A Cellular Automaton Lane-Changing Simulation Method for Introducing the Position of Maximum Attraction

Abstract

The present invention discloses a cellular automaton lane-changing simulation method introducing the maximum attraction position, comprising the steps of: judging whether a vehicle has a lane-changing requirement; calculating the maximum lateral vehicle speed; determining the lateral movement space of the vehicle; calculating the maximum longitudinal vehicle speed of the vehicle; determining the longitudinal movement space of the vehicle; searching for the maximum attraction position; and position updating. The present invention comprehensively considers the positional relationship between a vehicle and other vehicles at the same moment in both the lateral and longitudinal directions, quantifies the attraction of different possible arrival positions to the driver, can solve the problem of the influence of "the separation of lateral and longitudinal movements and the selection of lane-changing positions being obviously inconsistent with reality" on lane-changing in the study of vehicle lane-changing using the cellular automaton model, and has important research value for studying vehicle lane-changing problems using the cellular automaton model.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent traffic management and control, and particularly relates to a cellular automaton lane-changing simulation method introducing the maximum attractive position. Background Art

[0002] After statistical analysis of traffic accidents that occurred in recent years by the National Highway Traffic Safety Administration of the United States, it was found that among vehicle traffic accidents, about 27% of the accidents were caused by vehicle lane changes. Similar conclusions were also drawn from large-scale real vehicle road test experiments in China: among the traffic accidents that occurred in recent years, 23.91% were caused by lane changes. In the research of lane-changing models, the cellular automaton model is widely used in the research of vehicle lane changes due to its advantages such as well conforming to the discrete characteristics of traffic flow, considering lane-changing safety in the lane-changing model, and using fuzzy theory to simulate human thinking and decision-making.

[0003] Although the discrete cellular automaton can simplify the traffic flow problem into an intuitive simulation process, the space and speed of vehicles in the model change in a jumpy manner, which is disadvantageous in studying the continuous changes of vehicles and cannot specifically show the changes in the lateral speed and position transformation of vehicle lane changes and other changes during the lane-changing process. Therefore, researchers improve the accuracy by refining the cellular space to make the discrete process approach the continuous process, but usually mainly refine the longitudinal space and do not refine the lateral space, so the complex front-back relationships between vehicles in reality cannot be shown in the cellular automaton. In addition, the lane-changing model based on cellular automata always divides vehicle movement into lateral movement and longitudinal movement, and the two are asynchronous. Regardless of the order of the two, the movement that occurs first will change the relationships with other vehicles at the location of the vehicle, making the movement that occurs later make a judgment on the target position in the changed relationship with other vehicles. Similarly, the decision-making of the lane-changing position also becomes the superposition of two position selections in different surrounding vehicle environments. This separation of lateral and longitudinal movements and the selection of lane-changing positions are obviously inconsistent with reality. In reality, the movement position of a vehicle at the next moment is determined based on the comprehensive lateral and longitudinal position relationships with other vehicles at the same moment. Summary of the Invention

[0004] The purpose of the present invention is to, on the basis of refining the lateral cellular space, comprehensively consider the position relationships of vehicles with other vehicles in the lateral and longitudinal directions at the same moment, quantify the attractiveness of different possible arrival positions to drivers, and propose a cellular automaton simulation method for the vehicle lane-changing process introducing the maximum attractive position, and for the first time, a lane-changing model that drives to the best lane-changing position in the comprehensive lateral and longitudinal directions is implemented in the cellular automaton model, making the cellular automaton simulation process closer to the actual situation.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A cellular automaton lane-changing simulation method for introducing the maximum attractive position, comprising the following steps:

[0007] 1) Determine whether the vehicle has a lane-changing requirement;

[0008] By comparing the current vehicle speed and the distance to the vehicle ahead (the distance the vehicle advances at the next moment), if the distance to the vehicle ahead is less than the vehicle speed, a lane-changing requirement is generated; otherwise, no lane-changing requirement is generated, that is, it reflects whether the current lane can meet the driver's pursuit of speed. If the distance to the vehicle ahead is less than the vehicle speed, the driver cannot maintain or pursue a higher vehicle speed; because the time of the model is 1 s, the distance the vehicle advances at the next moment is the vehicle speed, so if the distance the vehicle advances at the next moment < the distance to the vehicle ahead, it is required that the vehicle speed < the distance to the vehicle ahead.

[0009] 2) Calculate the maximum lateral vehicle speed;

[0010] Calculate its maximum lateral vehicle speed from the current vehicle speed and the head angle of the vehicle; take the maximum angle between the vehicle head and the road cross-section during lane-changing as α, and the current vehicle speed as V, so the maximum lateral vehicle speed = V * sinα;

[0011] The current vehicle speed of the vehicle is a dynamic attribute of the vehicle and changes continuously with the evolution of the model. The evolution of the model is the continuous calculation of the dynamic attributes (speed, position, etc.) of the vehicle. Therefore, after giving the initial values of all attributes (the initial values are randomly generated), the attribute values at the next moment can be determined from the attribute values at the previous moment;

[0012] 3) Determine the lateral movement space of the vehicle;

[0013] Determine the lateral movement space from the maximum lateral vehicle speed. The lateral movement space is the range of the vehicle's lateral movement within the next time step and is a discrete set of lateral positions in the cellular automaton;

[0014] 4) Calculate the maximum longitudinal vehicle speed;

[0015] ① Vehicle speed = min(vehicle's current vehicle speed, distance to the vehicle ahead, maximum vehicle speed)

[0016] ② If a randomly generated decimal is less than the random deceleration probability, then:

[0017] Vehicle speed = max(1, vehicle speed - longitudinal acceleration)

[0018] 5) Determine the longitudinal movement space of the vehicle;

[0019] Determine the longitudinal movement space from the maximum longitudinal vehicle speed. The longitudinal movement space is the range that the vehicle can reach in the longitudinal movement within the next time step and is a discrete set of lateral positions in the cellular automaton;

[0020] 6) Search for the maximum attractive position;

[0021] All possible spatial positions that the vehicle may reach at the next time step are obtained from the lateral movement space and the longitudinal movement space. The attraction of each position to the driver is calculated in this set of spatial positions, and the position with the maximum attraction is the position with the maximum attraction;

[0022] 7) Position update;

[0023] The vehicle position is changed to the position with the maximum attraction. The model adopts periodic boundary conditions, that is, if a vehicle drives out from the right boundary, a vehicle with the same state drives in at the corresponding position of its left boundary to ensure the conservation of the number of vehicles on the road. In the cellular automaton model, the model update order is divided into synchronous update and asynchronous update. The longitudinal position update of the vehicle is divided into spatial order and random order. Through the combination of different model updates and the longitudinal position update order of the vehicle, it is concluded that the asynchronous update order of from front to back and giving priority to faster vehicles in the same column is more in line with the actual situation during simulation.

[0024] The present invention is further configured as follows: In step 1), it is judged whether the vehicle has a lane-changing requirement. Specifically,

[0025] It is judged whether the current lane can meet the driver's speed requirement. If the distance to the vehicle in front is less than the vehicle speed, the vehicle cannot pursue or maintain a higher speed, then the vehicle has a lane-changing requirement, otherwise the vehicle continues to drive;

[0026] The present invention is further configured as follows: In step 3), the lateral movement space of the vehicle is determined. Specifically,

[0027] The vehicle in the model is in the form of an object occupying a two-dimensional rectangular area, with four attributes: position, length, width, and speed. Among them, the position is the position (m, n) of the midpoint of the vehicle area, m is the lateral position, and n is the longitudinal position. In the cellular space, taking (m, n) as the midpoint and making a rectangle with the length and width of the vehicle, the space occupied by the vehicle in the cellular space can be obtained;

[0028] The lateral movement space is determined by the maximum lateral vehicle speed. The lateral movement space is the range of lateral movement of the vehicle in the next time step, which is a discrete set of lateral positions in the cellular automaton. The steps for determining the lateral space by the maximum lateral vehicle speed include:

[0029] 3-1) Minimum value of the lateral space = max(1, vehicle lateral position - lateral vehicle speed)

[0030] 3-2) Maximum value of the lateral space = min(maximum lateral position of the lane, vehicle lateral position + lateral vehicle speed)

[0031] The lateral space is an integer set with an interval of 1 from the minimum value of the lateral space to the maximum value of the lateral space, representing the range of lateral movement positions of the vehicle in the next time step.

[0032] The present invention is further configured such that in step 4), calculating the maximum longitudinal vehicle speed is specifically as follows:

[0033] 4-1) Vehicle speed = min(vehicle current speed, distance to the vehicle ahead)

[0034] 4-2) If a randomly generated decimal is less than the random deceleration probability, then:

[0035] Vehicle speed = max(1, vehicle speed - longitudinal acceleration)

[0036] The present invention is further configured such that in step 6), searching for the position with the maximum attraction; specifically, obtaining a search space S from the lateral movement space and the longitudinal movement space, S being a two-dimensional spatial position matrix, row(S) being the number of rows of S, traversing each row in S, i.e., a lateral space A, and searching for the target lateral position of the lane-changing vehicle within the lateral space range, that is, calculating the attraction of each position to the driver within the lateral space range, and the position with the maximum attraction in the entire space is the position with the maximum attraction, including:

[0037] 6-1) Initializing search variables:

[0038] j = 1

[0039] Loc = 0

[0040] d_max = 0

[0041] where j is the loop variable for traversing the lateral space;

[0042] Loc is used to store the position with the maximum attraction;

[0043] d_max is used to store the distance to the vehicle ahead of the position with the maximum attraction;

[0044] 6-2) Updating the lateral space:

[0045] A = S[j, :]

[0046] where A is the j-th row in S, i.e., a lateral space;

[0047] 6-3) Initializing intermediate variables:

[0048] i = min(A)

[0049] m_aim = m

[0050] d_aim = 0

[0051] i_range = m

[0052] where i is the detected lateral space position, i.e., the loop variable when traversing the lateral space;

[0053] m_aim is used to store the target lateral position;

[0054] m is the current lateral position of the vehicle;

[0055] d_aim is used to store the distance to the vehicle in front at the target lateral position;

[0056] i_range is used to distinguish whether the loop variable is on the left or right side of the vehicle;

[0057] 6-4) Determine whether the safety condition is met at i, that is, determine whether the distance to the vehicle in front and the distance to the vehicle behind at i meet the safe distance; if it is met, go to 6-5, if not, go to step 6-6;

[0058] 6-5) Determine whether the distance to the vehicle in front at i is greater than d_aim. If so, then:

[0059] m_aim = i

[0060] d_aim = d(i)

[0061] where d(i) is the distance to the vehicle in front at i;

[0062] If not, go to 6-6;

[0063] 6-6) Determine whether i < i_range? If so, i = i + 1, and go to step 6-4. If not, go to 6-7;

[0064] 6-7) Determine whether i_range = m? If so, then:

[0065] i = m + 1

[0066] i_range = max(A)

[0067] Go to step 6-4

[0068] If not, go to step 6-8;

[0069] 6-8) Update the position of the maximum attraction;

[0070] Determine whether d_max < d_aim? If so, then:

[0071] Loc = m_aim;

[0072] Determine whether j > row(S)? If not, then:

[0073] j = j + 1

[0074] Go to step 6-2;

[0075] If so, the search is completed, and Loc is the position with the maximum attraction force.

[0076] The present invention is further configured that: the position update in the step 7) is specifically as follows

[0077] The model adopts periodic boundary conditions, that is, if a vehicle drives out from the right boundary, a vehicle with the same state drives in at the corresponding position on the left boundary, ensuring the conservation of the number of vehicles on the road. In the cellular automaton model, the model update sequence is divided into synchronous update and asynchronous update, and the longitudinal position update of the vehicle is divided into spatial order and random order. Through the combination of different model updates and the longitudinal position update order of the vehicle, it is concluded that the asynchronous update order of from front to back and giving priority to the fast vehicle in the same column is more in line with the actual situation during simulation;

[0078] The longitudinal position update of the vehicle, that is, the position of the vehicle at the next moment = the position of the vehicle at this moment + the vehicle speed;

[0079] Compared with the prior art, the beneficial effects of the present invention are as follows

[0080] On the basis of refining the horizontal cell space, the present invention comprehensively considers the positional relationship between vehicles in the same moment in the horizontal and vertical directions, quantifies the attraction force of different possible arrival positions on the driver, and proposes a cellular automaton simulation method for the vehicle lane-changing process introducing the maximum attraction force position, and for the first time realizes a lane-changing model that drives to the best lane-changing position in the comprehensive horizontal and vertical directions in the cellular automaton model. Therefore, it can solve the problem of the influence of the separation of horizontal and vertical movements and the selection of lane-changing positions that are obviously inconsistent with reality in the study of the vehicle lane-changing process using the cellular automaton model, and has important research value for using the cellular automaton model to study the vehicle lane-changing problem.

[0081] The present invention promotes and improves the application of the cellular automaton model in traffic flow simulation, and plays a promoting role in the application of the cellular automaton model in traffic flow simulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 is the vehicle lane-changing flow chart of the present invention;

[0083] Figure 2 is the flow chart for searching the maximum attraction force position in the step 6) of the present invention;

[0084] Figure 3 is the schematic diagram of the relationship between the vehicle width refinement and the vehicle in front in the embodiment of the present invention, where 3a shows that there is a vehicle in the left front of vehicle i, 3b shows that there is a vehicle directly in front of vehicle i, and 3c shows that there is a vehicle in the right front of vehicle i;

[0085] Figure 4 is the schematic diagram of the step-by-step lane-changing process of the vehicle in the embodiment of the present invention;

[0086] Figure 5 It is a schematic diagram of a fast car overtaking and changing lanes in an embodiment of the present invention;

[0087] Figure 6 It shows a schematic diagram of the head angle. Specific implementation manners

[0088] The present invention will be further described below in conjunction with the accompanying drawings of the specification.

[0089] A cellular automaton lane-changing simulation method introducing the maximum attraction position in the present invention includes the following steps:

[0090] 1) Determine whether the vehicle has a lane-changing requirement;

[0091] By comparing the current vehicle speed and the distance to the vehicle ahead, if the distance to the vehicle ahead is less than the vehicle speed, a lane-changing requirement is generated, otherwise no lane-changing requirement is generated, that is, it reflects whether the current lane can meet the driver's pursuit of speed. If the distance to the vehicle ahead is less than the vehicle speed, the driver cannot maintain or pursue a higher vehicle speed;

[0092] Determine whether the current lane can meet the driver's speed requirement. If the current distance to the vehicle ahead is less than the vehicle speed of the vehicle, making the vehicle unable to maintain or pursue a higher vehicle speed, the vehicle generates a lane-changing requirement, otherwise the vehicle continues to travel at the current lateral position.

[0093] 2) Calculate the maximum lateral vehicle speed;

[0094] Calculate its maximum lateral vehicle speed from the current vehicle speed and the maximum head angle; take the maximum inclination angle of the vehicle head during lane-changing as α, and the current vehicle speed as V, so the maximum lateral vehicle speed = V * sinα;

[0095] 3) Determine the lateral movement space of the vehicle;

[0096] Determine the lateral movement space from the maximum lateral vehicle speed. The lateral movement space is the range that the vehicle can reach laterally in the next time step, and in the cellular automaton, it is a discrete set of lateral positions;

[0097] The vehicle in the model is in the form of an object occupying a two-dimensional rectangular area and has four attributes: position, length, width, and speed. Among them, the position is the position of the midpoint of the vehicle area. In the cellular space, a rectangle with the length and width being the vehicle length and width respectively is made with this position as the midpoint, that is, the space occupied by the vehicle in the cellular space is obtained;

[0098] The steps of determining the lateral movement space from the maximum lateral vehicle speed of the vehicle include:

[0099] 3-1) The minimum value of the lateral movement space = max(1, vehicle lateral position - lateral vehicle speed);

[0100] 3 - 2) Maximum value of lateral movement space = min(maximum lateral position of the lane, vehicle lateral position + lateral vehicle speed);

[0101] 3 - 3) The lateral movement space is the set of integers at intervals of 1 from the minimum value of the lateral movement space to the maximum value of the lateral movement space, representing the range of lateral movement positions of the vehicle in the next time step.

[0102] 4) Calculate the maximum longitudinal vehicle speed;

[0103] 4 - 1) Vehicle speed = min(current vehicle speed, distance to the vehicle ahead, maximum vehicle speed)

[0104] 4 - 2) If a randomly generated decimal is less than the random deceleration probability, then:

[0105] Vehicle speed = max(1, vehicle speed - longitudinal acceleration)

[0106] 5) Determine the longitudinal movement space of the vehicle;

[0107] The longitudinal movement space is determined by the maximum longitudinal vehicle speed. The longitudinal movement space is the range that the vehicle can reach in longitudinal movement in the next time step, which is a discrete set of lateral positions in the cellular automaton;

[0108] The longitudinal movement space is the set of integers at intervals of 1 from the minimum value of the longitudinal movement space to the maximum value of the longitudinal movement space, representing the range of longitudinal movement positions of the vehicle in the next time step.

[0109] 6) Search for the position with the maximum attraction;

[0110] All possible spatial positions that the vehicle can reach in the next time step are obtained from the lateral movement space and the longitudinal movement space. The attraction of each position to the driver is calculated in this set of spatial positions, and the position with the maximum attraction is the position with the maximum attraction;

[0111] The search space S is obtained from the lateral movement space and the longitudinal movement space. S is a two - dimensional spatial position matrix, and row(S) is the number of rows of S. Each row in S, that is, a lateral space A, is traversed. The target lateral position of the lane - changing vehicle is searched within the lateral space range, that is, the attraction of each position within the lateral space range to the driver is calculated, and the position with the maximum attraction in the entire space is the position with the maximum attraction.

[0112] The calculation of the attraction specifically includes:

[0113] ① If the safety condition is not met, the attraction is 0;

[0114] ② If the safety condition is met, the attraction is the distance to the vehicle ahead;

[0115] The specific process includes:

[0116] 6-1) Initialization of search variables:

[0117] j = 1

[0118] Loc = 0

[0119] d_max = 0

[0120] Among them, j is the loop variable for traversing the lateral space;

[0121] Loc is used to store the position of the maximum attraction;

[0122] d_max is used to store the distance to the vehicle in front at the position of the maximum attraction;

[0123] 6-2) Update the lateral space:

[0124] A = S[j, :]

[0125] Among them, A is a lateral space in the j-th row of S;

[0126] 6-3) Initialization of intermediate variables:

[0127] i = min(A)

[0128] m_aim = m

[0129] d_aim = 0

[0130] i_range = m

[0131] Among them, i is the position for detecting the lateral space, that is, the loop variable when traversing the lateral space;

[0132] m_aim is used to store the target lateral position;

[0133] m is the current lateral position of the vehicle;

[0134] d_aim is used to store the distance to the vehicle in front at the target lateral position;

[0135] i_range is used to distinguish whether the loop variable is on the left or right side of the vehicle;

[0136] 6-4) Determine whether the safety condition is met at i, that is, determine whether the distance to the vehicle in front and the distance to the vehicle behind at i meet the safe distance; if it is met, go to 6-5, if not, go to step 6-6;

[0137] 6-5) Determine whether the distance to the vehicle in front at i is greater than d_aim, if so, then:

[0138] m_aim = i

[0139] d_aim = d(i)

[0140] Among them, d(i) is the distance to the vehicle in front at position i;

[0141] If not, go to 6-6;

[0142] 6-6) Determine whether i < i_range? If so, i = i + 1, and go to step 6-4; if not, go to 6-7;

[0143] 6-7) Determine whether i_range = m? If so, then:

[0144] i = m + 1

[0145] i_range = max(A)

[0146] Go to step 7-4

[0147] If not, go to step 7-8;

[0148] 6-8) Update the position of the maximum attraction;

[0149] Determine whether d_max < d_aim? If so, then:

[0150] Loc = m_aim;

[0151] Determine whether j > row(S)? If not, then:

[0152] j = j + 1

[0153] Go to step 6-2;

[0154] If so, the search is completed, and Loc is the position of the maximum attraction.

[0155] 7) Position update;

[0156] The vehicle position is changed to the position of the maximum attraction. The model adopts periodic boundary conditions, that is, if a vehicle exits from the right boundary, a vehicle with the same state will enter from the corresponding position on the left boundary to ensure the conservation of the number of vehicles on the road.

[0157] Example:

[0158] A further illustration of the cellular automaton lane-changing simulation method introducing the maximum attraction position of the present invention is given through an example.

[0159] By refining the lateral space, the front-back relationship between vehicles is extended compared with the traditional model. For example, when the vehicle width is 2 cells, there are three cases where there is a vehicle in front of vehicle i, as Figure 3As shown in the figure. The difference among these three cases is that if vehicle i wants to change lanes to get rid of the influence of the front vehicle distance caused by vehicle j in front, due to different lateral movement distances, the vehicle lane change takes into account the actual spatial position relationship between the front and rear vehicles. The front and rear relationship is no longer only the two cases of directly in front and not directly in front, which is obviously more in line with the actual situation.

[0160] After introducing the maximum attractive position, the vehicle lane change becomes more flexible. When the vehicle changes lanes, the lateral movement space is obtained according to the maximum lateral vehicle speed, corresponding to the driver's estimation of the lane change range in actual driving, and the possible movement range of the vehicle at the next moment is obtained from the longitudinal vehicle speed, corresponding to the driver's estimation of the position that the vehicle can reach at the next moment in actual driving. In the model, the lane change process is described as the search for the maximum attractive position within all possible reachable spaces, and the attraction is the front vehicle distance under safe conditions. If the safe conditions are not met, the attraction is 0, corresponding to the driver's pursuit of a larger road space on the premise of safety in actual driving.

[0161] As Figure 4 shown in the figure, vehicle i will choose different lane change strategies within the lateral movement space in the face of different spatial occupancy situations of the vehicle in front. The dotted line represents the lateral position of vehicle i at the next time step. Therefore, after introducing the maximum attractive position, the vehicle lane change becomes a process affected by the vehicle's own speed and the spatial occupancy situation of surrounding vehicles, rather than the result described as a one-step process in the traditional model. And the vehicle lane change will have different strategies due to different vehicle speeds and the spatial occupancy situation of surrounding vehicles, making the decision-making more flexible.

[0162] Now an example is given to illustrate that this model is more in line with the real driving situation. As Figure 5 shown in the figure, in the real world, when two slow vehicles are in front of a fast vehicle, the fast vehicle will look for opportunities to overtake. However, in the scenario described in the figure, the fast vehicle cannot complete the overtaking in the traditional model. Among them, the squares are cells, the red lines are lane dividers, the rectangular blocks are vehicles, and the letters in the rectangles are vehicle labels, and t is the time step of model evolution. In the figure, the speed of vehicle i is much faster than the vehicles j and k in front of it, and the longitudinal distance between vehicles j and k is very small. For the traditional cellular automaton model, vehicle i can only change lanes to follow vehicle k, and is restricted by the too small front and rear vehicle distances on both sides and cannot continue to change lanes. In this model, there is an unoccupied lateral space between vehicles j and k, which is enough for vehicle i to complete a straight movement. Therefore, vehicle i can give full play to its speed advantage, change lanes to the unoccupied lateral space in the middle at t = 2, exceed vehicles j and k in the longitudinal position at t = 3 and make the longitudinal distance between it and vehicles j and k safe enough, and at t = 4, change lanes back to the original lane to complete the overtaking, which is obviously more in line with the lane change law of the fast vehicle facing the slow vehicle in the real world.

Claims

1. A cellular automaton lane-changing simulation method for introducing the maximum attractive position, characterized in that It includes the following steps: 1) Determine whether the vehicle has a lane-changing requirement; By comparing the current vehicle speed and the distance to the vehicle ahead, if the distance to the vehicle ahead is less than the current vehicle speed, a lane-changing requirement is generated; otherwise, no lane-changing requirement is generated, that is, it reflects whether the current lane can meet the driver's pursuit of speed. If the distance to the vehicle ahead is less than the current vehicle speed, the driver cannot maintain or pursue a higher vehicle speed; 2) Calculate the maximum lateral vehicle speed; Calculate its maximum lateral vehicle speed from the current vehicle speed and the maximum head angle; take the maximum angle of the vehicle's head turning during lane change as α, and the current vehicle speed as V, so the maximum lateral vehicle speed = V * sinα; 3) Determine the lateral movement space of the vehicle; Determine the lateral movement space from the maximum lateral vehicle speed. The lateral movement space is the range that the vehicle can reach laterally in the next time step, and in the cellular automaton, it is a discrete set of lateral positions; 4) Calculate the maximum longitudinal vehicle speed; 4-1) Vehicle speed = min(current vehicle speed of the vehicle, distance to the vehicle ahead, maximum vehicle speed) 4-2) If a randomly generated decimal is less than the random deceleration probability, then: Vehicle speed = max(1, vehicle speed - longitudinal acceleration) 5) Determine the longitudinal movement space of the vehicle; Determine the longitudinal movement space from the maximum longitudinal vehicle speed. The longitudinal movement space is the range that the vehicle can reach longitudinally in the next time step, and in the cellular automaton, it is a discrete set of lateral positions; 6) Search for the position with the maximum attraction; Obtain all possible spatial positions that the vehicle can reach in the next time step from the lateral movement space and the longitudinal movement space. Calculate the attraction of each position to the driver in this set of spatial positions, and the position with the maximum attraction is the position with the maximum attraction; specifically: Obtain the search space S from the lateral movement space and the longitudinal movement space. S is a two-dimensional spatial position matrix, and row(S) is the number of rows of S. Traverse each row in S, that is, a lateral space A. Search for the target lateral position of the lane-changing vehicle within the lateral space range, that is, calculate the attraction of each position to the driver within the lateral space range. The position with the maximum attraction searched throughout the space is the position with the maximum attraction; 7) Update the position; Change the vehicle position to the position with the maximum attraction. The model adopts periodic boundary conditions, that is, if a vehicle drives out from the right boundary, a vehicle with the same characteristics drives in at the corresponding position on the left boundary to ensure the conservation of the number of vehicles on the road.

2. The cellular automaton lane-changing simulation method for introducing the maximum attraction position according to claim 1, wherein: The specific method for step 1) to determine whether the vehicle has a lane-changing requirement is as follows: Judge whether the current lane can meet the driver's speed requirement. If the current distance to the vehicle ahead is less than the vehicle speed, making the vehicle unable to maintain or pursue a higher vehicle speed, then the vehicle generates a lane-changing requirement; otherwise, the vehicle continues to drive at the current lateral position.

3. The cellular automaton lane-changing simulation method for introducing the maximum attraction position according to claim 2, characterized in that: The specific method for step 3) to determine the lateral movement space of the vehicle is as follows: In the model, the vehicle is in the form of an object occupying a two-dimensional rectangular area, with four attributes: position, length, width, and speed. Among them, the position is the position of the midpoint of the vehicle area. In the cellular space, with this position as the midpoint, make a rectangle with the length and width being the vehicle length and width respectively, that is, obtain the space occupied by the vehicle in the cellular space; The steps to determine the lateral movement space from the maximum lateral vehicle speed of the vehicle include: 3-1) Minimum value of lateral movement space = max(1, vehicle lateral position - maximum lateral vehicle speed); 3-2) Maximum value of lateral movement space = min(maximum lateral position of the lane, vehicle lateral position + maximum lateral vehicle speed); 3-3) The lateral movement space is the set of integers with an interval of 1 from the minimum value of the lateral movement space to the maximum value of the lateral movement space, representing the range of the vehicle's lateral movement position in the next time step.

4. A cellular automaton lane-changing simulation method for introducing the maximum attraction position according to claim 3, characterized in that: Step 5) determines the longitudinal movement space of the vehicle, specifically: The longitudinal movement space is the set of integers with an interval of 1 from the minimum value of the longitudinal movement space to the maximum value of the longitudinal movement space, representing the range of the vehicle's longitudinal movement position in the next time step.

5. A cellular automaton lane-changing simulation method for introducing the maximum attraction position according to claim 1, characterized in that: The calculation of the attraction force includes: ① If the safety condition is not met, the attraction force is 0; ② If the safety condition is met, the attraction force is the distance to the vehicle ahead; The specific process includes: 6-1) Initialization of search variables: j＝1 Loc = 0 d_max = 0 where j is the loop variable for traversing the lateral space; Loc is used to store the position of the maximum attraction force; d_max is used to store the distance to the vehicle ahead at the position of the maximum attraction force; 6-2) Update the lateral space: A = S[j, :] where A is the first lateral space in the j-th row of S; 6-3) Initialization of intermediate variables: i = min(A) m_aim = m d_aim = 0 i_range = m where i is the position for detecting the lateral space, that is, the loop variable when traversing the lateral space; m_aim is used to store the target lateral position; m is the current lateral position of the vehicle; d_aim is used to store the distance to the vehicle ahead at the target lateral position; i_range is used to distinguish whether the loop variable is on the left or right side of the vehicle; 6-4) Determine whether the safety condition is met at position i, that is, determine whether the distance to the vehicle ahead and the distance to the vehicle behind at position i meet the safe distance; if it is met, go to 6-5, if not, go to step 6-6; 6-5) Determine whether the distance to the vehicle ahead at position i is greater than d_aim, if so, then: m_aim = i d_aim = d(i) where d(i) is the distance to the vehicle ahead at position i; If not, go to 6-6; 6-6) Determine whether i < i_range? If so, i = i + 1, and go to step 6-4, if not, go to 6-7; 6-7) Determine whether i_range = m? If so, then: i = m + 1 i_range = max(A) Go to step 7-4 If not, go to step 7-8; 6-8) Update the position of the maximum attraction force; Determine whether d_max < d_aim? If so, then: Loc = m_aim; Determine whether j > row(S)? If not, then: j = j + 1 Go to step 6-2; If so, the search is completed, and Loc is the position of the maximum attraction force.

Images