Dynamic vehicle path planning method based on offline calculation, terminal and storage medium

Abstract

The application discloses a kind of dynamic vehicle path planning method based on offline calculation, terminal and storage medium, method includes: obtaining the interval information of road traffic cost in preset area;According to interval information, road traffic cost is randomly sampled, forms M static vehicle path planning scene, and according to the evolution operation of intersection and variation, M static vehicle path planning scene is solved, and the vehicle path planning scheme set of offline is obtained;The scheme set obtained offline is corrected to current vehicle planning scheme, and according to the scheme after correction and the vehicle path planning scheme of random initial iteration, the optimal vehicle path planning scheme found is output;The information of road traffic cost can be optimized based on the vehicle path planning scheme obtained in advance when it changes, so as to make quick response to the change of road traffic information in vehicle service process, and give better vehicle path adjustment scheme in limited time.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle path planning, and in particular to a dynamic vehicle path planning method, terminal and storage medium based on offline calculation. Background Art

[0002] Vehicle Routing Problems (VRPs) are an important type of optimization problem in logistics. In basic VRPs, a group of vehicles needs to serve a group of customers. Each vehicle departs from the same starting point (e.g., a sorting point), serves a certain number of customers, and then returns to the starting point. Each customer is served only once. In real-world scenarios, customers, their demands, and traffic information along each route can change dynamically. Logistics companies need to adapt vehicle routing plans to these dynamic changes, thereby saving costs and improving service quality.

[0003] Dynamic Vehicle Routing Problems (DVRPs) have been proven to be NP-hard problems. Commonly used solutions include heuristic and meta-heuristic methods. Heuristic methods make local adjustments to the current path when information changes. For example, when there is a new customer demand, the new customer is placed on the optimal feasible path using the insertion method, thereby obtaining an effective solution within a certain period of time. Meta-heuristic methods use some population-based optimization algorithms (such as ant colony algorithms, particle swarm algorithms, etc.) to generate multiple paths, and then perform operations such as crossover and mutation on these paths when information changes to obtain new path planning. However, heuristic methods based on local adjustments are prone to falling into local optimal solutions, while meta-heuristic methods based on global optimization have the problems of slow convergence and long time consumption, making them unable to respond in a timely manner.

[0004] With the rapid development of information technology, information processing systems allow for the collection of extensive historical road traffic information and customer demand data. Statistical techniques are then used to process this data, yielding statistical information that provides useful random information for optimization algorithms. Currently, some methods use information about random variations in customer demand to pre-determine a set of solutions, then make local adjustments to these solutions as the problem changes. However, this approach considers a scenario where a fixed set of vehicles repeatedly serves a group of customers and is not applicable to scenarios where information changes during the process of vehicles serving customers. Therefore, existing technologies need improvement. Summary of the Invention

[0005] The technical problem to be solved by the present invention is that, in response to the defects of the existing technology, the present invention provides a dynamic vehicle path planning method, terminal and storage medium based on offline calculation to solve the technical problem of low response efficiency of traditional vehicle path planning methods.

[0006] The technical solutions adopted by the present invention to solve the technical problems are as follows:

[0007] In a first aspect, the present invention provides a dynamic vehicle path planning method based on offline calculation, comprising:

[0008] Obtaining the interval information of road travel costs within a preset area;

[0009] Randomly sampling the road travel cost based on the interval information to form M static vehicle routing scenarios, and solving the M static vehicle routing scenarios based on evolutionary operations of crossover and mutation to obtain an offline vehicle routing solution set; wherein the vehicle routing solution set is a set of multiple chromosomes with the minimum travel cost in each scenario;

[0010] The set of vehicle path planning schemes obtained offline and the current vehicle path planning scheme are revised, and the revised schemes and the random initial vehicle path planning schemes are iterated to output the optimal vehicle path planning scheme found.

[0011] In one implementation, obtaining the interval information of the road travel cost in a preset area includes:

[0012] Obtaining interval information of road travel costs from the vehicle departure station to each customer point within the preset area based on historical information;

[0013] The interval information of the road travel cost between each customer point in the preset area is obtained according to the historical information.

[0014] In one implementation, randomly sampling the road travel cost according to the interval information to form M static vehicle path planning scenarios includes:

[0015] Random sampling is performed based on the interval information of the road travel cost from the vehicle departure station to each customer point and the interval information of the road travel cost between each customer point to form M static vehicle path planning scenarios.

[0016] In one implementation, solving the M static vehicle path planning scenarios according to the evolutionary operations of crossover and mutation to obtain an offline vehicle path planning solution set includes:

[0017] Initialize the service path encoding of the customer points that need service, and obtain a population consisting of P chromosomes. Each chromosome is a random ordering of all customer points, and each chromosome node represents a customer.

[0018] Evaluate each chromosome in the population under each vehicle path planning scenario to obtain an evaluation cost of a single chromosome under a single scenario;

[0019] Using crossover and mutation operations on the P chromosomes in the population to generate P child chromosomes, evaluating each child chromosome in each static vehicle path planning scenario to obtain a travel cost of a single chromosome in each scenario;

[0020] Perform population merging and selection operations on the current population and the generated offspring population;

[0021] The crossover operation, the mutation operation, and the population merging operation are iteratively performed to output multiple chromosomes with the minimum driving distance cost in each scenario, thereby obtaining an offline vehicle path planning solution set.

[0022] In one implementation, merging the current population and the generated offspring population includes:

[0023] Select the chromosome with the lowest travel cost in each scenario, and select the chromosome with the second lowest travel cost in each scenario;

[0024] Delete identical chromosomes until the number of selected chromosomes is P.

[0025] In one implementation, the modifying of the offline acquired vehicle path planning solution set and the current vehicle planning solution includes:

[0026] When planning a vehicle service route online, the vehicle path encoding of the customer points to be served is initialized to obtain a population consisting of N chromosomes. Each chromosome is a random ordering of all customer points, and each chromosome node represents a customer.

[0027] Based on the N initialized chromosomes and the offline vehicle path planning scheme, a cost evaluation is performed under the current scenario, and N chromosomes that meet the conditions are selected;

[0028] According to the mutation operation, local search optimization is performed on N chromosomes that meet the conditions to obtain N optimized chromosomes.

[0029] In one implementation, performing local search optimization on the selected N chromosomes according to the mutation operation to obtain N optimized chromosomes may then include:

[0030] Perform crossover operation on N optimized chromosomes to obtain N daughter chromosomes;

[0031] Select N chromosomes with the lowest travel cost from N parent chromosomes and N child chromosomes;

[0032] Determine whether the N chromosomes with the lowest travel cost meet the vehicle path planning time constraints.

[0033] If the time constraint is met, the optimal chromosome is output as the latest vehicle path planning solution.

[0034] In one implementation, iterating based on the revised solution and the random initial vehicle path planning solution and outputting the found optimal vehicle path planning solution includes:

[0035] Plan the next service customer for the current vehicle according to the latest vehicle route planning plan;

[0036] If the cost of traveling from the starting station or the customer point where the vehicle is currently located to a customer point that the vehicle has not served and has not yet visited changes, and / or the cost of traveling between customer points changes, the current route planning solution and the set of route planning solutions obtained by offline calculations are updated;

[0037] The updated path planning scheme and the random initial vehicle path planning scheme are subjected to local search optimization and cross-operation iterative optimization to output the optimal vehicle path plan.

[0038] In a second aspect, the present invention also provides a terminal comprising: a processor and a memory, wherein the memory stores a dynamic vehicle path planning program based on offline calculation, and when the dynamic vehicle path planning program based on offline calculation is executed by the processor, it is used to implement the operation of the dynamic vehicle path planning method based on offline calculation as described in the first aspect.

[0039] In a third aspect, the present invention also provides a storage medium, which is a computer-readable storage medium, and which stores a dynamic vehicle path planning program based on offline calculation. When the dynamic vehicle path planning program based on offline calculation is executed by a processor, it is used to implement the operation of the dynamic vehicle path planning method based on offline calculation as described in the first aspect.

[0040] The present invention adopts the above technical solution to achieve the following effects:

[0041] The present application can obtain the interval information of road traffic cost, randomly sample the road traffic cost according to the interval information, form M static vehicle path planning scenes, solve the M static vehicle path planning scenes according to the evolution operation of crossover and mutation, obtain the vehicle path pre-planning scheme set under different change conditions, so as to obtain the set of multiple chromosomes with the minimum driving distance cost in each scene; and when the vehicle path is planned online, the vehicle path planning scheme set obtained offline and the current vehicle planning scheme are corrected, the group of vehicle path planning schemes obtained in advance is updated when the relevant road traffic cost changes, and the current path planning scheme and the randomly initialized path planning scheme are combined as the starting point to re-plan a new vehicle path scheme; the present application can quickly respond to the change of road traffic information in the vehicle service process, and give a better vehicle path adjustment scheme in a limited time. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained according to the structures shown in the drawings without creative labor.

[0043] Figure 1 is a flowchart of a dynamic vehicle path planning method based on offline calculation in an implementation manner of the present application.

[0044] Figure 2 is an initial state diagram of a vehicle path planning problem in an implementation manner of the present application.

[0045] Figure 3 is an encoding schematic diagram of a vehicle path planning problem in an implementation manner of the present application.

[0046] Figure 4 is a schematic diagram of four groups of chromosomes that can be obtained by offline calculation in an implementation manner of the present application.

[0047] Figure 5 is a schematic diagram of the current optimal path in an implementation manner of the present application.

[0048] Figure 6 is a schematic diagram of the change of the distance cost of customer points v2 to v4 in an implementation manner of the present application.

[0049] Figure 7 is a schematic diagram of the path after re-planning in an implementation manner of the present application.

[0050] Figure 8It is a functional schematic diagram of a terminal in an implementation manner of the application.

[0051] The implementation, functional features and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0052] To make the objectives, technical solutions and advantages of the application clearer and more explicit, the application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application and not used to limit the application.

[0053] Exemplary method

[0054] For the dynamic vehicle routing problem, the existing vehicle routing schemes include a heuristic method based on local adjustment and a meta-heuristic method based on global optimization, both of which have the problem of low adjustment efficiency. A new method is as follows: for the vehicle routing problem with dynamic changes in customer demand, a set of solutions is obtained in advance using random change information of customer demand, and then local adjustment is performed on the basis of the set of solutions when the problem changes. However, this method considers the scenario in which a set of fixed vehicles repeatedly serve a set of customers, and is not applicable to the problem of how to respond in a timely manner to changes in information during vehicle service.

[0055] To solve the above technical problems, the embodiment of the application provides a dynamic vehicle routing method based on offline calculation. The embodiment of the application can quickly respond to changes in road traffic information during vehicle service, and give a more optimal vehicle routing adjustment scheme within a limited time.

[0056] As shown in Figure 1 The embodiment of the application provides a dynamic vehicle routing method based on offline calculation, which comprises the following steps:

[0057] In step S100, interval information of road traffic costs in a preset area is obtained.

[0058] In this embodiment, the dynamic vehicle routing method based on offline calculation is applied to a terminal, and the terminal includes but is not limited to a computer, a mobile terminal and other devices.

[0059] This example focuses on the dynamic vehicle routing problem of how to respond promptly to changes in road travel costs (e.g., time) when a logistics company dispatches vehicles to serve a group of customers. A common application scenario for this type of problem is when a courier company dispatches vehicles from a starting point to collect packages from various post stations and returns to the starting point. Road travel costs from the starting point to the post stations, as well as between the post stations, can change over time. For example, if a road becomes congested or a road becomes unobstructed, the courier company needs to promptly adjust its vehicle routing plan to minimize the vehicle routing costs.

[0060] To ensure timely response to information changes, the present invention pre-samples possible changes based on the random variation of the cost (e.g., time) spent on each road. A set of vehicle routing solutions is then acquired offline for these sampled scenarios. Based on this set of vehicle routing solutions, the current vehicle routing plan is solved using this set of vehicle routing solutions as part of the initial solution each time information changes, resulting in a timely updated vehicle routing solution.

[0061] Specifically, in one implementation of this embodiment, step S100 includes the following steps:

[0062] Step S101, obtaining interval information of road travel costs from a vehicle departure station to each customer point in the preset area based on historical information;

[0063] Step S102: obtaining interval information of road travel costs between customer points in the preset area based on the historical information.

[0064] In this embodiment, when obtaining interval information on road travel costs, a dynamic vehicle routing problem model can be constructed. In this model, a group of vehicles with limited capacity are responsible for serving a group of customers. The demand for each customer is known in advance. Each vehicle departs from a starting station, serves the corresponding customer, and then returns to the starting station. Each customer can only be served once by a vehicle, and the total load of the vehicles cannot exceed their capacity. A service path must be found for each vehicle to minimize the vehicle routing cost. After a vehicle has served a customer, if the travel cost between the starting station or the customer point where the vehicle is currently located and the customer points that the vehicle has not served and is not currently traveling to changes, as well as between these customer points, the current vehicle routing is adjusted to minimize the total vehicle travel cost for serving the remaining unvisited customers.

[0065] The definition of the dynamic vehicle path planning problem is shown in the following formula. The vehicle path planning problem is defined on an undirected complete graph (such as Figure 2 shown).

[0066] Assume that for any two customers i and j, the cost of the journey from i to j is equal to the cost of the journey from j to i, that is, C i，j =C j，i , u represents the number of customers, k represents the number of vehicles, P i (i=1,2,...,k) represents the service path of the i-th vehicle, including the customer number to be served and the service sequence information, n i (i=1,2,...,k) represents the path P i The number of customers on the road will change, as the road information and the number of unserved customers will change, so we use C i,j (t) and u(t) represent the distance cost from i to j and the number of unserved customers, l(P i )(i=1,2,...,k) represents vehicle i to serve P i The maximum capacity required by all customers on the platform, Q represents the maximum capacity of each vehicle, v 0,i Represents vehicle i service P i The starting point of all vehicles, v0 represents the starting point P i,j Representative path P i The j-th customer on .

[0067]

[0068]

[0069]

[0070]

[0071] l i ≤Q, i=1, 2, ..., k

[0072] This embodiment utilizes information about randomly changing road travel costs from the starting station to each customer point, and also between each customer point, to pre-account for different variations and obtain corresponding vehicle routing solutions. When this information changes, it can be optimized based on the pre-acquired vehicle routing solution, thereby enabling rapid response to changes in road traffic information during vehicle service. It can provide a more optimal vehicle routing adjustment solution within a limited time, thereby improving service quality.

[0073] like Figure 1 As shown, in one implementation of the embodiment of the present invention, the dynamic vehicle path planning method based on offline calculation further includes the following steps:

[0074] In step S200 , the road travel cost is randomly sampled according to the interval information to form M static vehicle path planning scenarios, and the M static vehicle path planning scenarios are solved according to the evolutionary operations of crossover and mutation to obtain an offline vehicle path planning solution set.

[0075] In this embodiment, an offline optimization method is designed to obtain a set of vehicle path planning schemes in advance. In this stage, a random sampling method is used to sample the travel cost of a route with random information to obtain a set of static VPR scenarios (i.e., static vehicle path planning scenarios). Then, a crossover operation and a mutation evolution operation are used to solve the set of static VPR scenarios, thereby obtaining a set of vehicle path planning schemes in advance, i.e., obtaining an offline vehicle path planning scheme set, wherein the vehicle path planning scheme set is a set of multiple chromosomes with the minimum travel distance cost in each scenario.

[0076] Specifically, in one implementation of this embodiment, step S200 includes the following steps:

[0077] Step S201 : Randomly sampling the interval information of the road passage cost from the vehicle departure station to each customer point and the interval information of the road passage cost between each customer point to form M static vehicle path planning scenarios.

[0078] In this embodiment, the intervals of the road travel costs from the vehicle departure station to each customer point and the intervals of the road travel costs between each customer point are first obtained based on historical information. Then, the travel time from the vehicle departure station to each customer point and the travel time between each customer point are randomly sampled from the obtained intervals to form M different vehicle path planning scenarios.

[0079] like Figure 2 As shown ( Figure 2 In the example, v0 represents the starting point and the final return station of all vehicles, and the maximum load of each vehicle is 300; the numbers next to the customer points represent the customer's demand, v1: 125, v2: 125, v3: 200, v4: 50; the numbers on each edge represent the distance cost of the vehicle traveling between the two points). Assuming that the travel cost from customer point v2 to v4 varies in the range [2, 7], if an integer is taken, at most 6 vehicle path planning scenarios can be obtained, with travel costs from v2 to v4 of 2, 3, 4, 5, 6, and 7 respectively. When M <= 6, M of these scenarios can be obtained.

[0080] In this embodiment, after obtaining M scenarios, the following population initialization, population individual evaluation, generation of offspring populations, population selection, and population iteration are performed based on the formed M static vehicle path planning scenarios, and the obtained solution set is used as a pre-processed offline set of vehicle path planning solutions.

[0081] Specifically, in an implementation manner of the embodiment, step S200 further includes the following steps:

[0082] Step S202, initializing the service path coding of the customer points in need of service to obtain a population composed of P chromosomes; wherein each chromosome is a random order of all customer points, and each node of each chromosome represents a customer.

[0083] In the embodiment, in the process of population initialization, natural numbers can be used to code the customers, and the natural numbers are sequentially numbered in natural number order, as shown in the following table: Figure 3 As shown in the table, there are 4 customers, and a sequence of numbers 1-4 is generated, and the sequence is randomly arranged. Such a sequence is called a chromosome. The size of the population is set to P, and P different chromosomes are randomly generated.

[0084] Step S203, evaluating each chromosome in the population under each vehicle path planning scenario to obtain the evaluation cost of a single chromosome under a single scenario.

[0085] In the embodiment, in the process of population individual evaluation, each chromosome in the population is evaluated under each vehicle path planning scenario. For a single chromosome under a single scenario, the Split method is first used to determine the optimal vehicle path scheme, and then the driving distance cost of the path scheme is calculated, that is, the evaluation cost of the chromosome under the scenario.

[0086] For the chromosome as shown in the following table: Figure 3 When the passing costs of v2 to v4 are in the range of [2, 5), the optimal vehicle path planning scheme is to send out two vehicles, one of which serves v1, v2, and v4 in order from v0 and returns to v0, and the other of which serves v3 from v0 and returns to v0. When the passing costs of v2 to v4 are in the range of (5, 7], the optimal vehicle path planning scheme is to send out two vehicles, one of which serves v1 and v2 in order from v0 and returns to v0, and the other of which serves v4 and v3 from v0 and returns to v0. When the passing costs of v2 to v4 are 5, both of the above vehicle path schemes are optimal.

[0087] Step S204, generating P offspring chromosomes using the crossover operation and the mutation operation on the P chromosomes in the population, and evaluating each offspring chromosome under each static vehicle path planning scenario to obtain the driving distance cost of a single chromosome under each scenario.

[0088] In this embodiment, during the process of generating the offspring population, crossover and mutation strategies are used on the chromosomes in the population to generate P offspring chromosomes. The crossover strategy is the Order Crossover strategy, and the mutation strategy is the Exchange strategy. Then, each offspring chromosome is evaluated in each VRP scenario to obtain the vehicle travel cost of the chromosome in each scenario.

[0089] Step S205: performing population merging and selection operations on the current population and the generated offspring population.

[0090] In this embodiment, during the population selection process, the current population and the generated offspring population are merged, and the chromosome with the lowest vehicle travel cost in each scenario is first selected, and then the chromosome with the second lowest vehicle travel cost in each scenario is selected, and so on. The same chromosomes are deleted during the selection process until the number of selected chromosomes is P.

[0091] That is, in one implementation of this embodiment, step S205 includes the following steps:

[0092] Step S205a, selecting the chromosome with the lowest travel distance cost in each scenario, and selecting the chromosome with the second lowest travel distance cost in each scenario;

[0093] Step S205b: Delete identical chromosomes until the number of selected chromosomes is P.

[0094] In this embodiment, there are P chromosomes in the parent population, and the generated offspring population also has P chromosomes. Through population merging and population selection, the P chromosomes that meet the conditions are selected as the population to continue the next round of iteration.

[0095] Step S206 , iteratively executing the crossover operation, the mutation operation, and the population selection operation, outputting multiple chromosomes with the minimum driving distance cost in each scenario, and obtaining an offline vehicle path planning solution set.

[0096] In this embodiment, during the population iteration process, crossover operations, mutation operations, and population selection operations are iteratively performed until the computing resources allowed by offline are exhausted, and multiple chromosomes with the minimum vehicle travel cost in each scenario are output for use in online path planning.

[0097] For Figure 2 For the problem shown in the figure, when the travel cost from customer point v2 to v4 varies in the range [2,7], the chromosome groups that can be obtained are as follows Figure 4 The four situations shown.

[0098] The embodiment pre-processes the random change information of road travel cost from the starting site to each customer site and between each customer site, and can obtain the path with minimum vehicle travel cost in each scene in an offline state, thereby serving as offline data for online path planning.

[0099] As shown in the embodiment of the application, in one implementation mode, the dynamic vehicle path planning method based on offline calculation further comprises the following steps: Figure 1

[0100] Step S300, the vehicle path planning scheme set obtained offline and the current vehicle planning scheme are corrected, and iteration is performed according to the corrected scheme and a randomly initialized vehicle path planning scheme, and the optimal vehicle path planning scheme found is output.

[0101] In the embodiment, the online dynamic vehicle path planning scheme based on offline solution set is designed, and when the vehicle is online planned for the first time or the road traffic cost changes, the offline obtained solution set and the current vehicle planning scheme are corrected, and a randomly initialized vehicle path planning scheme is added as an initial solution, local search and crossover operation are used to perform iteration on the initial solution, and then the optimal vehicle path planning scheme found is output, and the next customer served by each vehicle is determined according to the scheme and then served.

[0102] Specifically, in one implementation mode of the embodiment, step S300 comprises the following steps:

[0103] Step S301, when the vehicle service route is planned online, the vehicle path of the customer to be served is coded and initialized to obtain a population composed of N chromosomes; wherein each chromosome is a random order of all customer sites, and each node of the chromosome represents a customer;

[0104] Step S302, according to the N initialized chromosomes and the offline vehicle path planning scheme, the cost in the current scene is evaluated, and N chromosomes meeting the conditions are selected.

[0105] Step S303, according to the mutation operation, the N chromosomes meeting the conditions are locally searched and optimized to obtain N optimized chromosomes.

[0106] In the embodiment, when the vehicle service route is planned online, the customer points to be served are first coded and initialized for path, the customers are coded using natural numbers, and the natural numbers are sequentially numbered, the sequence is randomly arranged to obtain a chromosome (for example, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10). Figure 3 ​As shown, a total of N chromosomes are initialized. Then, a cost evaluation is performed on the N initialized chromosomes and the offline calculation portion (i.e., the set of offline vehicle routing chromosomes output in step S200) in the current scenario. The Split method is also used to determine the optimal vehicle routing solution. The driving cost of this solution is then calculated, representing the estimated cost of the chromosome in this scenario. The N best-performing chromosomes (i.e., the N chromosomes that meet the lowest cost) are selected. Finally, a local search optimization is performed on these N best-performing chromosomes using the mutation operation Exchange, resulting in the N optimized chromosomes.

[0107] Step S304, performing a crossover operation on the N optimized chromosomes to obtain N offspring chromosomes;

[0108] Step S305 , selecting N chromosomes with the lowest travel cost from the N parent chromosomes and the N child chromosomes;

[0109] Step S306: determine whether the selected N chromosomes with the lowest travel distance cost meet the vehicle path planning time limit.

[0110] Step S307: If the time constraint condition is met, the optimal chromosome is output as the latest vehicle path planning solution.

[0111] In this embodiment, after obtaining N optimized chromosomes, an Order Crossover operation is performed on the N optimized chromosomes to obtain N child chromosomes; and N chromosomes with the lowest travel distance cost are selected from the N parent chromosomes and the N child chromosomes. Thereafter, it is determined whether the selected N chromosomes meet the time constraint of the vehicle routing plan. If the time constraint is met, the optimal chromosome is provided as the latest vehicle routing plan.

[0112] For Figure 2 For the problem shown in the figure, one of the optimal vehicle path planning solutions is as follows Figure 5 As shown, another optimal path is to provide reverse service to each customer point on the path. Figure 5 The path planning solution in [1] can be obtained from the first set of chromosomes obtained from offline calculations, while the reverse service solution can be obtained from any of the 2nd to 4th sets of chromosomes obtained from offline calculations. After obtaining the vehicle path planning solution, the next service customer for the current vehicle is planned according to the latest vehicle path planning solution. If the time limit is not met, the local search and crossover operation iterative optimization are continued on the currently selected N chromosomes until the final path planning solution that meets the time limit is found.

[0113] Step S308: planning the next service customer for the current vehicle according to the latest vehicle route planning solution;

[0114] Step S309: If the cost of traveling from the starting station or the customer point where the vehicle is currently located to a customer point that the vehicle has not yet served and has not yet visited changes, and / or the cost of traveling between customer points changes, the current route planning solution and the set of route planning solutions obtained by offline calculation are updated;

[0115] Step S310 , performing local search optimization and cross-operation iterative optimization on the updated path planning solution and the random initial vehicle path planning solution, and outputting the optimal vehicle path planning solution.

[0116] In this embodiment, after obtaining the latest vehicle route planning solution that meets the time limit, the next service customer is planned for the current vehicle according to the latest vehicle route planning solution. When one vehicle has served a customer, if the travel cost between the starting station or the customer point where the vehicle is located and the customer point that the vehicle has not served and is not heading to, as well as between these customer points, changes (such as Figure 6 As shown, the cost from v2 to v4 becomes 7, and the path needs to be replanned). Then, the current path planning scheme and a set of path planning schemes obtained by offline calculation are updated, and the customer points that have been served in the chromosome and the vehicle is not currently there are deleted (for Figure 6 In the path planning shown, since one vehicle is heading to v1 and the other is staying at v3, there is no need to delete the customer points in the chromosome at this time). At the same time, N chromosomes are initialized according to the customer points to be served next. Each chromosome is a random order of the customer points where the current vehicle is located and the unserved stations.

[0117] For the updated chromosomes and the reinitialized chromosomes, the Split method is used to evaluate the chromosomes and the crossover and mutation operations are used for iterative optimization. Here, the travel cost of the path plan is calculated by considering that some vehicles start from the current customer point instead of the starting point. The following decoding method is used:

[0118] For Figure 6In the path planning shown, one car starts from v1 and one car starts from v3. If the chromosome sequence is (1, 4, 3, 2), it means that the two vehicle paths are v1 and v4 back to v0, and v3 and v2 back to v0 respectively; if the chromosome sequence is (2, 4, 1, 3), the front (2, 4) uses the split method to represent sending one more car from v0 to serve v2 and v4 and then return to v0, and then the car going to v1 returns to v0, and the car going to v3 returns to v3, which is the evaluation cost of the chromosome in this scenario. The N best-performing chromosomes are selected from them, and then the mutation operation Exchange is used to perform local search and crossover operation iterative optimization on these N best-performing chromosomes. When the vehicle path planning time limit is met, the optimal chromosome is output as the updated vehicle planning path.

[0119] like Figure 7 As shown in the figure, the updated optimal vehicle planning path can be obtained directly from the second group of path solutions obtained offline, or by performing one exchange operation on the first and third groups and two exchange operations on the fourth group. It can be seen that the efficiency of online adjustment based on the offline solution set is very high. If all customers are served, the remaining vehicles directly return to the starting station and the vehicle path planning algorithm terminates.

[0120] This embodiment achieves the following technical effects through the above technical solution:

[0121] This embodiment obtains interval information of road travel costs and can randomly sample road travel costs based on the interval information to form M static vehicle path planning scenarios. The M static vehicle path planning scenarios are then solved based on evolutionary operations of crossover and mutation to obtain vehicle path pre-planning schemes under different change conditions, thereby obtaining a set of multiple chromosomes with the minimum travel distance cost under each scenario. Furthermore, during online vehicle path planning, the set of vehicle path planning schemes obtained offline and the current vehicle planning scheme are corrected. When the travel distance cost of the relevant roads changes, the pre-acquired set of vehicle path planning schemes is updated, and on this basis, a new vehicle path plan is replanned using the current path planning scheme and the randomly initialized path planning scheme as starting points. This embodiment can quickly respond to changes in road traffic information during vehicle service and provide a more optimal vehicle path adjustment scheme within a limited time.

[0122] Exemplary devices

[0123] Based on the above embodiments, the present invention also provides a terminal, comprising: a processor, a memory, an interface, a display screen, and a communication module connected via a system bus; wherein the processor is used to provide computing and control capabilities; the memory includes a storage medium and an internal memory; the storage medium stores an operating system and a computer program; the internal memory provides an environment for the operation of the operating system and computer program in the storage medium; the interface is used to connect to external devices, such as mobile terminals and computers; the display screen is used to display corresponding information; and the communication module is used to communicate with a cloud server or a mobile terminal.

[0124] When the computer program is executed by the processor, it is used to implement the operation of a dynamic vehicle path planning method based on offline calculation.

[0125] It will be understood by those skilled in the art that Figure 8 The principle block diagram shown in the figure is only a block diagram of a partial structure related to the solution of the present invention, and does not constitute a limitation on the terminal to which the solution of the present invention is applied. The specific terminal may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0126] In one embodiment, a terminal is provided, comprising: a processor and a memory, wherein the memory stores a dynamic vehicle path planning program based on offline calculation, and when the dynamic vehicle path planning program based on offline calculation is executed by the processor, it is used to implement the operation of the dynamic vehicle path planning method based on offline calculation as described above.

[0127] In one embodiment, a storage medium is provided, wherein the storage medium stores a dynamic vehicle path planning program based on offline calculation, and when the dynamic vehicle path planning program based on offline calculation is executed by the processor, it is used to implement the operation of the dynamic vehicle path planning method based on offline calculation as described above.

[0128] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware through a computer program. The computer program can be stored in a non-volatile storage medium. When executed, the computer program can include the processes in the above-described method embodiments. Any reference to memory, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory.

[0129] In summary, the present invention provides a dynamic vehicle path planning method, terminal and storage medium based on offline calculation, the method comprising: obtaining interval information of road travel costs in a preset area; randomly sampling the road travel costs according to the interval information to form M static vehicle path planning scenarios, and solving the M static vehicle path planning scenarios according to evolutionary operations of crossover and mutation to obtain an offline vehicle path planning scheme set; correcting the offline obtained vehicle path planning scheme set and the current vehicle planning scheme, and iterating according to the corrected scheme and the random initial vehicle path planning scheme to output the found optimal vehicle path planning scheme; when the information of the road travel distance cost changes, the present invention can optimize based on the pre-acquired vehicle path planning scheme, thereby quickly responding to changes in road traffic information during the vehicle service process and providing a more optimal vehicle path adjustment scheme within a limited time.

[0130] It should be understood that the application of the present invention is not limited to the above examples. For those skilled in the art, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A dynamic vehicle path planning method based on offline calculation, characterized in that: include: Obtaining the interval information of road travel costs within a preset area; Randomly sampling the road travel cost based on the interval information to form M static vehicle routing scenarios, and solving the M static vehicle routing scenarios based on evolutionary operations of crossover and mutation to obtain an offline vehicle routing solution set; wherein the vehicle routing solution set is a set of multiple chromosomes with the minimum travel cost in each scenario; The set of offline vehicle routing solutions and the current vehicle routing solution are revised, and the revised solution and the random initial vehicle routing solution are iterated to output the optimal vehicle routing solution found. The modifying of the offline acquired vehicle path planning solution set and the current vehicle planning solution includes: When planning a vehicle service route online, the vehicle path encoding of the customer points to be served is initialized to obtain a population consisting of N chromosomes. Each chromosome is a random ordering of all customer points, and each chromosome node represents a customer. Based on the N initialized chromosomes and the offline vehicle path planning solution set, a cost evaluation is performed in the current scenario, and N chromosomes that meet the conditions are selected; Perform local search optimization on N chromosomes that meet the conditions according to the mutation operation to obtain N optimized chromosomes; The method performs local search optimization on the selected N chromosomes according to the mutation operation to obtain N optimized chromosomes, and then includes: Perform crossover operation on N optimized chromosomes to obtain N daughter chromosomes; Select N chromosomes with the lowest travel cost from N parent chromosomes and N child chromosomes; Determine whether the N chromosomes with the lowest travel cost meet the vehicle path planning time constraints. If the time constraint is met, the optimal chromosome is output as the latest vehicle path planning solution.

2. The dynamic vehicle path planning method based on offline calculation according to claim 1 is characterized in that: The obtaining of interval information of road travel costs within a preset area includes: Obtaining interval information of road travel costs from the vehicle departure station to each customer point within the preset area based on historical information; The interval information of the road travel cost between each customer point in the preset area is obtained according to the historical information.

3. The dynamic vehicle path planning method based on offline calculation according to claim 2 is characterized in that: The randomly sampling the road travel cost according to the interval information to form M static vehicle path planning scenarios includes: Random sampling is performed based on the interval information of the road travel cost from the vehicle departure station to each customer point and the interval information of the road travel cost between each customer point to form M static vehicle path planning scenarios.

4. The dynamic vehicle path planning method based on offline calculation according to claim 3 is characterized in that: The M static vehicle path planning scenarios are solved according to the evolutionary operations of crossover and mutation to obtain an offline vehicle path planning solution set, including: Initialize the service path encoding of the customer points that need service, and obtain a population consisting of P chromosomes. Each chromosome is a random ordering of all customer points, and each chromosome node represents a customer. Evaluate each chromosome in the population under each vehicle path planning scenario to obtain an evaluation cost of a single chromosome under a single scenario; Using crossover and mutation operations on the P chromosomes in the population to generate P child chromosomes, evaluating each child chromosome in each static vehicle path planning scenario to obtain a travel cost of a single chromosome in each scenario; Perform population merging and selection operations on the current population and the generated offspring population; The crossover operation, the mutation operation, and the population merging operation are iteratively performed to output multiple chromosomes with the minimum driving distance cost in each scenario, thereby obtaining an offline vehicle path planning solution set.

5. The dynamic vehicle path planning method based on offline calculation according to claim 4 is characterized in that: The population merging and selection operations on the current population and the generated offspring population include: Select the chromosome with the lowest travel cost in each scenario, and select the chromosome with the second lowest travel cost in each scenario; Delete identical chromosomes until the number of selected chromosomes is P.

6. The dynamic vehicle path planning method based on offline calculation according to claim 1, characterized in that: The iterating according to the revised plan and the random initial vehicle path planning plan and outputting the found optimal vehicle path planning plan includes: Plan the next service customer for the current vehicle according to the latest vehicle route planning plan; If the cost of traveling from the starting station or the customer point where the vehicle is currently located to a customer point that the vehicle has not served and has not yet visited changes, and / or the cost of traveling between customer points changes, the current route planning solution and the set of route planning solutions obtained by offline calculations are updated; The updated path planning scheme and the random initial vehicle path planning scheme are subjected to local search optimization and cross-operation iterative optimization to output the optimal vehicle path plan.

7. A terminal, characterized in that: include: A processor and a memory, wherein the memory stores a dynamic vehicle path planning program based on offline calculation, and when the dynamic vehicle path planning program based on offline calculation is executed by the processor, it is used to implement the operation of the dynamic vehicle path planning method based on offline calculation as described in any one of claims 1 to 6.

8. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, which stores a dynamic vehicle path planning program based on offline calculation. When the dynamic vehicle path planning program based on offline calculation is executed by a processor, it is used to implement the operation of the dynamic vehicle path planning method based on offline calculation as described in any one of claims 1 to 6.

Images