CN109978274A - Dispatch the planing method in path - Google Patents
Dispatch the planing method in path Download PDFInfo
- Publication number
- CN109978274A CN109978274A CN201910262670.3A CN201910262670A CN109978274A CN 109978274 A CN109978274 A CN 109978274A CN 201910262670 A CN201910262670 A CN 201910262670A CN 109978274 A CN109978274 A CN 109978274A
- Authority
- CN
- China
- Prior art keywords
- subregion
- objective function
- path
- client
- solution
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
- G06Q10/0835—Relationships between shipper or supplier and carriers
- G06Q10/08355—Routing methods
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to scheduling routing problem fields, and in particular to a kind of planing method for dispatching path, it is intended to solve the problems, such as to be easily trapped into local optimum when subregion is more.This method comprises: all clients are divided into multiple subregions and then construct a R tree by step S1 according to the geographical location of each client；Step S2 uses initial solution of the greedy algorithm for the block planning vehicle scheduling path as corresponding objective function to each subregion；Step S3 finds out the optimal solution that the subregion corresponds to objective function to each subregion；Step S4, it merges the subregion for possessing identical father node is bottom-up, and root node is punished the optimal solution in area as the final result of vehicle scheduling path planning until the root node of arrival R tree by the optimal solution for asking the subregion to correspond to objective function the subregion after merging.The present invention expands search range using subregion duplication, is merged using subregion from bottom to top, avoids search process and lacks global information, effectively increases the quality of the solution of acquisition.
Description
Technical field
The present invention relates to scheduling routing problem fields, and in particular to a kind of planing method for dispatching path.
Background technique
Vehicle scheduling routing problem, which refers to, gives a certain number of clients, the cargo demand of their each own different numbers,
Warehouse provides cargo transport service using a certain number of vehicles for these clients.In order to solve vehicle scheduling routing problem,
Traffic route appropriate is needed to arrange, meets the needs of all clients in the case where not violating certain constraint, and make mesh
Scalar functions (path total length, cost or time-consuming etc.) are minimum.
Paper " the A spatial parallel heuristic approach for solving delivered in 2017
Very large-scale vehicle routing problems " in, author proposes one for handling extensive vehicle
Dispatch the parallel algorithm of routing problem.The main thought of algorithm is that entire data are carried out cutting, will be right similar in space length
As being placed in a subregion, to realize that by a large-scale vehicle scheduling routing problem cutting be multiple small-scale vehicles
Dispatch routing problem.Spatial object in data is divided into multiple mutually disjoint subregions according to spatial position by algorithm.Fig. 1 is
The subregion schematic diagram of the algorithm.As shown in Figure 1, the dot of black represents customer data, data are divided into 4 subregions, Mei Gexu
Line boxed area represents a subregion, and gray area therein is borderline region.One basic operation unit of algorithm is divided into two
A step, first step handle 4 subregions using 4 threads simultaneously, and second step handles 3 using 3 threads simultaneously
Borderline region.Circulation executes basic operation unit several times until being unable to get better solution.
If, in order to improve degree of parallelism, needing to improve the usage quantity of thread using the algorithm in above-mentioned paper, to need
Increase number of partitions.But when the subregion divided is more and more, the data volume for including in each subregion is also just fewer and fewer, then
The information that per thread obtains is also just fewer and fewer.Since algorithm can not obtain the global information of data, optimization space is very
It is small, easily fall into local optimum.
In view of this, the present invention is specifically proposed.
Summary of the invention
In order to solve the above problem in the prior art, the invention proposes a kind of planing methods for dispatching path, are adding
On the basis of fast planning speed, the quality of understanding is improved.
An aspect of of the present present invention proposes a kind of planing method for dispatching path, the described method comprises the following steps:
Step S1, according to the geographical location of each client and preset each node maximum client's number, by all clients
Multiple subregions are divided into, and then construct a R tree；Wherein, the corresponding subregion of each leaf node of R tree, what leaf node included
It is client, what intermediate node included is the minimum circumscribed rectangle of all clients positioned at the node；
Step S2, to each subregion, according to the geographical location of client each in the subregion and cargo demand and warehouse
Geographical location, use greedy algorithm for the block planning vehicle scheduling path, correspond to the initial of objective function as the subregion
Solution, is accessed once each client, and each car is no more than preset maximum load；
Step S3, to each subregion, according to the subregion correspond to the initial solution of objective function, each client geographical location and
Cargo demand finds out the optimal solution that the subregion corresponds to objective function；
Step S4 is merged the subregion for possessing identical father node is bottom-up, and after merging each time, equal pairing
The optimal solution that subregion after and asks the subregion to correspond to objective function finds out root node punishment area pair until the root node of arrival R tree
The optimal solution for answering objective function, the final result as vehicle scheduling path planning.
Preferably, in step S3 " to each subregion, the initial solution of objective function, each client are corresponded to according to the subregion
Geographical location and cargo demand find out the optimal solution that the subregion corresponds to objective function " it specifically includes:
Step S31, replicates subregion, so that the corresponding one or more subregion examples of each subregion, and it is all described
The total number of subregion example is equal to preset Thread Count；For one thread of each subregion example allocation；
Step S32, according to the geographical location of each client and cargo demand, using distributed computing method to each institute
It states subregion example and corresponds to the initial solution of objective function and optimize, obtain the optimization solution that each subregion example corresponds to objective function；
Step S33 corresponds to the optimization of objective function according to the corresponding each subregion example of the subregion to each subregion
Solution, selects the subregion to correspond to the optimal solution of objective function.
Preferably, in step S4 it " merges the subregion for possessing identical father node is bottom-up, and merges each time
Afterwards, the optimal solution for asking the subregion to correspond to objective function the subregion after merging finds out root section until the root node of arrival R tree
Point punishment area corresponds to the optimal solution of objective function, the final result as vehicle scheduling path planning " it specifically includes:
The number of plies of subregion is denoted as L by step S41, and L=1 is the bottom；
The subregions for possessing identical father node all in L layers are merged upwards and obtain L+1 layers of subregion by step S42, and
The subregions for possessing identical father node all in described L layers are corresponded into the union of the optimal solution of objective function as corresponding L+
1 layer of subregion corresponds to the initial solution of objective function；
Step S43 seeks the optimal solution of corresponding objective function to L+1 layers of each subregion；
Step S44, L=L+1；If the quantity of subregion is greater than 1 in L layers, step S42 is gone to；Otherwise, step is gone to
S45；
Step S45, judge L layers of subregion correspond to the total number of paths for including in the optimal solution of objective function whether be greater than it is default
Vehicle fleet, if so, by being reinserted into other paths comprising the client on the least path of customer quantity, Zhi Daolu
Diameter sum is equal to the preset vehicle fleet.
Preferably, the method also includes:
Step S5, judges whether repartition number is less than preset threshold value, if so, going to step S6；
Step S6, the root node punishment area obtained according to step S4 correspond to the optimal solution of objective function, take this optimal
Corresponding each paths are solved as repartition object；According to the first client in each path in the R tree that step S1 is constructed institute
The leaf node position of category carries out repartition, and the path that the first client belongs to same leaf node is divided into same subregion；It goes to
Step S3.
Preferably, " according to the geographical location of each client and cargo demand, using distributed computing side in step S32
The initial solution that method corresponds to objective function to each subregion example optimizes, and obtains each subregion example and corresponds to objective function
Optimization solution " specifically includes:
Step S321 sets the range of choice of each parameter of TABU search, and randomly chooses respectively in the range of choice of setting
The value of parameter constitutes the parameter set of the TABU search of current partition example；
Step S322 randomly selects a kind of operation in these three operations of cross, insertion and swap, is passed through
All movements that the operation generates, the variation in path where the movement is used to record client；And therefrom select one it is optimal
And the movement in taboo list is not to generate current solution, or selection one is located at the solution in taboo list and generated than current
The optimal solution that subregion example corresponds to objective function more preferably moves to generate current solution, updates taboo list and optimal solution；
Step S323 repeats step S322, when the execution number of step S322 is more than preset TABU search number
When threshold value, scatter searching is executed.
Preferably, the objective function is the letter for calculating path total length in corresponding subregion, totle drilling cost or total time-consuming
Number；
Correspondingly, corresponding vehicle scheduling path when the optimal solution is the target function value minimum.
Preferably, the objective function is the function for calculating path total length in corresponding subregion；
Correspondingly, the step of " executing scatter searching " in step S323 specifically includes:
Step S3231 executes 2-opt operation to each paths, obtains a plurality of different new route；From the new route
One paths of middle selection replace executing the original path before 2-opt, and the wherein lower path of target function value, what is selected is general
Rate is bigger；
Step S3232, to each client in current partition example, calculate the client and its in the paths forerunner away from
From obtaining first distance；The distance for calculating the shortest client with a distance from it in the client and current partition example, obtain second away from
From；Clients all in current partition example are subjected to descending according to the corresponding first distance and the difference of the second distance
Sequence；According to the sequence, successively the client in current partition is reinserted or is exchanged, until currently available solution with
The difference between solution when step S3232 starts meets preset condition；
Step S3233 reinserts the client in current partition example in each path in the path, until
The target function value in the path no longer reduces.
Preferably, preset condition described in step S2332 are as follows:
Wherein, costnewFor the currently available solution, costpreSolution when starting for the step S3232, α are default
Parameter.
The second aspect of the present invention proposes a kind of storage equipment, wherein be stored with a plurality of program, described program be suitable for by
Reason device is loaded and is executed, to realize the planing method in scheduling path recited above.
The third aspect of the present invention proposes a kind of processing equipment, comprising:
Processor is suitable for loading procedure；And
Memory is suitable for storing said program；
Described program is suitable for being loaded and being executed by the processor, to realize the planning side in scheduling path recited above
Method.
Compared with the immediate prior art, the invention has the following beneficial effects:
The planing method in scheduling path proposed by the present invention, use space divides, by large scale vehicle scheduling routing problem
The speed solved the problems, such as can be accelerated with the independent small-scale vehicle scheduling routing problem of parallel processing by being converted to several；It adopts
Merged with subregion from bottom to top in layer, avoids search process and lack global information, overcome and be easily trapped into part most
Excellent problem improves the quality of understanding；It is replicated using subregion, realizes at least one example for each subregion, avoid thread sky
Not busy situation, while search range can be expanded；Multiple bottom-up merging may be implemented using repartition, take full advantage of complete
Office's information, further increases the quality of solution.The present invention can not only well solve large scale vehicle scheduling routing problem, also together
Sample is suitable for small-scale problem.In the case where nowadays hardware resource is no longer major limitation, the present invention program can be sufficiently sharp
Vehicle scheduling routing problem is better solved with these resources.
Detailed description of the invention
Fig. 1 is the schematic diagram for dispatching path planning algorithm in the prior art；
Fig. 2 is the key step schematic diagram of the planing method example one in scheduling path of the invention；
Fig. 3 is the method schematic diagram that subregion duplication is carried out in the embodiment of the present invention one；
Fig. 4 is the bottom-up schematic diagram for carrying out subregion merging in the embodiment of the present invention one；
Fig. 5 is the key step schematic diagram of the planing method example two in scheduling path of the invention.
Specific embodiment
The preferred embodiment of the present invention described with reference to the accompanying drawings.It will be apparent to a skilled person that this
The a little technical principles of embodiment for explaining only the invention, it is not intended that limit the scope of the invention.
It should be noted that in the description of the present invention, term " first ", " second " are intended merely to facilitate description, and
It is not the relative importance of indication or suggestion described device, element or parameter, therefore is not considered as limiting the invention.
The invention solves vehicle scheduling routing problem be described as follows: give a warehouse, warehouse shares M vehicle, often
The maximum load of vehicle is Q, has N clients, each client respectively has cargo demand.Each car from warehouse for client into
Then row delivery service returns to warehouse, it is desirable that all clients are accessed once and at most access once, and at any time
All limited without violating the maximum load of vehicle.
" objective function " mentioned in the embodiment of the present invention refers to for calculating path total length, totle drilling cost in corresponding subregion
Or the function of total time-consuming；" optimal solution " corresponding vehicle scheduling path when being target function value minimum.In the embodiment of the present invention
Distributed algorithm mainly has following technical point: using R tree (R-tree) to all clients progress space partition zone, to subregion into
Row is bottom-up to be merged, is made full use of using partition data duplication CPU, using TABU search to vehicle scheduling routing problem
It is solved.
Fig. 2 is the key step schematic diagram of the planing method example one in scheduling path of the invention.As shown in Fig. 2, this reality
The scheduling paths planning method applied in example includes step S1-S4:
Step S1, according to the geographical location of each client and preset each node maximum client's number, by all clients
Multiple subregions are divided into, and then construct a R tree.Wherein, the corresponding subregion of each leaf node of R tree, what leaf node included
It is client, what intermediate node included is the minimum circumscribed rectangle of all clients positioned at the node.Geographical location falls in same
The client of leaf node has been in the same subregion.
Step S2, to each subregion, according to the geographical location of client each in the subregion and cargo demand and warehouse
Geographical location, use greedy algorithm for the block planning vehicle scheduling path, correspond to the initial of objective function as the subregion
Solution, is accessed once each client, and each car is no more than preset maximum load.
In this step, for each Paralleled build initial solution is position using greedy algorithm for each subregion
A distribution project is generated in all clients of this subregion: a selection one client x1 nearest from warehouse, it is initial for him
Change a paths, a client x2 nearest apart from existing customer is then selected from client remaining in same subregion, attempts
X2 is discharged to x1 to below, if extending obtained new path in this way is feasible (not violate the maximum load limit of vehicle
System), then it realizes and x2 is discharged to behind x1 (i.e. planning is from warehouse to x1 again to the path of x2), otherwise attempt an optionally client and insert
Enter to behind x1.If a feasible path can not be obtained by having attempted all remaining clients all, current path is added
To initial solution, and in addition initialize a new path.When clients all in the subregion have a paths to pass through, initially
Solve construction complete.
Step S3, to each subregion, according to the subregion correspond to the initial solution of objective function, each client geographical location and
Cargo demand finds out the optimal solution that the subregion corresponds to objective function.Specifically, step S3 may include step S31-S33:
Step S31, replicates subregion, so that the corresponding one or more subregion examples of each subregion, and all subregions
The total number of example is equal to preset Thread Count, is then one thread of each subregion example allocation.
In distributed computing, a subregion example can correspond to a thread.If the number of partitions is less than the Thread Count of distribution,
It then needs to replicate a part of subregion, duplication, which is meant, to be enabled a subregion there are multiple examples.In a replication process, first
A thread is distributed for each subregion, it is ensured that each subregion at least has an example, then followed by each subregion
Thread is distributed again, until all threads have been assigned.Ensure that the number of partitions is more than or equal to the thread of distribution by this measure
Number, thus the case where avoiding the occurrence of a large amount of idle threads.
Fig. 3 is the method schematic diagram that subregion duplication is carried out in the embodiment of the present invention one.In the example of fig. 3, the number of partitions 4
(subregion 1, subregion 2 ..., subregion 4), assignable Thread Count be 6, by duplication after, subregion 1 and subregion 2 possess 2 realities respectively
Example, subregion 3 and subregion 4 possess 1 example respectively.
Step S32, according to the geographical location of each client and cargo demand, using distributed computing method to each point
The initial solution that area's example corresponds to objective function optimizes, and obtains the optimization solution that each subregion example corresponds to objective function.The step
Step S321-S323 can be specifically included:
Step S321 sets the range of choice of each parameter of TABU search, and randomly chooses respectively in the range of choice of setting
The value of parameter constitutes the parameter set of the TABU search of current partition example.
Including most important two parameters: maximum number of iterations and Tabu Length, the two parameters, which can be, not to be had
Optimised mistake.
Step S322 is selected at random in these three operations of cross (intersection), insertion (insertion) and swap (exchange)
A kind of operation is taken, all movements generated by the operation are obtained, the variation in path where which is used to record client；And from
Middle selection one is optimal and is not at the movement in taboo list to generate current solution, or selection one is located in taboo list
And the optimal solution of the solution generated objective function more corresponding than current partition example more preferably moves to generate current solution, updates taboo list
And optimal solution.
For example, if select swap, and swap is client a and client b to be exchanged, then caused by this operation it is mobile just
It is (a, b), indicates the path where exchange client a and b.Every kind of operation can generate multiple movements, and each movement is ok
For generating a new solution.
Step S323 repeats step S322, when the execution number of step S322 is more than preset TABU search number
When threshold value, scatter searching (diversification) is executed to break a taboo and search for the situation for being chronically at local optimum.
It is illustrated below to the step of executing scatter searching in step S323:
Diversification mainly includes three phases: being upset, selection client optimizes, local tuning.If I
Require planning after each vehicle the shortest words of path summation, objective function is exactly total for calculating path in corresponding respective partition
The function of length.The step of scatter searching may include step S3231-S3233:
Step S3231, upset stage are primarily used to upset current solution: executing 2-opt operation to each paths, obtain
To a plurality of different new route；It replaces executing the original path before 2-opt from a paths are selected in new route, wherein target
The lower path of functional value, the probability selected are bigger.
Step S3232, the client for selecting some needs optimised reinsert or exchange: in current partition example
Each client calculates the client at a distance from its institute in the paths forerunner, obtains first distance；Calculate the client and current point
In area's example with a distance from it shortest client distance, obtain second distance；By clients all in current partition example according to right
The difference of the first distance and second distance answered carries out descending sort；According to sequence, the client in current partition is carried out again
Insertion or exchange, the difference between solution when currently available solution and step S3232 start meet preset condition.
Step S3233, the local tuning stage as unit of each path, by the visitor in current partition example in each path
Family is reinserted in the path, until the target function value in the path no longer reduces.
In the present embodiment, shown in preset condition such as formula (1):
Wherein, costnewFor currently available solution, costpreSolution when starting for step S3232, α are preset parameter.
Step S33 corresponds to the optimization solution of objective function according to the corresponding each subregion example of the subregion to each subregion,
The subregion is selected to correspond to the optimal solution of objective function.
Due to being previously noted each subregion, there are at least one examples, then executing taboo for the same subregion
After search, multiple and different optimization solutions may be obtained, selects an optimal solution to indicate this in the solution different from these
The arrangement path of subregion.
Step S4 is merged the subregion for possessing identical father node is bottom-up, and after merging each time, equal pairing
The optimal solution that subregion after and asks the subregion to correspond to objective function finds out root node punishment area pair until the root node of arrival R tree
The optimal solution for answering objective function, the final result as vehicle scheduling path planning.Specifically, step S4 may include:
The number of plies of subregion is denoted as L by step S41, and L=1 is the bottom；
The subregions for possessing identical father node all in L layers are merged upwards and obtain L+1 layers of subregion by step S42, and
The subregions for possessing identical father node all in L layers are corresponded into the union of the optimal solution of objective function as corresponding L+1 layers
Subregion corresponds to the initial solution of objective function.
As it can be seen that the initial solution of the new subregion after merging is one of all clients fallen in the spatial dimension of this subregion
Distribution project.
Step S43 seeks the optimal solution of corresponding objective function to L+1 layers of each subregion.
L+1 layers of each subregion is equally mutually indepedent, can independent similar S31-S33 the step of, has executed and has avoided searching
Demand to obtain optimal solution.
Step S44, L=L+1；If the quantity of subregion is greater than 1 in L layers, illustrate the root section for being merged into R-tree not yet
Point then goes to step S42 and continues subregion merging；Otherwise, step S45 is gone to.
Step S45, judge L layers of subregion correspond to the total number of paths for including in the optimal solution of objective function whether be greater than it is default
Vehicle fleet, if so, by being reinserted into other paths comprising the client on the least path of customer quantity, Zhi Daolu
Diameter sum is equal to the preset vehicle fleet.
Fig. 4 is the bottom-up schematic diagram for carrying out subregion merging in the embodiment of the present invention one.As shown in figure 4,12 clients
(i.e. C1, C2 ..., C12) is divided into 5 subregions (i.e. L1_P1, L1_P2 ..., L1_P5) by constructing R-tree.Work as first layer
5 subregions all independently executed TABU search after, the solution of the 1st, 2 and 3 subregion of this layer is stitched together to form
The initial solution of two layers of subregion (i.e. L2_P1), the solution of the 4th and 5 subregion of first layer are stitched together to form the second layer
The initial solution of another subregion (i.e. L2_P2), therefore the number of partitions becomes 2 from 5；Merging is continued up, has arrived third layer just only
There is a subregion (i.e. L3_P1), currently has reached the root node of R-tree, that is to say, that all clients are all located at same
In subregion, so merging next time can not be carried out again.At this moment the optimal solution of the second layer two subregions L2_P1 and L2_P2 are asked
Union is exactly the initial solution of third layer subregion L3_P1.
Fig. 5 is the key step schematic diagram of the planing method example two in scheduling path of the invention.As shown in figure 5, this reality
Applying and dispatching paths planning method in example includes step S1-S6:
Wherein, step S1-S4 and the S1-S4 in embodiment one are corresponding identical, and details are not described herein again.
Step S5, judges whether repartition number is less than preset threshold value, if so, going to step S6；
Step S6, the root node punishment area obtained according to step S4 correspond to the optimal solution of objective function, take the optimal solution pair
Each paths answered are as repartition object；According to the first client in each path in the R tree that step S1 is constructed belonging to
Leaf node position carries out repartition, and the path that the first client belongs to same leaf node is divided into same subregion；Then it goes to
Step S3.
In this step, corresponding path can be indicated with the ID of the first client in each path, if a few
The first client on path had once been divided into identical leaf node in step sl, then divided this several paths in this step
Into the same subregion, correspondingly, this several paths just becomes the initial solution of this new subregion.
Although each step is described in the way of above-mentioned precedence in above-described embodiment, this field
Technical staff is appreciated that the effect in order to realize the present embodiment, executes between different steps not necessarily in such order,
It (parallel) execution simultaneously or can be executed with reverse order, these simple variations all protection scope of the present invention it
It is interior.
Based on the planing method in above-mentioned scheduling path, the present invention also proposes a kind of storage equipment, wherein being stored with a plurality of journey
Sequence, described program are suitable for being loaded and being executed by processor, to realize the planing method in scheduling path recited above.
Further, the present invention also proposes a kind of processing equipment, comprising: processor and memory.Wherein, processor is suitable for
Loading procedure, memory are suitable for storing said program；Described program is suitable for being loaded and being executed by the processor, to realize above
The planing method in the scheduling path.
Those skilled in the art should be able to recognize that, side described in conjunction with the examples disclosed in the embodiments of the present disclosure
Method step, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate electronic hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is executed actually with electronic hardware or software mode, specific application and design constraint depending on technical solution.
Those skilled in the art can use different methods to achieve the described function each specific application, but this reality
Now it should not be considered as beyond the scope of the present invention.
So far, it has been combined preferred embodiment shown in the drawings and describes technical solution of the present invention, still, this field
Technical staff is it is easily understood that protection scope of the present invention is expressly not limited to these specific embodiments.Without departing from this
Under the premise of the principle of invention, those skilled in the art can make equivalent change or replacement to the relevant technologies feature, these
Technical solution after change or replacement will fall within the scope of protection of the present invention.
Claims (10)
1. a kind of planing method for dispatching path, which is characterized in that the described method comprises the following steps:
Step S1 divides all clients according to the geographical location of each client and preset each node maximum client's number
For multiple subregions, and then construct a R tree；Wherein, the corresponding subregion of each leaf node of R tree, that leaf node includes is visitor
Family, what intermediate node included is the minimum circumscribed rectangle of all clients positioned at the node；
Step S2, to each subregion, according to the ground in the geographical location of client each in the subregion and cargo demand and warehouse
Position is managed, uses greedy algorithm for the block planning vehicle scheduling path, the initial solution of objective function is corresponded to as the subregion, is made
Each client is accessed once, and each car is no more than preset maximum load；
Step S3 corresponds to the initial solution of objective function, the geographical location of each client and cargo according to the subregion to each subregion
Demand finds out the optimal solution that the subregion corresponds to objective function；
Step S4 is merged the subregion for possessing identical father node is bottom-up, and after merging each time, to merging after
Subregion ask the subregion to correspond to the optimal solution of objective function, the root node until reaching R tree finds out root node and punishes area and corresponds to mesh
The optimal solution of scalar functions, the final result as vehicle scheduling path planning.
2. the planing method in scheduling path according to claim 1, which is characterized in that " to each subregion, root in step S3
The initial solution of objective function, the geographical location of each client and cargo demand are corresponded to according to the subregion, the subregion is found out and corresponds to mesh
The optimal solution of scalar functions " specifically includes:
Step S31, replicates subregion, so that the corresponding one or more subregion examples of each subregion, and all subregions
The total number of example is equal to preset Thread Count；For one thread of each subregion example allocation；
Step S32, according to the geographical location of each client and cargo demand, using distributed computing method to each described point
The initial solution that area's example corresponds to objective function optimizes, and obtains the optimization solution that each subregion example corresponds to objective function；
Step S33 corresponds to the optimization solution of objective function according to the corresponding each subregion example of the subregion to each subregion,
The subregion is selected to correspond to the optimal solution of objective function.
3. the planing method in scheduling path according to claim 1, which is characterized in that " identical father will be possessed in step S4
The subregion of node is bottom-up to be merged, and after merging each time, asks the subregion to correspond to target the subregion after merging
The optimal solution of function, the root node until reaching R tree find out the optimal solution that root node punishment area corresponds to objective function, as vehicle
Scheduling path planning final result " specifically include:
The number of plies of subregion is denoted as L by step S41, and L=1 is the bottom；
The subregions for possessing identical father node all in L layers are merged upwards and obtain L+1 layers of subregion by step S42, and by institute
State all subregions for possessing identical father node in L layers correspond to objective function optimal solution union as corresponding L+1 layers
Subregion corresponds to the initial solution of objective function；
Step S43 seeks the optimal solution of corresponding objective function to L+1 layers of each subregion；
Step S44, L=L+1；If the quantity of subregion is greater than 1 in L layers, step S42 is gone to；Otherwise, step S45 is gone to；
Step S45 judges that L layers of subregion correspond to whether the total number of paths for including in the optimal solution of objective function is greater than preset vehicle
Sum, if so, by being reinserted into other paths comprising the client on the least path of customer quantity, until path is total
Number is equal to the preset vehicle fleet.
4. the planing method in scheduling path according to claim 1, which is characterized in that the method also includes:
Step S5, judges whether repartition number is less than preset threshold value, if so, going to step S6；
Step S6, the root node punishment area obtained according to step S4 correspond to the optimal solution of objective function, take the optimal solution pair
Each paths answered are as repartition object；According to the first client in each path in the R tree that step S1 is constructed belonging to
Leaf node position carries out repartition, and the path that the first client belongs to same leaf node is divided into same subregion；Go to step
S3。
5. the planing method in scheduling path according to claim 2, which is characterized in that " according to each client in step S32
Geographical location and cargo demand, the initial of objective function is corresponded to each subregion example using distributed computing method
Solution optimizes, and obtains the optimization solution that each subregion example corresponds to objective function " it specifically includes:
Step S321, sets the range of choice of each parameter of TABU search, and each parameter is randomly choosed in the range of choice of setting
Value constitute current partition example TABU search parameter set；
Step S322 randomly selects a kind of operation in these three operations of cross, insertion and swap, obtains and pass through the behaviour
Make all movements generated, the variation in path where the movement is used to record client；And therefrom select one it is optimal and do not have
There is the movement being located in taboo list to generate current solution, or selection one is located at the solution in taboo list and generated and compares current partition
The optimal solution that example corresponds to objective function more preferably moves to generate current solution, updates taboo list and optimal solution；
Step S323 repeats step S322, when the execution number of step S322 is more than preset TABU search frequency threshold value
When, execute scatter searching.
6. the planing method in scheduling path according to any one of claims 1-5, which is characterized in that the objective function
For the function for calculating path total length in corresponding subregion, totle drilling cost or total time-consuming；
Correspondingly, corresponding vehicle scheduling path when the optimal solution is the target function value minimum.
7. the planing method in scheduling path according to claim 5, which is characterized in that
The objective function is the function for calculating path total length in corresponding subregion；
Correspondingly, the step of " executing scatter searching " in step S323 specifically includes:
Step S3231 executes 2-opt operation to each paths, obtains a plurality of different new route；It is selected from the new route
A paths are selected to replace the original path before execution 2-opt, wherein the lower path of target function value, the probability selected are got over
Greatly；
Step S3232 calculates the client at a distance from its institute in the paths forerunner to each client in current partition example,
Obtain first distance；The distance for calculating the shortest client with a distance from it in the client and current partition example, obtains second distance；
Clients all in current partition example are subjected to descending row according to the corresponding first distance and the difference of the second distance
Sequence；According to the sequence, successively the client in current partition is reinserted or exchanged, until currently available Xie Yubu
The difference between solution when rapid S3232 starts meets preset condition；
Step S3233 reinserts the client in current partition example in each path, until the road in the path
The target function value of diameter no longer reduces.
8. the planing method in scheduling path according to claim 7, which is characterized in that preset described in step S2332
Condition are as follows:
Wherein, costnewFor the currently available solution, costpreSolution when starting for the step S3232, α are preset ginseng
Number.
9. a kind of storage equipment, wherein being stored with a plurality of program, which is characterized in that described program is suitable for being loaded and being held by processor
Row, to realize the planing method in scheduling path of any of claims 1-8.
10. a kind of processing equipment, comprising:
Processor is suitable for loading procedure；And
Memory is suitable for storing said program；
It is characterized in that, described program is suitable for being loaded and being executed by the processor, to realize any one of claim 1-8 institute
The planing method in the scheduling path stated.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910262670.3A CN109978274A (en) | 2019-04-02 | 2019-04-02 | Dispatch the planing method in path |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910262670.3A CN109978274A (en) | 2019-04-02 | 2019-04-02 | Dispatch the planing method in path |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109978274A true CN109978274A (en) | 2019-07-05 |
Family
ID=67082474
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910262670.3A Pending CN109978274A (en) | 2019-04-02 | 2019-04-02 | Dispatch the planing method in path |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109978274A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110361017A (en) * | 2019-07-19 | 2019-10-22 | 西南科技大学 | A kind of full traverse path planing method of sweeping robot based on Grid Method |
| CN111199369A (en) * | 2019-12-18 | 2020-05-26 | 南方科技大学 | Addressing route-finding planning method, device, equipment and computer readable storage medium |
| CN112395288A (en) * | 2020-09-25 | 2021-02-23 | 浙江大学 | R-tree index merging and updating method, device and medium based on Hilbert curve |
| CN112541627A (en) * | 2020-12-10 | 2021-03-23 | 赛可智能科技(上海)有限公司 | Method, device and equipment for planning path and optimizing performance of electric logistics vehicle |
| CN113298284A (en) * | 2020-02-05 | 2021-08-24 | 富士通株式会社 | Information processing apparatus, recording medium, information processing method, and information processing system |
| CN116629586A (en) * | 2023-07-24 | 2023-08-22 | 青岛民航凯亚系统集成有限公司 | Airport guarantee vehicle scheduling method and system based on ALNS |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101944095A (en) * | 2009-07-08 | 2011-01-12 | 广东融讯信息科技有限公司 | Path planning method and system |
| CN106156897A (en) * | 2016-08-22 | 2016-11-23 | 武汉轻工大学 | Optimum path planning analog systems in logistics distribution |
| CN108876024A (en) * | 2018-06-04 | 2018-11-23 | 清华大学深圳研究生院 | Path planning, path real-time optimization method and device, storage medium |
| CN109034481A (en) * | 2018-07-31 | 2018-12-18 | 北京航空航天大学 | A kind of vehicle routing problem with time windows modeling and optimization method based on constraint planning |
-
2019
- 2019-04-02 CN CN201910262670.3A patent/CN109978274A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101944095A (en) * | 2009-07-08 | 2011-01-12 | 广东融讯信息科技有限公司 | Path planning method and system |
| CN106156897A (en) * | 2016-08-22 | 2016-11-23 | 武汉轻工大学 | Optimum path planning analog systems in logistics distribution |
| CN108876024A (en) * | 2018-06-04 | 2018-11-23 | 清华大学深圳研究生院 | Path planning, path real-time optimization method and device, storage medium |
| CN109034481A (en) * | 2018-07-31 | 2018-12-18 | 北京航空航天大学 | A kind of vehicle routing problem with time windows modeling and optimization method based on constraint planning |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110361017A (en) * | 2019-07-19 | 2019-10-22 | 西南科技大学 | A kind of full traverse path planing method of sweeping robot based on Grid Method |
| CN110361017B (en) * | 2019-07-19 | 2022-02-11 | 西南科技大学 | Grid method based full-traversal path planning method for sweeping robot |
| CN111199369A (en) * | 2019-12-18 | 2020-05-26 | 南方科技大学 | Addressing route-finding planning method, device, equipment and computer readable storage medium |
| CN111199369B (en) * | 2019-12-18 | 2023-12-12 | 南方科技大学 | Addressing routing method, apparatus, device and computer readable storage medium |
| CN113298284A (en) * | 2020-02-05 | 2021-08-24 | 富士通株式会社 | Information processing apparatus, recording medium, information processing method, and information processing system |
| CN112395288A (en) * | 2020-09-25 | 2021-02-23 | 浙江大学 | R-tree index merging and updating method, device and medium based on Hilbert curve |
| CN112541627A (en) * | 2020-12-10 | 2021-03-23 | 赛可智能科技(上海)有限公司 | Method, device and equipment for planning path and optimizing performance of electric logistics vehicle |
| CN112541627B (en) * | 2020-12-10 | 2023-08-01 | 赛可智能科技(上海)有限公司 | Method, device and equipment for planning path and optimizing performance of electric logistics vehicle |
| CN116629586A (en) * | 2023-07-24 | 2023-08-22 | 青岛民航凯亚系统集成有限公司 | Airport guarantee vehicle scheduling method and system based on ALNS |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109978274A (en) | Dispatch the planing method in path | |
| Morabit et al. | Machine-learning–based column selection for column generation | |
| Defryn et al. | A fast two-level variable neighborhood search for the clustered vehicle routing problem | |
| Mester et al. | Active guided evolution strategies for large-scale vehicle routing problems with time windows | |
| Subramanian et al. | A hybrid algorithm for a class of vehicle routing problems | |
| Gendreau et al. | Vehicle routing: modern heuristics | |
| CN109948855A (en) | A kind of isomery harmful influence Transport route planning method with time window | |
| Li et al. | A multistart adaptive memory-based tabu search algorithm for the heterogeneous fixed fleet open vehicle routing problem | |
| CN103686865B (en) | The decision making device and method of Internet usage | |
| JP6243045B2 (en) | Graph data query method and apparatus | |
| Baldi et al. | Branch-and-price and beam search algorithms for the variable cost and size bin packing problem with optional items | |
| Hu et al. | A storage allocation algorithm for outbound containers based on the outer–inner cellular automaton | |
| CN103246941A (en) | Scheduling method for export container wharf pile-up space | |
| De Armas et al. | A hybrid GRASP-VNS for ship routing and scheduling problem with discretized time windows | |
| CN108921468A (en) | A kind of logistic distribution vehicle intelligence wire arranging method | |
| Crainic | Parallel solution methods for vehicle routing problems | |
| Zhang et al. | A scatter search for multi-depot vehicle routing problem with weight-related cost | |
| CN113098714A (en) | Low-delay network slicing method based on deep reinforcement learning | |
| CN101650805B (en) | Search method for path planning | |
| CN110098964A (en) | A kind of disposition optimization method based on ant group algorithm | |
| Gholami et al. | Solving parallel machines job-shop scheduling problems by an adaptive algorithm | |
| Dasari et al. | Multi-start heuristics for the profitable tour problem | |
| Alvelos et al. | Combining column generation and metaheuristics | |
| Chaovalitwongse et al. | GRASP with a new local search scheme for vehicle routing problems with time windows | |
| Gacias et al. | Scheduler-oriented algorithms to improve human–machine cooperation in transportation scheduling support systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20230811 |
|
| AD01 | Patent right deemed abandoned |