Airport departure layer lane traffic control decision method and system based on fuzzy cellular automaton

Abstract

The invention discloses an airport departure layer lane traffic control decision method and system based on a fuzzy cellular automaton. The method comprises the following steps: constructing a fuzzy cellular automaton model suitable for an airport departure layer lane; collecting historical traffic flow data of an airport departure layer lane; performing time-phased prediction on future traffic flow based on historical traffic flow data of airport departure layer lanes by using a long-short-term memory neural network; based on the predicted traffic flow, the fuzzy cellular automaton model is used for carrying out safety index evaluation on an airport departure layer lane, and whether traffic signal control needs to be started or not is determined; and if traffic signal control is needed, determining optimal signal timing based on a traffic delay index obtained by simulation. The invention aims to optimize the passenger-vehicle mixed traffic flow of the airport departure layer lane and improve the traffic efficiency and safety.

Description

Technical Field

[0001] The invention belongs to the technical field of traffic control, and in particular relates to a traffic control decision method and system for a lane on a departure level of an airport based on a fuzzy cellular automaton. Background Art

[0002] The airport departure lane is a key area where the airport and the urban transportation system meet. Frequent human-vehicle interactions have a significant impact on traffic safety and efficiency. Currently, most airport departure lanes do not implement signal control. Instead, airport staff conduct on-site command when traffic volume is heavy and conflicts between people and vehicles are frequent. This method relies on manual experience to a certain extent and may not be able to respond to complex and changing traffic conditions in real time, leading to safety risks and traffic delays.

[0003] When evaluating whether to implement a signal control strategy on the departure lanes of an airport, strictly following existing specifications and standards may encounter some specific scenarios. Although these scenarios do not meet the strict standards for enabling signal control, they pose significant safety risks. In addition, after enabling signal control, using a fixed timing scheme may cause significant traffic delays. Therefore, a method for dynamically determining the start and stop and timing scheme of signal control is urgently needed to minimize traffic delays while ensuring safety.

[0004] At present, when evaluating the impact of signal control on the traffic system, simulation methods are usually used to evaluate safety and efficiency. However, traditional traffic behavior modeling methods do not fully study the complex interaction dynamics between people and vehicles, and fail to fully consider the decision-making process of vehicles and pedestrians during the interaction process, as well as the impact of the attributes of the airport departure lane on the traffic behavior of people and vehicles. Therefore, the model built based on the traditional modeling method is difficult to accurately simulate the real human-vehicle interaction scenario, and cannot provide reliable decision support for the traffic control of the airport departure lane. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes a traffic control decision method and system for the airport departure lane based on fuzzy cellular automata, and proposes a fuzzy cellular automata model suitable for the airport departure lane to simulate the movement, decision-making and interaction of pedestrians and vehicles on the road section. The long short-term memory network is used to predict the traffic volume of vehicles and pedestrians in the road section in different time periods, and the traffic volume is input into the fuzzy cellular automata model for safety index evaluation to determine whether to enable signal control. If signal control is enabled, the traffic delay index under different timing schemes is further obtained based on model simulation, and the appropriate signal timing scheme is determined, so as to improve the system operation efficiency of the airport departure lane as much as possible while ensuring safety.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] The airport departure lane traffic control decision method based on fuzzy cellular automation includes the following steps:

[0008] Construct a fuzzy cellular automaton model suitable for airport departure lanes;

[0009] Collect historical traffic flow data for airport departure lanes;

[0010] Using long short-term memory neural network, based on the historical traffic flow data of the airport departure lane, the future traffic flow is predicted by time period;

[0011] Based on the predicted traffic flow, the fuzzy cellular automaton model is used to evaluate the safety index of the airport departure lane to determine whether traffic signal control needs to be enabled; if traffic signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation.

[0012] Preferably, the fuzzy cellular automaton model is used to simulate the human-vehicle interaction behavior occurring in the crosswalk area of ​​the road section, including: a pedestrian motion model, a vehicle motion model, a pedestrian decision model, a vehicle decision model and a vehicle lane changing model.

[0013] Preferably, the method for constructing a vehicle decision model includes:

[0014] For the video clips of the vehicle in front of the crosswalk, the time gap between the pedestrians who have not passed the lane where the vehicle is located and the conflict area is recorded. 1 , the speed of pedestrians v 1 , and the corresponding driving decisions;

[0015] GAP according to the driving decision 1 and v 1 The levels are divided into

[0016] Calculate the GAP of a specific value 1 and v 1 The frequency of the class to which it belongs is used as the degree of membership;

[0017] GAP 1 and v 1 The numerical value and the corresponding membership degree are plotted into a scatter plot, and a function is used for fitting to determine the type and parameters of the membership function that meet the preset requirements;

[0018] Based on the membership function type and parameters, the vehicle decision model is constructed.

[0019] Preferably, the method for constructing a pedestrian decision model includes:

[0020] For video clips of pedestrians crossing the street, the GAP of the vehicle closest to the conflict zone in the lane ahead is recorded when the pedestrian crosses the lane ahead.2 and speed v 2 , and the specific decision of the pedestrian;

[0021] Calculating GAP based on pedestrian decisions 2 With v 2 The type of

[0022] Calculate the GAP of a specific value 2 and v 2 The frequency of the type to which it belongs is used as the degree of membership;

[0023] GAP 2 and v 2 The numerical value and the corresponding membership degree are plotted into a scatter plot, and a function is used to fit it to determine the type and parameters of the membership function that meet the preset requirements, and finally calibrate the pedestrian decision model.

[0024] Preferably, based on the predicted traffic flow, the method for evaluating the safety index of the airport departure lane using the fuzzy cellular automaton model includes:

[0025]

[0026] Among them, TDTC p，v (t) is the time difference between the vehicle v and the pedestrian p reaching the conflict area at the original speed within the time step t; T p (t), T v (t) is the time when pedestrian p and vehicle v arrive at the conflict area; L p，v (t) is the position of the potential conflict point between pedestrian p and vehicle v; L p (t) and L v (t) represent the positions of pedestrians and vehicles respectively; and Represent the speeds of pedestrians and vehicles respectively.

[0027] Preferably, the method for determining whether traffic signal control needs to be activated includes:

[0028] Input several groups of traffic flow in the specification into the fuzzy cellular automaton model and compare the three N TDTC The smallest value is taken as the safety limit value A;

[0029] The predicted traffic flow value is input into the fuzzy cellular automation model. If the output N TDTC If the value is greater than the safety index A, the signal light at the pedestrian crossing of the airport departure lane needs to be turned on. TDTC If the value is less than the safety index A, there is no need to turn on the signal light.

[0030] Preferably, if traffic signal control is required, the method for determining the optimal signal timing based on the traffic delay index obtained by simulation includes:

[0031] The goal of setting up signal control on a road section is to maximize the overall traffic benefit. The optimized timing plan should ensure the minimum total delay, that is, the weighted sum W of pedestrian delay and vehicle delay is minimized, expressed as:

[0032] W=p×d kw +k×q×D kw ,

[0033] Where W is the weighted sum of pedestrian and vehicle delays; p and q are the pedestrian and vehicle traffic volumes, respectively; d kw , D kw are the average delays of pedestrians and vehicles under the signal control condition of the road section; k is the conversion coefficient of pedestrian and vehicle delays.

[0034] The present invention also provides an airport departure level lane traffic control decision system based on fuzzy cellular automata, comprising: a model building module, a data collection module, a flow prediction module and a signal timing module;

[0035] The model building module is used to build a fuzzy cellular automaton model suitable for the airport departure lane;

[0036] The data collection module is used to collect historical traffic flow data of the airport departure lane;

[0037] The traffic prediction module is used to use a long short-term memory neural network to predict future traffic flow by time period based on the historical traffic flow data of the airport departure lane;

[0038] The signal timing module is used to evaluate the safety index of the airport departure lane based on the predicted traffic flow using the fuzzy cellular automaton model to determine whether traffic signal control needs to be enabled; if traffic signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] By constructing a fuzzy cellular automaton model, the present invention can simulate the traffic conditions on the airport departure lane more realistically. In the process of simulating human-vehicle interaction, it can not only simulate normal vehicle following, lane changing and pedestrian movement, but also simulate the decision-making behavior of drivers and pedestrians when encountering crosswalks. In addition, the present invention uses safety index evaluation to determine whether to enable signal control, and improves the operating efficiency of the airport departure lane section as much as possible under the premise of effectively ensuring safety. Finally, by combining the traffic prediction results of each time period, a corresponding signal timing plan is formulated to optimize the signal control effect of the airport departure lane, thereby improving the overall safety and operating efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0042] Figure 1 It is a flow chart of a method for traffic control decision of a departure level lane of an airport based on fuzzy cellular automation according to an embodiment of the present invention;

[0043] Figure 2 This is a flow chart of the first lane-changing behavior of the lane-changing model according to an embodiment of the present invention;

[0044] Figure 3 This is a flow chart of the second lane-changing behavior of the lane-changing model according to an embodiment of the present invention;

[0045] Figure 4 Schematic diagram of a cellular space according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] Embodiment 1

[0049] like Figure 1As shown in the figure, the entire process of modeling and analyzing the human-vehicle interaction behavior on a road section is described, which mainly includes model construction, data processing, calculation of evaluation indicators, and determination of the start and stop of signal control and timing scheme.

[0050] The present invention provides an airport departure lane traffic control decision method based on fuzzy cellular automata, comprising the following steps:

[0051] Construct a fuzzy cellular automaton model suitable for airport departure lanes;

[0052] Collect historical traffic flow data for airport departure lanes;

[0053] Using long short-term memory neural network, based on the historical traffic flow data of the airport departure lane, the future traffic flow is predicted by time period;

[0054] Based on the predicted traffic flow, the fuzzy cellular automaton model is used to evaluate the safety index of the airport departure lane to determine whether traffic signal control needs to be enabled; if traffic signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation.

[0055] In this embodiment, a fuzzy cellular automaton model is constructed. The model is divided into five parts:

[0056] (1) Vehicle motion model.

[0057] The vehicle's movement uses the Nasch model to update its position. The specific process includes acceleration, deceleration, random slowing, and position update:

[0058] Acceleration: If the vehicle speed is less than the maximum speed V max , and there is enough space ahead (the number of cells between the vehicle and the vehicle in front is greater than the vehicle's current speed), the vehicle speed increases by 1.

[0059] Slow Down: If the space in front of the vehicle (in cells) is smaller than the current speed, the vehicle slows down to the size of the space in front.

[0060] Random deceleration: Each car has a probability p of decelerating by 1, simulating the random deceleration behavior of the driver.

[0061] Position update: Each vehicle moves forward the corresponding number of cells based on the updated speed.

[0062] (2) Vehicle lane-changing model.

[0063] The airport departure lane has a drop-off function. The lane near the terminal is the drop-off lane, and the other two are driving lanes. If a vehicle on the driving lane needs to drop off passengers, it must first change lanes to the drop-off lane before stopping to drop off passengers. If the vehicle that has completed the drop-off in the drop-off area is not the lead vehicle, it will change lanes to the driving lane and continue driving, while the lead vehicle will continue driving in the drop-off lane.

[0064] The present invention divides lane changing behaviors into two types:

[0065] like Figure 2 As shown in the figure, the first type is changing lanes from the driving lane to the drop-off lane. The specific process is to slow down first, and then determine whether there is an empty space in the drop-off lane. If there is no empty space, the vehicle continues to drive, and continues to determine whether there is an empty space in the drop-off lane. If there is an empty space, it is determined whether the size of the empty space meets the lane change requirements. If the empty space is larger than the vehicle length, the vehicle changes lanes to the empty space in the drop-off lane; if the empty space is smaller than the vehicle length, the vehicle continues to drive, and continues to determine whether there is an empty space larger than the vehicle length.

[0066] like Figure 3 As shown in the figure, the second lane change behavior is because the vehicle in the drop-off lane needs to change lanes to the driving lane to continue driving after completing the drop-off. Since the target vehicle in this case changes from stationary to moving, its speed along the driving lane can be ignored, so the safety distance determination formula is:

[0067] d safe ≥2V j +l,

[0068] Among them, d safe is the safe distance; V j is the speed of the target vehicle immediately behind it on the lane; l is the length of the target vehicle.

[0069] The target vehicle will judge every gap in the lane and change lanes when the above formula is met.

[0070] (3) Vehicle decision model.

[0071] The fuzzy reasoning model can be used to simulate the process of drivers making decisions based on common sense. In the model, the factors affecting decision-making are selected as the time gap between pedestrians and the conflict zone. 1 and the pedestrian's speed v 1 The present invention divides the value of each influencing factor into three levels, namely "near", "medium" and "far", and "low", "medium" and "high".

[0072] The fuzzy rules are set as: 1 If v is "far", 1 For any value, the driver will choose to continue driving; if GAP 1 If v is "near",1 For any value, drivers will choose to slow down and give way; if GAP 1 If it is "middle", then v 1 Continue driving when v is "low". 1 When the speed is "Medium" or "High", slow down and give way. See Table 1.

[0073] Table 1

[0074]

[0075] The rule for setting the output variable is: if the result of the fuzzy calculation of the input variable is to continue driving, the value of the output variable should be less than 2, otherwise it should be greater than 2. The input variables are: the time gap between the pedestrian and the conflict zone 1 and the pedestrian's speed v 1 Therefore, in the model, the driver's driving decision can be determined based on the value of the output variable.

[0076] (4) Pedestrian decision model.

[0077] The simulation of pedestrian decision-making can be achieved in the same way as the vehicle decision-making model, except that the timing of the judgment becomes before the pedestrian passes through each lane, and the fuzzy variable is the time gap between the vehicle and the conflict area. 2 and the vehicle's speed v 2 The formulation of fuzzy rules and the value of output variables are exactly the same as those of the vehicle decision model, but if the value of the output variable is less than 2 but very close to 2, the pedestrian will accelerate through this lane. If both the vehicle and the pedestrian decide to give way or continue driving (walking), the vehicle will remain stopped while the pedestrian will accelerate through the lane in front.

[0078] (5) Pedestrian motion model

[0079] To avoid collisions between pedestrians walking in the same direction, the maximum speed of pedestrians is required to be less than or equal to the sum of the speed of the pedestrian in front and the distance between them. To avoid collisions between pedestrians walking in opposite directions, the crosswalk is divided into two parts, left and right, and pedestrians are required to walk on the right.

[0080] The movement of pedestrians uses the Nasch model to update their positions. The specific process includes acceleration, deceleration, random slowing down and position update:

[0081] Acceleration: When there is no obstacle ahead and the current speed is lower than the maximum speed, the pedestrian accelerates with a deterministic probability of 1.

[0082] Slow down: To avoid collision with pedestrians ahead, pedestrians slow down to the size of the space ahead when the distance ahead is less than the current speed.

[0083] Random deceleration: During the movement process, pedestrians randomly decelerate by 1 with a certain probability to simulate the occasional speed reduction phenomenon.

[0084] Position update: Pedestrians move the corresponding number of cells in the target direction according to the updated speed until they encounter obstacles or other pedestrians.

[0085] In this embodiment, data processing. Extract the segments where the vehicle passes through the crosswalk and there are pedestrians who have not walked into the conflict area, as well as the parts of pedestrians crossing the street, from the surveillance video, and process them separately.

[0086] For the video segment of the vehicle in front of the crosswalk, it is necessary to record the time interval GAP between the pedestrians who have not passed through the lane where the vehicle is located in front of the crosswalk and the conflict area 1 , the speed v of the pedestrians 1 , and the corresponding driving decisions.

[0087] According to the driving decision, classify the levels of GAP 1 and v 1 , that is, if the driver chooses to continue driving, then GAP 1 (or v 1 ) is "far" (or "small"); if the driver chooses to stop and give way, then GAP 1 (or v 1 ) is "near" (or "big"); if some drivers choose to continue driving and some drivers choose to stop and give way, then GAP 1 (or v 1 ) is "medium" (or "medium").

[0088] Calculate the frequencies of the levels to which specific values of GAP 1 and v 1 belong, and use them as membership degrees. The formula for calculating the membership degree of GAP 1 is:

[0089]

[0090]

[0091] where P f , P m and P s are the frequencies belonging to far, medium and near respectively; F f , F m and F s are the frequencies belonging to far, medium and near respectively.

[0092] The calculation method of the speed v 1 is the same as that of GAP 1 .

[0093]

[0094] Among them, P f1 , P m1 and P s1 are the frequencies belonging to fast, medium and slow respectively; F f1 、F m1 and F s1 are the frequencies belonging to fast, medium and slow respectively.

[0095] Then the membership function is fitted and GAP 1 and v 1 The values ​​and their membership are plotted into a scatter plot, and a Gaussian function is used for fitting to determine the mean, standard deviation and amplitude parameters of the Gaussian function. The fitted Gaussian function is used as the membership function of the vehicle decision model, with GAP 1 and v 1 As input, the corresponding membership value is calculated through the Gaussian membership function. Based on these membership values, the system applies the preset fuzzy rules for reasoning and generates specific driving operation instructions through the defuzzification process, thereby determining the final driving decision and executing it.

[0096] For video clips of pedestrians crossing the street, the gap between the vehicle closest to the conflict zone in the lane in front of the pedestrian should be recorded. 2 and speed v 2 , and the specific decision of the pedestrian. Calculate GAP based on pedestrian decision 2 With v 2 The level of 2 For far, middle and near, v 2 The process is similar to that of vehicles. The membership calculation process of the driver's decision is repeated, and the function fitting and parameter determination are performed to finally calibrate the pedestrian decision model.

[0097] Through the surveillance video data of the airport departure level road section, the required time distance between pedestrians and vehicles and the conflict zone, the speed of pedestrians and vehicles and the corresponding decision data are extracted. The extracted data are used to calibrate the parameters in the pedestrian and vehicle decision-making models to establish a fuzzy decision-making model.

[0098] Collect data needed for traffic volume forecasting. Collect historical traffic volume data for the airport departure lanes, including vehicle and pedestrian flow data.

[0099] In this embodiment, the traffic flow prediction of the airport departure lane is performed. The traffic volume prediction is performed using a long short-term memory neural network model. Specifically, it includes:

[0100] 1. Data preprocessing:

[0101] The collected data are preprocessed, including data cleaning and normalization, to ensure the quality and consistency of the data.

[0102] 2. Model construction and training:

[0103] A deep learning model based on the long short-term memory (LSTM) network is constructed to process time series data and predict future traffic flow. The LSTM network is composed of several LSTM units connected in series, and each unit includes three main gating mechanisms, namely the input gate, the forget gate, and the output gate, which are used to control the flow of information and the update of memory.

[0104] Input layer: The input layer accepts multi-dimensional historical traffic flow data and inputs it into the LSTM network as a time series.

[0105] LSTM layer: The core of the network consists of multiple LSTM layers, each of which contains several LSTM units that can capture long-term dependencies in the data and maintain the state information of the time series.

[0106] Fully connected layer: The output of the LSTM layer is processed by a fully connected layer to integrate the features and generate the final prediction results.

[0107] Output layer: The output layer converts the results of the fully connected layer into predicted values, i.e., future traffic flow.

[0108] The LSTM model is trained using the collected historical traffic flow data. During the training process, the back propagation algorithm is used to adjust the model parameters and optimize the model performance by minimizing the prediction error. The dataset is divided into a training set and a validation set to ensure the generalization ability of the model.

[0109] 3. Verification of prediction results:

[0110] The prediction results are compared with the actual data, the prediction error is calculated, and the prediction accuracy test is performed to evaluate the prediction performance of the model.

[0111] In this embodiment, whether to start signal control is determined based on the safety index. The conflict time difference TDTC can better simulate and analyze the potential conflict between pedestrians and vehicles. Its value specifically refers to the difference in time between the vehicle and the pedestrian maintaining the original speed and entering the conflict area. Its specific calculation formula is:

[0112]

[0113] Among them, TDTC p，v (t) is the time difference between the vehicle v and the pedestrian p reaching the conflict area at the original speed within the time step t; T p (t), T v(t) is the time when pedestrian p and vehicle v arrive at the conflict area; L p，v (t) is the position of the potential conflict point between pedestrian p and vehicle v; L p (t) and L v (t) represent the positions of pedestrians and vehicles respectively; and Represent the speeds of pedestrians and vehicles respectively.

[0114] The higher the absolute value of TDTC, the safer the interaction behavior is; and the closer the TDTC value is to zero, the more dangerous the interaction behavior is. The value of TDTC can be positive or negative. A positive TDTC means that the vehicle v arrives at the potential conflict area before the pedestrian p, and a negative TDTC means that the vehicle v arrives at the potential conflict area later than the pedestrian p.

[0115] like Figure 4 As shown, in order to explore the total number of interactive behaviors with potential conflict risks, the present invention considers that when the absolute value of TDTC is less than or equal to 3, there is a potential conflict in the human-vehicle interactive behavior. When the absolute value of TDTC is greater than 3, it is considered that there is no potential conflict. The specific formula is as follows:

[0116]

[0117] Where: X i Indicates the classification of the i-th TDTC value: When the absolute value of the i-th TDTC value is less than or equal to 3, X i =1, indicating that there is a potential conflict in the interaction behavior; when the absolute value of the i-th TDTC value is greater than 3, X i = 0, indicating that the interaction behavior has no potential conflict. NTDTC is the total number of TDTC values ​​less than or equal to 3 among n TDTC values.

[0118] In order to determine whether to use signal control, the present invention refers to the peak hour flow conditions for setting up pedestrian crossing signals on road sections specified in the "Road Traffic Signal Setting and Installation Specifications" (GB14886-2016), inputs several groups of traffic flows in the specifications into the fuzzy cellular automation model, and runs traffic simulation.

[0119] During the simulation process, when there is no conflict between vehicles and pedestrians, the vehicle motion model, vehicle lane change model and pedestrian motion model in the fuzzy cellular automation model are used to simulate the movement of vehicles and pedestrians in the road section; when there is a conflict between vehicles and pedestrians (at the crosswalk of the road section), the vehicle decision model and pedestrian decision model are used to generate the decision of whether to give way made by vehicles and pedestrians when facing a conflict between vehicles and pedestrians.

[0120] A safety index is introduced into the simulation to record the collision time difference of pedestrian-vehicle conflicts at crosswalks, and the total number of collision risks in the simulation cycle (denoted as NTDTC) is counted. Several NTDTC values ​​corresponding to different traffic flows in the specification are obtained, and the minimum value is selected as the safety threshold A through comparison.

[0121] The predicted traffic flow is input into the model, and the traffic simulation is run to obtain the corresponding NTDTC value, which is compared with A. If the NTDTC value is greater than A, traffic signal control needs to be enabled, otherwise traffic signal control does not need to be enabled.

[0122] As shown in Table 2, several groups of traffic flow in the specification are input into the fuzzy cellular automaton model, and the three N TDTC The smallest value is taken as the safety limit value A.

[0123] The predicted traffic flow value is input into the fuzzy cellular automation model. If the output N TDTC If the value is greater than the safety index A, the signal light at the pedestrian crossing of the airport departure lane needs to be turned on. TDTC If the value is less than the safety index A, there is no need to turn on the signal light.

[0124] Table 2

[0125]

[0126] In this embodiment, the signal timing scheme is optimized based on the total delay of the road section. The signal control is set on the road section with the goal of maximizing the overall traffic benefit. The optimized timing scheme should ensure the minimum total delay, that is, the weighted sum W of pedestrian delay and vehicle delay is minimized, and the expression is:

[0127] W=p×d kw +k×q×D kw ,

[0128] Where W is the weighted sum of pedestrian and vehicle delays; p and q are the pedestrian and vehicle traffic volumes, respectively; d kw , D kw are the average delays of pedestrians and vehicles under the signal control condition of the road section; k is the conversion coefficient of pedestrian and vehicle delays.

[0129] In the signal timing optimization stage, the initial signal timing plan is first given and input into the simulation model (fuzzy cellular automaton model) together with the predicted traffic volume to calculate the pedestrian and vehicle delay indicators of the road section. The signal timing plan is optimized by adjusting the signal cycle and green signal ratio. For example, the green signal ratio can be increased by arithmetic progression, with a fixed time increment each time, and then the delay indicator is recalculated, and the length of the signal cycle is gradually adjusted to find the optimal signal cycle.

[0130] Each adjusted signal timing scheme needs to be input into the simulation model to recalculate the pedestrian and vehicle delay indicators. Through multiple simulations and adjustments, the weighted sum W of pedestrian and vehicle delays under different schemes is compared, and the signal timing scheme that minimizes W is finally selected.

[0131] Through the above method, the traffic operation efficiency of the airport departure lane can be improved as much as possible while ensuring safety.

[0132] Embodiment 2

[0133] The present invention also provides an airport departure level lane traffic control decision system based on fuzzy cellular automata, comprising: a model building module, a data collection module, a flow prediction module and a signal timing module;

[0134] The model building module is used to build a fuzzy cellular automaton model suitable for the airport departure lane;

[0135] The data collection module is used to collect historical traffic flow data of the airport departure lanes;

[0136] The traffic prediction module is used to use the long short-term memory neural network to predict the future traffic flow by time period based on the historical traffic flow data of the airport departure lanes;

[0137] The signal timing module is used to evaluate the safety index of the airport departure lane based on the predicted traffic flow using the fuzzy cellular automaton model to determine whether traffic signal control needs to be enabled; if traffic signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation.

[0138] The present invention constructs a fuzzy cellular automaton model that can accurately simulate the human-vehicle interaction behavior on the airport departure lane. The fuzzy decision model is used to characterize the decision-making process of pedestrians and vehicles when encountering traffic conflicts, and the impact of the roadside parking and dropping-off behavior of vehicles on the airport departure lane on the traffic flow is particularly considered. Based on the collected human-vehicle conflict data, the fuzzy decision-making model of pedestrians and vehicles when encountering conflicts is calibrated. Based on the historical traffic flow of the airport departure lane, the long short-term memory neural network is used to predict the future traffic flow in different time periods. Based on the predicted flow, the constructed fuzzy cellular automaton model is used to evaluate the safety index of the airport departure lane to determine whether traffic signal control needs to be enabled. If signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation. This method provides a scientific basis and data support for the optimization of the start-stop and timing schemes of signal control, thereby improving the safety and traffic efficiency of the airport departure lane.

[0139] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A traffic control decision method for airport departure lane based on fuzzy cellular automation, characterized in that: The following steps are involved: Construct a fuzzy cellular automaton model suitable for airport departure lanes; Collect historical traffic flow data for airport departure lanes; Using long short-term memory neural network, based on the historical traffic flow data of the airport departure lane, the future traffic flow is predicted by time period; Based on the predicted traffic flow, the fuzzy cellular automaton model is used to evaluate the safety index of the airport departure lane to determine whether traffic signal control needs to be activated; If traffic signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation.

2. The airport departure lane traffic control decision method based on fuzzy cellular automata according to claim 1 is characterized in that: The fuzzy cellular automaton model is used to simulate the human-vehicle interaction behavior occurring in the crosswalk area of ​​the road section, including: a pedestrian motion model, a vehicle motion model, a pedestrian decision model, a vehicle decision model and a vehicle lane changing model.

3. The airport departure lane traffic control decision method based on fuzzy cellular automata according to claim 2 is characterized in that: Methods for building vehicle decision models include: For the video clip of the vehicle in front of the crosswalk, the time distance GAP1 between the vehicle and the conflict area and the pedestrian who has not passed the lane where the vehicle is located, the speed v1 of the pedestrian, and the corresponding driving decision are recorded; Classifying the levels of GAP1 and v1 according to the driving decision; The frequency of the level to which GAP1 and v1 with specific values ​​belong is calculated and used as the degree of membership; The values ​​of GAP1 and v1 and the corresponding membership are plotted into a scatter plot, and fitted with a function to determine the type and parameters of the membership function that meet the preset requirements; Based on the membership function type and parameters, the vehicle decision model is constructed.

4. The airport departure lane traffic control decision method based on fuzzy cellular automata according to claim 2 is characterized in that: Methods for building pedestrian decision models include: For the video clips of pedestrians crossing the street, the GAP2 and speed v2 of the vehicle closest to the conflict zone in the lane ahead, as well as the specific decision of the pedestrian, are recorded; The types of GAP2 and v2 are estimated based on pedestrian decisions; Calculate the frequency of the type of GAP2 and v of a specific value and use it as the degree of membership; The values ​​of GAP2 and v2 and the corresponding membership are plotted into a scatter plot, and a function is used to fit them to determine the type and parameters of the membership function that meet the preset requirements, and finally the pedestrian decision model is calibrated.

5. The airport departure lane traffic control decision method based on fuzzy cellular automata according to claim 1 is characterized in that: Based on the predicted traffic flow, the method for evaluating the safety index of the airport departure lane using the fuzzy cellular automaton model includes: Among them, TDTC p，v (t) is the time difference between the vehicle v and the pedestrian p reaching the conflict area at the original speed within the time step t; T p (t), T v (t) is the time when pedestrian p and vehicle v arrive at the conflict area; L p，v (t) is the position of the potential conflict point between pedestrian p and vehicle v; L p (t) and L v (t) represent the positions of pedestrians and vehicles respectively; and Represent the speeds of pedestrians and vehicles respectively.

6. The airport departure lane traffic control decision method based on fuzzy cellular automata according to claim 5 is characterized in that: Methods for determining whether traffic signal control needs to be activated include: Input several groups of traffic flow in the specification into the fuzzy cellular automaton model and compare the three N TDTC The smallest value is taken as the safety limit value A; The predicted traffic flow value is input into the fuzzy cellular automation model. If the output N TDTC If the value is greater than the safety index A, the signal light at the pedestrian crossing of the airport departure lane needs to be turned on. TDTC If the value is less than the safety index A, there is no need to turn on the signal light.

7. The airport departure lane traffic control decision method based on fuzzy cellular automata according to claim 1 is characterized in that: If traffic signal control is required, the method for determining the optimal signal timing based on the traffic delay index obtained by simulation includes: The goal of setting up signal control on a road section is to maximize the overall traffic benefit. The optimized timing plan should ensure the minimum total delay, that is, the weighted sum W of pedestrian delay and vehicle delay is minimized, expressed as: W=p×d kw +k×q×D kw , Where W is the weighted sum of pedestrian and vehicle delays; p and q are the pedestrian and vehicle traffic volumes, respectively; d kw , D kw are the average delays of pedestrians and vehicles under the signal control condition of the road section; k is the conversion coefficient of pedestrian and vehicle delays.

8. The airport departure lane traffic control decision system based on fuzzy cellular automation is characterized by: include: Model building module, data collection module, traffic prediction module and signal timing module; The model building module is used to build a fuzzy cellular automaton model suitable for the airport departure lane; The data collection module is used to collect historical traffic flow data of the airport departure lane; The traffic prediction module is used to use a long short-term memory neural network to predict future traffic flow by time period based on the historical traffic flow data of the airport departure lane; The signal timing module is used to evaluate the safety index of the airport departure lane based on the predicted traffic flow and use the fuzzy cellular automation model to determine whether traffic signal control needs to be enabled; If traffic signal control is required, the optimal signal timing is determined based on the traffic delay index obtained by simulation.

Images