A bus trunk multi-intersection conditional signal priority control method

Abstract

The application discloses a kind of bus trunk multi-intersection conditional signal priority control method, first, determine optimization target, i.e. improve the punctuality performance of BRT system and minimize the adverse effects on private car;Second, establish conditional TSP model, the delay constraint of vehicle, traffic priority time constraint and saturation and queue length constraint, ensure that vehicle arrives on time, balance the relationship between bus punctuality and private car delay;Third, collect the motor vehicle flow demand of station and station section intersection, intersection signal timing, upstream bus station departure time and downstream bus station arrival time data;Finally, the model is solved using mixed integer nonlinear programming method.The application solves the problem of unexpected arrival time of bus vehicles at continuous intersection stop lines affecting the reliability of the operation of the bus rapid transit system, improves the punctuality performance of the bus rapid transit system, minimizes the adverse effects on private cars, and improves the overall efficiency of the transportation system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent traffic signal control, and in particular to a method for conditional signal priority control at multiple intersections along a bus trunk. BACKGROUND

[0002] The rapid development of urbanization and the rapid growth of urban population have made the traffic congestion problem in China's urban areas more and more serious. As an effective measure to alleviate traffic congestion, public transportation, especially bus operation, has always been considered as an effective means to reduce the burden of urban roads due to its high passenger carrying capacity. In order to reduce the interference of private cars on the road to bus service, the Bus Rapid Transit (BRT) system is widely used. The BRT system uses dedicated bus lines to provide a more efficient travel mode. However, similar to traditional bus operation, BRT vehicles often encounter red lights when approaching intersections, resulting in unnecessary delays for passengers.

[0003] In existing research, Transit Signal Priority (TSP) is considered as an effective and economic solution to improve the service level of BRT system. There are mainly three kinds of TSP control strategies: passive control, active control and adaptive control. Passive TSP control is mainly based on historical bus information for offline optimization, while active TSP control is based on real-time detection of bus priority request at the front intersection, taking measures such as early green light or extended green light. Adaptive TSP control combines advanced vehicle networking, traffic sensing and signal control technology, which can optimize the traffic performance of buses and private cars in real time.

[0004] Although TSP has achieved remarkable results in reducing bus vehicle delays, as a time-tabled transportation tool, BRT system should pay more attention to the punctuality of its operation. For passengers, bus punctuality is the most direct indicator to evaluate the reliability of bus operation. The unexpected arrival time of BRT vehicles at consecutive intersections will inevitably affect the operational reliability of BRT system. Therefore, when optimizing BRT operation, not only the TSP of intersections should be considered, but also the TSP between road sections should be considered to improve the overall operational reliability. SUMMARY

[0005] The present application proposes a method for conditional signal priority control at multiple intersections along a bus trunk, which optimizes signal priority time to improve the punctuality of the Bus Rapid Transit (BRT) system and reduce the negative impact on private car traffic. It is not only suitable for BRT systems in specific cities or regions, but also suitable for other urban transportation systems that need to balance the improvement of bus service level and the reduction of private car delays.

[0006] Technical solution: The method for conditional signal priority control at multiple intersections along a bus trunk of the present application comprises the following steps:

[0007] Step (1): Data acquisition, obtain the signal timing plan of the intersection, the timetable of the bus, the phase difference between intersections, the distance between the stop and the intersection, the vehicle running speed between the stop and the intersection, the distance between intersections, and the speed between intersections.

[0008] Step (2): Set a vehicle detector at the intersection described in step (1), the distance between the vehicle detector and the intersection is greater than the maximum queue length, and the arrival time of the bus is obtained through the vehicle detector.

[0009] Step (3): Determine the objective function, set two priority targets to coordinate the competition between arterial bus vehicles and private cars, and give different weights to obtain the overall objective function, the process is:

[0010] Step (3.1), determine the first priority target: reduce the bus vehicle's schedule delay at the downstream bus stop, and eliminate the problem of invalid bus signal priority.

[0011] Step (3.2), determine the second priority target: reduce the total traffic priority time allocated to the intersection, and thereby reduce the side effects of TSP on private cars.

[0012] Step (3.3), give the final objective function;

[0013]

[0014] Wherein, P1 is the weight coefficient for distinguishing the priority of the two control targets, P2 is the weight coefficient for distinguishing the priority of the two control targets, P1>>P2; ΔT is the deviation of the downstream bus station timetable; G ib is a decision variable, representing the bus priority time generated by the early green light at intersection i; G ie is a decision variable, representing the bus priority time generated by the green light extension strategy at intersection i; N is the number of intersections on the stop link.

[0015] Step (4): Establish a conditional TSP model, set constraints on the delay relative to the bus timetable, the bus arrival time, the bus departure time, the bus phase start / end time and the bus priority time. Among them, the upper limit of the bus priority time in the bus priority time constraint is obtained based on the queue overflow constraint and the saturation constraint.

[0016] Step (4.1), set the delay constraint relative to the bus schedule: when the actual arrival time of the bus exceeds the scheduled arrival time in the downstream station schedule, the delay relative to the bus schedule is equal to the difference between the actual arrival time and the scheduled arrival time of the bus at the downstream station; when the actual arrival time of the bus is earlier than the scheduled time of the downstream station, there is no delay for passengers on the bus or waiting at the station, and the delay relative to the bus schedule is equal to 0;

[0017] The relative bus schedule delay constraint is expressed as:

[0018]

[0019] Where, t f is the actual arrival time of the bus at the downstream station, is the scheduled arrival time of the bus at the downstream station.

[0020] Step (4.2), set the traffic arrival time constraint: due to the phase sequence and signal state of each intersection, the time when the bus arrives at intersection i determines whether it encounters a red light or a green light, which needs to follow the following traffic arrival time constraint:

[0021] Step (4.3), set the bus departure time constraint: according to the arrival time, the departure time of the bus at each intersection is divided into two cases: if the bus arrives during the green light at intersection i, it will pass through the intersection directly; if the bus arrives during the red light at intersection i, it will wait for the start time of the bus phase in the next signal cycle.

[0022] The bus departure time constraint formula is as follows:

[0023]

[0024] Where, t ip is the time when the bus passes through intersection i;

[0025] Step (4.4), set the bus phase start or end time constraint.

[0026] Step (4.5), TSP will cause queue overflow problem on the branch due to oversaturation condition. In order to determine the bus priority time, it is necessary to ensure that the queue in each direction does not overflow to the adjacent downstream intersection; according to the queue length required for each non-priority traffic flow to return to the initial state and the maximum queue vehicles of the jth phase of the ith intersection, the upper limit of the bus priority time under the queue overflow constraint is calculated. For example, Figure 1As shown, the maximum queued vehicles only appear in the second signal cycle, therefore, the maximum transit priority time for each intersection is limited by the maximum queued vehicles that appear in the second signal cycle for each intersection phase.

[0027] The maximum queued vehicles for the ith intersection jth phase in the case is calculated as follows:

[0028]

[0029] Where, N ijmax is the maximum queued vehicles for phase j at intersection i; is the number of trapped vehicles in the first signal cycle; Δt ij is the green loss time for phase j at intersection i in the case of transit signal priority; is the red duration for phase j at intersection i in the case of no transit signal priority.

[0030] To prevent the queued vehicles of non-priority flow from overflowing to adjacent intersections, the maximum queue length is less than the minimum distance between two intersections, and the constraint formula is as follows:

[0031]

[0032] Where, l s is the queue length of each standard vehicle; is the spatial queue length limit of the upstream road segment of phase j at intersection i.

[0033] Step (4.6), set the transit priority time constraint: the additional green time allocated for the priority phase is actually at the expense of reducing the green duration of other phases at intersection i. The maximum transit priority time generated under the early green and green extension strategy should be appropriately limited; considering that the TSP green provided for transit vehicles has a negative impact on private vehicles in other directions (conflicting flow), the saturation and queue overflow are considered to limit the upper limit of the transit priority time at intersection i;

[0034] The formula of the transit priority time constraint is as follows:

[0035]

[0036]

[0037] Where, is the upper limit of the transit priority time at intersection i under the saturation constraint; is the upper limit of the transit priority time at intersection i under the queue overflow constraint.

[0038] The upper limit of the bus priority time under the saturation constraint at intersection i is calculated by the following formula The upper limit of the bus priority time under the queue spill constraint at intersection i is calculated by the following formula

[0039]

[0040] where g ij is the green time of phase j at intersection i under the standard signal control scheme without considering bus priority or other special traffic management strategies, and represents the maximum additional green time provided by phase j.

[0041] The upper limit of the bus priority time under the saturation constraint at intersection i is calculated by the following formula The upper limit of the bus priority time under the queue spill constraint at intersection i is calculated by the following formula

[0042] The upper limit of the bus priority time under the queue spill constraint at intersection i is calculated by the following formula The upper limit of the bus priority time under the queue spill constraint at intersection i is calculated by the following formula

[0043]

[0044] where represents the maximum green loss time of phase j at intersection i.

[0045] Step (4.7), according to the saturation definition, the minimum required green time of phase j at intersection i under the maximum saturation is calculated, the formula is as follows:

[0046]

[0047] where is the minimum required green time of phase j at intersection i under the maximum acceptable saturation; q ij is the traffic flow of phase j at intersection i; s ij is the saturation flow rate of phase j at intersection i; c i is the signal cycle length of intersection i; X i is the maximum acceptable saturation of intersection i.

[0048] Step (5): The conditional TSP model is a mixed integer nonlinear programming problem, which needs to be converted into a linear problem. By introducing binary variables, the nonlinear constraints in the conditional TSP model are linearized to facilitate the solution of the conditional TSP model.

[0049] Step (5.1), introducing binary variables to linearize nonlinear constraints, adding binary variables: δ represents the condition that the bus arrives at the downstream stop later than scheduled (1-yes, 0-no); β i represents the condition that the bus arrives at the intersection i during the green light (1-yes, 0-no); γ i represents whether the current saturation is lower than the preset maximum saturation (1-yes, 0-no).

[0050] Step (5.2), linearizing nonlinear constraints: linearizing the delay constraints relative to the schedule, bus arrival time constraints, bus departure time constraints and saturation constraints in the conditional TSP model,

[0051] Step (6): solving the conditional TSP model: the conditional TSP model is a mixed integer linear programming model after linearization, and the model is solved by using the intlinprog function in MATLAB, which is a function in MATLAB for solving mixed integer linear programming MILP problems.

[0052] The process of step (4.2) is as follows:

[0053] Step (4.2.1), calculating the arrival time of the bus at the intersection: the time when the bus arrives at the first intersection from the upstream bus stop is related to the departure time from the upstream bus stop and the travel time between the upstream bus stop and the first intersection; the time when the bus arrives at other intersections along the trunk is determined by the time of departure from the upstream intersection and the travel time between two adjacent intersections;

[0054] The arrival time of the bus at the intersection is as follows:

[0055]

[0056] Where, t ia represents the time when the bus arrives at the i-th intersection; t s represents the time of departure from the upstream stop; t (i-1)p represents the time when the bus passes through the i-1-th intersection; d (i-1)i represents the distance of each section on the trunk; v (i-1)i represents the speed of the bus on each section of the trunk; represents the travel time of the bus from the i-1-th intersection to the i-th intersection;

[0057] Step (4.2.2), constraint the signal state when bus arrives: since the signal of intersection runs in a cycle, the bus always arrives in a certain signal cycle, if the bus arrives at the ith intersection earlier than the end time of the transition phase, the bus arrives in the effective green time; otherwise, it arrives in the red time, since the concept of effective green time is used, the yellow time is not calculated separately;

[0058] The signal state constraint formula when the bus arrives is as follows:

[0059]

[0060] Where g ib is the start of the effective green time of the transition phase at intersection i; α i is an integer variable representing the number of signal cycles the bus experiences at intersection i; C i is the signal cycle length at intersection i; g ie is the end of the effective green time of the transition phase at intersection i; β i is a binary variable that determines whether the bus arrives at the green light at intersection i, 1 means yes, 0 means no;

[0061] Step (4.2.3), calculate the actual arrival time of the downstream bus station: the actual arrival time of the downstream station is related to the passing time of the bus at the last intersection on the main road, and the formula is as follows:

[0062]

[0063] Where d if represents the distance between the last intersection and the downstream station; v if represents the speed of the bus between the last intersection and the downstream station.

[0064] In step (4.4), in order to reduce the waiting time of the bus at each intersection and improve the punctuality performance of the bus at the downstream station, the TSP strategy is introduced in the model to provide priority for approaching buses, considering the following two aspects:

[0065] Step (4.4.1), calculate the actual start time of the transition phase at the intersection under the bus priority signal control strategy: if the bus arrives during the red light and the arrival time is close to the start time of the transition phase in the next signal cycle, introduce the early green light strategy to realize the bus signal priority. In order to adapt to this situation, the actual start time of the transition phase is several seconds earlier than the original signal time plan.

[0066] The actual start time of the transition phase of the intersection under the above-mentioned condition is as follows:

[0067]

[0068] where g ib is the actual start time of the transition phase at intersection i; is the original start time of the transition phase at intersection i; G ib is the priority time allocated for the early green strategy at intersection i;

[0069] Step (4.4.2), the actual end time of the transition phase at the intersection under the transit priority signal control strategy is calculated, when the arrival time of the bus lags behind, but approaches the end time of the transition phase in the current signal cycle, the green extension strategy is introduced to ensure that the bus does not need to stop to pass through the intersection;

[0070] The formula of the actual end time of the transition phase at the intersection under the transit priority signal control strategy is as follows:

[0071]

[0072] where g ie is the actual end time of the transition phase at intersection i; is the original end time of the transition phase at intersection i; G ie represents the priority time generated by the green extension strategy at intersection i.

[0073] The process of step (5.2) is as follows:

[0074] Step (5.2.1), linearize the delay constraint relative to the schedule: introduce a binary variable δ and linearize it with a number M, the formula is as follows:

[0075]

[0076]

[0077] -Mδ≤ΔT≤Mδ(1.54)

[0078] In the formula, when δ is equal to 1; when δ is equal to 0; when δ=1, activate formula (1.53); the schedule deviation is defined as When δ=0, activate formula (1.54); the schedule deviation is defined as ΔT=0.

[0079] Step (5.2.2), linearize the bus arrival time constraint: introduce a binary variable β i and linearize it with a number M, the formula is as follows:

[0080]

[0081] If the bus arrives at the intersection i before the end of the transition phase, i.e., β i = 1, as shown in equation (1.25). Otherwise, the bus arrives during the red phase (β i = 0), as shown in equation (1.26).

[0082] Step (5.2.3), linearize the bus departure time constraint: Introduce a binary variable β i to linearize with a number M, as follows:

[0083]

[0084] If the bus encounters a green light at intersection i (β i = 1), no delay is incurred, and the bus departure time is defined as t ip = t ia . Otherwise, the bus departure time is the start of the green light in the next signal cycle, defined as t ip = g ib + (a i + 1)C i . Equations (1.27) and (1.28) are activated when β i are equal to 1 and 0, respectively.

[0085] Step (5.2.4), linearize the saturation constraint: Introduce a binary variable γ i to linearize, as follows:

[0086]

[0087] Equation (1.29) is activated when γi= 1. When γi= 0, the upper bound of the priority time is 0, and no priority is provided. In addition, equation (1.30) is added to distinguish the cases. If denotes that there is still room to generate a bus priority green light, the upper bound of the priority time is valid and γi= 1. Otherwise, γi= 0.

[0088] Working principle: the process of the bus trunk multi-intersection conditional signal priority control method is as follows: firstly, the optimization target is determined, that is, the punctuality performance of the BRT system is improved and the adverse effect on private cars is minimized; secondly, the conditional TSP model is established, the delay constraint of the traffic vehicle, the traffic priority time constraint, and the saturation and queue length constraint are considered, so that the traffic vehicle can arrive at the destination on time, and the relationship between the punctuality of the bus vehicle and the delay of the private car is balanced; thirdly, relevant data are collected, including the motor vehicle flow demand of the intersection between the station and the station section, the signal timing of the intersection, the departure time of the upstream bus station and the arrival time of the downstream bus station; finally, the model is solved by using the mixed integer nonlinear programming method.

[0089] The specific process is as follows: the data processing and traffic signal control system are used to optimize the signal priority control of the bus trunk multi-intersection, the real-time detection and optimization model are used to improve the punctuality of the bus, and the negative effect on the private car traffic is reduced. Firstly, the road data and vehicle data are collected; secondly, the data are processed to obtain the required data; thirdly, the data are substituted into the conditional TSP model to obtain the optimized scheme; and finally, the optimized scheme is implemented.

[0090] The conditional priority strategy is provided as follows: only the bus with the arrival time later than the schedule is provided with the priority time; only when the bus punctuality performance is improved, the priority time is provided; and only when the saturation and queue overflow constraint condition is met, the priority time is provided.

[0091] Meanwhile, the conditional TSP model is constructed by considering the constraints of the schedule deviation, the bus arrival time, the bus passing time, the bus phase start / end time and the priority time, and the mixed integer nonlinear programming (MINLP) method is used to solve the model by using the branch and bound technique.

[0092] The implementation module of the present application comprises a data acquisition and detection module, a data processing module, an optimized signal priority control module and a terminal signal control module, which respectively process the data acquisition, the data processing, the model solving and the scheme implementation process. The specific process is as follows:

[0093] Firstly, the data acquisition and detection module detects the actual arrival time of the bus in real time through the detector installed in front of the upstream bus station, and the detector sends the collected data to the traffic signal control system.

[0094] Secondly, the data processing module calculates the time deviation based on the actual arrival time and the scheduled arrival time of the bus. The control system calculates the time deviation of the bus arriving at the intersection and determines whether to take signal priority measures: if the actual arrival time of the bus is later than the scheduled arrival time, there is a time delay, and signal priority measures are taken to reduce the delay; if the actual arrival time of the bus is earlier than the scheduled arrival time, no signal priority control is performed.

[0095] Thirdly, the optimized signal priority control module substitutes the data into the calculation to determine the specific implementation of the signal priority strategy. Finally, the terminal signal control module adjusts the priority time of the intersection signal light in real time based on the optimization result to ensure that the bus passes through the intersection on time while reducing the impact on private car traffic.

[0096] Advantages: Compared with the prior art, the present application has the following advantages:

[0097] (1) Improve the punctuality of the bus: In view of the influence of the unexpected arrival time of the bus at the continuous intersection on the operation reliability of the BRT system, the signal priority time is optimized to reduce the delay of the bus at the intersection, so that the bus arrives at each station on schedule according to the predetermined timetable, thereby improving the punctuality of the BRT and improving the quality of public transportation services.

[0098] (2) Reduce the negative impact on private car traffic: While providing signal priority for buses, the signal priority strategy is optimized to reduce the delay of private cars by considering the flow and driving conditions of private cars, so as to improve the traffic efficiency of buses without affecting the traffic efficiency of private cars, achieving the purpose of balancing public transportation and private car traffic.

[0099] (3) Improve the efficiency and intelligent level of traffic management: By establishing a mixed integer nonlinear programming MINLP model, considering the constraints of arrival time deviation, saturation and queue length, the signal priority time is optimized in real time to enhance the accuracy and response speed of traffic signal control and improve the intelligent level of the traffic management system.

[0100] (4) Reduce the operating cost of the bus: By reducing the delay of the bus and improving its punctuality, the operating cost of the bus is reduced, including fuel consumption and driver working time, thereby improving the economic benefit of the bus company.

[0101] (5) Improve the urban traffic environment: By optimizing the traffic efficiency of the bus, reducing the dwell time of the bus at the intersection, it helps to reduce vehicle emissions, reduce air pollution and improve the urban traffic environment, providing a greener and healthier travel condition for urban residents. BRIEF DESCRIPTION OF DRAWINGS

[0102] Figure 1 Flow chart of the bus trunk multi-intersection conditional signal priority control method of the present application;

[0103] Figure 2 Maximum queuing vehicle number in the supersaturation condition of the present application;

[0104] Figure 3 Intersection geometric configuration in the embodiment of the present application. DETAILED DESCRIPTION

[0105] Embodiment:

[0106] The bus trunk multi-intersection conditional signal priority control method of the present application obtains the maximum flow of each phase of the intersection, the saturation flow rate of each phase of the intersection, the signal light duration of each phase of the intersection, the cycle duration of the intersection, the shortest road section length of each phase of the intersection, the phase difference of adjacent intersections, the maximum queuing length, the schedule of the bus, the driving distance of the bus, and the speed data of the bus through the data acquisition and detection module, then transmits the data into the data processing module for data processing, substitutes the processed data into the optimization signal priority control module for calculation and solution to obtain an optimal scheme, and finally implements through the terminal signal control module.

[0107] As Figure 3 , the present application selects a continuous three-intersection, and the saturation flow of the three intersections is 1800 vehicles / hour. The signal cycle of all intersections is 100 seconds. The flow from the first phase to the fourth phase of the three intersections is 270 pcu / h, 126 pcu / h, 216 pcu / h, and 108 pcu / h, respectively. The green light duration from the first phase to the fourth phase of the three intersections is 30s, 14s, 24s, and 12s, respectively. The one-cycle green light loss time of each phase of each intersection is 5s. The shortest road section length of each phase of each intersection is 100m for the second phase, 200m for the third phase, and 100m for the fourth phase. The maximum queuing length is 6m / vehicle. The phase difference from the first intersection to the second intersection is -13s, and the phase difference from the second intersection to the third intersection is -33s. The bus travel distance of each segment of the trunk is known, which is 150m, 300m, 300m, and 150m, respectively. The bus driving speed is also known.

[0108] According to the known data, first, the bus priority time upper limit is obtained by processing: the bus priority time upper limit of each intersection determined according to the shortest road length and the bus priority time upper limit of each intersection determined according to the saturation degree are obtained, and the minimum value among them is the maximum green light early start and green light extension time. The bus priority time upper limit of the intersection determined according to the shortest road length and the saturation degree is as follows:

[0109]

[0110] The sum of the maximum green light early start and the maximum green light delay time of each intersection, i.e., the sum of the maximum bus priority time of each intersection. According to the known data, the maximum bus priority time of the intersection determined according to the shortest road length is 148.3s; assuming that the saturation is less than or equal to 0.85, the maximum bus priority time of each intersection determined according to the saturation is 20.588s. By comparing the two, it is analyzed that the maximum bus priority time of each intersection determined according to the saturation is smaller, and therefore it is selected as the maximum green light early start time and the maximum green light delay time of each intersection. The sum of the maximum green light early start and the maximum green light delay time of each intersection, i.e., the sum of the maximum bus priority time of each intersection, is 123.528s.

[0111] Then, the processed data is substituted into the optimized signal priority control module for solving the model. In this example, the mixed integer optimization method in MATLAB is adopted, and the intlinprog command is called for solving to obtain the optimized scheme. The final result of the solving is as follows: the time difference between the actual arrival of the bus and the schedule is 31.8s; the actual arrival time of the bus at the downstream stop is 181.8s; the arrival time of the bus at intersection 1 is 113.6s; the arrival time of the bus at intersection 2 is 140.9s; the arrival time of the bus at intersection 3 is 168.2s; the green light start time of the bus phase at intersection 1 is 69s; the green light start time of the bus phase at intersection 2 is 40s; the green light start time of the bus phase at intersection 3 is 23s; the green light end time of the bus phase at intersection 1 is 114s; the green light end time of the bus phase at intersection 2 is 86s; the green light end time of the bus phase at intersection 3 is 69s; the green light early start (red light early break) time at intersection 1 is 0s; the green light early start (red light early break) time at intersection 2 is 16s; the green light early start (red light early break) time at intersection 3 is 0s; the green light late break (green light extension) time at intersection 1 is 15s; the green light late break (green light extension) time at intersection 2 is 0s; the green light late break (green light extension) time at intersection 3 is 16s; the total priority time of each intersection is 47s. Finally, the terminal signal control module adjusts the priority time of the intersection signal in real time according to the optimization result, so as to ensure that the bus passes through the intersection on time and reduce the influence on private cars.

Claims

1. A bus trunk multi-intersection conditional signal priority control method, characterized in that: The method comprises the following steps: Step (1): data acquisition, acquiring signal timing of intersections, bus timetable, phase difference between intersections, distance from bus stop to intersection, vehicle running speed from bus stop to intersection, distance between intersections, and speed between intersections; Step (2): setting a vehicle detector at the intersection in step (1), the distance between the vehicle detector and the intersection being greater than the maximum queue length, and acquiring the arrival time of the bus through the vehicle detector; Step (3): determining a target function, the process being as follows: Step (3.1), determining a first priority target: reducing the delay of the bus vehicle at the downstream bus station to the timetable; Step (3.2), determining a second priority target: reducing the total traffic priority time allocated to the intersection; Step (3.3), giving the target function, the formula being as follows: Wherein, P1 is the weight coefficient for distinguishing the priority of two control targets, P2 is the weight coefficient for distinguishing the priority of two control targets, P1 >> P2; ΔT is the deviation of the downstream bus stop schedule; G ib is the decision variable, which represents the bus priority time generated by the early green light at intersection i; G ie is the decision variable, which represents the bus priority time generated by the green light extension strategy at intersection i; N is the number of intersections on the stop point link; Step (4): establishing a conditional TSP model, the process being as follows: Step (4.1), setting a delay constraint relative to the bus timetable: when the actual arrival time of the bus is later than the scheduled arrival time in the plan of the downstream station, the delay relative to the bus timetable is equal to the difference between the actual arrival time and the scheduled arrival time of the bus at the downstream station; when the actual arrival time of the bus is earlier than the scheduled time of the downstream station, the delay relative to the bus timetable is equal to 0; The delay constraint relative to the bus timetable is as follows: where t f denotes the actual arrival time of the bus at the downstream stop, denotes the planned arrival time of the bus at the downstream stop; Step (4.2), setting a traffic arrival time constraint condition; comprising the following process: Step (4.2.1), calculating the arrival time of the bus at the intersection, the formula being as follows: where t ia denotes the time of the bus arriving at the ith intersection; t s denotes the time of departure from the upstream station; t (i-1)p denotes the time of the bus passing the ith-1 intersection; d (i-1)i denotes the distance of each segment on the trunk line; v (i-1)i denotes the speed of the bus for each segment distance on the trunk line; denotes the bus travel time from the ith-1 intersection to the ith intersection; Step (4.2.2), constraining the state of the signal at the arrival time of the bus, the signal state constraint formula of the bus being as follows: where g ib is the beginning of the effective green time for the transit phase at intersection i; a i is an integer variable representing the number of signal cycles experienced by the bus at intersection i; C i is the length of the signal cycle at intersection i; g ie is the end of the effective green time for the transit phase at intersection i; β i is a binary variable determining whether the bus arrives at the green light at intersection i, 1 for yes and 0 for no. Step (4.2.3), calculating the actual arrival time of the downstream bus station, the formula being as follows: where d if denotes the distance of the road segment between the last intersection and the downstream station; v if denotes the speed of the bus between the last intersection and the downstream station; Step (4.3), setting a bus departure time constraint condition, the formula being as follows: where t ip is the time of the bus passing the intersection i; Step (4.4), setting a bus phase start or end time condition; Step (4.5), determining the bus priority time and avoiding queue overflow of each direction to the adjacent downstream intersection; according to the time queue length required for the non-priority traffic flow to return to the initial state and the maximum number of queued vehicles of the jth phase of the ith intersection, the upper limit of the bus priority time under the queue overflow constraint is calculated; The maximum number of queued vehicles of the jth phase of the ith intersection is calculated according to the following formula: where N ijmax is the maximum number of queued vehicles for phase j at intersection i; is the number of trapped vehicles in the first signal cycle; Δt ij is the green loss time for phase j at intersection i in the case of bus signal priority; is the red duration for phase j at intersection i in the case of no bus signal priority; The maximum queue length is less than the minimum distance between two intersections, and the constraint formula is as follows: where l s is the queue length of each standard vehicle; is the space queue length limit of the upstream link of phase j at intersection i; Step (4.6), setting a bus priority time constraint condition, the formula being as follows: wherein, is the upper limit of the transit priority time at intersection i under the saturation constraint; is the upper limit of the transit priority time at intersection i under the queue spill constraint; Step (4.7), according to the definition of the saturation degree, calculating the minimum required green light duration of the phase j at the intersection i under the maximum saturation degree, the formula being as follows: wherein, is the minimum green duration required for phase j at intersection i at maximum accepted saturation; q ij is the traffic flow for phase j at intersection i; s ij is the saturated flow rate for phase j at intersection i; c i is the signal cycle length for intersection i; X i is the maximum accepted saturation for intersection i; Step (5): introducing a binary variable to linearize the nonlinear constraint in the TSP model, the process being as follows: Step (5.1), introducing binary variables: δ represents the status of the bus arriving late at the downstream stop relative to the schedule (1 - yes, 0 - no); β i represents the status of the bus arriving at the intersection i during the green light (1 - yes, 0 - no); γ i represents whether the current saturation is lower than the preset maximum saturation (1 - yes, 0 - no); Step (5.2), linearizing the nonlinear constraint: linearizing the delay constraint relative to the timetable, the bus arrival time constraint, the bus departure time constraint and the saturation degree constraint in the TSP model; Step (6): solving the conditional TSP model.

2. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: In step (4.3), the bus departure time constraint is defined: according to the arrival time, the bus departure time at each intersection is divided into two cases: if the bus arrives at the intersection during the green light, the bus passes through the intersection; if the bus arrives at the intersection during the red light, the bus waits for the start time of the bus phase in the next signal cycle.

3. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: The process of step (4.4) is as follows: In step (4.4.1), the actual start time of the transition phase of the intersection under the bus priority signal control is calculated, and the formula is as follows: where g ib is the start of the effective green time of the transition phase at intersection i; is the original start time of the transition phase at intersection i; G ib is a decision variable representing the bus priority time resulting from the early green at intersection i; In step (4.4.2), the actual end time of the transition phase of the intersection under the bus priority signal control strategy is calculated, and the formula is as follows: where g ie is the end of the effective green time of the transition phase at intersection i; is the original end time of the transition phase at intersection i; G ie is the decision variable, representing the transit priority time generated by the green extension strategy at intersection i.

4. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: In step (4.5), the maximum queuing vehicles appear in the second signal cycle, and the queuing vehicles in the second signal cycle are used to limit the maximum bus priority time of the intersection.

5. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: The upper limit of the bus priority time under the saturation constraint at the intersection i in step (4.6) The formula is as follows: where g ij is the green time of the standard signal control plan at the intersection i for phase j without considering bus priority or other special traffic management strategies, denotes the maximum additional green time provided by phase j.

6. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: The upper limit of the bus priority time under the queue spill constraint at the intersection i in step (4.6) is obtained by inserting the formulas (1.42) and (1.43) into the formula (1.44) to obtain the maximum green loss limit of a phase, and is determined by the queue spill constraint. The bus priority time upper limit under the queue overflow constraint at the intersection i The formula is as follows: wherein, represents the maximum green loss time for phase j at intersection i.

7. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: The process of step (5.2) is as follows: In step (5.2.1), the delay constraint relative to the schedule is linearized: a binary variable δ and a number M are introduced for linearization, and the formula is as follows: -Mδ≤ΔT≤Mδ(1.54) wherein, when δ is equal to 1 ; when δ is equal to 0; when δ = 1, the formula (1.53) is activated; the schedule deviation is defined as when δ = 0, the formula (1.54) is activated; the schedule deviation is defined as ΔΤ = 0; Step (5.2.2), Linearization of the bus arrival time constraints: Introduce binary variable β i Linearization with number M, formula as follows: Wherein, if the arrival time of the bus at intersection i is earlier than the end time of the transition phase, the bus arrives during the green light, as shown in formula (1.25); otherwise, the bus arrives during the red light, as shown in formula (1.26); Step (5.2.3), linearization of the bus departure time constraint: Introduce binary variable β i Linearization with number M, formula as follows: where if the bus encounters a green light (β i = 1) at the intersection i, no delay is incurred and the bus departure time is defined as t ip = t ia ; otherwise, the bus departure time is the green start time of the next signal cycle, defined as t ip = g ib + (a i + 1)C i ; equations (1.27) and (1.28) are activated when β i are equal to 1 and 0, respectively; Step (5.2.4), linearization of saturation constraints: Introduce binary variable γ i Linearization is performed, formula as follows: wherein formula (1.29) is activated when γi = 1; when γi = 0, the upper limit of the priority time is 0; in addition, formula (1.30) is added to distinguish the case; if represents that there is a space to generate a bus priority green light, the upper limit of the priority time is valid and γi = 1; otherwise, γi = 0.

8. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: In step (6), the conditional TSP model is a mixed integer linear programming model after linearization.

9. The bus trunk multiple intersection conditional signal priority control method according to claim 1, characterized by: In step (6), when solving the conditional TSP model, the intlinprog function in MATLAB is used, which is a function in MATLAB for solving mixed integer linear programming problems.

Images