Unmanned station scheduling method and system

Abstract

The present invention belongs to the technical field of unmanned bus station dispatching, and specifically relates to an unmanned station dispatching method and system. The present invention constructs a cloud control platform that can communicate with unmanned buses and roadside equipment, and performs dispatching control through the cloud control platform. When formulating the SOC charging strategy, the actual needs of passengers and the remaining SOC power and driving status of the bus are fully considered, so that the formulation of the SOC charging strategy can meet the passenger-carrying task; at the same time, the present invention plans the driving route within the station for unmanned buses based on the distribution of parking spaces and charging spaces within the station and the characteristics of the station roads, and combines the SOC charging strategy to improve the efficiency of bus dispatching at the station. In addition, the present invention also takes into account the problems in the case of meeting, and performs meeting control with the goals of safety benefits, smoothness benefits and high efficiency benefits, thereby ensuring the safety, smoothness and efficiency of vehicle passage.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned bus station dispatching, and in particular relates to an unmanned bus station dispatching method and system. Background Art

[0002] With the continuous advancement of autonomous driving technology research, countries around the world have achieved significant success in the field of driverless public transportation. In recent years, cities in my country, including Suzhou, Chongqing, Shenzhen, Zhengzhou, Guangzhou, Haikou, and Tianjin, have deployed driverless bus routes on public roads, with operations and trial runs currently underway in many locations. Unmanned bus stations serve as the dispatching centers for driverless buses, and the trend toward intelligent and efficient operation is inevitable. However, the traditional bus station dispatching model still suffers from low traffic efficiency, untimely charging, and irregular dispatching, resulting in wasted time and space resources, loss of charging facilities, and a lack of consideration for passenger needs.

[0003] To solve the above problems, a more efficient scheduling method is needed, which can realize an efficient station operation mode through cloud control platform and 5G and other technologies. The Chinese patent application document with application publication number CN110641302A discloses a scheduling method, device and method for charging stations. This solution solves the problem of uneven use of chargers in charging stations by allocating charging guns with the longest idle time or the lowest temperature to vehicles to be charged. However, this method only improves the utilization rate of chargers, and does not consider the SOC remaining power of the vehicle to determine the specific charging mode. The Chinese patent application document with application publication number CN114387973A discloses a new energy station scheduling method and device. This method automatically generates a scheduling voice when it is determined that a preset target fault exists in a new energy station based on voice information, and feeds it back to the scheduling platform, and the scheduling platform schedules the new energy station. This method and device improve the efficiency of station scheduling to a certain extent, but the scheduling platform is highly dependent on manual participation and has a low degree of automation.

[0004] Therefore, the current unmanned station scheduling either fails to consider the remaining SOC power of the vehicle, resulting in unreasonable scheduling results or even failure to meet the needs of passengers; or relies on manual labor, with a low degree of automation and low efficiency. Summary of the Invention

[0005] The purpose of the present invention is to provide an unmanned station scheduling method and system to solve the problems of unreasonable scheduling and low efficiency in existing unmanned station scheduling.

[0006] To solve the above technical problems, the present invention provides an unmanned station scheduling method, comprising the following steps:

[0007] 1) Collect historical bus route records, classify passenger demand into several levels based on passenger volume, and formulate departure schedules based on the levels;

[0008] 2) Obtain the remaining SOC power and driving status of buses inside and outside the unmanned station, as well as the occupancy of charging stations inside the station. Adjust the SOC charging strategy according to the departure schedule to ensure that at each moment there are at least two buses in the station with SOC levels sufficient to complete the passenger transportation task;

[0009] 3) Obtain the distribution of parking spaces and charging stations within the station, as well as the characteristics of the station roads, and plan the driving routes within the station for the unmanned bus based on the SOC charging strategy;

[0010] 4) Control the operation of the unmanned bus within the station according to the planned route within the station and based on at least one of the following benefits: safety, ride comfort, and efficiency. Safety benefit refers to overtaking and lane changing during vehicle operation, ride comfort refers to the uniform speed of the vehicle, and efficiency benefit refers to the average speed of the vehicle.

[0011] When formulating the SOC charging strategy, the present invention fully considers the actual needs of passengers, the remaining SOC power of the bus, and the driving status, ensuring that the SOC charging strategy can meet the passenger carrying mission. Furthermore, the present invention plans the driving routes for unmanned buses within the station based on the distribution of parking and charging spaces within the station, as well as the characteristics of the station roads, and combines this with the SOC charging strategy to improve the efficiency of bus scheduling at the station. Furthermore, the present invention also considers the issues of meeting vehicles, and conducts meeting control based on safety, smoothness, and efficiency gains as the goals, ensuring the safety, smoothness, and efficiency of vehicle traffic.

[0012] Furthermore, the driving route planning process in step 3) is as follows:

[0013] A. Build a high-precision model of the unmanned station based on the road dimensions, parking spaces, and the distribution of charging stations.

[0014] B. Convert the GPS coordinates of buses, roads, parking spaces, and charging stations into two-dimensional plane coordinates;

[0015] C. Determine the starting coordinates of the bus's entire entry process based on the charging strategy, and plan the bus's route within the station using the shortest distance method.

[0016] The present invention establishes a high-precision model of an unmanned station, accurately displays buses, roads, parking spaces, and charging positions in the model, uses a charging strategy to determine the starting coordinates of the entire bus entry process, and determines the driving route according to the short-distance method. The route planning process is simple and easy to implement, and can accurately and efficiently formulate a driving route that meets the requirements.

[0017] Furthermore, the step C is to perform the shortest distance planning using a directed graph.

[0018] The present invention adopts a directed graph method to perform the shortest distance planning, which further improves the planning efficiency.

[0019] Furthermore, in step 4), a linear combination of safety benefit, ride comfort benefit, and efficiency benefit is used as the total driving benefit. A game interaction model of the bus is constructed with the total driving benefit as the goal, and the corresponding meeting strategy is solved.

[0020] Furthermore, the profit function used for the total driving profit is:

[0021]

[0022] W1+W2+W3=1

[0023] Where W1, W2 and W3 represent the weights of safety benefit, smoothness benefit and efficiency benefit respectively, U Lmin represents the normalized security benefit, U S represents the normalized smoothness gain, represents the normalized efficiency gain.

[0024] The present invention takes the linear combination of safety benefit, ride comfort benefit and high efficiency benefit as the goal, and can quickly obtain a meeting strategy that takes safety benefit, ride comfort benefit and high efficiency benefit into account through game theory.

[0025] Furthermore, the safety benefit is expressed as the shortest lateral distance or the shortest longitudinal distance between the two vehicles; the expression for the ride benefit is:

[0026]

[0027] a i represents the optimal acceleration of the unmanned bus in the i-th sampling, a j represents the acceleration taken in the j-th game, is the average value of the optimal riding acceleration that meets certain safety indicators during the historical game process, is the number of sampling times, and n is the number of games;

[0028] The expression of high efficiency benefit is:

[0029]

[0030] T i is the time from the end of the i-1th game to the end of the i-th game.

[0031] Furthermore, the SOC charging strategy in step 2) includes: when the remaining SOC power is greater than or equal to a first set threshold, the unmanned bus does not need to be charged; when the remaining SOC power is greater than or equal to a second set threshold and less than the first set threshold, if the charging position is free, the charging task is executed; if the charging position is not free, the optimal parking position is selected to wait for departure; when the remaining SOC power is less than the second set threshold, if the charging position is free, the charging task is executed; if the charging position is not free, the optimal waiting parking position closest to the charging position is selected.

[0032] Furthermore, the SOC charging strategy in step 2) further includes: when a certain vehicle is performing a charging task, if there is an unmanned bus with a remaining SOC power less than a second set threshold in the station, the vehicle is controlled to charge to the first set threshold and then enter the parking space to wait, so that the unmanned bus with a remaining SOC power less than the second set threshold enters the charging space to charge; if the remaining SOC power of all unmanned buses in the station is greater than the second set threshold, the vehicle is charged to the remaining SOC power of the third set threshold to complete the charging, and the third set threshold is greater than the first set threshold;

[0033] When the remaining SOC power is less than the second set threshold, the unmanned bus will no longer execute the departure command regardless of whether there is a free charging space. Once a free charging space is available, the charging task will be executed immediately. If the power of the next unmanned bus entering the station is less than the second set threshold, it will drive into the current best waiting parking space and its charging order will be after that vehicle.

[0034] Furthermore, the SOC charging strategy in step 2) further includes: when there are vacant charging positions, if the remaining SOC power of all unmanned buses in the station is greater than a second set threshold but there is an unmanned bus with a remaining SOC power less than the first set threshold, then the unmanned buses with a remaining SOC power between the second set threshold and the first set threshold will execute the charging instruction in order of the remaining SOC power from low to high. After the charging positions are full, the remaining unmanned buses will wait for departure or the next charging position will be vacant.

[0035] When formulating the charging strategy, the present invention fully considers the remaining SOC power and driving status of buses inside and outside the unmanned station, as well as the occupancy of charging positions inside the station, to meet the charging needs in various situations.

[0036] The present invention also provides an unmanned station dispatching system, which includes a cloud control platform, which is used to communicate with unmanned buses and roadside equipment, and execute the unmanned station dispatching method of the present invention to dispatch unmanned stations.

[0037] The present invention constructs a cloud control platform that can communicate with unmanned buses and roadside equipment, and performs dispatching and control through the cloud control platform. When formulating the SOC charging strategy, the actual needs of passengers and the remaining SOC power and driving status of the bus are fully considered, so that the formulated SOC charging strategy can meet the passenger-carrying task; at the same time, the present invention plans the driving route within the station for unmanned buses based on the distribution of parking spaces and charging spaces within the station and the characteristics of the station roads, and combines it with the SOC charging strategy, thereby improving the efficiency of bus scheduling at the station. In addition, the present invention also takes into account the problems in the case of meeting, and performs meeting control with safety benefits, smoothness benefits and high efficiency benefits as the goals, ensuring the safety, smoothness and efficiency of vehicle passage. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a flow chart of the unmanned station scheduling method of the present invention;

[0039] Figure 2 This is a schematic diagram of the architecture of the unmanned station dispatching system of the present invention;

[0040] Figure 3 A schematic diagram of the charging strategy formulated in the unmanned station scheduling method of the present invention;

[0041] Figure 4 The figure is a simplified diagram of the meeting conditions that may be encountered during the station dispatching process of the present invention. DETAILED DESCRIPTION

[0042] In order to make the objectives, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0043] Method Example:

[0044] The present invention first counts the historical riding records of bus routes, divides the passenger demand into several levels according to the statistical results and passenger flow, and formulates a departure schedule according to the divided levels; then obtains the remaining SOC power and driving status of the buses inside and outside the unmanned station and the occupancy of the charging positions inside the station, and adjusts the SOC charging strategy according to the departure schedule to ensure that the SOC status of at least two buses inside the station at each moment is sufficient to complete the passenger-carrying task; then obtains the distribution of parking spaces and charging positions inside the station and the characteristics of the station roads, and plans the driving route within the station for the unmanned bus based on the SOC charging strategy; finally, according to the planned driving route within the station, and based on at least one of the safety benefit, smoothness benefit and high efficiency benefit, the unmanned bus is controlled to operate within the station; wherein the safety benefit refers to the overtaking and lane-changing problems during the vehicle operation process, the smoothness benefit refers to the uniform speed problem of the vehicle, and the high efficiency benefit refers to the average speed problem of the vehicle. The implementation process of this method is as follows: Figure 1 As shown, the following is a detailed description.

[0045] The unmanned station scheduling method of the present invention is implemented based on the cloud control platform. Figure 2 As shown, it consists of an unmanned bus, a cloud control center and an unmanned station. The unmanned bus is an intelligent entity with autonomous obstacle avoidance and decision-making capabilities, and can share information with each other, with single-vehicle intelligence and multi-party coordination capabilities; the cloud control platform is responsible for planning the itinerary and issuing instructions. It is also a centralized processing platform for the real-time status of the unmanned bus and parking and charging information within the station. It is an internal information receiving and sending center and a scheduling decision-making platform for the unmanned bus and the station; the unmanned station is the basic scene for unmanned bus parking and charging. It is an adjustment platform, parking place and charging center for unmanned buses, and includes various functional facilities and service facilities. The unmanned bus station dispatching system adopts an unmanned station dispatching method based on passenger needs:

[0046] Step 1: Summarize the passenger demand patterns from bus card swiping records and corresponding times. Through cluster analysis, obtain the passenger flow characteristics at different stations, weather conditions, and time periods. Based on the different passenger flows, passenger demand is divided into five levels: super peak period (A), normal peak period (B), off-peak period (C), normal low peak period (D), and super low peak period (E). Then, adjustments are made to the departure schedule under the existing scheduling plan.

[0047] Specifically, factors that may influence passenger travel are clustered into three categories: time, weather, and region. Time is categorized as morning, noon, and afternoon (nighttime trips by autonomous buses are not considered for now); weather is primarily categorized as sunny, rainy, and snowy; and region is primarily categorized as urban and suburban areas. This classification primarily considers people's travel habits and clustering patterns. Taking a single bus route as an example, a cluster analysis of these factors, using the ratio of the sum of the number of passengers expected to be accommodated and received on a given bus route to the number of seats, is used as the evaluation indicator J for passenger congestion. Passenger traffic meeting these characteristics is then divided into five levels: super-peak (A), normal peak (B), off-peak (C), normal low-peak (D), and super-low-peak (E). For level A (super-peak), additional bus trips or shorter intervals are recommended; for level B (normal peak), shorter intervals are recommended; for levels C (off-peak) and D (normal low-peak), no adjustments are required; and for level E (super-low), intervals can be appropriately extended. Adjust the departure schedule under the existing scheduling plan to reduce the waste of resources caused by unreasonable scheduling, while ensuring that the frequency of bus departures is within the expected range of passengers' waiting time.

[0048] Step 2: The unmanned bus will share its remaining SOC status, speed, location and other information with the cloud control platform in real time. The cloud control platform will provide charging strategies and parking strategies for buses parked in the station and buses entering the station from outside in real time based on the parking information inside the unmanned station (including the occupancy status of parking spaces and charging spaces), and will issue departure instructions to the unmanned bus.

[0049] The charging strategy for unmanned bus station developed by the present invention is as follows: Figure 3 As shown, specifically including:

[0050] To ensure balanced use of charging facilities, when the remaining SOC of an incoming unmanned bus is greater than or equal to 50%, there is no need to execute a charging command, and the charging opportunity is left to other unmanned buses with lower remaining power. When its remaining SOC is greater than or equal to 30% and less than 50%, if there is an available charging position, the charging task is executed. If there is no available charging position, the optimal parking position is selected to wait for departure. When its remaining SOC is less than 30%, charging is required. If there is an available charging position, the charging task is executed. If there is no available charging position, the optimal waiting parking position closest to the charging position is selected.

[0051] While the bus is waiting to charge, if the battery level of the next unmanned bus entering the station is less than 30%, it will drive into the best waiting parking space when it enters the station. The charging order is determined by the remaining SOC level, and the unmanned bus with the lowest remaining battery level will be charged first.

[0052] To avoid wasting charging resources, if there is an unmanned bus with a remaining SOC of less than 30% in the station, the unmanned bus will be charged to 50% of the remaining SOC and then drive into the parking space to wait. The unmanned bus with a remaining SOC of less than 30% will drive into the charging space to charge; if the remaining SOC of all unmanned buses in the station is greater than 30%, the unmanned bus will be charged to 70% of the remaining SOC to complete the charging.

[0053] When there are vacant charging positions, if the remaining SOC power of all unmanned buses in the station is greater than 30% but there are unmanned buses with a remaining SOC power of less than 50%, the unmanned buses with a remaining SOC power between 30% and 50% will execute the charging instructions in order of the remaining SOC power from low to high. After the charging positions are full, the remaining unmanned buses will wait for departure or the next charging position will be vacant.

[0054] In order to ensure that the remaining SOC power of the unmanned bus meets the passenger carrying conditions under the new departure schedule, before the unmanned bus enters the station, the cloud control platform will determine whether it needs to execute a parking or charging command based on the parking information inside the station and the current remaining SOC power of the vehicle.

[0055] Step 3: Plan an optimal driving route for the unmanned bus based on the distribution of parking spaces and charging spaces within the station, the characteristics of the station roads, and the SOC charging strategy. After the unmanned bus enters the corresponding parking space or charging space according to the instructions, the parking information is updated and fed back to the next unmanned bus about to enter the station.

[0056] After determining the target charging or parking space based on the SOC charging strategy, the cloud control platform uses the Dijkstra algorithm to plan the driving trajectory for the unmanned bus. The steps are as follows:

[0057] ① Based on the road size and functional setting distribution of the unmanned station, use PreScan software to establish a high-precision model of the unmanned station.

[0058] ②Convert the GPS coordinates of buses, roads, and station facilities into two-dimensional plane coordinates, and mark each parking space and charging space as a1, a2, ..., a n and b1, b2, ..., b n .

[0059] ③ The starting coordinates of the entire bus entry process are s(x0,y0), p(x n ,y n ) (point p is the coordinate of the center of a charging station or parking space). From the starting point to the parking point, there are a finite number of n points. Let G = (V, E) be a weighted directed graph, where V represents the set of points the bus passes through. This set of points can be divided into two groups: one, denoted by S, represents the set of vertices for which the shortest path has been determined; the other, denoted by U, represents the set of vertices for which the shortest path has not yet been determined.

[0060] ④ Add the vertices in set U to set S in increasing order of the shortest path. The length of the shortest path from the source point v to each vertex in S is no greater than the length of the shortest path from the source point v to any vertex in U.

[0061] ⑤ Initially, S only contains the starting point s; U contains all vertices except S. Set the weighted adjacency matrix of the planar road model to map[, where when there is an edge from vertex i to vertex j, the weight of map[i][j] is less than i and greater than j, otherwise map[i][j] = ∞. Initialize the shortest distance array dist[i] = map[s][i], and dist[u] = 0.

[0062] ⑥ Select the nearest vertex k1 from U, remove k1 from U and add it to S.

[0063] ⑦Update the distance from each vertex in U to the starting point s(x0,y0).

[0064] ⑧Repeat steps ⑥ and ⑦ until all vertices are merged into set S.

[0065] ⑨ The set of path points from point p to the exit can be obtained by logic, and the best of both is the optimal path for the bus to travel inside the unmanned station according to a specific scheduling plan.

[0066] Step 4: Control the operation of the unmanned bus within the station according to the planned route within the station, and based on the goals of safety benefits, smoothness benefits, and efficiency benefits, to ensure safety, smoothness, and efficiency when meeting other vehicles.

[0067] like Figure 4 As shown, typical interaction scenarios include same-direction interaction and opposite-direction interaction. Same-direction interaction can be divided into general overtaking and in-parking overtaking; opposite-direction interaction includes typical lane-changing and out-parking interaction. The dimensions of a bus are closer to a rectangle, with basic defining parameters including the bus's length and width, position coordinates, speed, acceleration, the angle between its direction of travel and the road centerline, and the lateral and longitudinal distances between the two vehicles. While ensuring that the two vehicles do not collide, the smoothness of both vehicles is improved, and the game process is optimized to enable efficient and orderly completion of the interaction task.

[0068] In this embodiment, a linear combination of safety benefit, smoothness benefit, and efficiency benefit is used as the total driving benefit. A bus game interaction model is constructed with the total driving benefit as the goal, and the corresponding meeting strategy is solved.

[0069] The basic elements that constitute the game model:

[0070] Participants:

[0071] P={ε1,ε2…εi …ε n-1 ,ε n}

[0072] Strategy Space:

[0073] Ω=M1×M2…M i …M n-1 ×M n ,

[0074] x1∈M1,x2∈M2…x i ∈M i …x n-1 ∈M n-1 ,x n ∈M n

[0075] Profit function:

[0076]

[0077]

[0078] ......

[0080]

[0081] balanced:

[0082]

[0083] Where ε1,ε2…ε i …ε n-1 ,ε n represents all participants in the game event, Ω represents the strategy space composed of the participants choosing actions in their own way, x1, x2, x3...x n Indicates the possible actions that the participants may take, U i (x1, x2...x i ...x n-1 , x n ) means that in the current behavior combination (x1, x2...x i ...x n-1 , x n ), the payoff function of the i-th participant taking action is,

[0084] In different combinations of participant behaviors (x1, x2...x i ...x n-1 , x n ), the payoff function of the i-th participant is, represents the income of the i-th participant, A* is the optimal strategy combination of all participants in the Nash equilibrium state, It represents the optimal strategy adopted by each participant in the equilibrium state considering the maximization of comprehensive benefits.

[0085] The total benefit is composed of safety benefits, smoothness, and efficiency, as follows:

[0086] 1) Safety Benefits. Considering the characteristics of unmanned terminal dispatching systems, vehicle-to-vehicle interaction is concentrated primarily at parking spaces and charging station exits. This safety benefit is analyzed by combining two typical scenarios. Multi-vehicle interaction scenarios can be viewed as a combination and reconstruction of several two-vehicle interaction scenarios. The following analysis combines the two-vehicle interaction problem to analyze vehicle safety benefits.

[0087] When two vehicles, one going straight and the other turning, encounter each other, this phenomenon often occurs at parking spaces, charging station exits, and T-intersections. Consider the bus as a rectangle with length a and width b. Let the angle between vehicle A and the centerline of the road be α, and the angle between vehicle B and the centerline of the road be β. If the two vehicles encounter each other before vehicle A turns, the shortest lateral distance between them is:

[0088]

[0089] The minimum center distance between the two vehicles should be:

[0090]

[0091] If the two vehicles meet while vehicle A is turning, the shortest lateral distance between the two vehicles is:

[0092]

[0093] When two vehicles are turning in opposite directions, they encounter each other. This phenomenon often occurs at T-junctions. Consider the bus as a rectangle with length a and width b. At this time, we should focus on the speed components of the two vehicles in the direction perpendicular to the center line of the road. If car A is turning and car B is in its radar blind spot, and the distance between the two cars is close to the minimum safe distance, car A should brake urgently and wait until car B passes before restarting. If car A has no blind spot during the turn and the distance is sufficient, the minimum longitudinal distance between the two cars should meet:

[0094]

[0095] Where γ is the deviation angle between vehicle A and the perpendicular line of the road centerline.

[0096] 2) Smoothness benefit. Under the premise of meeting the safety of other vehicles, the bus should be kept as close to a constant speed as possible to avoid the bad riding experience caused by frequent acceleration and deceleration. is the average value of the optimal riding acceleration that meets certain safety indicators in the historical game process. During the meeting process, the unmanned bus undergoes m sampling times and n games. m represents the optimal acceleration of the unmanned bus in the mth sampling, a n represents the acceleration adopted in the nth game. Then the smoothness benefit of the bus meeting process is:

[0097]

[0098] 3) Efficiency benefit. In a certain stage of a game, under the condition of meeting certain safety and smoothness requirements, the average speed of the bus should be increased. Correspondingly, the maximum acceleration during acceleration and deceleration will increase. Let the total distance traveled by car A from the end of the i-1th game to the beginning of the i-th game be The acceleration of car A during the i-th game is The time from the end of the i-1th game to the end of the i-th game is T i The condition for vehicles to enter the interaction area is that the longitudinal distance between the two vehicles is less than X h Or the lateral distance is less than Y h (X h and Y h It is related to the current speed, acceleration, and braking characteristics of the unmanned bus, and a certain distance is reserved). When the distance between two vehicles intending to meet is less than a certain safe distance, the two vehicles begin to compete. When the distance between the two vehicles is greater than the safe distance again, the competition ends.

[0099] Then the expected distance of the game process is:

[0100]

[0101] The entire game process takes:

[0102]

[0103] 4) Total driving benefit. The total benefit function is a linear combination of safety benefit, ride comfort benefit, and efficiency benefit, and the sum of the weighted coefficients of the three is 1, that is, W1+W2+W3=1. in, U S 、U T They represent the security benefits L after parameter normalization. min , ride comfort benefit S, and efficiency benefit T, adopt a unified dimension to facilitate the calculation of total driving benefit. In actual driving analysis, vehicle safety performance is generally prioritized, so the value of W1 is greater than the ride comfort benefit coefficient W2 and the efficiency benefit coefficient W3.

[0104] As another implementation method, one or two of the above may be selected as targets for game control.

[0105] System Implementation:

[0106] The unmanned station dispatching system includes a cloud control platform, which is used to communicate with unmanned buses and roadside equipment and execute the unmanned station dispatching method of the present invention to dispatch unmanned stations. Figure 2 As shown, the cloud control platform is based on 5G network communication and communicates with the on-board unit and roadside RSU unit of each unmanned bus to obtain the unmanned bus's own remaining SOC status, speed, and location information, and performs station scheduling based on the above information. The specific scheduling method has been described in detail in the method embodiment and will not be repeated here.

Claims

1. An unmanned station scheduling method, characterized in that: The steps include: 1) Collect historical bus route records, classify passenger demand into several levels based on passenger volume, and formulate departure schedules based on the levels; 2) Obtain the remaining SOC power and driving status of buses inside and outside the unmanned station, as well as the occupancy of charging stations inside the station. Adjust the SOC charging strategy according to the departure schedule to ensure that at each moment there are at least two buses in the station with SOC levels sufficient to complete the passenger transportation task; 3) Obtain the distribution of parking spaces and charging stations within the station, as well as the characteristics of the station roads, and plan the driving routes within the station for the unmanned bus based on the SOC charging strategy; 4) Based on the planned driving routes within the station, a linear combination of safety benefits, ride comfort benefits, and efficiency benefits is used as the total driving benefit. A bus game interaction model is constructed with the total driving benefit as the goal, and the corresponding meeting strategy is solved to control the operation of the unmanned bus within the station. The safety benefit refers to the overtaking and lane-changing problems during vehicle operation, which is expressed by the shortest lateral distance or the shortest longitudinal distance between two vehicles. The smoothness benefit refers to the uniform speed of the vehicle, which is expressed as: a i represents the optimal acceleration of the unmanned bus in the i-th sampling, a j represents the acceleration taken in the j-th game, is the average value of the optimal riding acceleration that meets certain safety indicators during the historical game process, m is the number of sampling times, and n is the number of games; The efficiency gain refers to the average speed of the vehicle, which is expressed as: T i is the time from the end of the i-1th game to the end of the i-th game.

2. The unmanned station scheduling method according to claim 1, characterized in that: The driving route planning process in step 3) is as follows: A. Build a high-precision model of the unmanned station based on the road dimensions, parking spaces, and the distribution of charging stations. B. Convert the GPS coordinates of buses, roads, parking spaces, and charging stations into two-dimensional plane coordinates; C. Determine the starting coordinates of the bus's entire entry process based on the charging strategy, and plan the bus's route within the station using the shortest distance method.

3. The unmanned station scheduling method according to claim 2, characterized in that: The step C is to perform the shortest distance planning using a directed graph.

4. The unmanned station scheduling method according to claim 1, characterized in that: The profit function used for the total driving profit is: W1+W2+W3=1 Where W1, W2 and W3 represent the weights of safety benefit, smoothness benefit and high efficiency benefit respectively. represents the normalized security benefit, U S represents the normalized smoothness gain, represents the normalized efficiency gain.

5. The unmanned station scheduling method according to claim 1, characterized in that: The SOC charging strategy in step 2) includes: when the remaining SOC power is greater than or equal to a first set threshold, the unmanned bus does not need to be charged; when the remaining SOC power is greater than or equal to a second set threshold and less than the first set threshold, if there is an idle charging position, the charging task is executed; if there is no idle charging position, the optimal parking position is selected to wait for departure; when the remaining SOC power is less than the second set threshold, if there is an idle charging position, the charging task is executed; if there is no idle charging position, the optimal waiting parking position closest to the charging position is selected.

6. The unmanned station scheduling method according to claim 5, characterized in that: The SOC charging strategy in step 2) further includes: when a certain vehicle is performing a charging task, if there is an unmanned bus with a remaining SOC less than a second set threshold in the station, the vehicle is controlled to charge to the first set threshold and then drive into a parking space to wait, so that the unmanned buses with a remaining SOC less than the second set threshold drive into the charging space to charge; if the remaining SOC of all unmanned buses in the station is greater than the second set threshold, the vehicle is charged to a third set threshold to complete the charging, and the third set threshold is greater than the first set threshold; When the remaining SOC power is less than the second set threshold, the unmanned bus will no longer execute the departure command regardless of whether there is a free charging space. Once a free charging space is available, the charging task will be executed immediately. If the power of the next unmanned bus entering the station is less than the second set threshold, it will drive into the current best waiting parking space and its charging order will be after that vehicle.

7. The unmanned station scheduling method according to claim 5, characterized in that: The SOC charging strategy in step 2) further includes: when there are vacant charging positions, if the remaining SOC power of all unmanned buses in the station is greater than a second set threshold but there is an unmanned bus with a remaining SOC power less than the first set threshold, then the unmanned buses with a remaining SOC power between the second set threshold and the first set threshold will execute charging instructions in order of the remaining SOC power from low to high. After the charging positions are full, the remaining unmanned buses will wait for departure or the next vacant charging position.

8. An unmanned station dispatching system, characterized in that: The dispatching system includes a cloud control platform, which is used to communicate with unmanned buses and roadside equipment, and execute the unmanned station dispatching method according to any one of claims 1 to 7 to dispatch the unmanned station.

Images