Multi-vehicle joint scheduling method, device and equipment

Abstract

The application discloses a multi-vehicle joint scheduling method, device and equipment, relates to the technical field of Internet of Vehicles, and comprises the following steps: clustering orders in a first time period based on the starting positions of the orders to obtain an order set; matching vehicles for at least part of the orders in the order set based on the principle that the average waiting time loss of each user in the order set is minimum; and planning a path for each vehicle to carry the orders matched with the vehicle. The scheme solves the problems of low vehicle scheduling efficiency, low scheduling accuracy and long user waiting time in the existing vehicle scheduling scheme.

Description

Technical Field

[0001] This application relates to the field of vehicle networking technology, and in particular to a method, apparatus and equipment for joint scheduling of multiple vehicles. Background Technology

[0002] Smart parks, as one of the application scenarios for 5G+vehicle-to-everything (V2X) technology, are a crucial entry point for smart park construction. With the nation's significant investment in 5G-based smart transportation and smart city development, the prospects for combining smart parks with autonomous driving are vast. Many smart parks operate autonomous vehicles as shuttle buses. Therefore, how to schedule autonomous vehicles has become a key technical issue in this field.

[0003] Currently, in smart park autonomous vehicle dispatching solutions, existing buses are basically based on a fixed route and fixed stop operation mode. This does not allow for the most efficient vehicle dispatching, and passengers can only place orders and wait at fixed stops, resulting in inconvenience and a poor user experience. Existing vehicle dispatching is based on an algorithmic approach for each ride request, which can only meet the reasonable allocation at the current moment and does not comprehensively consider the overall optimal vehicle dispatching over a period of time, resulting in low dispatching accuracy. Summary of the Invention

[0004] The purpose of this application is to provide a multi-vehicle joint scheduling method, apparatus and equipment to solve the problems of low vehicle scheduling efficiency, low scheduling accuracy and long user waiting time in the existing vehicle scheduling scheme.

[0005] To achieve the above objectives, embodiments of this application provide a multi-vehicle joint scheduling method, including:

[0006] Based on the starting position of the orders, the orders within the first time period are clustered to obtain an order set;

[0007] Based on the principle of minimizing the average waiting time loss for each user within the order set, vehicles are matched for at least a portion of the orders in the order set.

[0008] Plan routes for transporting orders that match each of the vehicles.

[0009] Optionally, the method further includes:

[0010] Receive order requests from users in real time;

[0011] The orders are stored in the corresponding area of ​​the order pool according to their order type;

[0012] Orders are periodically extracted from the order pool, wherein the period for extracting orders is the length of the first time period;

[0013] Based on the starting position of the orders, the orders within the first time period are clustered to obtain an order set, including:

[0014] Based on the starting position, the orders extracted from the order pool within the first time period are clustered to obtain the order set.

[0015] Optionally, the step of clustering orders within a first time period based on their starting positions to obtain an order set includes:

[0016] Orders of the same type and whose starting locations are within the same region are clustered to obtain the order set.

[0017] Optionally, matching vehicles for at least a portion of the orders in the order set based on the principle of minimizing the average waiting time loss for each user within the order set includes:

[0018] Candidate vehicles are filtered according to the first condition to obtain a candidate vehicle set corresponding to the order set;

[0019] Based on the principle of minimizing the average waiting time loss for each user within the order set, vehicles are matched for at least a portion of the orders in the candidate vehicle set.

[0020] The first condition includes:

[0021] The candidate vehicle is located within a first search range, which is related to the area range where the starting position of each order in the order set is located;

[0022] The remaining capacity of the candidate vehicles is greater than or equal to the minimum requested capacity of the orders in the order set;

[0023] When the order type is ride-hailing, the candidate vehicle did not receive the order;

[0024] When the order type is shared, the candidate vehicle has either not received the order or has received a shared type order.

[0025] Optionally, the step of matching vehicles for at least a portion of the orders in the candidate vehicle set based on the principle of minimizing the average waiting time loss for each user within the order set includes:

[0026] Construct a first time loss matrix, the elements of which are the time losses between each candidate vehicle and each order;

[0027] Based on the first time loss matrix and the remaining capacity of the candidate vehicles, the KM algorithm is used to match vehicles for at least a portion of the orders in the order set from the candidate vehicle set.

[0028] Optionally, if the order type in the order set is a shared type, the method further includes:

[0029] Adjust the remaining capacity of candidate vehicles for matched orders;

[0030] Traverse the candidate vehicle set, and based on the remaining capacity of the candidate vehicles, the first time loss matrix, and the first time loss of the matched orders, correct the candidate vehicles that match the orders, and adjust the remaining capacity of the corresponding candidate vehicles, wherein the time loss of the updated orders is less than the first time loss.

[0031] Iterate through the candidate vehicle set and match vehicles for at least some of the orders that have not been matched with candidate vehicles, based on the remaining capacity of the candidate vehicles.

[0032] Optionally, the time loss between each candidate vehicle and each of the orders is related to at least one of the following:

[0033] The distance between the candidate vehicle location and the starting location of the order;

[0034] How well the order's origin and destination locations are connected to the planned routes of the candidate vehicles;

[0035] The ratio of the waiting time for an order to the average waiting time for the entire order set.

[0036] Optionally, the method further includes:

[0037] For orders that have not been matched with a vehicle, if the order lifecycle has ended, a dispatch failure notification message will be sent to the user.

[0038] Alternatively, for orders that are not matched with a vehicle, if the order lifecycle has not been reached, the order is stored in the order pool.

[0039] Optionally, planning a route for each of the vehicles to transport orders matching those vehicles includes:

[0040] When the order type is a ride-hailing service, the route is planned for the vehicle based on the starting location of the order;

[0041] Alternatively, if the order type is shared, the shortest path is planned for each vehicle based on a dynamic programming algorithm.

[0042] Optionally, based on a dynamic programming algorithm, the shortest path is planned for each of the vehicles, including:

[0043] A second loss matrix is ​​constructed for each of the vehicles, where the elements of the second loss matrix are the current position of the vehicle and the time loss between any two positions of the start and end positions of the order matched with the vehicle.

[0044] Based on the second loss matrix and the dynamic programming equation, the path with the minimum time loss is determined as the shortest path, wherein the shortest path includes the order in which the vehicle passes through the start and end positions of each order matched with the vehicle.

[0045] Optionally, the dynamic programming equation is:

[0046]

[0047] Where d(i, V) represents the sequence of orders starting from point i and passing through the order set V = {V1, V2, ..., V}. n The path length with minimum loss at each position in}, A j Indicates the starting position of order j, B j V represents the termination position of order j. j Let j be the set of points representing the start and end positions, or the start and end positions of order j. This represents the time loss from point i to the starting position of order j. V represents the time loss from point i to the termination position of order j. j ={A j B j} indicates that the vehicle has not reached the starting position of order j, V j ={B j} indicates that the vehicle has arrived at the starting position of order j. V indicates that the vehicle has reached the end position of order j, and V represents the set of start and end positions of each order that matches the vehicle.

[0048] Optionally, the method further includes:

[0049] Send transport information to the vehicle, the transport information including at least one of the following:

[0050] path;

[0051] Path distance;

[0052] The order in which the vehicle passes through each destination point on the route;

[0053] Orders corresponding to each destination.

[0054] This application also provides a multi-vehicle joint dispatching device, including:

[0055] The acquisition module is used to cluster orders within a first time period based on their starting position to obtain an order set;

[0056] The first processing module is used to match vehicles for at least a portion of the orders in the order set based on the principle of minimizing the average waiting time loss of each user in the order set.

[0057] The second processing module is used to plan routes for each of the vehicles to transport orders that match the vehicles.

[0058] This application embodiment also provides a multi-vehicle joint scheduling device, including: a processor, a memory, and a program stored in the memory and executable on the processor. When the program is executed by the processor, it implements the steps of the multi-vehicle joint scheduling method as described above.

[0059] This application also provides a readable storage medium, characterized in that the readable storage medium stores a program, which, when executed by a processor, implements the steps of the multi-vehicle joint scheduling method described above.

[0060] The above-mentioned technical solution of this application has at least the following beneficial effects:

[0061] The multi-vehicle joint scheduling method of this application firstly clusters orders within a first time period based on their starting positions to obtain an order set; secondly, based on the principle of minimizing the average waiting time loss for each user within the order set, vehicles are matched for at least a portion of the orders in the order set; finally, routes are planned for each vehicle to transport the orders matched with that vehicle. This application's solution addresses the problems of low vehicle scheduling efficiency, low scheduling accuracy, and long user waiting times in existing vehicle scheduling schemes. Attached Figure Description

[0062] Figure 1 This is one of the flowcharts illustrating the multi-vehicle joint scheduling method according to an embodiment of this application;

[0063] Figure 2 This is a second schematic flowchart of the multi-vehicle joint scheduling method according to an embodiment of this application;

[0064] Figure 3 This is a schematic diagram of the structure of the multi-vehicle joint dispatching device according to an embodiment of this application;

[0065] Figure 4 This is a schematic diagram of the structure of a multi-vehicle joint dispatching device according to an embodiment of this application. Detailed Implementation

[0066] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0067] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0068] The multi-vehicle joint scheduling method, apparatus, and equipment provided in this application will be described in detail below with reference to the accompanying drawings and through specific embodiments and application scenarios.

[0069] like Figure 1 The diagram shown is one of the flowcharts of a multi-vehicle joint scheduling method according to an embodiment of this application. The method includes:

[0070] Step 101: Based on the starting position of the orders, cluster the orders within the first time period to obtain an order set;

[0071] In other words, this step involves classifying the orders obtained in the first time period according to their starting positions. Orders located at the same starting position or whose starting positions are within the same area form an order set, which facilitates the subsequent joint scheduling of multiple vehicles.

[0072] It should be noted that an order includes at least one of the following information: order type, order start location, order end location, order request time, and requested transportation capacity. Specifically, the order type can include ride-hailing, car-sharing, etc.

[0073] Step 102: Based on the principle of minimizing the average waiting time loss for each user within the order set, match vehicles for at least a portion of the orders in the order set;

[0074] In other words, when matching vehicles to orders, the average waiting time of users for each order is fully considered. This allows for overall optimization and matching of orders over a period of time, improving vehicle dispatch efficiency.

[0075] Step 103: Plan a route for each vehicle to transport orders that match the vehicle.

[0076] The multi-vehicle joint scheduling method of this application firstly clusters orders within a first time period based on their starting positions to obtain an order set; secondly, based on the principle of minimizing the average waiting time loss for each user within the order set, vehicles are matched for at least a portion of the orders in the order set; finally, routes are planned for each vehicle to transport the orders matched to that vehicle. This application's solution achieves optimal overall scheduling of a batch of orders within a certain period, solving the problems of low vehicle scheduling efficiency, low scheduling accuracy, and long user waiting times in existing vehicle scheduling schemes.

[0077] Furthermore, as an optional implementation, the method also includes:

[0078] Receive order requests from users in real time;

[0079] The orders are stored in the corresponding area of ​​the order pool according to their order type;

[0080] As mentioned earlier, order types can include ride-hailing, car-sharing, etc. In other words, when storing orders in the order pool, they can be categorized and stored in different containers according to their order type.

[0081] Orders are periodically extracted from the order pool, wherein the period for extracting orders is the length of the first time period;

[0082] Based on this, in step 101, orders within the first time period are clustered according to their starting positions to obtain an order set, including:

[0083] Based on the starting position, the orders extracted from the order pool within the first time period are clustered to obtain the order set.

[0084] In other words, the apparatus or equipment that executes the multi-vehicle joint scheduling method of this application embodiment receives user requests in real time. After receiving a user-requested order, it does not process the order immediately, but pushes it into an order pool to find a better allocation time. For example, the execution frequency of the scheduling algorithm can be preset, such as once every 5 seconds. The specific frequency can be set according to the busyness of the business in the park. That is, every 5 seconds, an order is extracted from the order pool, and then the overall scheduling of multiple orders is optimized.

[0085] In this optional implementation, the orders received in real time are first stored in an order pool, and orders are periodically extracted from the order pool for scheduling of multiple vehicles and multiple orders. In this way, delayed scheduling is achieved to seek better order allocation timing and improve the allocation accuracy.

[0086] As an optional implementation, step 101 involves clustering orders within the first time period based on their starting positions to obtain an order set, including:

[0087] Orders of the same type and whose starting locations are within the same region are clustered to obtain the order set.

[0088] In other words, in this optional implementation, the orders pulled from the order pool are classified according to the order type (ride-hailing type, car-sharing type), and then the orders under each category are clustered according to the starting position of the orders to obtain multiple order sets in different regions. The orders in each order set are of the same type and the starting positions of each order are within the same region. This can improve the efficiency of vehicle dispatching.

[0089] As an optional implementation, step 102, based on the principle of minimizing the average waiting time loss for each user within the order set, matches vehicles to at least a portion of the orders in the order set, including:

[0090] Candidate vehicles are filtered according to the first condition to obtain a candidate vehicle set corresponding to the order set;

[0091] Based on the principle of minimizing the average waiting time loss for each user within the order set, vehicles are matched for at least a portion of the orders in the candidate vehicle set.

[0092] The first condition includes:

[0093] The candidate vehicle is located within a first search range, which is related to the area range where the starting position of each order in the order set is located. It should be noted that the first search range is greater than or equal to the area range where the starting position of the order is located, that is, the candidate vehicle is searched within the area range where the starting position of the order is located or its surrounding area.

[0094] The remaining capacity of the candidate vehicle is greater than or equal to the minimum requested capacity of the orders in the order set; that is, the remaining capacity of the candidate vehicle is required to be able to carry the orders with the minimum requested capacity in the order set, i.e., the candidate vehicle can carry at least one order.

[0095] When the order type is ride-hailing, the candidate vehicle did not receive the order; it should be noted that ride-hailing means that a vehicle carrying this type of order can only carry one order.

[0096] When the order type is shared, the candidate vehicle has either not received the order or has received a shared type order.

[0097] It should be noted that when a vehicle meets the first condition mentioned above, the vehicle's current location information, remaining capacity, received orders, and planned routes are obtained, and the vehicle is added to the candidate vehicle set to facilitate multi-vehicle, multi-order joint matching and scheduling of orders within the order set.

[0098] In this optional implementation, firstly, a candidate vehicle set corresponding to the order set is selected based on the first condition, and then vehicles are matched for at least some orders from the candidate vehicle set. In this way, the overall optimization matching of multiple orders and multiple vehicles is realized, and the overall optimal scheduling of a batch of orders within a certain period of time is achieved, thereby improving the efficiency of vehicle scheduling.

[0099] As a specific implementation, in the candidate vehicle set, based on the principle of minimizing the average waiting time loss for each user within the order set, vehicles are matched for at least a portion of the orders in the order set, including:

[0100] (1) Construct a first time loss matrix, wherein the elements of the first time loss matrix are the time loss between each candidate vehicle and each order;

[0101] For example, assuming the total number of candidate vehicles is M and the number of orders in the order set is N, the first-time loss matrix is ​​as follows:

[0102]

[0103] Among them, the element C in the first time loss matrix ij Let be the time loss between vehicle i and order j.

[0104] Specifically, the time loss between each candidate vehicle and each of the aforementioned orders is related to at least one of the following:

[0105] The distance between the candidate vehicle location and the starting location of the order;

[0106] How well the order's origin and destination locations are connected to the planned routes of the candidate vehicles;

[0107] The ratio of the waiting time for an order to the average waiting time for the entire order set.

[0108] For example, the time loss between each candidate vehicle and each order can be calculated using the following formula:

[0109] C ij =f(d ij P j , λ ij (2)

[0110] d ijThis represents the distance from vehicle i to order j, which is the same as the user's waiting time.

[0111] λ ij The λ factor represents the route alignment penalty between the start and end positions of order j and the existing planned route of vehicle i. This penalty occurs when the existing planned route of vehicle i includes the start and end positions of order j (i.e., adding order j does not affect the total travel distance of vehicle i, and the order and the vehicle's planned route are perfectly aligned) or when vehicle i does not yet have a planned route (i.e., vehicle i is empty). ij =1, otherwise, λ ij >1;

[0112] P j This is represented as the ratio of the waiting time for order j to the average waiting time for all orders. This term is introduced to prioritize users who submit order requests earlier, thereby increasing the success rate of order allocation.

[0113] (2) Based on the first time loss matrix and the remaining capacity of the candidate vehicles, the KM algorithm is used to match vehicles for at least a portion of the orders in the order set from the candidate vehicle set.

[0114] It should be noted that this step specifically involves matching orders and candidate vehicles based on the KM algorithm. The matching result is one candidate vehicle matched with one order, meaning there is a one-to-one correspondence between orders and candidate vehicles. As such, there may be orders that do not find a matching vehicle. Therefore, the matching result of this step includes matching pairs of vehicles and orders, and may also include unmatched orders.

[0115] It should also be noted that the matching results in this step should meet the requirement of minimizing the average waiting time loss for each user, that is, satisfying the requirements of the following formula:

[0116]

[0117] The constraints on each of the above elements are as follows:

[0118]

[0119] j, m i ≥n j , where m i n represents the remaining vehicle capacity. j Requesting capacity for an order means that the remaining vehicle capacity is greater than the requested capacity.

[0120] Each order can only be accepted by one vehicle.

[0121] In the ride-hailing mode, each vehicle is only allowed to accept one order.

[0122] In the sharing model, each vehicle can receive multiple orders, but the total capacity of the received orders is less than or equal to the total capacity of the vehicle.

[0123] Furthermore, as an optional implementation, if the order type in the order set is a shared type, the method further includes:

[0124] (1) Adjust the remaining capacity of candidate vehicles for matched orders;

[0125] In other words, after matching orders and vehicles based on the KM algorithm, the remaining capacity of the vehicles in the matching pair is adjusted, that is: the remaining capacity of the vehicles in the matching pair is adjusted to the difference between the remaining capacity before matching and the capacity requested by the order in the matching pair.

[0126] (2) Traverse the candidate vehicle set, and based on the remaining capacity of the candidate vehicles, the first time loss matrix and the first time loss of the matched orders, correct the candidate vehicles that match the orders and adjust the remaining capacity of the corresponding candidate vehicles, wherein the time loss of the updated orders is less than the first time loss.

[0127] In other words, this step involves performing a global search again for each order in the matching pair. Specifically, this includes: obtaining the first time loss corresponding to the matching pair, traversing all candidate vehicles, finding the vehicle with the minimum time loss that meets the capacity requirements, and if the time loss is less than the first time loss, forming a new matching pair between the order and the current vehicle (the traversed vehicle that meets the requirements), and adjusting the remaining capacity of the current vehicle and the original vehicle (adjusting the remaining capacity of the original vehicle to the remaining capacity before matching, and adjusting the remaining capacity of the current vehicle to the difference between the remaining capacity before matching and the time loss of the order in the matching pair).

[0128] (3) Traverse the candidate vehicle set and match vehicles for at least some of the orders that have not been matched with candidate vehicles, based on the remaining capacity of the candidate vehicles.

[0129] In other words, for orders that are not matched, a global search is performed again, which includes finding vehicles that meet the capacity requirements and have the least time loss, forming new matching pairs, and updating the vehicle capacity information.

[0130] It should be noted that after the above steps, the final order and vehicle matching pair is the optimal matching pair (with the minimum average user waiting time loss). In this way, the overall optimization matching of multiple orders and multiple vehicles is achieved, realizing the overall optimal scheduling of a batch of orders within a certain period of time, and improving the efficiency of vehicle scheduling.

[0131] It should also be noted that when matching orders and candidate vehicles, an optimization model is constructed to minimize the average user waiting time by combining vehicle location, remaining vehicle capacity, planned vehicle routes, the start and end points of the order, the number of order requests, and the degree of route compatibility between the orders already received by the vehicle. This model enables joint matching and scheduling of multiple vehicles and multiple orders, increasing the success rate of order matching and achieving the goal of optimizing the overall scheduling of the park over a period of time.

[0132] The specific vehicle dispatching process of this application embodiment will be described below:

[0133] For shared orders, after obtaining the order set and the candidate vehicles corresponding to the order set, the KM algorithm is used to obtain the preliminary optimal (minimum user average waiting time loss) matching results between vehicles and orders: successfully matched pairs and unmatched orders; where the matched pairs are one-to-one between vehicles and orders;

[0134] For each order in the matching pair, perform a global search again: obtain the first time loss of the order in the matching pair; traverse all candidate vehicles, find the vehicle with the lowest time loss and the capacity that meets the requirements; if the time loss is less than the first time loss, form a new matching pair with the current vehicle, and update the capacity of the current vehicle and the original vehicle.

[0135] For orders that are not matched, a global search is performed again, traversing all candidate vehicles to find the vehicle with the required capacity and the least time loss, forming a new matching pair, updating the capacity of existing vehicles, and finally making the matching result meet the requirements of the above formula (3), minimizing the average waiting time loss of each user, so as to achieve the overall optimal scheduling of a batch of orders within a certain time and improve the efficiency of vehicle scheduling.

[0136] It should be emphasized that for special vehicle orders, since each vehicle can only carry one order, the KM algorithm can be used for preliminary optimal matching.

[0137] Furthermore, as an optional implementation, the method also includes:

[0138] For orders that have not been matched with a vehicle, if the order lifecycle has ended, a dispatch failure notification message will be sent to the user.

[0139] Alternatively, for orders that are not matched with a vehicle, if the order lifecycle has not been reached, the order is stored in the order pool.

[0140] In other words, after executing a scheduling algorithm, on the one hand, orders that fail to match can be processed differently based on their lifecycle. If the lifecycle has not been reached (e.g., 3 minutes), the order is pushed into the order pool to await the next round of scheduling. If the lifecycle has been reached, a scheduling failure message is returned to the user to remind them to resend the request. On the other hand, for orders that match successfully, a scheduling success message can be sent to the user to notify them.

[0141] As an optional implementation, step 103, planning a route for each vehicle to transport orders matching the vehicle, includes:

[0142] When the order type is a ride-hailing service, the route is planned for the vehicle based on the starting location of the order;

[0143] Alternatively, if the order type is shared, the shortest path is planned for each vehicle based on a dynamic programming algorithm.

[0144] In other words, when planning routes for vehicles, different planning methods can be used depending on the order type. For example, when the order type is a private car service, a vehicle can only carry one order, so the vehicle's route can be planned as starting from the vehicle's current location, traveling to the order's starting location, and finally traveling to the order's ending location. When the order type is a shared car service, the vehicle will contain multiple order starting and ending locations, which is a multi-starting-point, multi-endpoint route. The planned route should meet the following requirements: each destination must be reached (at least once); some points need to be visited in an orderly manner, that is, for the same order's starting and ending locations, the starting location must be reached first, and then the ending location must be visited. Using a dynamic programming algorithm to solve this problem, the globally optimal solution can be obtained while ensuring the algorithm's performance (time consumption within 20ms).

[0145] During the planning process, the shortest path should be planned based on the average waiting time of users.

[0146] As an optional implementation, based on the dynamic programming algorithm, the shortest path is planned for each of the vehicles, including:

[0147] A second loss matrix is ​​constructed for each of the vehicles, where the elements of the second loss matrix are the current position of the vehicle and the time loss between any two positions of the start and end positions of the order matched with the vehicle.

[0148] The second loss matrix can be as follows:

[0149]

[0150] in:

[0151] This represents the time loss from the vehicle's current position 0 to the starting position of order j;

[0152] This represents the time loss from the vehicle's current position 0 to the termination position of order j;

[0153] This represents the time loss from the starting position of order 1 to the starting position of order 1; since it is the same point, the loss is 0.

[0154] This represents the time loss from the end position of order j to the end position of order j; since it is the same point, the loss is 0.

[0155] This represents the time loss from the starting point of order 1 to the ending point of order 1.

[0156] This represents the time loss from the end position of order j to the start position of order j. Due to the access order restriction between the start and end positions, for a certain order, it is necessary to first pass through the start position of the order and then go to the end position of the order; otherwise, the time loss is defined as ∞.

[0157] Based on the second loss matrix and the dynamic programming equation, the path with the minimum time loss is determined as the shortest path, wherein the shortest path includes the order in which the vehicle passes through the start and end positions of each order matched with the vehicle.

[0158] Specifically, the dynamic programming equation is:

[0159]

[0160] Where d(i, V) represents the sequence of orders starting from point i and passing through the order set V = {V1, V2, ..., V}. n The path length with minimum loss at each position (order start position, order end position) in}, A j Indicates the starting position of order j, B j V represents the termination position of order j. j Let j be the set of points representing the start and end positions, or the start and end positions of order j. This represents the time loss from point i to the starting position of order j. V represents the time loss from point i to the termination position of order j. j ={A j B j} indicates that the vehicle has not reached the starting position of order j, V j ={B j} indicates that the vehicle has arrived at the starting position of order j. V indicates that the vehicle has reached the end position of order j, and V represents the set of start and end positions of each order that matches the vehicle.

[0161] In this optional implementation, based on the time loss between the starting position and / or ending position of the vehicle and each order matched with the vehicle, a dynamic programming algorithm is used to plan the order in which the vehicle passes through the starting position of each order when transporting each order, so as to minimize the average time loss of users of each order, realize the joint matching and scheduling of multiple vehicles and multiple orders, and achieve the overall optimal scheduling of the park within a certain period of time.

[0162] In other words, for vehicles that receive multiple orders at the same time, dynamic programming algorithms are used to perform multi-start and multi-end point path planning to obtain the global optimal solution with the shortest travel distance, thereby improving vehicle operating efficiency. This eliminates the need for vehicles to travel along fixed routes and at fixed stops. Instead, dynamic vehicle scheduling and path planning are performed in real time based on user order requests.

[0163] Furthermore, as an optional implementation, the method also includes:

[0164] Send transport information to the vehicle, the transport information including at least one of the following:

[0165] path;

[0166] Path distance;

[0167] The order in which the vehicle passes through each destination point on the route;

[0168] Orders corresponding to each destination.

[0169] In other words, after determining the route for each vehicle to transport orders, this application embodiment sends route-related information, such as the optimal planned route, the shortest distance of the route, the order in which vehicles arrive at the destination (the start and / or end points of each order), and the order to which each destination belongs, to the vehicle so that the vehicle can pick up and drop off the user's order at the corresponding location according to the dynamically optimal planned route.

[0170] Below, in conjunction with Figure 2 The specific implementation process of the multi-vehicle joint scheduling method in the embodiments of this application will be described in detail below:

[0171] (1) Receive orders requested by users;

[0172] (2) Push the order into (store) the order pool to delay the dispatch of vehicles for the order;

[0173] (3) Retrieve orders from the order pool periodically based on an intelligent scheduling algorithm;

[0174] (4) Query reasonable vehicles or search candidate vehicles, perform joint matching of multiple orders and multiple vehicles, dynamically allocate the best vehicle to this batch of orders, and achieve the overall optimal solution for the current batch of orders;

[0175] (5) Perform multi-task path planning for the matched vehicles to plan the optimal driving path for the vehicles and minimize the driving cost of the vehicles.

[0176] (6) For orders that fail to match, determine whether the order has reached its lifecycle. If it has, send an order failure message. If it has not, push the order back into the order pool.

[0177] (7) Output vehicle path.

[0178] Here, it should also be noted that, as Figure 2 As shown, the method in this application embodiment can be applied to services such as shuttle buses, patrol cars, and cleaning vehicles.

[0179] like Figure 3 As shown in the figure, this application embodiment also provides a multi-vehicle joint dispatching device, including:

[0180] The acquisition module 301 is used to cluster orders within a first time period based on the starting position of the orders to obtain an order set;

[0181] The first processing module 302 is used to match vehicles for at least a portion of the orders in the order set based on the principle of minimizing the average waiting time loss of each user in the order set.

[0182] The second processing module 303 is used to plan routes for each of the vehicles to transport orders that match the vehicles.

[0183] The multi-vehicle joint scheduling device of this application embodiment firstly, the acquisition module 301 clusters orders within a first time period based on the starting position of the orders to obtain an order set; secondly, the first processing module 302 matches vehicles for at least a portion of the orders in the order set based on the principle of minimizing the average waiting time loss of each user within the order set; finally, the second processing module 303 plans routes for each vehicle to transport the orders matched with the vehicle. The solution of this application achieves overall optimal scheduling of a batch of orders within a certain period, solving the problems of low vehicle scheduling efficiency, low scheduling accuracy, and long user waiting times in existing vehicle scheduling schemes.

[0184] Optionally, the device further includes:

[0185] The receiving module is used to receive user-requested orders in real time;

[0186] A storage module is used to store the orders in the corresponding area of ​​the order pool according to the order type;

[0187] An extraction module is used to periodically extract orders from the order pool, wherein the period for extracting orders is the length of the first time period;

[0188] The acquisition module 301 is specifically used for:

[0189] Based on the starting position, the orders extracted from the order pool within the first time period are clustered to obtain the order set.

[0190] Optionally, the acquisition module 301 is specifically used to cluster orders of the same type and whose starting positions are within the same region to obtain the order set.

[0191] Optionally, the first processing module 302 includes:

[0192] The filtering submodule is used to filter candidate vehicles according to the first condition to obtain a candidate vehicle set corresponding to the order set;

[0193] The matching submodule is used to match vehicles for at least a portion of the orders in the order set, based on the principle of minimizing the average waiting time loss of each user in the order set.

[0194] The first condition includes:

[0195] The candidate vehicle is located within a first search range, which is related to the area range where the starting position of each order in the order set is located;

[0196] The remaining capacity of the candidate vehicles is greater than or equal to the minimum requested capacity of the orders in the order set;

[0197] When the order type is ride-hailing, the candidate vehicle did not receive the order;

[0198] When the order type is shared, the candidate vehicle has either not received the order or has received a shared type order.

[0199] Optionally, the matching submodule includes:

[0200] A construction unit is used to construct a first time loss matrix, wherein the elements of the first time loss matrix are the time loss between each candidate vehicle and each order;

[0201] The matching unit is used to match vehicles for at least a portion of the orders in the order set from the candidate vehicle set, based on the first time loss matrix and the remaining capacity of the candidate vehicles, using the KM algorithm.

[0202] Optionally, the matching submodule further includes:

[0203] The adjustment unit is used to adjust the remaining capacity of candidate vehicles that have been matched with orders;

[0204] The correction unit is used to traverse the candidate vehicle set, correct the candidate vehicles that match the order based on the remaining capacity of the candidate vehicles, the first time loss matrix and the first time loss of the matched order, and adjust the remaining capacity of the corresponding candidate vehicles, wherein the time loss of the updated order is less than the first time loss.

[0205] The second matching unit is used to traverse the candidate vehicle set and match vehicles for at least some orders that have not been matched with candidate vehicles, based on the remaining capacity of the candidate vehicles.

[0206] Optionally, the time loss between each candidate vehicle and each of the orders is related to at least one of the following:

[0207] The distance between the candidate vehicle location and the starting location of the order;

[0208] How well the order's origin and destination locations are connected to the planned routes of the candidate vehicles;

[0209] The ratio of the waiting time for an order to the average waiting time for the entire order set.

[0210] Optionally, the device further includes:

[0211] The first sending module is used to send a scheduling failure notification message to the user if an order that has not been matched with a vehicle reaches the end of its life cycle.

[0212] The third processing module is used to store orders that have not been matched with vehicles in the order pool if the order lifecycle has not been reached.

[0213] Optionally, the second processing module is specifically used for:

[0214] When the order type is a ride-hailing service, the route is planned for the vehicle based on the starting location of the order;

[0215] Alternatively, if the order type is shared, the shortest path is planned for each vehicle based on a dynamic programming algorithm.

[0216] Optionally, when the second processing module is used to plan the shortest path for each of the vehicles based on a dynamic programming algorithm, it is specifically used to:

[0217] A second loss matrix is ​​constructed for each of the vehicles, where the elements of the second loss matrix are the current position of the vehicle and the time loss between any two positions of the start and end positions of the order matched with the vehicle.

[0218] Based on the second loss matrix and the dynamic programming equation, the path with the minimum time loss is determined as the shortest path, wherein the shortest path includes the order in which the vehicle passes through the start and end positions of each order matched with the vehicle.

[0219] Optionally, the dynamic programming equation is:

[0220]

[0221] Where d(i, V) represents the sequence of orders starting from point i and passing through the order set V = {V1, V2, ..., V}. n The path length with minimum loss at each position in}, A j Indicates the starting position of order j, B j V represents the termination position of order j. j Let j be the set of points representing the start and end positions, or the start and end positions of order j. This represents the time loss from point i to the starting position of order j. V represents the time loss from point i to the termination position of order j. j ={A j B j} indicates that the vehicle has not reached the starting position of order j, V j ={B j} indicates that the vehicle has arrived at the starting position of order j. V indicates that the vehicle has reached the end position of order j, and V represents the set of start and end positions of each order that matches the vehicle.

[0222] Optionally, the device further includes:

[0223] The second sending module is used to send transportation information to the vehicle, the transportation information including at least one of the following:

[0224] path;

[0225] Path distance;

[0226] The order in which the vehicle passes through each destination point on the route;

[0227] Orders corresponding to each destination.

[0228] like Figure 4As shown, this application embodiment also provides a multi-vehicle joint scheduling device, including: a processor 400, a memory 420 and a program stored in the memory 420 and executable on the processor 400. When the program is executed by the processor, it implements the various processes of the multi-vehicle joint scheduling method embodiment described above and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0229] The transceiver 410 is used to receive and send data under the control of the processor 400.

[0230] Among them, Figure 4 In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits of one or more processors represented by processor 400 and memory represented by memory 420. The bus architecture can also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides an interface. Transceiver 410 can be multiple elements, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium. For different user equipment, user interface 430 can also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc.

[0231] The processor 400 is responsible for managing the bus architecture and general processing, while the memory 420 can store the data used by the processor 400 when performing operations.

[0232] This application also provides a readable storage medium storing a program. When executed by a processor, this program implements the various processes described in the above-described multi-vehicle joint scheduling method embodiment, achieving the same technical effect. To avoid repetition, it will not be described again here. The readable storage medium may be, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0233] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0234] The above description is the preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principles described in this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A multi-vehicle joint scheduling method, characterized in that, include: Based on the starting position of the orders, the orders within the first time period are clustered to obtain an order set; A first time loss matrix is ​​constructed, where each element represents the time loss between each candidate vehicle and each order. Based on the first time loss matrix and the remaining capacity of the candidate vehicles, the KM algorithm is used to match at least a portion of the orders and candidate vehicles in the candidate vehicle set corresponding to the order set. The matching result is one candidate vehicle matched with one order. The time loss between each candidate vehicle and each order is related to at least one of the following: the distance between the candidate vehicle's location and the order's starting location; the degree of route compatibility between the order's starting and ending locations and the candidate vehicle's planned route; and the ratio of the order's waiting time to the average waiting time of the order set. Plan routes for transporting orders that match each of the vehicles.

2. The method according to claim 1, characterized in that, The method further includes: Receive order requests from users in real time; The orders are stored in the corresponding area of ​​the order pool according to their order type; Orders are periodically extracted from the order pool, wherein the period for extracting orders is the length of the first time period; Based on the starting position of the orders, the orders within the first time period are clustered to obtain an order set, including: Based on the starting position, the orders extracted from the order pool within the first time period are clustered to obtain the order set.

3. The method according to claim 1, characterized in that, The process of clustering orders within a first time period based on their starting positions to obtain an order set includes: Orders of the same type and whose starting locations are within the same region are clustered to obtain the order set.

4. The method according to claim 1, characterized in that, The method further includes: Candidate vehicles are filtered according to the first condition to obtain a candidate vehicle set corresponding to the order set; The first condition includes: The candidate vehicle is located within a first search range, which is related to the area range where the starting position of each order in the order set is located; The remaining capacity of the candidate vehicles is greater than or equal to the minimum requested capacity of the orders in the order set; When the order type is ride-hailing, the candidate vehicle did not receive the order; When the order type is shared, the candidate vehicle has either not received the order or has received a shared type order.

5. The method according to claim 1, characterized in that, If the order type in the order set is a shared type, the method further includes: Adjust the remaining capacity of candidate vehicles for matched orders; Traverse the candidate vehicle set, and based on the remaining capacity of the candidate vehicles, the first time loss matrix, and the first time loss of the matched orders, correct the candidate vehicles that match the orders, and adjust the remaining capacity of the corresponding candidate vehicles, wherein the time loss of the updated orders is less than the first time loss. Iterate through the candidate vehicle set and match vehicles for at least some of the orders that have not been matched with candidate vehicles, based on the remaining capacity of the candidate vehicles.

6. The method according to claim 1, characterized in that, The method further includes: For orders that have not been matched with a vehicle, if the order lifecycle has ended, a dispatch failure notification message will be sent to the user. Alternatively, for orders that are not matched with a vehicle, if the order lifecycle has not been reached, the order is stored in the order pool.

7. The method according to claim 1, characterized in that, The step of planning routes for each of the vehicles to transport orders matched with those vehicles includes: When the order type is a ride-hailing service, the route is planned for the vehicle based on the starting location of the order; Alternatively, if the order type is shared, the shortest path is planned for each vehicle based on a dynamic programming algorithm.

8. The method according to claim 7, characterized in that, Based on the dynamic programming algorithm, the shortest path is planned for each of the vehicles, including: A second loss matrix is ​​constructed for each of the vehicles, where the elements of the second loss matrix are the current position of the vehicle and the time loss between any two positions of the start and end positions of the order matched with the vehicle. Based on the second loss matrix and the dynamic programming equation, the path with the minimum time loss is determined as the shortest path, wherein the shortest path includes the order in which the vehicle passes through the start and end positions of each order matched with the vehicle.

9. The method according to claim 8, characterized in that, The dynamic programming equation is: ； in, This indicates starting from point i and passing through the order set. The path length with minimum loss at each position in the middle. This indicates the starting position of order j. Indicates the ending position of order j. Let j be the set of points representing the start and end positions, or the start and end positions of order j. This represents the time loss from point i to the starting position of order j. This represents the time loss from point i to the termination position of order j. This indicates that the vehicle has not reached the starting position of order j. This indicates that the vehicle has arrived at the starting position of order j. V indicates that the vehicle has reached the end position of order j, and V represents the set of start and end positions of each order that matches the vehicle.

10. The method according to claim 1, characterized in that, The method further includes: Send transport information to the vehicle, the transport information including at least one of the following: path; Path distance; The order in which the vehicle passes through each destination point on the route; Orders corresponding to each destination.

11. A multi-vehicle joint dispatching device, characterized in that, include: The acquisition module is used to cluster orders within a first time period based on their starting position to obtain an order set; A first processing module is used to construct a first time loss matrix, the elements of which are the time losses between each candidate vehicle and each order. Based on the first time loss matrix and the remaining capacity of the candidate vehicles, the KM algorithm is used to match at least a portion of the orders and candidate vehicles in the candidate vehicle set corresponding to the order set, with the matching result being one candidate vehicle matched with one order. The time loss between each candidate vehicle and each order is related to at least one of the following: the distance between the candidate vehicle's location and the order's starting location; the degree of route compatibility between the order's starting and ending locations and the candidate vehicle's planned route; and the ratio of the order's waiting time to the average waiting time of the order set. The second processing module is used to plan routes for each of the vehicles to transport orders that match the vehicles.

12. A multi-vehicle joint dispatching device, characterized in that, include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the multi-vehicle joint scheduling method as described in any one of claims 1 to 10.

13. A readable storage medium, characterized in that, The readable storage medium stores a program that, when executed by a processor, implements the steps of the multi-vehicle joint scheduling method as described in any one of claims 1 to 10.

Images