; obtain vehicle data detected by the entrance vehicle detection system at the target single-point intersection at time T, the vehicle data including the ID assigned to each vehicle by the entrance vehicle detection system, as well as the speed and position of the vehicle; S1-2) updating the cumulative delay of each vehicle in the entrance lane based on the vehicle speed data.

4. The method according to claim 3, characterized in that Step S2, obtaining traffic light information of the target single-point intersection, and determining whether time T belongs to the solution update phase, if so, executing steps S3 to S6; otherwise, calculating the optimal solution after time T+1 based on the optimal timing solution generated by the genetic algorithm at the last moment of the previous solution update phase, and executing step S6, including: S2-1) Obtaining traffic light information at the target single-point intersection at time T, the traffic light information including the light color corresponding to each phase, the duration of the light color, and whether a green light or red light countdown is being displayed; S2-2) Determine whether time T belongs to the solution update phase. If so, execute step S2-4; otherwise, execute step S2-3; S2-3) When time T does not belong to the plan update phase, the optimal plan after time T+1 is calculated based on the optimal timing plan generated by the genetic algorithm at the last moment of the previous plan update phase, and step S6 is executed; S2-4) When time T belongs to the plan update stage, the processing of steps S3 to S6 is continued to complete the plan update and traffic light control.

5. The method according to claim 4, characterized in that The aforementioned S2-1) obtains traffic light information of the target single-point intersection at time T, the traffic light information including the light color corresponding to each phase, the duration of the light color, and whether the countdown to the green light or red light is being displayed, including: The judgment results of the target single-point intersection traffic status and whether the plan is updated corresponding to the different traffic light information at time T are divided into the following three cases: (1) A phase is in green light status, but the signal light has not yet displayed the green light countdown: the traffic flow corresponding to this phase has the right of way. This moment belongs to the scheme update stage, and the subsequent signal timing needs to be updated based on the vehicle data at this moment; (2) A phase is in green light state, and the traffic light is showing the green light countdown: the traffic flow corresponding to this phase has the right of way. This moment does not belong to the scheme update stage. When the green light countdown ends, the traffic state of the target single-point intersection enters situation (3); (3) No green light phase: Traffic in all phases has no right of way. This moment does not belong to the plan update stage, and the subsequent signal timing plan has been determined.

6. The method according to claim 4, characterized in that The step S3 dynamically segments each traffic flow and analyzes the spatiotemporal distribution characteristics of traffic flows in each phase of the target single-point intersection at time T, including: S3-1) Determine vehicle type: Based on the spatiotemporal trajectory data of traffic flow at time T and before, a vehicle speed equal to 0 is used as a sign that the vehicle has arrived at the target single-point intersection. Vehicles with a speed equal to 0 at time T or before are determined to be arriving vehicles, and the remaining vehicles are considered upstream vehicles. S3-2) Set up the traffic flow matrix: Based on the position of the vehicles at time T, follow the one-to-one principle and map all vehicles to specific cells in the entrance lane. Based on the cell matching results, for each traffic flow, set the position matrix M of the arrival traffic flow in the entrance lane. ap and the delay matrix M ad , where the upstream traffic flow is set at the position matrix M of the entrance lane up and velocity matrix M uv ; Position matrix M ap , delay matrix M ad , position matrix M up and velocity matrix M uv The size of the matrix M is m rows and n columns, where m = the number of lanes corresponding to the traffic flow, n = the length of the lane corresponding to the traffic flow within the study range / the length of the cell, and the first column represents the cell next to the stop line; the position matrix M ap With M up , velocity matrix M uv and the delay matrix M ad All are initialized to 0 matrices. If there is a vehicle in a cell, the element in the position matrix corresponding to the cell is changed to 1, the element in the speed matrix is changed to the speed of the vehicle, and the element in the delay matrix is changed to the accumulated delay of the vehicle. S3-3) Count vehicle arrivals: The distance between the last arriving vehicle in each traffic flow and the stop line is recorded as the queue length L at time T. 0 , the number of arriving vehicles in the traffic flow is recorded as The sum of the cumulative delays of the arriving vehicles in this traffic flow is recorded as L 0 、 and The calculation formula is as follows, where L pcu Indicates the cell length: L 0 =max(J a )×L pcu (1) J a (r,c)=M ap (r,c)×c (2) S3-4) Divide the upstream fleet: Determine the fleet division threshold L f , when the distance between two moving vehicles is less than L f , the two vehicles are classified into the same fleet; the cells corresponding to the elements in the same column of the upstream traffic flow position matrix are merged into a unit area, the number of vehicles in each unit area is calculated, and the unit area is marked as "no car" or "car" according to the number of vehicles; if the states of two adjacent unit areas are different, the dividing line between the two unit areas is used as the alternative dividing line for traffic flow segmentation; if the area between the two adjacent alternative dividing lines is "no car state" and the distance between the two is less than L f Delete these two alternative dividing lines; S3-5) Calculate the fleet parameters: Consider each fleet as a uniform entity, and use s for the xth fleet in the upstream traffic flow. x，1 Indicates the distance between the front of the convoy and the stop line, s x，2 represents the distance between the rear end of the convoy and the stop line, q x 、v x 、k x 、N x represents the flow, speed, density, and number of vehicles of the fleet, c x，1 、c x，2 Indicates the matrix column subscript corresponding to the cell where the first and last vehicles of the team are located, L pcu represents the space occupied by a standard vehicle, and the calculation formula for the relevant parameters of the fleet is as follows: s x，1 =L pcu ×(c x，1 ―1) (5) s x，2 =L pcu ×c x，2 (6) 7. The method according to claim 6, characterized in that The step S4, based on the spatiotemporal distribution characteristics of traffic flow in each phase of the target single-point intersection and the constraints of each phase, establishes an online signal control optimization model for the target single-point intersection, including: S4-1) Select the decision variables and target variables of the target single-point intersection signal online control optimization model, and take the n after time T. k time periods, numbered with k, and t k represents the duration of time period k, represents the status of the signal light in phase i during time period k, Indicates green light; Indicates yellow or red light, with t k and As the decision variable; D represents the number of all import vehicles detected at time T in n k The sum of the cumulative delays in each period, with D as the target variable; S4-2) Determine the expression of the objective function: Among them, K represents the time period set, P represents the phase set, represents the added delay of phase i in time period k, The calculation of is related to the decision variable values and the results of the spatiotemporal characteristics analysis of the traffic flow in phase i in S3; S4-3) Determine the constraints: 1≤k≤n k ,k∈Z (24) t k ∈Z + (25) Among them: (12) (13) In order to meet the basic conditions of the signal timing plan; (14) indicates that the status of the signal lights of each phase in period 1 should be consistent with time T, represents the signal light status of phase i at time T; (15) indicates that if there is a phase in the green light state in time period k, then t k It must be at least longer than the green light countdown time requirement, t c represents the duration requirement of the green light countdown, Δt represents the frequency of data acquisition; (16)(17) In order to meet the minimum green light time constraint, G 0 Represents the green light phase at time T, i=G 0 Indicates the duration of the green light in this phase. represents the minimum green light time of phase i; (18)(19) In order to meet the maximum red light time constraint, R 0 Represents the red light phase set at time T, i∈R 0 Indicates the duration of the red light phase up to time T. represents the maximum red light time of phase i, and M represents the maximum value; (20) In order to meet the green light interval time constraint, I is a matrix, and I(i, j) represents the green light interval time required for the green light phase to switch from i to j; (21) In order to ensure that all the vehicles entering the lane detected at time T can enter the lane n k Pass the target single-point intersection in the time period, represents the number of vehicles entering the lane detected at time T in phase i, represents the number of vehicles passing through the target single-point intersection in phase i during time period k; (22) In order to satisfy the maximum average vehicle delay constraint in each phase, represents the accumulated delay of phase i up to time T, represents the cumulative delay of phase i in time period k; (23) In order to meet the maximum queue length constraint of the corresponding traffic flow in each phase, represents the maximum queue length of vehicle flow m corresponding to phase i in time period k, represents the maximum queue length constraint of vehicle flow m corresponding to phase i; (24)(25)(26) are the values of the variables.

8. The method according to claim 7, characterized in that The step S5, solving the target single-point intersection signal online control optimization model based on a genetic algorithm to obtain the optimal signal timing plan after time T+1, includes: S5-1) Coding: Set two chromosomes for each individual in the genetic algorithm. Chromosome I represents the phase sequence, and chromosome II represents the corresponding duration. The length of both chromosomes is n. p ×n c , where n p is the number of phases, and is 1, 2,, n p As the phase number, n c is the preset number of cycles; the encoding method of chromosome I is: 1, 2,, n p ×n c Randomly shuffle the order and move the number of the current green light phase to the first position; the encoding method of chromosome II is: randomly generate n p ×n c A number greater than the countdown duration requirement and less than 60s; S5-2) Decoding: In chromosome Ⅰ, i, i+n p 、i+2×n p ,,i+(n c ―1)×n p Indicates the phase numbered i, according to the 1, 2, n p ×n c The order of the phases obtaining the right of way after time T is determined by the sequence of chromosomes II and I. Chromosome II corresponds to chromosome I and represents the green light time allocated to the phases of the corresponding order. If there are multiple adjacent numbers representing the same phase in chromosome I, the position of the first digit is recorded as the reserved position, the positions of the following digits are recorded as the deleted positions, and the corresponding digits are deleted. At the same time, the digits corresponding to the deleted positions in chromosome II are added to the reserved positions and deleted as the green light time allocated to the phase. S5-3) Calculate fitness: According to the decoded timing plan, calculate the objective function value D of the signal control optimization model under the plan; check whether each constraint condition is satisfied. If there is a constraint condition h that is not satisfied, calculate the corresponding penalty value e h ; The calculation formula of fitness is as follows: S5-4) Determine the termination condition: set the number of iterations as the termination condition; S5-5) Selection: Eliminate individuals whose fitness ranks in the bottom third of the population; S5-6) Crossover: Select individuals ranked in the top third of fitness and pair them up for crossover. Chromosomes I and II cross simultaneously, and each crossover produces two new individuals. S5-7) Mutation: The top 30 individuals in fitness are retained, and the remaining individuals mutate according to a certain probability. There are two types of mutation: one is to randomly swap the phase order of two phases, and the other is to randomly change the green light duration of one phase between 50% and 150%. When an individual needs to mutate, a certain number of offspring are generated according to the two mutation types, and the offspring with the highest fitness is selected as the mutation result of the individual; S5-8) Determine the optimal signal timing scheme after time T+1, the optimal signal timing scheme includes the remaining green light time t of the current green light phase at time T+1 g_after(T+1) , the next green light phase P g_next The selection of , and the initial value of the green light time t set for the next green light phase g_next After the iteration ends, select the individual with the largest fitness, decode it, and take the first and second digits in chromosome II as t g_after(T+1) and t g_next , take the phase corresponding to the second digit in chromosome I as P g_next .

Images