An optimization control method for intelligent connected bus operation based on heterogeneous route scenarios

Abstract

An operation optimization control method for an intelligent connected bus based on a heterogeneous line scenario relates to an operation optimization control method for an intelligent connected bus. The present invention aims to solve the technical problem that the coexistence of current manually driven bus lines and intelligent connected bus lines causes uneven headway time of intelligent connected buses traveling on bus lanes, as well as congestion at stations and intersections. The intelligent connected bus operation hierarchical optimization control model established by the present invention outputs the optimal speed control scheme and signal priority scheme under the condition of uncertainty in the operation of manually driven buses, improves the uniformity of headway time, and reduces the number of times of entering node queues. The present invention provides a new idea for the efficient operation of intelligent connected bus companies, and is of great significance to promoting the large-scale operation of intelligent connected buses from specific simple roads to complex roads.

Description

Technical Field

[0001] The present invention relates to an intelligent network-connected public transportation operation optimization control method. Background Art

[0002] As a crucial component of the transportation system, many cities are committed to improving the service quality and attractiveness of public transportation systems. The emergence of intelligent connected buses (ICBs) offers new opportunities for promoting efficient urban public transportation operations. They can provide a comprehensive, efficient, accurate, and reliable information service system, comprehensively elevating the level of intelligent public transportation. In recent years, various cities have been conducting pilot demonstrations of IBBs. In 2017, the first IBBs were successfully piloted in Shenzhen, China; in 2020, Changsha, China launched its first IBB route, "315"; and in 2023, Singapore conducted its first closed-road testing of IBBs in China. The application scenarios of IBBs are expanding from specific, simple roads to the complex routes of urban ground public transportation. Therefore, for a long time, public transportation will coexist with traditional manually driven bus routes (without any connectivity) and IBB routes. This has led to problems such as uneven headway times for IBBs operating on dedicated bus lanes and congestion at stops and intersections. Summary of the Invention

[0003] The present invention aims to solve the technical problems that the coexistence of manual bus lines and intelligent connected bus lines has caused uneven headway of intelligent connected buses traveling on bus lanes and congestion at stations and intersections, and to provide an intelligent connected bus operation optimization control method based on heterogeneous line scenarios.

[0004] The intelligent networked bus operation optimization control method based on heterogeneous line scenarios of the present invention is carried out in the following steps:

[0005] Step 1: Select a hybrid bus lane that includes an intelligent connected bus route and a manually driven bus route to build a heterogeneous route scenario. Establish an attribute set for the intelligent connected bus route in the heterogeneous route scenario to store route and node information. Establish an attribute set for the intelligent connected bus trip in the heterogeneous route scenario to store trip operation information.

[0006] The attribute set of the intelligent connected bus route includes the length of the non-queuing section, the length of the queuing section, the speed limit of the section, the location of nodes, and the timing information of the signal lights. The nodes include stations and signal lights. The queuing section is the section area before the stop line of the signal light, that is, the section area where buses need to queue when the signal light is in the red phase. The length of the queuing section is determined based on the historical maximum queue length. The non-queuing section is the section outside the queuing section.

[0007] The attribute set of intelligent connected bus trips includes the dispatch headway and initial departure schedule of the trips within the control period;

[0008] Step 2: Using V2X technology, the company can obtain real-time information on the location and speed of connected buses, passenger boarding and alighting data at stops, traffic light status, and the location and speed of manually driven buses.

[0009] Step 3: Focus on the operation status of intelligent connected buses and build a hierarchical optimization control model for intelligent connected buses:

[0010] 1. Description of the operating status of intelligent connected buses

[0011] The present invention mainly describes the operating status of intelligent connected buses from two aspects: whether the operating status of manually driven buses is considered;

[0012] 1.1 Not considering the operating status of manually driven buses

[0013] The running time T of the intelligent networked bus in the non-queuing section between nodes i k,j (1) Queuing section running time T i k,j (2) (This time requires that the car must be moving,) according to its corresponding running speed V i k,j (1) and V i k,j (2) Decision;

[0014]

[0015] Among them, L k,j (1) is the length of the non-queuing section; L k,j (2) is the length of the queue section, which is determined based on the longest queue length in history. There are M stations in total, k represents the kth station, k≤M, j is the jth signal light between station k and station k+1, and i is the train number.

[0016] Waiting time of vehicles at traffic lights and the time of reaching the stop line at the signal light The relationship between them is expressed as follows:

[0017]

[0018]

[0019] in, is the time when train number i leaves station k, The moment the green light turns on; C is the green light duration;k,j is the signal period; is the waiting time of train number i at the jth signal light between station k and station k+1; The time when train number i arrives at the jth signal stop line between station k and station k+1;

[0020] The time it takes for smart connected bus i to arrive at station k+1 Time to leave station k+1 The dwell time T at station k+1 i k+1,dwell The relationship between the three is shown in formula (7), the stay time at the station T i k+1,dwell The number of passengers boarding at station k+1 and the number of passengers getting off The specific formula is as follows:

[0021]

[0022] Where β is the average boarding and alighting time for each passenger;

[0023] 1.2 Considering the operating status of manually driven buses

[0024] Based on the overlapping characteristics of bus routes, when an intelligent connected bus operates on the non-queuing section and the queuing section between two stations k and k+1, the presence of multiple manually driven buses will affect the operating speed of the intelligent connected bus, further affecting the operating time of the intelligent connected bus, which will reduce the operating efficiency of the intelligent connected bus. Therefore, when describing the operating status of the intelligent connected bus, it is necessary to measure the impact of the uncertainty of the operation of manually driven buses:

[0025] 1.2.1 Description of the operating status of non-queuing sections

[0026] The operating time of a smart connected bus on each road section is determined by its operating speed, which is in turn affected by the traffic conditions on the road section. When a smart connected bus is traveling at a certain speed, if there is a manually driven bus ahead of it driving at a lower speed than that, and the manually driven bus is forced to reduce its speed on the current road section, it is defined as an obstructing bus. Therefore, when operating on a road section, the smart connected bus needs to determine whether there is an obstructing bus ahead.

[0027] (1) Determine whether there is an accessible bus scene ahead

[0028] like Figure 2As shown in the figure, assume that the distance between the end point of a certain road section and the starting point is s3. At the departure time a0, the distance between the target bus 1 and the starting point is s1, and the running speed is u1. At the time a0, the distance between the target bus 2 and the starting point is s2, and the running speed is u2. Bus 2 is running in front of the target bus 1. According to the positions and speeds of the two, there are two running states before the target bus 1 reaches the end point s3 of the road section:

[0029] Scenario I: When target bus 1 and bus 2 run at speed u1 and u2 respectively to the end point s3 of the road section, the distance between target bus 1 and bus 2 is always less than the safety distance s0 between the two vehicles. Bus 2 has no impact on the running state of target bus 1. There is no obstacle bus in front of target bus 1, and the following formula (8) is satisfied:

[0030]

[0031] a5 is the time when target bus 1 arrives at the destination, and a3 is the time when bus 2 arrives at the destination;

[0032] Scenario II: At time a1, target bus 1 reaches a distance s0 from bus 2. Limited by the speed of bus 2 ahead, target bus 1 cannot maintain its original speed u1 and is forced to slow down and follow bus 2 ahead to the end of the segment. Therefore, the time it takes for target bus 1 to reach the end of the segment increases by a4-a2. At this point, bus 2 becomes an obstacle bus for target bus 1:

[0033]

[0034] The above equations (8) and (9) can be used to determine whether the manually driven bus on the road section will affect the operation status of the intelligent network bus. If there is no barrier bus, the intelligent network bus can maintain a constant speed to the end of the road section; if there is an barrier bus, the intelligent network bus needs to adjust its speed at a certain position on the road section.

[0035] (2) Barrier-free bus scene ahead

[0036] If the smart connected bus is traveling on a non-queuing section and there is no barrier-free bus ahead, the speed is The time it takes to travel from the current position to the end of section j at a constant speed for:

[0037]

[0038] (3) Bus scene with obstacles ahead

[0039] When a smart connected bus is traveling on a non-queuing section, it is subject to a speed limit imposed by the obstructing bus. To prevent this speed limit from increasing the running time, the smart connected bus must adjust its speed to overtake the obstructing bus ahead. Therefore, the driving process of a smart connected bus on a non-queuing section can be divided into the following three stages:

[0040] Phase 1: The barrier bus is in front of the intelligent connected bus i and the distance is greater than the minimum safe following distance d0 of the vehicle, that is, exist Under the conditions of intelligent network bus i with speed Driving at a constant speed, the driving time in this stage is

[0041] Phase 2: Overtaking process; the intelligent networked bus i and the vehicle ahead at a speed of Distance of the bus traveling with obstacles When the speed Increase to overtaking speed After that, it starts to overtake; when the position of the intelligent network bus i is ahead of the obstacle bus, it means that the overtaking process is completed. The driving time of this stage is

[0042] Phase 3: The intelligent connected bus surpasses the obstacle bus and continues to travel at a speed of Driving at a constant speed, the driving time in this stage is

[0043] The driving process of the above three stages must meet the following requirements:

[0044]

[0045] According to formula (11), the time to travel to the destination is for:

[0046]

[0047] 1.2.2 Node operation status description

[0048] According to the operation characteristics of bus routes, nodes are the traffic lights and the stops where passengers get on and off the bus.

[0049] (1) Description of the operating status at the signal light

[0050] When smart connected bus i runs from the starting position of the queue section to the stop line of the signal light, it is necessary to consider whether there is a queue of manually driven buses. It is assumed that there are no other smart connected buses in the process of smart connected bus i running from the starting position of the queue section to the stop line of the signal light.

[0051] Use the following formula (16) to determine whether the signal phase state of the intelligent network-connected bus i when it arrives at the starting position of the queue section is red or green; n k,j The remaining red light duration of the current phase when the intelligent connected bus i arrives at the starting position of the queue section; is the red light duration of signal light j; n k,j If it is equal to 0, it is green light, otherwise it is red light;

[0052]

[0053] When the red light turns on, a queue of vehicles arriving at the queue section begins to form; when the green light turns on, the queue begins to dissipate. The number of vehicles q in the queue section ahead detected by the intelligent network-connected bus i is k,j Therefore, the time for the manual bus queue to dissipate is txs k,j According to the traffic wave theory, it can be calculated by formula (17): Refers to the saturation flow rate;

[0054]

[0055] The travel time on a queuing section is related to whether the vehicle joins the queue, the signal status, and the speed of the vehicle on the queuing section. Based on different influencing factors, the travel time of the queuing section mainly falls into the following three scenarios:

[0056] ① Traffic light j is red

[0057] When the traffic light j is red, It is necessary to determine whether the intelligent connected bus i joins the queue according to the dissipation time of the signal light queue, as shown in formula (18), where, is an integer variable between 0 and 1. When the smart connected bus i joins the queue, otherwise

[0058]

[0059] L k,j (2) is the length of the queue section; is the speed of smart connected bus i in the queue section before the jth signal light between station k and station k+1;

[0060] The operating time of the intelligent connected bus i in the queue section As shown in formula (19):

[0061]

[0062] The end time is when the vehicle passes the stop line of the signal light;

[0063] ② Traffic light j is green and the queue is dissipating:

[0064] When signal j is green and the queue is dissipating, n k,j =0 and txs k,j ≠0, and the queued vehicles are dissipating; the running time of the intelligent networked bus i in the queue section As shown in formulas (20) and (21):

[0065]

[0066] ③ Traffic light j is green and the queue has completely dissipated:

[0067] When the signal light j is green and the queue has completely dissipated, n k,j =0 and txs k,j = 0, the queued vehicles have completely disappeared; the running time of smart connected bus i in the queue section As shown in formula (22):

[0068]

[0069] For the above three situations, the decision on whether to activate the bus signal priority strategy is made based on the time when the smart connected bus i reaches the stop line of signal j and the signal phase state, so that the smart connected bus can pass;

[0070]

[0071] r i k,j is the time when the intelligent connected bus i arrives at the jth signal stop line between station k and station k+1;

[0072] Equations (24) and (25) determine whether to start the bus signal priority strategy;

[0073]

[0074] in, When the green light turns on, The green light duration, is the red light duration of signal light j; Indicates whether to start the signal priority strategy of early red light off. If the signal priority is started, it is 1, otherwise it is 0; Indicates whether to start the signal priority strategy of delaying the green light. If the signal priority is started, it is 1, otherwise it is 0;

[0075] When equations (26) and (27) are satisfied at the same time, it means that the intelligent connected bus i can pass the stop line of the signal light during the green light phase, which is determined by the relationship between the time when the intelligent connected bus i reaches the stop line and the phase duration;

[0076]

[0077] in, The time when the red light turns off early or the green light is delayed. Indicates whether to start the signal priority strategy of early red light off. If the signal priority is started, it is 1, otherwise it is 0; Indicates whether to start the signal priority strategy of delaying the green light. If the signal priority is started, it is 1, otherwise it is 0;

[0078] (2) Description of site operation status

[0079] The time it takes for smart connected bus i to arrive at station k+1 is Time to leave the site is The site stay time is The waiting time at the station due to the limited number of berths is when When it is less than 30s, it can be ignored and is equal to 0. As shown in formulas (28)-(30), J is the number of traffic lights between the current position of smart connected bus i and station k+1; β is the average boarding and alighting time for each passenger, is the number of passengers boarding at station k+1, is the number of passengers getting off at station k+1;

[0080]

[0081]

[0082] 2. Construction of hierarchical optimization control model for intelligent connected buses

[0083] From the perspectives of global optimization and local optimization, a first-level optimization control model and a second-level optimization control model are constructed, and a solution algorithm for the optimization control model is designed to obtain the optimal vehicle speed control scheme and signal control scheme. First, the first-level optimization control model is established by taking into account passenger demand and traffic signal control interference factors without considering the impact of manually driven buses. However, due to the interference of the uncertainty of the operation of manually driven buses in actual operation, the intelligent connected buses cannot maintain the optimal speed planning scheme output by the first-level optimization control model, which will reduce the optimization efficiency. Based on this, the second-level optimization control model is used to reasonably adjust the operating speed and signal priority time of the intelligent connected buses between each node.

[0084] 2.1 First-level optimization control model

[0085] The first-level optimization control model takes a global optimization perspective, ignoring the impact of manually driven buses and taking into account the constraints of economy and operation sequence. It aims to improve the uniformity of headway time between intelligent connected buses and restore the regularity of intelligent connected bus operation routes:

[0086] 2.1.1 Objective Function

[0087] During the operation of the intelligent network bus, the deviation of the headway from the dispatching headway is used as an indicator to measure the uniformity of the headway. The headway of two consecutive intelligent network buses i and i+1 arriving at the same station k+1 is calculated as follows: The optimization goal is to minimize the deviation from the dispatching headway h:

[0088]

[0089] is the time when smart connected bus i arrives at station k+1;

[0090] 2.1.2 Constraints

[0091] (1) Speed ​​restriction

[0092] Since the speed of the bus affects the energy consumption of the smart connected bus, it further affects the operating costs of the bus company. At the same time, passengers expect the bus to have a certain degree of smoothness. The present invention introduces the running speed under the greedy strategy as a criterion to ensure economy and smoothness. Here, the greedy strategy means that the smart connected bus running between two nodes can maintain the highest economy and smoothness, that is, its running speed is in the energy-saving driving speed range.

[0093]

[0094] V i k,j (1) is the speed of the intelligent connected bus i on the non-queuing section before the jth signal light between station k and station k+1; V i k,j (2) is the speed of the intelligent connected bus i in the queue section before the jth signal light between station k and station k+1; ΔV is the speed control threshold, which is a known quantity and its specific value depends on the working conditions;

[0095] According to the Urban Road Traffic Planning and Design Specifications, the operating speed of the queuing section should be less than or equal to the design speed of the section ρV designed ; Since the queue section is close to the traffic light, the speed V of the smart connected bus i in the queue section is i k,j (2) Less than or equal to ρV designed ;

[0096] V i k,j (2)≤ρV designed (33)

[0097] Where V designed is the design speed of the road section, which is a known quantity, and ρ is 0.7;

[0098] In summary, the vehicle speed constraint should satisfy the following equations (34) and (35):

[0099] max{V min ,V eco -ΔV}≤V i k,j (1)≤min{V max ,V eco +ΔV} (34)

[0100] max{V min ,V eco -ΔV}≤V i k,j (2)≤min{V max ,V eco +ΔV,ρV designed} (35)

[0101] Among them, V min and V max are the minimum and maximum operating speeds of the road section, respectively, both are known quantities;

[0102] (2) Run order constraints

[0103] The goal of the optimization model of the present invention is to restore the regularity of the routes of the intelligent connected buses. Therefore, when implementing the control strategy, it is necessary to ensure that the operating order of the front and rear intelligent connected buses does not change.

[0104]

[0105] The time when smart connected bus i leaves station k+1;

[0106] 2.2 Second-level optimization control model

[0107] The second-level optimization control model takes a local optimization approach, considering the output of the first-level optimization control model, the impact of manually driven buses, and the constraints of speed-varying overtaking. It establishes an optimization model with the dual objectives of minimizing the number of times intelligent connected buses enter node queues and minimizing the deviation between the speed sequence between nodes and the speed planned by the first-level optimization control model. This aims to reduce the impact of uncertainty caused by manually driven buses.

[0108] 2.2.1 Objective Function

[0109] (1) Node queuing times

[0110] Since my country's bus lanes have repeating route characteristics, when smart connected buses arrive at each node, they need to consider the operating status of manually driven buses on other routes. When at bus stops, the limited number of parking spaces can easily cause smart connected buses and manually driven buses to queue up to enter the station. At traffic lights, the duration of the signal phase can cause smart connected buses to enter the queue of manually driven buses that have not yet dissipated. Based on this, smart connected buses are prone to increasing the number of stops, resulting in frequent acceleration and deceleration, which reduces passenger comfort and increases the operating energy consumption of smart connected buses. The present invention aims to reduce the number of times smart connected buses enter the queues at each node, thereby reducing the impact of manually driven buses.

[0111] ①Number of queues at traffic lights

[0112] The phenomenon that the intelligent network-connected bus arrives at the end of the queue before the queue of manually driven buses at the signal light has not dissipated is defined as the intelligent network-connected bus joining the queue at the intersection, e.g. Figure 3 (a) is the number of queues of smart connected bus i at the jth signal light between station k and station k+1, which is 0 or 1;

[0113] ②Number of queues at the station

[0114] Station queuing refers to the phenomenon that when an intelligent network-connected bus and a manually driven bus enter the station at the same time, due to the limitation of the number of parking spaces, the manually driven bus in front is serving at the station, while the intelligent network-connected bus in the back has to queue up in front of the station. Figure 3 As shown in (b); when the queuing time of the intelligent network bus to the station is greater than t1, the intelligent network bus will enter the queuing waiting stage; A variable of 0 or 1, indicating whether smart connected bus i is waiting in line at stop k+1, 1 for queuing, 0 for not queuing; the present invention assumes that the smart connected bus i+1 at the rear will not be affected by the smart connected bus i in front and will not queue when it arrives at the stop;

[0115]

[0116] is the waiting time of smart connected bus i due to the berth limit at station k+1;

[0117] In summary, the optimization goal is to minimize the number of times that intelligent connected buses enter the node queue:

[0118]

[0119] (2) Operating speed deviation

[0120] The operating time of a smart connected bus on each road section is determined by its operating speed, which is in turn affected by the traffic conditions on the road section. When a smart connected bus is traveling at a certain speed, if there is a manually driven bus ahead of it driving at a lower speed than that, and this manually driven bus will be forced to reduce its speed on the current road section, it is defined as an obstructing bus. Therefore, when operating on a road section, the smart connected bus needs to determine whether there is an obstructing bus ahead.

[0121] Due to the interference of the obstructing bus, the intelligent connected bus generates a speed sequence by changing speeds and overtaking between nodes. The optimization goal is to minimize the deviation between the operating speeds in this sequence and the operating speed output by the first-level optimization control model. This improves the running smoothness while responding to the dynamic changes of the manually driven bus in real time.

[0122]

[0123] 2.2.2 Constraints

[0124] (1) Speed ​​restriction

[0125] On the basis of satisfying equations (34) and (35), the overtaking speed of the intelligent networked bus i is The speed must be greater than the speed of the obstructing bus The operating speed of each non-queue section and the operating speed of each queuing section in the second-level optimization control model of intelligent connected bus i should satisfy the following equations (40)-(43):

[0126]

[0127]

[0128]

[0129] ΔV is the speed control threshold, which is a known quantity and its specific value depends on the operating conditions;

[0130] Queuing section: is the speed of smart connected bus i in the queue section before the jth signal light between station k and station k+1;

[0131] Non-queue sections:

[0132] Phase 1: The barrier bus is in front of the intelligent connected bus i and the distance is greater than the minimum safe following distance d0 of the vehicle, that is, exist Under the conditions of intelligent network bus i with speed Driving at a constant speed, the driving time in this stage is

[0133] Phase 2: Overtaking process; the intelligent networked bus i and the vehicle ahead at a speed of Distance of the bus traveling with obstacles When the speed Increase to overtaking speed After that, it starts to overtake; when the position of the intelligent network bus i is ahead of the obstacle bus, it means that the overtaking process is completed. The driving time of this stage is

[0134] Phase 3: The intelligent connected bus surpasses the obstacle bus and continues to travel at a speed of Driving at a constant speed, the driving time in this stage is

[0135] (2) Overtaking restrictions

[0136] Since the queuing section of the urban road is close to the traffic light and overtaking is not possible, the second stage of overtaking by the intelligent connected bus i should be completed before the end of the non-queuing section:

[0137]

[0138] Among them, L k,j (1) is the length of the non-queuing section before the jth signal light between station k and station k+1. The meanings of the other physical quantities are the same as those in formulas (40)-(43);

[0139] (3) Signal control constraints

[0140] To avoid conflicts in bus priority applications, the maximum number of bus priority applications per signal cycle is one. At the same time, the green light duration during the non-priority phase of a bus, after the red light is turned off early or the green light is extended, must be sufficient to ensure safe crossing for pedestrians in all directions.

[0141]

[0142]

[0143] in, and Respectively, it indicates whether to start the signal priority strategy of early red light off and whether to start the signal priority strategy of delayed green light. If the signal priority is started, it is 1, otherwise it is 0;

[0144] μ k,j and are the traffic flow, saturation flow, green-to-signal ratio and pedestrian safe crossing time of the bus non-priority phase of signal light j; Minimum green light duration to avoid oversaturation of non-priority bus phases;

[0145] t0 is the vehicle startup loss time; is the length of the crosswalk; V p is the walking speed of pedestrians, which is 1.5m / s; I k,j The green light interval is set to 5s;

[0146] Step 4: To improve the accuracy and real-time performance of the intelligent connected bus operation control, the present invention solves the intelligent connected bus hierarchical optimization control model based on the concept of rolling optimization. At each trigger moment of the first-level rolling optimization, a genetic algorithm (GA) is used to solve the first-level single-objective optimization control model to obtain the planned speed of the intelligent connected bus.

[0147] At each triggering moment of the second-level rolling optimization, the non-dominated sorting genetic algorithm (NSGA-II) is used to solve the second-level optimization control model to obtain the optimal speed control scheme and signal priority adjustment scheme. The algorithm flow chart is shown in Figure 4 .

[0148] The headway is the time difference between two adjacent vehicles arriving at the same bus stop; the dispatch headway is the time difference between adjacent vehicles leaving the stop as determined in advance in the departure schedule and is a known quantity.

[0149] Bus tandem refers to the phenomenon that the arrival time difference of adjacent buses at a certain station is very small or they arrive at the station at the same time.

[0150] Advantages of the present invention:

[0151] This invention focuses on scenarios where traditional manually driven bus routes and intelligent connected bus routes are heterogeneous on bus lanes, aiming to solve the problems of intelligent connected bus trains colliding with each other and queuing at stops and traffic lights. The hierarchical optimization control model for intelligent connected bus operation established by this invention outputs the optimal speed control scheme and signal priority scheme under the conditions of uncertainty in the operation of manually driven buses, thereby improving the uniformity of headway time and reducing the number of times vehicles enter node queues. This invention provides new ideas for the efficient operation of intelligent connected bus companies and is of great significance in promoting the large-scale operation of intelligent connected buses from specific simple roads to complex roads. BRIEF DESCRIPTION OF THE DRAWINGS

[0152] Figure 1 This is a schematic diagram of the intelligent network bus route of the present invention. The yellow vehicles are non-buses.

[0153] Figure 2 Schematic diagram of two operation scenarios of target bus 1 on the road section;

[0154] Figure 3 Schematic diagram of node queuing phenomenon: (a) queuing at the traffic light; (b) queuing at the station; yellow vehicles are not buses;

[0155] Figure 4 This is a scroll optimization control flow chart in step 4 of the present invention;

[0156] Figure 5 This is a schematic diagram of the No. 41 bus route in Beijing for the first trial;

[0157] Figure 6 are the lighting parameters of each signal light in Test 1;

[0158] Figure 7 To test the speed parameters of each section of No. 41 bus route in Beijing;

[0159] Figure 8 This is the spatiotemporal trajectory diagram outputted by the no-control case in Experiment 1;

[0160] Figure 9 This is the spatiotemporal trajectory diagram output for the optimized control case in Experiment 1;

[0161] Figure 10 This is the data graph of the headway deviation value output for the uncontrolled case in Test 1;

[0162] Figure 11 This is the data diagram of the headway deviation value output for the optimization control case in Experiment 1.

[0163] The present invention is verified by the following test:

[0164] Experiment 1: This experiment is an optimization control method for intelligent connected bus operation based on heterogeneous route scenarios. It is carried out in the following steps:

[0165] This experiment establishes a simulation case based on the actual data of Beijing No. 41 intelligent network bus, and further compares and analyzes the results of the uncontrolled optimization scheme and the optimized control scheme proposed by the present invention. Figure 5 As shown in the figure, the section of Beijing Bus No. 41 from the first stop Dongdan Road Intersection to Pingleyuan is located on the bus lane and is not affected by social vehicles, which meets the research object of this invention. The total length of this section is 7067m, with a total of 12 stations and 18 traffic lights. The energy-saving driving speed range of Bus No. 41 is [30,40] km / h. The energy-saving driving speed V is selected by this invention. eco =30km / h; the standard dispatch headway is set to 360s.

[0166] The passenger arrival and alighting rates at each stop on Route 41 are shown in Table 1. The number of boardings at each stop is calculated by multiplying the headway between the previous and current connected buses at each stop by the passenger arrival rate at that stop. The number of alightings at each stop is calculated by multiplying the number of passengers on board the connected bus when it arrives by the passenger alighting rate at that stop. The average boarding and alighting time per person is assumed to be 2 seconds.

[0167] Table 1 Passenger arrival rate and alighting rate at each station

[0168]

[0169] Assuming that the first phase of each signal light is the phase for the intelligent connected bus to pass (i.e., the green light), the lighting parameters of each signal light are as follows: Figure 6 As shown in the figure, when smart connected bus No. 1 leaves stop 1, all traffic lights turn green. The length of the queue section is determined by the number of signal phases. The queue section length for phase 2 (a signal dedicated to pedestrian crossings with no cross traffic) is set to 10 meters, and the queue section length for other phases (with cross traffic) is set to 20 meters.

[0170] To verify that the optimization control model established in this invention can respond in real time to the impact of uncertainty caused by manually driven buses in mixed traffic environments, the intelligent connected bus system updates the operating status of manually driven buses on the route every 100 seconds:

[0171] ① Assume that there is always a manually driven bus on each non-queuing section, and its speed is in the range [0, 15] m / s (obtained from the historical data of bus route 41, such as Figure 7 As shown in FIG, a random integer is generated as the speed of the manually driven bus. Table 2 shows the position and speed parameters of some manually driven buses at the 100th second.

[0172] ② Randomly generate 18 integers in the interval [0,16] as the queue dissipation time of manually driven buses at each signal light when the green light is on in the signal cycle of the 100th second, as shown in Table 3;

[0173] ③ Randomly generate 12 integers as the arrival times of the manually driven buses at each station. For example, 113s, 273s, 357s, 467s, 575s, and 610s represent the arrival times of the manually driven buses at station 2 during the six updates, respectively. Here, any arrival time of two buses less than 30s is considered a queue at the station. The effects of vehicle dynamics are ignored.

[0174] Table 2 Parameters of some manually driven buses

[0175] Location (m) 48 267 343 840 1117 1305 1564 2199 2331 2524 Speed ​​(m / s) 9 15 9 6 4 5 6 11 9 12

[0176] Table 3 Queue dissipation time of manually driven buses

[0177] traffic light sequence 1 2 3 4 5 6 7 8 9 Queue dissipation time (s) 11 9 6 14 12 12 8 8 6 traffic light sequence 10 11 12 13 14 15 16 17 18 Queue dissipation time (s) 5 14 4 5 2 4 10 9 8

[0178] The specific algorithm process is as follows:

[0179] The first-level rolling optimization is based on event and time dual drive. The rolling optimization process is activated by the intelligent network bus departure event, and the station sequence k in each control time domain is identified. fl ={1…k…M}, where fl is determined by the number of stations. The state of manually driven buses on each road section and queuing section at the trigger moment (when the intelligent connected bus leaves the station), the phase state of each signal light, and the passenger flow data at each station (the operating state data of manually driven buses in 1.1 is not considered) are used as inputs to the optimization control model. The single-objective optimization model is solved to obtain the speed planning scheme for each non-queuing section and queuing section. When the intelligent connected bus arrives at the next station, the above process is reactivated and repeated multiple times until the terminal. The basic process of the first-level rolling optimization control of the intelligent connected bus embedded in GA can be expressed as:

[0180]

[0181] The second-level rolling optimization based on time-driven. Input the initial position s when the smart connected bus i leaves the station i,1 =p k (p k is the location of site k), initial time The first-level optimization control scheme outputs the planned vehicle speed, etc., and solves the dual-objective optimization model to determine the position s i,1 Speed ​​control scheme and priority scheme of each signal light to station k+1. Travel through rolling step t with the current control scheme DRHAfter that, the rolling optimization times are updated to b = b + 1, and the status information of the manually driven bus ahead is retrieved and the process is activated again. This process is repeated several times until the bus reaches station k + 1, and the first-level rolling optimization control process is reactivated. The basic process of the second-level rolling optimization of the intelligent connected bus embedded in NSGA-II can be expressed as follows:

[0182]

[0183] The number of stations can be used to determine the first-level rolling optimization times of each intelligent network bus schedule to be 11, and the second-level rolling step size t is set DRH = 30s. The minimum green light time of each signal light is set to 30% of the initial green light time, and the minimum safety distance is set to d0 = 5m. Other relevant parameters in the algorithm are set as follows: Crossing probability P C =0.7, mutation probability P M = 0.1, the initial population size N0 = 30 and the maximum number of iterations E = 200. The operation status of the intelligent connected bus is simulated through Matlab and the solution algorithm is written.

[0184] The following describes the specific hierarchical rolling optimization control process using the control results of bus 5 (i is 5) as an example. First, under the premise of minimizing the uniformity of headway, the first-level speed planning scheme is output, as shown in Table 4. The first sequence in the first-level optimization result, that is, the planned speed sequence between station 1 and station 2, is used as the input parameter of the second-level rolling optimization. Considering the two objectives of the number of node queues and the deviation from the first-level planned speed, the second-level speed control scheme and signal optimization scheme are output, but only the rolling step size t is actually executed. DRH The optimal control scheme within the time limit is t DRH Afterwards, the second-level rolling optimization control process is activated again, as shown in Table 5. The above process is repeated until the train arrives at station 2. The second-level optimization results are shown in Table 6. The time of arrival at station 2 is 1856 seconds.

[0185] Table 4 Bus 5 first level optimization results

[0186] Site 1 2 3 4 5 6 Arrival time(s) — 1847 1994 2130 2211 2246 Departure time(s) 1800 1848 1997 2133 2213 2247 Site S7 S8 S9 S10 S11 S12 Arrival time(s) 2360 2455 2495 2599 2676 2738 Departure time(s) 2365 2457 2496 2604 2677 —

[0187] Table 5bus 5 second level rolling optimization process

[0188]

[0189] Table 6 Bus 5 second level optimization results

[0190] Location (m) 0 88 96 177 325 Time(s) 1800 1809 1810 1830 1856

[0191] Table 7 Number of queues at stations

[0192] Site 1 2 3 4 5 6 7 8 9 10 11 12 total No control case 0 0 1 0 1 0 1 1 1 1 0 1 7 Optimization control case 0 0 0 2 0 0 0 0 0 0 0 0 2

[0193] Table 8 Number of queues at intersections

[0194]

[0195] Tables 7 and 8 show the number of times a total of 8 No. 41 buses queued at various stations and traffic lights during the study period. Without control, the number of queues at stations and traffic lights was 7 and 79, respectively. After hierarchical optimization control, the number of queues at stations and traffic lights was reduced by 71% and 73%, respectively.

[0196] Figure 8 and Figure 9 The spatiotemporal trajectory of each intelligent network bus during the study period is given. Without adjusting the operation status of the intelligent network bus on the route, the operation regularity of each bus is poor (e.g. Figure 8 ), and the speed control and signal adjustment have greatly improved the operation effect (such as Figure 9 ).

[0197] Figure 10 and Figure 11 The deviation of the headway between each intelligent networked bus and the preceding bus at each stop during the study period is given, which is equal to the ratio of the actual headway to the standard dispatched headway. Compared with the uncontrolled simulation case ( Figure 10 ), the optimization control method proposed in this invention ( Figure 11 ) reduces the headway deviation by 67%. Therefore, the optimization control method proposed in the present invention is effective in restoring line regularity and reducing node queues.

[0198] This paper establishes a hierarchical optimization control model for intelligent connected bus operations using two strategies: speed control and signal priority. Based on the model's nonlinear characteristics and the need for real-time algorithm performance, an optimization control algorithm based on rolling optimization was designed. Finally, two case studies were simulated using MATLAB, and the results of uncontrolled and optimized control were compared. The results showed a 67% reduction in headway deviation, and a 71% and 73% reduction in queues at various stations and intersections, respectively. Therefore, the proposed optimization control method is effective in restoring route regularity and reducing node queues.

Claims

1. An intelligent connected bus operation optimization control method based on heterogeneous line scenarios is characterized in that the intelligent connected bus operation optimization control method based on heterogeneous line scenarios is carried out according to the following steps: Step 1: Select a hybrid bus lane that includes an intelligent connected bus route and a manually driven bus route to build a heterogeneous route scenario. Establish an attribute set for the intelligent connected bus route in the heterogeneous route scenario to store route and node information. Establish an attribute set for the intelligent connected bus trip in the heterogeneous route scenario to store trip operation information. Step 2: Using V2X technology, the company can obtain real-time information on the location and speed of connected buses, passenger boarding and alighting data at stops, traffic light status, and the location and speed of manually driven buses. Step 3:

1. Based on the information established in step 1 and the data obtained in step 2, the operation status data of the intelligent connected bus is established. The process is as follows: 1.

1. The establishment of the operating status data of intelligent networked buses without considering manual driving of buses is as follows: Step 3.

1. Obtain the running time T of the intelligent networked bus in the non-queuing section between each node i k,j (1) and the running time T of the queue section i k,j (2), Where k represents the kth station, j represents the jth signal light between station k and station k+1, and i represents the train number; Step 32: Based on the waiting time of vehicle i at the jth signal light between station k and station k+1 The running time T of the intelligent networked bus in the non-queuing section between nodes i k,j (1) The running time T of the queue section i k,j (2) Calculate the time when train number i arrives at the jth signal stop line between station k and station k+1 Step 3: The time when train number i leaves station k The running time T of the intelligent networked bus in the non-queuing section between nodes i k,j (1) Queuing section running time T i k,j (2) The waiting time of train number i at the jth signal light between station k and station k+1 Get the time when smart connected bus i arrives at station k+1 Step 34: Number of passengers boarding based on station k+1 and the number of passengers getting off Get the stay time T of the site i k+1,dwell , the expression is: Where β is the average boarding and alighting time for each passenger; Step 35: Based on the time when smart connected bus i arrives at station k+1 and the dwell time T at the site i k+1,dwell , get the time when smart connected bus i leaves station k+1 1.2 Considering the establishment of the operating status data of intelligent connected buses under manual driving, the specific process is as follows: 1.2.1 The establishment of non-queuing section operation status data is as follows: (1) Barrier-free bus scene ahead Intelligent network bus i with speed The time it takes to travel from the current position to the end of section j at a constant speed for: (2) Bus scene with obstacles ahead The driving process of intelligent connected buses on non-queuing sections can be divided into the following three stages: Phase 1: The barrier bus is in front of the intelligent connected bus i Place, and Greater than the minimum safe following distance d0 of the vehicle, that is, exist Under the conditions of intelligent network bus i with speed Driving at a constant speed, the driving time of stage 1 is Phase 2: Overtaking process; Intelligent networked bus i and the front at speed Distance of the bus traveling with obstacles When the speed Increase to overtaking speed After that, it starts to overtake; when the position of the intelligent network bus i is ahead of the obstacle bus, it means that the overtaking process is completed, and the driving time of stage 2 is Phase 3: The intelligent connected bus surpasses the obstacle bus and continues to travel at a speed of Driving at a constant speed, the driving time of stage 3 is The driving process of the above three stages must meet the following requirements: Time to the destination for: 1.2.2 Establishment of node operation status data. The specific process is as follows: (1) The establishment of the operating status data at the traffic light is as follows: Assume that there are no other smart connected buses when smart connected bus i runs from the starting position of the queue section to the stop line of the signal light; Determine whether the signal phase state of the intelligent connected bus i when it arrives at the starting position of the queue section is red or green; When the red light turns on, a queue of vehicles arriving at the queuing section begins to form; when the green light turns on, the queue begins to dissipate; If the number of manually driven buses q in the process of intelligent connected bus i running from the starting position of the queue section to the signal stop line is k,j , then calculate the dissipation time txs of the manually driven bus queue k,j ; The running time of the queuing section mainly has the following three scenarios: ① Traffic light j is red ② Traffic light j is green and the queue is dissipating: ③ The signal light j is green and the queue has completely dissipated; For the above three situations, the decision on whether to activate the bus signal priority strategy is made based on the time when the smart connected bus i reaches the stop line of signal j and the signal phase state, so that the smart connected bus can pass; r i k,j is the time when the intelligent connected bus i arrives at the jth signal stop line between station k and station k+1; The following formula decides whether to start the bus signal priority strategy; in, When the green light turns on, The green light duration, is the red light duration of signal light j; Indicates whether to start the signal priority strategy of early red light off. If the signal priority is started, it is 1, otherwise it is 0; Indicates whether to start the signal priority strategy of delaying the green light. If the signal priority is started, it is 1, otherwise it is 0; When the following equations are satisfied simultaneously, it means that the smart connected bus i can pass the stop line of the signal light during the green light phase, which is determined by the relationship between the time when the smart connected bus i reaches the stop line and the phase duration; in, The time when the red light turns off early or the green light is delayed. Indicates whether to start the signal priority strategy of early red light off. If the signal priority is started, it is 1, otherwise it is 0; Indicates whether to start the signal priority strategy of delaying the green light. If the signal priority is started, it is 1, otherwise it is 0; (2) Establishment of site operation status data. The specific process is as follows: The time it takes for smart connected bus i to arrive at station k+1 is Time to leave the site is The site stay time is The waiting time at the station due to the limited number of berths is when When it is less than 30s, it can be ignored and is equal to 0; as shown in the following formula, J is the number of traffic lights between the current position of smart connected bus i and station k+1; β is the average boarding and alighting time for each passenger, is the number of passengers boarding at station k+1, is the number of passengers getting off at station k+1; 2. Based on the data of the intelligent connected bus operation status established in 1 above, a hierarchical optimization control model for intelligent connected buses is constructed. The specific process is as follows: The hierarchical optimization control model of intelligent connected public transportation includes the first-level optimization control model and the second-level optimization control model; 2.

1. Construct the first-level optimization control model, ignoring the impact of manually driven buses. The specific process is as follows: 2.1.

1. Construct the objective function. The specific process is as follows: During the operation of the intelligent network bus, the deviation of the headway from the dispatching headway is used as an indicator to measure the uniformity of the headway. The headway of two consecutive intelligent network buses i and i+1 arriving at the same station k+1 is calculated as follows: The optimization goal is to minimize the deviation from the dispatching headway h: 2.1.2 The constraints are as follows: The vehicle speed constraint satisfies the following formula: max{V min ,V eco -ΔV}≤V i k,j (1)≤min{V max ,V eco +ΔV} max{V min ,V eco -ΔV}≤V i k,j (2)≤min{V max ,V eco +ΔV,ρV designed } Among them, V min and V max are the minimum and maximum operating speeds of the road section, both of which are known quantities; V i k,j (1) is the speed of the intelligent connected bus i on the non-queuing section before the jth signal light between station k and station k+1; V i k,j (2) is the speed of the intelligent connected bus i in the queue section before the jth signal light between station k and station k+1; ΔV is the speed control threshold; V designed is the design speed of the road section, which is a known quantity, and ρ is 0.7; (2) The running order constraints are as follows: The time when smart connected bus i leaves station k+1; The time when the intelligent connected bus i+1 leaves the station k+1; 2.2 Construct the second-level optimization control model. The specific process is as follows: 2.2.1 Construct the objective function. The specific process is as follows: (1) Objective function of node queue times: The optimization goal is to minimize the number of times that intelligent connected buses enter the node queue: is the number of queues of smart connected bus i at the jth signal light between station k and station k+1, which is 0 or 1; A variable of 0 or 1, indicating whether smart connected bus i is waiting in line at station k+1, 1 means queuing, 0 means not queuing; (2) Construct the objective function of the running speed deviation: 2.2.2 The constraints are as follows: (1) Vehicle speed constraints are as follows: Overtaking speed of intelligent connected bus i The speed must be greater than the speed of the obstructing bus Queuing section: is the speed of smart connected bus i in the queue section before the jth signal light between station k and station k+1; Non-queue sections: Phase 1: The barrier bus is in front of the intelligent connected bus i and the distance is greater than the minimum safe following distance d0 of the vehicle, that is, exist Under the conditions of intelligent network bus i with speed The driving time in this stage is Phase 2: Overtaking process; the intelligent networked bus i and the vehicle ahead at a speed of Distance of the bus traveling with obstacles When the speed Increase to overtaking speed After that, it starts to overtake; when the position of the intelligent network bus i is ahead of the obstacle bus, it means that the overtaking process is completed. The driving time of this stage is Phase 3: The intelligent connected bus surpasses the obstacle bus and continues to travel at a speed of The driving time in this stage is (2) Overtaking restrictions are as follows: Since the queuing section of the urban road is close to the traffic light and overtaking is not possible, the second stage of overtaking of the intelligent connected bus i should be completed before the end of the non-queuing section; (3) Signal control constraints are as follows: In order to avoid conflicts in bus priority applications, the maximum number of bus priority applications in each signal cycle is one. At the same time, it is necessary to ensure that the green light duration of the bus non-priority phase after the red light is turned off early or the green light is extended meets the safe crossing needs of pedestrians in all directions. Step 4: Solve the hierarchical optimization control model of intelligent connected bus operation constructed in step 3 based on the rolling optimization method to obtain the planned speed, optimal speed control plan and signal priority adjustment plan of the intelligent connected bus.

2. The intelligent networked bus operation optimization control method based on heterogeneous line scenarios according to claim 1 is characterized in that The attribute set of the intelligent connected bus route described in step 1 includes the length of the non-queuing section, the length of the queuing section, the speed limit of the section, the location of the node and the timing information of the signal light. The nodes include stations and signal lights. The queuing section is the section area before the stop line of the signal light, that is, the section area where buses need to queue when the signal light is in the red phase. The length of the queuing section is determined based on the historical maximum queue length. The non-queuing section is the section outside the queuing section.

3. The intelligent networked bus operation optimization control method based on heterogeneous line scenarios according to claim 1 is characterized in that The intelligent connected bus trip attribute set described in step 1 includes the dispatching headway and initial departure schedule of the trip within the control period.

4. The intelligent networked bus operation optimization control method based on heterogeneous line scenarios according to claim 1 is characterized in that In step 4, the genetic algorithm is used to solve the first-level single-objective optimization control model at each triggering moment of the first-level rolling optimization to obtain the planned speed of the intelligent connected bus; At each triggering moment of the second-level rolling optimization, the non-dominated sorting genetic algorithm is used to solve the second-level optimization control model to obtain the optimal speed control scheme and signal priority adjustment scheme.

5. The intelligent networked bus operation optimization control method based on heterogeneous line scenarios according to claim 4 is characterized in that In step 4, when solving the first-level single-objective optimization control model using the genetic algorithm in the first-level rolling optimization, the data of 1.1 in step 3 on the operating status of the intelligent connected bus without considering the manually driven bus is substituted into it.

6. The intelligent networked bus operation optimization control method based on heterogeneous line scenarios according to claim 4 is characterized in that In step 4, when solving the second-level optimization control model using the non-dominated sorting genetic algorithm in the second-level rolling optimization, the data of the operating status of the intelligent connected bus under the manually driven bus in step 3 is substituted into it.

Images