CN107194513A - A kind of optimization method for solving full channel logistics distribution - Google Patents
A kind of optimization method for solving full channel logistics distribution Download PDFInfo
- Publication number
- CN107194513A CN107194513A CN201710382144.1A CN201710382144A CN107194513A CN 107194513 A CN107194513 A CN 107194513A CN 201710382144 A CN201710382144 A CN 201710382144A CN 107194513 A CN107194513 A CN 107194513A
- Authority
- CN
- China
- Prior art keywords
- mrow
- shops
- msub
- munder
- route
- 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.)
- Granted
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 invention discloses a kind of optimization method for solving full channel logistics distribution, the first stage solves LAP problems using Lagrangian relaxation technology.Second stage solves Multi-types vehicle routine problem using adaptive extensive neighborhood search, by one group it is simple destroy with algorithm for reconstructing can in solution space extensive search feasible solution, the situation for being absorbed in local optimum can be prevented effectively from.Simultaneously, the simulated annealing acceptance criteria realized in adaptive extensive neighborhood search can guarantee that the quality and convergence of solution, algorithm output result after the iterations specified is performed, can meet time-constrain requirement, preferable vehicle delivery scheme is tried to achieve for enterprise in limiting time.With reference to the simplicity and the validity of adaptive extensive neighborhood search of Lagrangian relaxation technology, hybrid algorithm integrated solution efficiency high can effectively solve full channel logistics distribution.
Description
Technical field
The present invention relates to the Location-Routing Problem of computer science, more particularly to a kind of full channel logistics of solution is matched somebody with somebody
Send the optimization method of problem.
Background technology
With the popularization of internet, the clothing, food, lodging and transportion -- basic necessities of life of people increasingly be unable to do without network.Historical background in " internet+"
Under, numerous traditional forms of enterprises seek transition one after another, and to realize industrial upgrading, enterprise preferably develops.Utilize internet platform, enterprise
Online and offline sale can be carried out, Sales Channel has been widened, has added the source of customers.But, melt on line with resource under line
During conjunction, enterprise is encountered by new challenge.By taking retail business as an example, current enterprise is met under line by the dis-tribution model of self-operation
The order demand of each shops, carries out item dispenser, it is impossible to accomplish by third-party logistics company to the customer that commodity are bought on line
The unified dispatching of goods is carried out to the client of online and offline.Solid shop/brick and mortar store is sold under line, such as convenience store, supermarket, with goods
Demand is big, the features such as a small range aggregation is a wide range of interior scattered on geographical position, and vehicle easily occurs in the dis-tribution model of self-operation
Charging ratio is low, the phenomenon such as no-load ratio height.On-line selling has that the source of customers is wide, scale is big, geographical position is scattered, goods demand
Small the features such as, dispensed by third-party logistics, the resource of enterprise itself can not be effectively utilized.Which results in enterprise
Logistics cost is higher, and profitability declines, so as to reduce the market competitiveness of enterprise, constrains the development of enterprise.
Full channel logistics distribution refers to uniformly carries out goods delivery to the Sales Channel of online and offline.Intend matching somebody with somebody using two-stage
Pattern is sent, as shown in Figure 1.Enterprise sets up shops in the region that client assembles：Supermarket, convenience store pick up by oneself cabinet, and client can be with
To closest shops voluntarily picking, shops is in addition to merchandising function, the also function with storage express mail.Two-stage
The flow of dispatching is as follows:The first order, article is uniformly dispensed into each shops by enterprise from the total storehouse of logistics, on the one hand meets door
The daily order demand in shop, on the other hand meets the electric business client cargo demand that the shops is serviced.The second level, courier from
Shops's picking is dispensed to region client, and client can also be selected voluntarily to shops's picking.For enterprise, full channel logistics
Shops's quantity will increase under dis-tribution model, and the daily demand of shops will rapidly increase, how using one group of vehicle to extensive door
Shop and client carry out goods delivery, are urgent problems to be solved.
Full channel logistics distribution substantially needs to solve a Location-Routing Problem (Location-routing
Problem, LRP), it is a NP-hard problem, in the case of given Customer Location and dispatching terminal possible position, really
Determine the transit route of quantity, position and the distribution vehicle of terminal, realize that totle drilling cost minimizes target.LRP be integrated with addressing-point
Send problem (Location-allocation Problem, LAP) and Vehicle Routing Problems (Vehicle-routing
Problem, VRP), therefore, totle drilling cost by solving the problems, such as that LAP expense and the expense of VRP problems are constituted, feasible scheme also by
The solution of LAP problems and the solution of VRP problems are constituted.Under full channel logistics distribution model, terminal is shops, not only with storage
Express mail function, can also produce order demand.Fleet of enterprise possesses different vehicles, therefore, and the VRP problems being related to are specially one
Multi-types vehicle routine problem.Common solution LRP algorithms are broadly divided into two classes：Exact algorithm and heuritic approach.Exact algorithm
Including：Branch and bound method, integer programming method, Nonlinear Programming Method.Heuritic approach is mainly a combined method, by simple structure
Algorithm and intelligent algorithm combine composition.Simple structure algorithm includes：Savings method, insertion and scanning method.Intelligent algorithm includes：Prohibit
Avoid searching algorithm, simulated annealing, ant group algorithm and genetic algorithm.Exact algorithm can try to achieve the optimal solution of problem, but calculate
Time is long, is only suitable for solving small-scale problem.In existing a combined method, dispatching terminal only has storage express mail work(
Can, the vehicle from the total storehouse of logistics to dispatching terminal is same vehicle, it is impossible to meet enterprise's multi-vehicle-type dispatching, to on-line off-line
The demand of the unified dispatching of channel, it is difficult to applied in actual scene.
The content of the invention
Because the algorithm that current logistics distribution is used is difficult to meet actual demand, the present invention provides a kind of based on Lagrange
Loose adaptive large neighborhood search algorithm solves full channel logistics distribution, under the conditions of a variety of realistic constraint conditions, energy
In limiting time effective distribution project is provided for enterprise.
In order to realize above-mentioned technical purpose, the technical scheme is that,
A kind of optimization method for solving full channel logistics distribution, comprises the following steps：
Step 1, the use of Lagrangian Relaxation is each client distribution one according to Customer Location and the position of alternative shops
Individual shops is serviced, i.e., concentrated from alternative shops as connection expense apart from summation using all clients to service shops and determine one
The position of group shops so that the demand of all clients is all satisfied and sets up expense with being connected expense sum at least, so gives
The shops of each client's specified services and the position for determining shops；
Step 2, according to the order demand of shops and the customer demand serviced, calculated using adaptive extensive neighborhood search
Method determines that one group is used to dispense goods to the vehicle of each shops and the travel route of each car：Built first by greedy insertion
The initial distribution project distributed including shops and vehicle route, is then moved using the destruction for removing the shops in initial distribution project
Division forms feasible dispatching to be destroyed, then by the way that the shops of removal is inserted into the reconstruction method in distribution route again
Scheme, vehicle fleet size needed for making vehicle delivery scheme is minimum, and driving path is most short；
Step 3, shops is built to the path of its services client, realizes that shops is dispensed into client's using greedy insertion
The optimization of distribution route.
In a kind of optimization method of the full channel logistics distribution of described solution, described step 1, to each client
The shops of specified services and determine mathematical modeling expression formula that the position of shops used for：
Wherein I gathers for alternative shops, and I={ 1,2,3 ..., m }, J is client set, J={ 1,2,3 ..., n }, OiFor
Alternative shops i's sets up expense, WiFor alternative shops i storing capacity, djFor client j demand,
In a kind of optimization method of the full channel logistics distribution of described solution, described step 1, Lagrange is used
Relaxing techniques solve mathematical modeling the step of be：
Step 1) introduce non-negative Lagrange multiplier λiLoose constraint condition (2) is obtained：
Step 2) object function that adds (5) after must being relaxed in object function (1)：
Wherein constraints is formula (3) and formula (4)；
Step 3) for each shops j, f in (1) is tried to achieve using dynamic programmingLAPLower bound, by solve (6) in
Dual problem：
Try to achieve most suitable λiTo obtain problem fLAPOptimal solution；
Step 4) using most suitable λ in Subgradient optimization Algorithm for Solving dual problem (7)i, initialization vector λ0
For 0, in t times and t+1 iteration, vectorial λtCalculation is as follows：
Wherein xij(λt) it is object function LR (λj) in λtUnder the optimal solution tried to achieve, htFor controlling elements step-length, work as subgradientFor 0 when stop iteration, or work as htLevel off to 0 or it is infinitely great when, LR (λt) level off to D, calculate and stop；
Step 5) step-length htComputational methods are as follows：
Wherein, 0≤βt≤ 2, in LR (λt) rise when, βtKeep constant, be otherwise the half of currency, fUP(t) it is repeatedly
LR (the λ recorded in generationt) upper bound, fLB(t) it is LR (λt) a lower bound,For the subgradient of object function (6).
In a kind of optimization method of the full channel logistics distribution of described solution, described step 2, greediness insertion is used
Method builds the step of including initial distribution project that shops and vehicle route distribute and is：
Use following insertion cost calculation formula：
Wherein, u represents the shops for needing to insert,Expression is inserted on the route between shops i and shops j, road
The corresponding vehicle of line is k；Value be made up of two parts expense, be the expense that u inserts i and j respectively, and newly-increased one
Car services u expense；It is vehicle by shops i to shops u expense, every kilometer of k vehicles is multiplied by by the distance between two shops
Oil consumption is obtained；By the way that controlling elements γ is in insertion current route or increases a car acquirement balance newly, random value 0.00,0.05,
0.10,...,1.65,1.70}；
First, it is sky to initialize M car, and M represents the quantity that fleet of enterprise possesses vehicle；Then shops's insertion is calculated one by one
The insertion cost of distribution route corresponding to vehicle, the minimum route of selection cost is inserted, if it is maximum that vehicle is violated after insertion
Capacity-constrained is reprinted, then increases a same type car newly；Do not stop iteration to perform untill all clients are inserted into route.
It is initial using removing in a kind of optimization method of the full channel logistics distribution of described solution, described step 2
The destruction of shops in distribution project removes the step of method is to be destroyed, and is random by randomly choose that shops removed
Remove, calculate shops's overhead value removed from current distribution project and the worst shifting removed according to order from big to small
Remove, or random clustered to any distribution route using the progress of Density Clustering technology obtains the cluster shifting that either cluster is removed after gathering is closed
A kind of removal mode in removing is removed.
The step of a kind of optimization method of the full channel logistics distribution of described solution, the worst described removal, includes：
The overhead functions that a shops removes from current distribution project are defined, when carrying out removal iteration every time, institute are first calculated
There is the overhead value for not removing shops, and by small order sequence is arrived greatly, then select a shops from sequence according to random function
Removed, random function ensures easier to be selected in the more forward position of sequence.
A kind of optimization method of the full channel logistics distribution of described solution, the step of described Density Clustering is removed is wrapped
Include：
First, the shops's number that need to be removed is set as q, and q takes min { 0.4n, 60 }, and wherein n represents the sum of client, then
One distribution route of random selection, which using Density Clustering technology cluster, obtains gathering conjunction, and one is randomly choosed from gathering conjunction
Cluster is removed, and is still needed to remove shops number if the shops possessed in cluster is less than, is directly removed the cluster, otherwise, from the cluster with
Machine generates a subset so that shops's number in subset, which is equal to, to be still needed to remove shops's number, is performed repeatedly until q shops's quilt
Untill removal, when realizing Density Clustering, designated field radius parameter EpsDistance and kernel object number parameter are needed
MinPts, wherein EpsDistance determine by following formula (11),
(davg-dmin)*ξ(11)
Wherein, davgFor the average distance between 10 shops randomly selecting, dminFor the most short distance between 10 shops
From controlling elements ξ takes the value between 0 to 1；The each random values of MinPts { 2,3,4 }, according to the shops assembled in actual scene
Number is set.
In a kind of optimization method of the full channel logistics distribution of described solution, described step 2, by that will move again
The shops removed is inserted into the reconstruction method in distribution route to form feasible distribution project, be using greedy insertion or
Regret-2 insertions are rebuild：
Wherein greedy insertion is：Shops is removed for one find the minimum route of insertion cost (10) in each iteration
Inserted, with Δ fi,kRepresent to insert shops i route k cost function, if can not be by because vehicle capacity constraint is violated
Shops i inserts route k, then Δ fi,k=∞, each iteration finds the combination of following shops and route during reconstruction：
Regret-2 insertions are：With Δ fi 1Represent the cost that shops i inserts optimal route, Δ fi 2Represent shops i insertions
The cost of secondary major path, finds the maximum shops i of following difference and carries out insertion process in each iteration during reconstruction：
In a kind of optimization method of the full channel logistics distribution of described solution, described step 2, use is additionally included in
Reconstruction method is formed after feasible distribution project, the step of by simulated annealing acceptance criterion to carry out receiving new explanation：
Using receive new explanation X ' probability as：
Wherein, f (x) is weighs the overhead function of the distribution project quality obtained by step 2, and T represents temperature, by TinitOpen
Begin to cool, cooling formula is that T=T ω, ω are cooldown rate, span 0<ω<1, TinitObtain public by solving equation (15)
Formula (16)：
Wherein, finit(x) initial solution structural scheme overhead is represented.
In a kind of optimization method of the full channel logistics distribution of described solution, described step 2, in addition to basis is ruined
Go out stage and phase of regeneration chooses the step of algorithm combination is to improve situation about currently solving formed by algorithms of different, according to difference
The situation for the solution that algorithm combination is tried to achieve to assign reciprocal fraction to respective algorithms combination, before next iteration, uses roulette
Back-and-forth method selection algorithm is combined, the maximum iteration that algorithm end condition sets for execution.
In a kind of optimization method of the full channel logistics distribution of described solution, described step 3, build shops and arrive it
The process in the path of institute's services client is：TSP paths between shops and its services client are built by greedy insertion, will
The insertion cost that client k is inserted between client i and j is：dik+dkj-dij, wherein dijThe distance between client i and j are represented, often
Secondary is that the position of shops's selection insertion Least-cost is inserted, until in all clients insertion distribution route.
The technical effects of the invention are that, invent a kind of three phase algorithm and solve full channel logistics distribution, can be in limit
In fixing time preferably distribution project is tried to achieve for enterprise.First stage solves LAP problems using Lagrangian relaxation technology, passes through
The complicated constraints of relaxation is simultaneously added in object function so that problem solving difficulty is substantially reduced.Meanwhile, Lagrange relaxation
Technology can provide a lower bound of challenge, and trying to achieve problem by Subgradient optimization algorithm maximization lower bound preferably solves, and solves
Quality is high and algorithm is easily achieved, and is that second stage determines preferable shops position and client's allocative decision.Second stage is used
Adaptive extensive neighborhood search solves Multi-types vehicle routine problem, according to dispatching shops scale is big and geographically small range
The characteristics of scattered in interior aggregation is a wide range of, one group of design is simple to be destroyed and algorithm for reconstructing so that adaptive extensive neighborhood is searched
Rope algorithm can in solution space extensive search feasible solution, the situation for being absorbed in local optimum can be prevented effectively from.Meanwhile, the present invention exists
The simulated annealing acceptance criteria realized in adaptive extensive neighborhood search can guarantee that the quality and convergence of solution, algorithm
The output result after the iterations specified is performed, can meet time-constrain requirement, be tried to achieve preferably for enterprise in limiting time
Vehicle delivery scheme.With reference to the simplicity and the validity of adaptive extensive neighborhood search of Lagrangian relaxation technology, mix
Hop algorithm integrated solution efficiency high, can effectively solve full channel logistics distribution.
The invention will be further described below in conjunction with the accompanying drawings.
Brief description of the drawings
Fig. 1 is full channel logistics distribution model schematic diagram；
Fig. 2 is adaptive large neighborhood search algorithm schematic diagram；
Fig. 3 is greedy insertion schematic diagram.
Embodiment
The present invention is directed to full channel logistics distribution, proposes a kind of three stages solution annual reporting law.First, using Lagrange
Relaxing techniques solves shops's addressing and client's assignment problem.Then, logistics is solved by adaptive large neighborhood search algorithm
Total storehouse finally, the path between shops and its services client is solved using greedy insertion to the path planning problem of each shops
Planning problem.
The present invention point three phases solve Location-Routing Problem：First stage, according to Customer Location and the position of alternative shops
Put, be the shops of each client's specified services, determine the position of shops, be a Facility Location Problem, using Lagrangian pine
Relaxation technology is solved.Second stage, sends one group of vehicle delivery goods to each shops, meets the order need of shops from the total storehouse of logistics
Sum the customer demand that it is serviced, build each car travel route, be a Multi-types vehicle routine problem, using adaptive
Large neighborhood search algorithm is solved.Phase III, shops is built to the path of its services client, client, which can visit, to ask for soon
Part, can also be dispensed by courier, be a traveling salesman problem (Traveling Salesman Problem, TSP), be used
Greedy insertion is solved.
First stage mainly solves the problem of shops's addressing is assigned with client.Each alternative shops's storage express mail capacity has
Limit, and with setting up expense.For each client distribute a shops serviced, if all clients to service shops apart from summation
For connection expense.Phase targets be from alternative shops collect I in determine the position of one group of shops so that the demand of all clients all by
Meet and set up expense with being connected expense sum at least, shown in founding mathematical models such as formula (1-4).
Symbol is defined as follows with variable parameter in formula：
I:Alternative shops's set, I={ 1,2,3 ..., m }
J:Client set, J={ 1,2,3 ..., n }
Oi:Alternative shops i's sets up expense.
Wi：Alternative shops i storing capacity.
dj：Client j demand
Wherein, constraints (2) represents that each client can only be serviced by a shops；Condition (3) represents shops's service
Customer demand sum is no more than its storing capacity；If representing, client i is serviced condition (4) by shops j, in position candidate
Shops is set up at j.Mathematical modeling is solved by Lagrangian relaxation technology, Lagrange multiplier loose constraint condition (2) is introduced,
So that each client can be serviced by multiple shops, problem (1) is converted into knapsack problem, is solved using dynamic programming algorithm,
Obtain the lower bound of problem (1).Then, the lagrange duality problem of (1) is solved, using Subgradient optimization algorithm, problem is obtained
(1) solution, it is determined that setting up position and its client serviced of shops.
Second stage solves Multi-types vehicle routine problem and obtains vehicle delivery scheme.Fleet of enterprise possesses various,
Every kind of vehicle fixed overhead, variable overhead, maximum reprinting amount are different.One group of each shops of vehicle service is sent from the total storehouse of logistics, is made each
Shops's demand is met, and shops's demand includes shops's order demand and its services client demand summation.Phase targets are vehicles
Vehicle fleet size needed for distribution project is minimum, and driving path is most short.Solved using adaptive extensive field searching algorithm.From
Adapt to extensive field searching algorithm and perform an iteration mainly including producing initial solution, destroying (ruin) currently solution, reconstruction
(recreate) step such as feasible solution, adaptively selected strategy, as shown in Fig. 2 wherein aiAlgorithm combination is represented, is ruined comprising one kind
Go out removal algorithm and a kind of algorithm for reconstructing, πiFor its fraction, a is determinediIt is chosen the probability used.First, using a kind of simple structure
Make algorithm and produce subgraph a in initial distribution project, such as Fig. 2.Then, a is used in the destruction stageiIn destruction remove algorithm will
Q shops removes from distribution project, subgraph b in such as Fig. 2.Then, a is used in phase of regenerationiIn algorithm for reconstructing by q not
Distribution shops is reinserted on distribution route, forms subgraph c in feasible distribution project, such as Fig. 2.Finally, according to acceptance criteria
Judge whether to adopt the distribution project newly produced, update the algorithm combination a for destroying and being used with phase of regenerationiFraction πi.Adaptively
The preferably solution that extensive field searching algorithm can currently find output after the iterations specified is performed is matched somebody with somebody as final
Send scheme.
Phase III mainly builds the TSP paths between shops and its services client.Client asks for express delivery to shops, also may be used
Dispensed by courier, the access route of courier dispatching is produced by greedy insertion.
The calculating process in above three stage specifically includes following steps：
(1) mathematical modeling is solved using Lagrangian relaxation technology in the first stage to comprise the following steps that：
1) non-negative Lagrange multiplier λ is introducediLoose constraint condition (2) is obtained：
2) object function for adding (5) after must being relaxed in object function (1)：
Constraints is (3) and (4)
3) for each shops j, object function (6) is a knapsack problem, and problem can be tried to achieve using dynamic programming
(1) f inLAPLower bound, pass through solve (6) in dual problem：
Try to achieve most suitable λiTo obtain problem fLAPOptimal solution.
4) using most suitable λ in Subgradient optimization Algorithm for Solving dual problem (7)i.Initialization vector λ0For 0, at t times
With in t+1 iteration, vectorial λtCalculation is as follows：
Wherein xij(λt) it is object function LR (λj) in λtUnder the optimal solution tried to achieve, htFor controlling elements step-length.Work as subgradientFor 0 when algorithm stop iteration, or work as htLevel off to 0 or it is infinitely great when, LR (λt) leveling off to D, algorithm stops
Only.
5) step-length htComputational methods are as follows：
Wherein, 0≤βt≤ 2, typically take β0=2, in LR (λt) rise when, βtKeep constant, otherwise one for currency
Half.fUP(t) LR (λ recorded for algorithm in iterationt) upper bound, fLB(t) it is LR (λt) a lower bound.Formula denominator is time ladder
Square of degree.
(2) second stage solves one group of special-shaped vehicle from the total storehouse of logistics to each using adaptive large neighborhood search algorithm
The distribution route of shops, is comprised the following steps that：
1) initial distribution project is built using greedy insertion, for Multi-types vehicle routine problem design insertion cost meter
Calculate formula as follows：
Wherein, u represents the shops for needing to insert,Expression is inserted on the route between shops i and shops j, road
The corresponding vehicle of line is k.Value determined by two parts cost, be respectively u insertion i and j expense and increase newly a car
Service u expense.It is vehicle by shops i to shops u expense, every kilometer of oil of k vehicles is multiplied by by the distance between two shops
Consumption expense is obtained.By controlling elements γ, balance, random value can be obtained in insertion current route or a newly-increased car
{0.00,0.05,0.10,...,1.65,1.70}。
Greedy insertion process：First, it is sky to initialize M car, and M represents the quantity that fleet of enterprise possesses vehicle.Then,
The insertion cost of distribution route corresponding to shops's insertion vehicle is calculated one by one, the minimum route of selection cost is inserted, if inserting
Vehicle maximum is violated after entering and reprints capacity-constrained, then increases a same type car newly.Finally, do not stop iteration to perform until all clients
Untill being inserted into route.
2) the stage definitions neighborhood map mode of adaptive large neighborhood search algorithm is destroyed, using a kind of simple
Destroy removal algorithm to destroy current distribution project, remove q shops.Destroying removal algorithm has following 3 kinds：
A) it is random to remove：Q shops of selection is removed in the random set from shops
B) it is the worst to remove：The overhead functions that a shops removes from current distribution project are defined, overhead value is bigger, the shops
It should more be removed from affiliated route.First calculate all overhead values for not removing shops during each iteration of algorithm, and by arriving small suitable greatly
Sequence sorts.Then a shops is selected to be removed from sequence according to random function, random function ensures more forward in sequence
Position it is easier to be selected.
C) cluster is removed：A kind of new neighborhood map mode is defined, iteration performs following steps.First, randomly choose
One distribution route, which using Density Clustering technology cluster, obtains gathering conjunction, and then one cluster of random selection enters from gathering conjunction
Row is removed, and is still needed to remove shops's number if the shops possessed in cluster is less than, is directly removed the cluster.Otherwise, given birth at random from the cluster
Into a subset so that shops's number in subset, which is equal to, to be still needed to remove shops's number.Step is performed repeatedly until q shops's quilt
Untill removal.When realizing Density Clustering, designated field radius parameter EpsDistance and kernel object number parameter are needed
MinPts.EpsDistance determines by formula (11), wherein, davgFor the average distance between 10 shops randomly selecting,
dminFor the beeline between 10 shops, controlling elements ξ is taking the value between 0 to 1 according to actual conditions.MinPts is each
Random value { 2,3,4 }, can assemble shops's number according to actual scene and be set.In the situation that shops's scale is big and assembles
Under, when carrying out node removal to distribution project, destroy removal algorithm and removed a shops every time, shops may be from difference
Aggregation block, this is caused, and the former aggregation block cost of shops's insertion removed is minimum, and in phase of regeneration, the removal shops will be inserted again
Enter former aggregation block.Destroyed and reconstruction operation due to performing a wheel, distribution project does not change, adaptive extensive neighborhood is searched
Rope algorithm, which need to perform many wheel destruction and reconstruction operation, could remove whole aggregation block, so as to add the overall iterations of algorithm
With perform the time, when being dispensed in face of extensive shops, adaptive large neighborhood search algorithm can not be tried to achieve in finite time
More excellent solution.When being removed based on Density Clustering, distribution route is clustered, during removal in units of cluster, will entirely be assembled
Block is removed, and the effect of other removal algorithm successive ignitions can be reached in an iteration, meanwhile, density clustering algorithm realizes letter
Single, time complexity is relatively low.Therefore, the destruction removal algorithm based on Density Clustering not only reduces adaptive extensive neighborhood and searched
The iterations of rope algorithm performs, also reduces the algorithm overall execution time.
(davg-dmin)*ξ (11)
3) the q shops that the destruction stage removes is reinserted into distribution route by algorithm for reconstructing, forms feasible dispatching
Scheme.Algorithm for reconstructing has following 2 kinds：
A) greedy insertion：Shops is removed for one find the minimum route progress of insertion cost (10) in each iteration
Insertion.Assuming that Δ fi,kRepresent shops i insertion routes k cost function, if can not be by door because vehicle capacity constraint is violated
Shop i inserts route k, Δ fi,k=∞.The each iteration of algorithm finds the combination of following shops and route：
B) Regret-2 insertions:Assuming that Δ fi 1Represent the cost that shops i inserts optimal route, Δ fi 2Shops i is represented to insert
Enter the cost of time major path, in the bigger route for illustrating the more unsuitable insertion suboptimum of the shops of difference between the two.Regret-
2 insertions find the maximum shops i of following difference and carry out insertion process in each iteration：
4) acceptance criterion is calculated using simulated annealing, algorithm is received to solve more preferable new explanation than currently, also received with certain probability
Worse new explanation is solved than currently.The probability for receiving new explanation X ' is：
Wherein, f (x) is the overhead function of distribution project, and T represents temperature, by TinitStart cooling, cooling formula is T
=T ω, cooldown rate is ω, span 0<ω<1.Initial temperature TinitSetting on algorithm influence it is larger, to make algorithm
Adopt the new explanation than current optimal guards escorting prisoners 5%, solution equation (15) obtains TinitFormula (16):
Wherein, finit(x) initial solution structural scheme overhead is represented.Equation (14) left side is 0.5, is to ensure algorithm
When just starting iteration, poorer than initial scheme 50% new explanation is subjected to.
If 5) new explanation is received, improve 3 grades of situation point currently solved according to algorithm combination and update algorithm combination correspondence
Fraction：A) when algorithm combination tries to achieve current optimal solution, composite score increases by 30 points；B) when new explanation is more excellent than current solution, combination
Fraction increases by 10 points；C) when new explanation quality is less as preceding solution, composite score increases by 6 points.Before next iteration, roulette is used
Back-and-forth method selection algorithm is combined, and composite score is higher, and selected probability is bigger.Algorithm end condition changes to perform the maximum of setting
Generation number, the destruction stage removes shops q and is set to min { 0.4n, 60 }, and wherein n represents the sum of client, and this is an experience public affairs
Formula, small-scale problem is set to 0.4n, and extensive problem is 60.Because for different scenes, optimal algorithm combination is different.In order to
Adaptive large neighborhood search algorithm is set to automatically select optimal algorithm combination application in different scenes, it is necessary to increase adaptive
Answer selection strategy.Each algorithm combination corresponds to a fraction, and fraction is higher, more can illustrate that the combination is suitably applied currently
Scene, so, the combination should by more high probability selection use.New explanation be modeled annealing acceptance criterion adopt in the case of,
Fraction updates 3 grades of rule point.
Phase III builds the TSP paths between shops and its services client using greedy insertion.Client k is inserted
It is to the insertion cost between client i and j：dik+dkj-dij, wherein dijRepresent the distance between client i and j.Algorithm is every time
The position of one shops's selection insertion Least-cost is inserted, until in all clients insertion distribution route.In Fig. 3 example
In, when will number in the client for being 5 insertion route, due to being inserted between client 3 and client 2, cost is minimum, and client 5 will be inserted
Enter between 3 and 2.
Claims (11)
1. a kind of optimization method for solving full channel logistics distribution, it is characterised in that comprise the following steps：
Step 1, the use of Lagrangian Relaxation is that each client distributes a door according to Customer Location and the position of alternative shops
Shop is serviced, i.e., concentrated from alternative shops as connection expense apart from summation using all clients to service shops and determine one group of door
The position in shop so that the demand of all clients is all satisfied and sets up expense with being connected expense sum at least, so to each
The shops of client's specified services and the position for determining shops；
Step 2, it is true using adaptive large neighborhood search algorithm according to the order demand of shops and the customer demand serviced
Fixed one group is used to dispense goods to the vehicle of each shops and the travel route of each car：Being built first by greedy insertion includes
Shops and the initial distribution project of vehicle route distribution, then remove method using the destruction for removing the shops in initial distribution project
To be destroyed, then by the way that the shops of removal is inserted into the reconstruction method in distribution route again form feasible distribution side
Case, vehicle fleet size needed for making vehicle delivery scheme is minimum, and driving path is most short；
Step 3, shops is built to the path of its services client, realizes that shops is dispensed into the dispatching of client using greedy insertion
The optimization of route.
2. a kind of optimization method for solving full channel logistics distribution according to claim 1, it is characterised in that described
Step 1 in, come the shops to each client's specified services and determine mathematical modeling expression formula that the position of shops used for：
<mrow>
<msub>
<mi>minf</mi>
<mrow>
<mi>L</mi>
<mi>A</mi>
<mi>P</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<msub>
<mi>O</mi>
<mi>i</mi>
</msub>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>&Element;</mo>
<mi>J</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<msub>
<mi>d</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>s</mi>
<mo>.</mo>
<mi>t</mi>
<mo>.</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>&Element;</mo>
<mi>J</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<msub>
<mi>d</mi>
<mi>i</mi>
</msub>
<mo>&le;</mo>
<msub>
<mi>W</mi>
<mi>i</mi>
</msub>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>&le;</mo>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
</mrow>
</mtd>
<mtd>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
</mrow>
Wherein I gathers for alternative shops, and I={ 1,2,3 ..., m }, J is client set, J={ 1,2,3 ..., n }, OiFor alternative door
Shop i's sets up expense, WiFor alternative shops i storing capacity, djFor client j demand,
3. a kind of optimization method for solving full channel logistics distribution according to claim 2, it is characterised in that described
Step 1 in, using Lagrangian relaxation technology solve mathematical modeling the step of be：
Step 1) introduce non-negative Lagrange multiplier λiLoose constraint condition (2) is obtained：
<mrow>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>&Element;</mo>
<mi>J</mi>
</mrow>
</munder>
<msub>
<mi>&lambda;</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 2) object function that adds (5) after must being relaxed in object function (1)：
<mrow>
<mi>L</mi>
<mi>R</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>&lambda;</mi>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>&Element;</mo>
<mi>J</mi>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msub>
<mi>d</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>+</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<msub>
<mi>O</mi>
<mi>i</mi>
</msub>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>&Element;</mo>
<mi>J</mi>
</mrow>
</munder>
<msub>
<mi>&lambda;</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein constraints is formula (3) and formula (4)；
Step 3) for each shops j, f in (1) is tried to achieve using dynamic programmingLAPLower bound, pass through solve (6) in pair
Even problem：
<mrow>
<mi>D</mi>
<mo>=</mo>
<mi>max</mi>
<munder>
<mrow>
<mi>L</mi>
<mi>R</mi>
</mrow>
<mrow>
<mi>&lambda;</mi>
<mo>&GreaterEqual;</mo>
<mn>0</mn>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msub>
<mi>&lambda;</mi>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
Try to achieve most suitable λiTo obtain problem fLAPOptimal solution；
Step 4) using most suitable λ in Subgradient optimization Algorithm for Solving dual problem (7)i, initialization vector λ0For 0, at t times
With in t+1 iteration, vectorial λtCalculation is as follows：
<mrow>
<msup>
<mi>&lambda;</mi>
<mrow>
<mi>t</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
</msup>
<mo>=</mo>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mo>{</mo>
<mn>0</mn>
<mo>,</mo>
<msup>
<mi>&lambda;</mi>
<mi>t</mi>
</msup>
<mo>+</mo>
<msup>
<mi>h</mi>
<mi>t</mi>
</msup>
<mrow>
<mo>(</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mo>(</mo>
<msup>
<mi>&lambda;</mi>
<mi>t</mi>
</msup>
<mo>)</mo>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>}</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein xij(λt) it is object function LR (λj) in λtUnder the optimal solution tried to achieve, htFor controlling elements step-length, work as subgradientFor 0 when stop iteration, or work as htLevel off to 0 or it is infinitely great when, LR (λt) level off to D, calculate and stop；
Step 5) step-length htComputational methods are as follows：
<mrow>
<msup>
<mi>h</mi>
<mi>t</mi>
</msup>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>f</mi>
<mrow>
<mi>U</mi>
<mi>P</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msub>
<mi>f</mi>
<mrow>
<mi>L</mi>
<mi>B</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mo>|</mo>
<mo>|</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>I</mi>
</mrow>
</munder>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<msup>
<mi>&lambda;</mi>
<mi>t</mi>
</msup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mn>1</mn>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
</mrow>
</mfrac>
<msub>
<mi>&beta;</mi>
<mi>t</mi>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>9</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, 0≤βt≤ 2, in LR (λt) rise when, βtKeep constant, be otherwise the half of currency, fUP(t) it is in iteration
LR (the λ of recordt) upper bound, fLB(t) it is LR (λt) a lower bound,For the subgradient of object function (6).
4. a kind of optimization method for solving full channel logistics distribution according to claim 1, it is characterised in that described
Step 2 in, using greedy insertion build include shops and vehicle route distribution initial distribution project the step of be：
Use following insertion cost calculation formula：
<mrow>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>u</mi>
<mo>,</mo>
<msubsup>
<mi>r</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>C</mi>
<mrow>
<mi>i</mi>
<mi>u</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>C</mi>
<mrow>
<mi>u</mi>
<mi>j</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>C</mi>
<mrow>
<mi>i</mi>
<mi>j</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>&gamma;</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>C</mi>
<mrow>
<mn>0</mn>
<mi>u</mi>
</mrow>
<mi>k</mi>
</msubsup>
<mo>+</mo>
<msubsup>
<mi>C</mi>
<mrow>
<mi>u</mi>
<mn>0</mn>
</mrow>
<mi>k</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>10</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, u represents the shops for needing to insert,Expression is inserted on the route between shops i and shops j, route correspondence
Vehicle be k；Value be made up of two parts expense, be respectively u insert i and j expense, and increase newly a car service
U expense；It is expense of the vehicle by shops i to shops u, being multiplied by the every gas mileage of k vehicles by the distance between two shops obtains
Arrive；By the way that controlling elements γ is in insertion current route or increases a car acquirement balance newly, random value 0.00,0.05,
0.10,...,1.65,1.70}；
First, it is sky to initialize M car, and M represents the quantity that fleet of enterprise possesses vehicle；Then shops's insertion vehicle is calculated one by one
The insertion cost of corresponding distribution route, the minimum route of selection cost is inserted, and is reprinted if violating vehicle maximum after insertion
Capacity-constrained, then increase a same type car newly；Do not stop iteration to perform untill all clients are inserted into route.
5. a kind of optimization method for solving full channel logistics distribution according to claim 4, it is characterised in that described
Step 2 in, the step of method is to be destroyed is removed using the destruction for removing the shops in initial distribution project, is by random
The random removal that is removed of selection shops, overhead value that calculating shops removes from current distribution project and according to from big to small
The worst removal that order is removed, or random clustered to any distribution route using the progress of Density Clustering technology obtain gathering conjunction
A kind of removal mode removed afterwards in the cluster removal of either cluster is removed.
6. a kind of optimization method for solving full channel logistics distribution according to claim 5, it is characterised in that described
The worst removal the step of include：
Define the overhead functions that are removed from current distribution project of a shops, every time during progress removal iteration, first calculate it is all not
The overhead value of shops is removed, and by small order sequence is arrived greatly, then selects shops's progress from sequence according to random function
Remove, random function ensures easier to be selected in the more forward position of sequence.
7. a kind of optimization method for solving full channel logistics distribution according to claim 5, it is characterised in that described
Density Clustering include the step of remove：
First, the shops's number that need to be removed is set as q, and q takes min { 0.4n, 60 }, and wherein n represents the sum of client, then at random
One distribution route of selection, which using Density Clustering technology cluster, obtains gathering conjunction, and one cluster of random selection enters from gathering conjunction
Row is removed, and is still needed to remove shops's number if the shops possessed in cluster is less than, is directly removed the cluster, otherwise, given birth at random from the cluster
Into a subset so that shops's number in subset, which is equal to, to be still needed to remove shops's number, is performed repeatedly until that q shops is removed
Untill, when realizing Density Clustering, designated field radius parameter EpsDistance and kernel object number parameter MinPts are needed,
Wherein EpsDistance determines by following formula (11),
(davg-dmin)*ξ(11)
Wherein, davgFor the average distance between 10 shops randomly selecting, dminFor the beeline between 10 shops, control
Factor ξ processed takes the value between 0 to 1；The each random values of MinPts { 2,3,4 }, according to the shops's number assembled in actual scene
Set.
8. a kind of optimization method for solving full channel logistics distribution according to claim 4, it is characterised in that described
Step 2 in, form feasible distribution project by the way that the shops of removal is inserted into the reconstruction method in distribution route again, be
Rebuild using greedy insertion or Regret-2 insertions：
Wherein greedy insertion is：Shops is removed for one find the minimum route progress of insertion cost (10) in each iteration
Insertion, with Δ fi,kRepresent shops i insertion routes k cost function, if can not be by shops because vehicle capacity constraint is violated
I inserts route k, then Δ fi,k=∞, each iteration finds the combination of following shops and route during reconstruction：
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
<mo>=</mo>
<mi>arg</mi>
<munder>
<mi>min</mi>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>S</mi>
<mo>,</mo>
<mi>k</mi>
<mo>&Element;</mo>
<mi>R</mi>
</mrow>
</munder>
<msub>
<mi>&Delta;f</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
</mrow>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>12</mn>
<mo>)</mo>
</mrow>
</mrow>
Regret-2 insertions are：With Δ fi 1Represent the cost that shops i inserts optimal route, Δ fi 2Represent shops i insertion suboptimums
The cost of route, finds the maximum shops i of following difference and carries out insertion process in each iteration during reconstruction：
<mrow>
<mi>i</mi>
<mo>=</mo>
<mi>arg</mi>
<munder>
<mi>max</mi>
<mrow>
<mi>i</mi>
<mo>&Element;</mo>
<mi>S</mi>
</mrow>
</munder>
<mrow>
<mo>(</mo>
<msubsup>
<mi>&Delta;f</mi>
<mi>i</mi>
<mn>2</mn>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>&Delta;f</mi>
<mi>i</mi>
<mn>1</mn>
</msubsup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>13</mn>
<mo>)</mo>
</mrow>
<mo>.</mo>
</mrow>
9. a kind of optimization method for solving full channel logistics distribution according to claim 4, it is characterised in that described
Step 2 in, be additionally included in and formed after feasible distribution project using reconstruction method, connect by simulated annealing acceptance criterion
The step of by new explanation：
Using receive new explanation X ' probability as：
<mrow>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mfrac>
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<msup>
<mi>x</mi>
<mo>&prime;</mo>
</msup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
</mrow>
<mi>T</mi>
</mfrac>
</mrow>
</msup>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>14</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, f (x) is weighs the overhead function of the distribution project quality obtained by step 2, and T represents temperature, by TinitStart drop
Temperature, cooling formula is that T=T ω, ω are cooldown rate, span 0<ω<1, TinitFormula is obtained by solving equation (15)
(16)：
<mrow>
<mn>0.5</mn>
<mo>=</mo>
<msup>
<mi>e</mi>
<mfrac>
<mrow>
<mo>-</mo>
<msub>
<mi>f</mi>
<mrow>
<mi>i</mi>
<mi>n</mi>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mo>&CenterDot;</mo>
<mn>0.05</mn>
</mrow>
<msub>
<mi>T</mi>
<mrow>
<mi>i</mi>
<mi>n</mi>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
</mfrac>
</msup>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>15</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>T</mi>
<mrow>
<mi>i</mi>
<mi>n</mi>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>f</mi>
<mrow>
<mi>i</mi>
<mi>n</mi>
<mi>i</mi>
<mi>t</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>)</mo>
</mrow>
<mn>.0.05</mn>
</mrow>
<mrow>
<mi>ln</mi>
<mn>2</mn>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>16</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, finit(x) initial solution structural scheme overhead is represented.
10. a kind of optimization method for solving full channel logistics distribution according to claim 4, it is characterised in that institute
In the step 2 stated, in addition to the algorithm combination according to formed by destruction stage and phase of regeneration choose algorithms of different is worked as to improve
The step of situation of preceding solution, the situation for the solution tried to achieve is combined according to algorithms of different to assign reciprocal fraction to respective algorithms combination,
Before next iteration, combined using roulette wheel selection selection algorithm, the greatest iteration that algorithm end condition sets for execution
Number of times.
11. a kind of optimization method for solving full channel logistics distribution according to claim 4, it is characterised in that institute
In the step 3 stated, build shops is to the process in the path of its services client：Shops and its institute are built by greedy insertion
TSP paths between services client, be by the client k insertion costs being inserted between client i and j：dik+dkj-dij, wherein dij
The distance between client i and j are represented, the position for shops's selection insertion Least-cost is inserted every time, until all
In client's insertion distribution route.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710382144.1A CN107194513B (en) | 2017-05-26 | 2017-05-26 | Optimization method for solving problem of whole-channel logistics distribution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710382144.1A CN107194513B (en) | 2017-05-26 | 2017-05-26 | Optimization method for solving problem of whole-channel logistics distribution |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107194513A true CN107194513A (en) | 2017-09-22 |
| CN107194513B CN107194513B (en) | 2020-09-29 |
Family
ID=59875003
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710382144.1A Active CN107194513B (en) | 2017-05-26 | 2017-05-26 | Optimization method for solving problem of whole-channel logistics distribution |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107194513B (en) |
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107705043A (en) * | 2017-10-31 | 2018-02-16 | 顺丰科技有限公司 | A kind of Bus stop planning method, apparatus, equipment, storage medium |
| CN107833002A (en) * | 2017-11-28 | 2018-03-23 | 上海海洋大学 | Multistage low-carbon logistics distribution network planing method based on collaboration multi-objective Algorithm |
| CN108108847A (en) * | 2017-12-29 | 2018-06-01 | 合肥工业大学 | A kind of paths planning method of electric business logistics last one kilometer dispatching |
| CN108492068A (en) * | 2018-03-22 | 2018-09-04 | 北京百度网讯科技有限公司 | Method and apparatus for path planning |
| CN108492020A (en) * | 2018-03-16 | 2018-09-04 | 浙江工商大学 | Pollution vehicle dispatching method and system based on simulated annealing and branch's cutting optimization |
| CN108573326A (en) * | 2018-04-17 | 2018-09-25 | 武汉理工大学 | Video server site selecting method based on genetic-ant colony hybrid algorithm |
| CN108596469A (en) * | 2018-04-19 | 2018-09-28 | 中南大学 | A kind of quick self-adapted extensive neighborhood search method towards extensive Vehicle Routing Problems |
| CN108665115A (en) * | 2018-05-21 | 2018-10-16 | 北京百度网讯科技有限公司 | Method for optimizing scheduling and device |
| CN108665307A (en) * | 2018-05-07 | 2018-10-16 | 苏州工业职业技术学院 | Chain solid shop/brick and mortar store dispatching optimization method and system under electric business background |
| CN108764777A (en) * | 2018-04-26 | 2018-11-06 | 浙江工商大学 | Electronic logistic car dispatching method with time window and system |
| CN108985597A (en) * | 2018-06-29 | 2018-12-11 | 华南理工大学 | A kind of dynamic logistics dispatching method |
| CN109165886A (en) * | 2018-07-16 | 2019-01-08 | 顺丰科技有限公司 | A kind of logistics vehicles paths planning method and device, equipment, storage medium |
| CN109272151A (en) * | 2018-09-06 | 2019-01-25 | 中南大学 | A kind of vehicle path planning algorithm optimization method based on Spark |
| CN109583650A (en) * | 2018-11-30 | 2019-04-05 | 浙江工商大学 | A kind of method of the addressing of electric vehicle electrical changing station and logistics distribution combined dispatching |
| CN109916404A (en) * | 2017-12-12 | 2019-06-21 | 顺丰科技有限公司 | A kind of paths planning method, device, storage medium and equipment |
| CN109919530A (en) * | 2017-12-12 | 2019-06-21 | 顺丰科技有限公司 | A kind of Distribution path destroys method for reconstructing, device, storage medium and equipment |
| CN109919348A (en) * | 2017-12-12 | 2019-06-21 | 顺丰科技有限公司 | A kind of method for optimizing route, device, equipment, storage medium |
| CN110097218A (en) * | 2019-04-18 | 2019-08-06 | 北京邮电大学 | Unmanned commodity distribution method and system under changing environment when a kind of |
| CN110390413A (en) * | 2018-04-17 | 2019-10-29 | 菜鸟智能物流控股有限公司 | Processing method and device for combined optimization problem |
| CN110555640A (en) * | 2018-05-31 | 2019-12-10 | 北京京东振世信息技术有限公司 | Method and device for route planning |
| CN110674967A (en) * | 2019-08-02 | 2020-01-10 | 重庆大学 | Express vehicle path robust optimization method under uncertain driving time |
| CN110705798A (en) * | 2019-10-09 | 2020-01-17 | 四川大学 | Warehouse assembly and assembly integrated product distribution route and technician scheduling optimization method |
| CN111144647A (en) * | 2019-12-25 | 2020-05-12 | 华院数据技术(上海)有限公司 | General vehicle path planning method and system based on large-scale neighborhood search algorithm |
| CN111178730A (en) * | 2019-12-24 | 2020-05-19 | 中国航空工业集团公司西安飞机设计研究所 | Method and device for planning supply of oiling machine |
| CN111428902A (en) * | 2019-01-10 | 2020-07-17 | 顺丰科技有限公司 | Method and device for determining transport route |
| CN111428991A (en) * | 2020-03-20 | 2020-07-17 | 北京百度网讯科技有限公司 | Method and device for determining delivery vehicles |
| CN111598516A (en) * | 2020-05-20 | 2020-08-28 | 圆通速递有限公司 | Method and system for posthouse site selection |
| CN111985698A (en) * | 2020-07-30 | 2020-11-24 | 中国外运股份有限公司 | Method and device for determining sharing of kitchen and bathroom vehicles for home delivery |
| CN112686458A (en) * | 2021-01-05 | 2021-04-20 | 昆明理工大学 | Optimized scheduling method for multi-vehicle fleet cargo delivery process |
| CN113762667A (en) * | 2020-08-13 | 2021-12-07 | 北京京东振世信息技术有限公司 | Vehicle scheduling method and device |
| CN113762886A (en) * | 2021-08-06 | 2021-12-07 | 浙江浙石油综合能源销售有限公司 | Oil product secondary logistics distribution optimization method and system based on variable neighborhood search algorithm |
| CN113898403A (en) * | 2021-10-09 | 2022-01-07 | 中国矿业大学(北京) | Pure electric trackless rubber-tyred vehicle path planning model and method |
| CN115271354A (en) * | 2022-06-24 | 2022-11-01 | 湖南湘邮科技股份有限公司 | Service electronic fence dynamic planning method and system based on delivery unit |
| CN116341781A (en) * | 2023-03-28 | 2023-06-27 | 暨南大学 | Path planning method based on large-scale neighborhood search algorithm and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030078802A1 (en) * | 2001-10-22 | 2003-04-24 | International Business Machines Corporation | Method, system and program for creating a delivery plan |
| CN102880798A (en) * | 2012-09-20 | 2013-01-16 | 浪潮电子信息产业股份有限公司 | Variable neighborhood search algorithm for solving multi depot vehicle routing problem with time windows |
| CN103473612A (en) * | 2013-09-06 | 2013-12-25 | 周伟华 | Site selection and transportation optimization method for super-large scale logistics distribution |
| CN105976298A (en) * | 2016-06-02 | 2016-09-28 | 广东工业大学 | Logistics transport scheduling method based on discrete bat algorithm |
| CN106447121A (en) * | 2016-10-12 | 2017-02-22 | 上海节点供应链管理有限公司 | Intelligent optimization scheduling method based on city delivery |
-
2017
- 2017-05-26 CN CN201710382144.1A patent/CN107194513B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030078802A1 (en) * | 2001-10-22 | 2003-04-24 | International Business Machines Corporation | Method, system and program for creating a delivery plan |
| CN102880798A (en) * | 2012-09-20 | 2013-01-16 | 浪潮电子信息产业股份有限公司 | Variable neighborhood search algorithm for solving multi depot vehicle routing problem with time windows |
| CN103473612A (en) * | 2013-09-06 | 2013-12-25 | 周伟华 | Site selection and transportation optimization method for super-large scale logistics distribution |
| CN105976298A (en) * | 2016-06-02 | 2016-09-28 | 广东工业大学 | Logistics transport scheduling method based on discrete bat algorithm |
| CN106447121A (en) * | 2016-10-12 | 2017-02-22 | 上海节点供应链管理有限公司 | Intelligent optimization scheduling method based on city delivery |
Non-Patent Citations (2)
| Title |
|---|
| 苗国强等: "基于自适应大规模邻域搜索算法的带时间窗的车辆路径问题", 《物流技术》 * |
| 魏占阳等: "基于自适应大规模邻域搜索算法的两级车辆路径问题", 《物流科技》 * |
Cited By (53)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107705043A (en) * | 2017-10-31 | 2018-02-16 | 顺丰科技有限公司 | A kind of Bus stop planning method, apparatus, equipment, storage medium |
| CN107833002B (en) * | 2017-11-28 | 2021-10-22 | 上海海洋大学 | Multi-stage low-carbon logistics distribution network planning method based on cooperative multi-objective algorithm |
| CN107833002A (en) * | 2017-11-28 | 2018-03-23 | 上海海洋大学 | Multistage low-carbon logistics distribution network planing method based on collaboration multi-objective Algorithm |
| CN109919348A (en) * | 2017-12-12 | 2019-06-21 | 顺丰科技有限公司 | A kind of method for optimizing route, device, equipment, storage medium |
| CN109919530A (en) * | 2017-12-12 | 2019-06-21 | 顺丰科技有限公司 | A kind of Distribution path destroys method for reconstructing, device, storage medium and equipment |
| CN109916404A (en) * | 2017-12-12 | 2019-06-21 | 顺丰科技有限公司 | A kind of paths planning method, device, storage medium and equipment |
| CN108108847A (en) * | 2017-12-29 | 2018-06-01 | 合肥工业大学 | A kind of paths planning method of electric business logistics last one kilometer dispatching |
| CN108108847B (en) * | 2017-12-29 | 2021-05-04 | 合肥工业大学 | Route planning method for last kilometer distribution of E-commerce logistics |
| CN108492020A (en) * | 2018-03-16 | 2018-09-04 | 浙江工商大学 | Pollution vehicle dispatching method and system based on simulated annealing and branch's cutting optimization |
| CN108492020B (en) * | 2018-03-16 | 2021-02-02 | 浙江工商大学 | Polluted vehicle scheduling method and system based on simulated annealing and branch cutting optimization |
| CN108492068A (en) * | 2018-03-22 | 2018-09-04 | 北京百度网讯科技有限公司 | Method and apparatus for path planning |
| US11282026B2 (en) | 2018-03-22 | 2022-03-22 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for planning route |
| CN110390413A (en) * | 2018-04-17 | 2019-10-29 | 菜鸟智能物流控股有限公司 | Processing method and device for combined optimization problem |
| CN108573326B (en) * | 2018-04-17 | 2021-07-06 | 武汉理工大学 | Video server site selection method based on genetic ant colony hybrid algorithm |
| CN108573326A (en) * | 2018-04-17 | 2018-09-25 | 武汉理工大学 | Video server site selecting method based on genetic-ant colony hybrid algorithm |
| CN110390413B (en) * | 2018-04-17 | 2024-04-16 | 菜鸟智能物流控股有限公司 | Method and device for processing combination optimization problem |
| CN108596469B (en) * | 2018-04-19 | 2021-11-30 | 中南大学 | Rapid self-adaptive large-scale neighborhood searching method for large-scale vehicle path problem |
| CN108596469A (en) * | 2018-04-19 | 2018-09-28 | 中南大学 | A kind of quick self-adapted extensive neighborhood search method towards extensive Vehicle Routing Problems |
| CN108764777A (en) * | 2018-04-26 | 2018-11-06 | 浙江工商大学 | Electronic logistic car dispatching method with time window and system |
| CN108764777B (en) * | 2018-04-26 | 2021-03-30 | 浙江工商大学 | Electric logistics vehicle scheduling method and system with time window |
| CN108665307A (en) * | 2018-05-07 | 2018-10-16 | 苏州工业职业技术学院 | Chain solid shop/brick and mortar store dispatching optimization method and system under electric business background |
| CN108665115A (en) * | 2018-05-21 | 2018-10-16 | 北京百度网讯科技有限公司 | Method for optimizing scheduling and device |
| CN110555640A (en) * | 2018-05-31 | 2019-12-10 | 北京京东振世信息技术有限公司 | Method and device for route planning |
| CN110555640B (en) * | 2018-05-31 | 2023-08-08 | 北京京东振世信息技术有限公司 | Route planning method and device |
| CN108985597A (en) * | 2018-06-29 | 2018-12-11 | 华南理工大学 | A kind of dynamic logistics dispatching method |
| CN108985597B (en) * | 2018-06-29 | 2021-11-19 | 华南理工大学 | Dynamic logistics scheduling method |
| CN109165886B (en) * | 2018-07-16 | 2022-06-03 | 顺丰科技有限公司 | Logistics vehicle path planning method and device, equipment and storage medium |
| CN109165886A (en) * | 2018-07-16 | 2019-01-08 | 顺丰科技有限公司 | A kind of logistics vehicles paths planning method and device, equipment, storage medium |
| CN109272151B (en) * | 2018-09-06 | 2021-06-25 | 中南大学 | Spark-based vehicle path planning algorithm optimization method |
| CN109272151A (en) * | 2018-09-06 | 2019-01-25 | 中南大学 | A kind of vehicle path planning algorithm optimization method based on Spark |
| CN109583650B (en) * | 2018-11-30 | 2021-03-30 | 浙江工商大学 | Electric vehicle battery replacement station site selection and logistics distribution joint scheduling method |
| CN109583650A (en) * | 2018-11-30 | 2019-04-05 | 浙江工商大学 | A kind of method of the addressing of electric vehicle electrical changing station and logistics distribution combined dispatching |
| CN111428902A (en) * | 2019-01-10 | 2020-07-17 | 顺丰科技有限公司 | Method and device for determining transport route |
| CN110097218B (en) * | 2019-04-18 | 2021-04-13 | 北京邮电大学 | Unmanned commodity distribution method and system in time-varying environment |
| CN110097218A (en) * | 2019-04-18 | 2019-08-06 | 北京邮电大学 | Unmanned commodity distribution method and system under changing environment when a kind of |
| CN110674967A (en) * | 2019-08-02 | 2020-01-10 | 重庆大学 | Express vehicle path robust optimization method under uncertain driving time |
| CN110705798A (en) * | 2019-10-09 | 2020-01-17 | 四川大学 | Warehouse assembly and assembly integrated product distribution route and technician scheduling optimization method |
| CN111178730A (en) * | 2019-12-24 | 2020-05-19 | 中国航空工业集团公司西安飞机设计研究所 | Method and device for planning supply of oiling machine |
| CN111144647A (en) * | 2019-12-25 | 2020-05-12 | 华院数据技术(上海)有限公司 | General vehicle path planning method and system based on large-scale neighborhood search algorithm |
| CN111428991A (en) * | 2020-03-20 | 2020-07-17 | 北京百度网讯科技有限公司 | Method and device for determining delivery vehicles |
| CN111428991B (en) * | 2020-03-20 | 2022-08-02 | 北京百度网讯科技有限公司 | Method and device for determining delivery vehicles |
| CN111598516A (en) * | 2020-05-20 | 2020-08-28 | 圆通速递有限公司 | Method and system for posthouse site selection |
| CN111598516B (en) * | 2020-05-20 | 2023-08-22 | 圆通速递有限公司 | Post site selection method and system |
| CN111985698B (en) * | 2020-07-30 | 2024-03-08 | 中国外运股份有限公司 | Method and device for determining split loading of kitchen and bathroom type vehicles conveyed and installed at home |
| CN111985698A (en) * | 2020-07-30 | 2020-11-24 | 中国外运股份有限公司 | Method and device for determining sharing of kitchen and bathroom vehicles for home delivery |
| CN113762667A (en) * | 2020-08-13 | 2021-12-07 | 北京京东振世信息技术有限公司 | Vehicle scheduling method and device |
| CN112686458B (en) * | 2021-01-05 | 2023-03-07 | 昆明理工大学 | Optimized dispatching method for multi-vehicle fleet delivery process |
| CN112686458A (en) * | 2021-01-05 | 2021-04-20 | 昆明理工大学 | Optimized scheduling method for multi-vehicle fleet cargo delivery process |
| CN113762886A (en) * | 2021-08-06 | 2021-12-07 | 浙江浙石油综合能源销售有限公司 | Oil product secondary logistics distribution optimization method and system based on variable neighborhood search algorithm |
| CN113898403A (en) * | 2021-10-09 | 2022-01-07 | 中国矿业大学(北京) | Pure electric trackless rubber-tyred vehicle path planning model and method |
| CN115271354A (en) * | 2022-06-24 | 2022-11-01 | 湖南湘邮科技股份有限公司 | Service electronic fence dynamic planning method and system based on delivery unit |
| CN115271354B (en) * | 2022-06-24 | 2023-08-25 | 湖南湘邮科技股份有限公司 | Service electronic fence dynamic planning method and system based on delivery unit |
| CN116341781A (en) * | 2023-03-28 | 2023-06-27 | 暨南大学 | Path planning method based on large-scale neighborhood search algorithm and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107194513B (en) | 2020-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107194513A (en) | A kind of optimization method for solving full channel logistics distribution | |
| Surekha et al. | Solution to multi-depot vehicle routing problem using genetic algorithms | |
| CN108764777B (en) | Electric logistics vehicle scheduling method and system with time window | |
| Archetti et al. | An optimization-based heuristic for the split delivery vehicle routing problem | |
| Kontoravdis et al. | A GRASP for the vehicle routing problem with time windows | |
| CN107220731A (en) | A kind of logistics distribution paths planning method | |
| CN109299810A (en) | A kind of goods stock stowage method | |
| CN111798067B (en) | Automatic driving automobile distribution path planning method based on self-adaptive large neighborhood search algorithm | |
| Yao et al. | Improved ant colony optimization for seafood product delivery routing problem | |
| CN109559062A (en) | A kind of task distribution of cooperative logistical problem and paths planning method | |
| Li et al. | Deploying autonomous mobile lockers in a two-echelon parcel operation | |
| Bakas et al. | A flexible transportation service for the optimization of a fixed-route public transport network | |
| CN111860957B (en) | Multi-vehicle-type vehicle path planning method considering secondary distribution and balancing | |
| Guo et al. | Heuristic approaches for a vehicle routing problem with an incompatible loading constraint and splitting deliveries by order | |
| Corberán et al. | The Chinese Postman Problem with load-dependent costs | |
| Chen et al. | A hybrid two-stage sweep algorithm for capacitated vehicle routing problem | |
| Casazza et al. | The multiple vehicle balancing problem | |
| Minis et al. | Real-time management of vehicle breakdowns in urban freight distribution | |
| CN116151497A (en) | Multi-compartment vehicle path planning method and system with picking and delivering functions | |
| Nguyen et al. | GRASP with learning process for a two-echelon location routing problem | |
| CN108416482A (en) | One kind is based on regional shifty logistics distribution paths planning method | |
| Baklagis et al. | The team orienteering pick-up and delivery problem with time windows and its applications in fleet sizing | |
| Cui et al. | A Time‐Dependent Vehicle Routing Problem for Instant Delivery Based on Memetic Algorithm | |
| Corberán et al. | The multidepot drone general routing problem with duration and capacity constraints | |
| Song et al. | Operational redistribution model for a large-scale bicycle-sharing system: Case Study in China |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |