CN116797126A - Rural terminal logistics site selection-path planning method based on double-layer planning - Google Patents
Rural terminal logistics site selection-path planning method based on double-layer planning Download PDFInfo
- Publication number
- CN116797126A CN116797126A CN202310542905.0A CN202310542905A CN116797126A CN 116797126 A CN116797126 A CN 116797126A CN 202310542905 A CN202310542905 A CN 202310542905A CN 116797126 A CN116797126 A CN 116797126A
- Authority
- CN
- China
- Prior art keywords
- express
- point
- rural
- points
- express delivery
- 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
- 238000013439 planning Methods 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000002068 genetic effect Effects 0.000 claims abstract description 21
- 238000012384 transportation and delivery Methods 0.000 claims description 116
- 230000006870 function Effects 0.000 claims description 16
- 210000004027 cell Anatomy 0.000 claims description 15
- 239000000427 antigen Substances 0.000 claims description 6
- 102000036639 antigens Human genes 0.000 claims description 6
- 108091007433 antigens Proteins 0.000 claims description 6
- 210000000349 chromosome Anatomy 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 230000009193 crawling Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000010367 cloning Methods 0.000 claims description 3
- 230000005764 inhibitory process Effects 0.000 claims description 3
- 230000014759 maintenance of location Effects 0.000 claims description 3
- 230000035772 mutation Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000035755 proliferation Effects 0.000 claims description 3
- 230000006978 adaptation Effects 0.000 claims 1
- 238000010276 construction Methods 0.000 abstract description 3
- 230000000694 effects Effects 0.000 abstract 1
- 239000010410 layer Substances 0.000 description 15
- 238000010586 diagram Methods 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000002355 dual-layer Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a rural end logistics site selection-path planning method based on double-layer planning, which comprises the steps of acquiring rural POIs and path information between the rural POIs, and replacing rural logistics demand distribution with rural building distribution; building a rural end logistics site selection model by taking rural residents and express points as decision-making subjects, and building a rural end logistics path planning model by taking logistics companies as decision-making subjects; and solving the site selection model by adopting an immune algorithm, solving the path planning model by adopting a hybrid genetic algorithm, obtaining the approximately optimal site selection number of the express points by adopting the immune-hybrid genetic algorithm on the overall double-layer planning, and then carrying out more accurate site selection and path planning solving by adopting a nested double-layer planning algorithm. The invention has a certain effect on the construction and distribution of the terminal logistics network of rural areas with wide and thin land and multipoint express demand distribution, and also has self-adaption and expansibility for the distribution characteristics of different rural areas and the application of the distribution network when the express quantity is changed.
Description
Technical Field
The invention belongs to the field of logistics and operation management, and particularly relates to a rural tail end logistics site selection-path planning method based on double-layer planning.
Background
Under the promotion of online shopping consumption hot tide, rural electric merchants can also develop rapidly. The potential logistics demands in rural areas are large, but the problems of wide areas, low population density, relatively lagging logistics infrastructure and the like cause that the current distribution of rural express delivery cannot meet the higher pursuit of residents on life quality, and the vast rural areas cannot share the convenience of electronic commerce. The key to rural logistics is end distribution, "last mile" generally refers to the distance of an item from the end sorting center to the customer, which is the last loop of the entire distribution chain, which may be less than 1% of the total distance of the logistics distribution, but the distribution cost is 30% of the total distribution cost.
The network nodes of rural logistics are imperfect, unreasonable in distribution, resource is not intensive, functions are incomplete, the rural logistics are low in efficiency and high in operation cost, and the method is a weak link of short boards and agricultural modernization construction for restricting the healthy development of the logistics industry in China. Firstly, compared with the urban area, the rural area is inconvenient in traffic, the road is rugged and narrow, partial road sections can only run unidirectionally or cannot pass through, and even some areas are not communicated with the road. The three-level logistics distribution system of county-town (village) -village is long in time consumption and difficult to distribute. Secondly, rural areas population density is sparse, express delivery demand dispersion points are multiple-sided wide, addresses are relatively fuzzy, logistics distribution transportation cost is high, distribution efficiency is low, and logistics enterprises are difficult to profit. Finally, since the above characteristics and demands of rural logistics are obviously affected by seasons, it is difficult to determine the positions and construction scales of the logistics distribution center and the warehouse, and the positions of the logistics distribution center and the warehouse have great influence on the distribution cost and efficiency. The rural logistics distribution site selection and operation are very complicated, the profit is very thin, the intention of setting up express points in rural areas is not strong for most logistics enterprises, and the daily demands of rural residents are hardly met. Therefore, the problem of 'last kilometer' of rural logistics distribution is to flourish rural electric business, promote peasant consumption and upgrade, and improve peasant life quality. The method has the advantages that a proper rural logistics 'last kilometer' distribution mode is constructed, the problems of site selection and path planning are solved, support is provided for system operation and control cost, profit space is created for enterprises, and the method is an important problem for smooth rural logistics and promoting domestic large circulation to be solved urgently.
Disclosure of Invention
In order to solve the problems, the invention provides a rural end logistics site selection-path planning method based on double-layer planning, which comprises the steps of acquiring rural logistics demand information, adopting a common distribution-collection mode without establishing a distribution center for a rural end logistics distribution mode, respectively establishing operation study models for the problems of site selection and distribution of rural end express points and vehicle path planning, and designing a nested double-layer planning model based on an immune-hybrid genetic algorithm for settlement of the established models.
The technical scheme of the invention is as follows:
a rural end logistics site selection-path planning method based on double-layer planning comprises the following specific steps:
first, data preprocessing. The method comprises the steps of obtaining information of alternative express points, namely crawling rural POI (interest points) information, removing 'financial insurance service', 'logistics' and the like from the POIs, wherein the POIs are large categories which are not used as express point conditions, such as place names or professional places, and the place names are provided with specific places such as 'wine', 'country', 'charging', 'endowment' and the like, and removing POIs distributed by express points in the central zone of the town and nearby, and the rest POIs are used as rural alternative express points; for logistics demand distribution information, building distribution is adopted to simulate the characteristics of rural logistics demand multipoint distribution, and a density grading strategy based on building distribution is adopted to simulate different logistics demands.
And secondly, establishing an operation study model for site selection and path planning problems. Acquiring path information among rural POIs, and searching related rural express logistics distribution information from the path information; due to the specificity of rural roads, small-sized vehicles are adopted for carrying out multiple delivery until delivery of all rural express points is completed; during distribution, a common distribution center is not established at this time, vehicles directly go to express delivery points of different logistics companies to collect express delivery, and then the vehicles are distributed to rural express delivery points in batches; the rural residents directly go nearby to part of public places selected as express points in rural villages to take the express.
The method comprises the following steps:
(1) Upper layer site selection model establishment
The overall objective function is as follows:
wherein C is s The total cost of the rural express collection points is calculated; c (C) b The total cost of residents in rural areas; m is an alternative express point, and M e M, set m= { m|m=1, 2, …, f }; j is building, J e J, set j= { j|j=1, 2, …, g }; gamma is the unit express operation cost; c is the fixed cost of constructing an express delivery point; alpha is the weight of the express delivery cost; θ is the direct proportionality coefficient of the actual capacity and cost of the express point; μ is the weight of the express points biasing toward crowd density; u (u) mj The coverage of the express delivery point m comprises a building j which is 1, otherwise, the coverage of the express delivery point m is 0; u (u) m And the express point m which indicates the alternative express point is selected as the express point is 1, otherwise, the express point m is 0.
d mj Is the distance from the building j to the express point m. Wherein x is m 、y m The positions of the alternative express points converted in longitude and latitude are respectively x j 、y j The positions of the buildings converted in longitude and latitude are respectively; v is administrative village, and V e V, set v= { v|v=1, 2, …, h }. J (J) v Representing a building J within the coverage area of an administrative village v, and J v ＝{1,2,…,g v }，g v ≤g；M v Alternative express point i in administrative village v coverage range and M v ＝{1,2,…,f v }，f v ≤f。
The constraint conditions are as follows:
a convenience store not selected as an express delivery point cannot cover a building:
each building can only be assigned a coverage area belonging to one express point:
the number of express points selected needs to be within a certain number range:
r represents the maximum applicable capacity of the express point m, s j The actual express delivery of the building j is provided with capacity constraint for the express delivery of the coverage range of each express delivery point:
r is the furthest distance between the express point and the building, and the distance from the building to the express point has path length limitation:
each designated express delivery point needs to have coverage, at least more than one building:
0, 1 constraint of decision variables:
u m ∈{0,1},m∈M
u mj ∈{0,1},m∈M,j∈J
(2) Lower path planning model establishment
Objective function F d The following are provided:
where O represents the town-center express point, O e O, o= { o|o=1, 2, …, p }; i is the selected express point in rural area, and I e I, i= { i|i=1, 2, …, q }; r is the number of vehicle path plans, and R e R, r= { r|r=1, 2, …, s }; w (w) i The actual demand of the rural express delivery point i is obtained; t (T) i The time when the vehicle arrives at the rural express point i;the time when the vehicle arrives at the express point o of the town center in the r-th delivery; />For vehicle r-th from rural express delivery collection point i 1 Travel to rural express delivery collection point i 2 1, otherwise 0;express point o from town center for vehicle r time 1 Travel to rural express delivery collection point o 2 1, otherwise 0; />The vehicle is driven from the town center express delivery point o to the rural express delivery collection point i for the r time, and if not, the vehicle is 0; />The vehicle is driven from the express delivery point i to the town center rural express delivery collection point o for the r time, and if not, the vehicle is 0; />Is selected for rural areaExpress delivery point i 1 Travel to rural express delivery point i 2 Is a distance of (2); />Center express point o 1 Travel to town center express point o 2 Is a distance of (2); d, d oi The distance from the town center express delivery point o to the rural express delivery collection point i is set; d, d io The distance from the rural express delivery collection point i to the town center express delivery point o is travelled; ρ is the time satisfaction weight.
Selected express point i for rural area 1 Travel to another selected express point i 2 Time of (2); />Express point o for town center 1 Travel to town center express point o 2 Time of (2); t is t oi The time for the town center express delivery point o to travel to the rural express delivery collection point i; t is t io And (5) the time for the rural express delivery collection point i to travel to the town center express delivery point o. Express point o at center of town 2 Waiting time when collecting express delivery +.>The method comprises the following steps:
for rural express delivery point i 2 Rural express delivery point i during delivery 2 Is to wait for a period of
Constraint conditions:
each rural express point can only be traversed by the vehicle once in total, including one entry and one exit:
each point can only be traversed once when the vehicle traverses the central express point of the town each time, including entering once and exiting once:
the load capacity constraint of the vehicle is q, and the sum of the express delivery of the vehicle leaving the town center to the rural express delivery point can not exceed the actual capacity constraint of the vehicle:
if the vehicle goes to rural areas, the vehicle must be delivered to more than one express delivery point:
0, 1 constraint of decision variables:
thirdly, designing a nested double-layer planning algorithm based on an immune-hybrid genetic algorithm to solve the model.
Step 1: the immune algorithm flow is specifically as follows:
step 1.1: initializing parameters. The codes represent the numbers of the rural alternative express points and the buildings by natural numbers. The codes are divided into a plurality of rows, wherein the first row represents the number of the selected express points, and the rest is the codes allocated to the building. If the number of the express points in the first row is N, the operation is to select N express points. The (i+1) th row represents a building label collection obtained by distributing the selected (i) th express delivery point in the first row. The first row represents the length of the express point selection unchanged, the number of the row number codes allocated to the building remains unchanged, but the number of the representative buildings in each row of codes is variable.
Step 1.2: and generating an express point selection scheme of the initial solution. The method comprises the following steps: randomly extracting one point u from alternative express point set 1 The method comprises the steps of carrying out a first treatment on the surface of the For each express point x which is not found, calculatingCalculating the distance between x and the selected express point; with weighted probability distance +.>Selecting the next express point for probability address; repeating the first two steps until the required number of express points is reached.
Step 1.3: the iteration number is defined, i.e. the iteration is started.
Step 1.4: performing antibody diversity evaluation calculations including affinity between antibody and antigen Affinity of antibody->Antibody concentrationDesired reproduction Rate->Wherein P represents a penalty factor for the distance constraint penalty function, Q represents a penalty factor for the capacity constraint inverse function, R v,s The number of the v-th antibody and the s-th antibody is the same; len represents the number of bits representing each antibody; t is a set threshold; m is the total number of antibodies; alpha is a diversity evaluation parameter.
Step 1.5: antibodies are preferentially selected from the structures as memory cells. Selecting by elite retention strategy, storing the antibody with higher affinity with antigen into memory cell, and storing the antibody with high desired proliferation rate of the rest antibody into memory cell.
Step 1.6: selecting parent population by the same method as selecting memory cells, and performing immune selection, cloning, mutation and clone inhibition operation.
Step 1.7: combining the memory cells with the new population of individuals to form a new structure.
Step 1.8: judging whether the iteration times are reached, if not, turning to the step 1.4, and if so, turning to the step 1.9.
Step 1.9: and outputting the optimal express point selection and resident allocation scheme, the optimal fitness value, the objective function and other information.
Step 2: the flow of the hybrid genetic algorithm is specifically as follows:
step 2.1: reading data and setting parameters of the hybrid genetic algorithm. The coding mode is represented by a plurality of lines of natural number codes. The first line is the path planning sequence of the rural express points, the path planning of the express points at the center of the town consists of a second line and a plurality of codes with the same length, and each code represents the sequence of collecting the express at the center of the town each time. Fitness is defined as
Step 2.2: the improved path shortest greedy strategy generates an initial solution for path planning. Calculating the number of times of distribution according to the total express delivery and the maximum load of the vehicle; the greedy strategy of finding out the express point closest to the current point generates a closed path formed by the center express point of the inner ring town and the rural express point of the outer ring respectively. Dividing the rural express points of the outer ring into a plurality of batches according to the limit of the vehicle-mounted capacity. When the vehicle is in the town center, the vehicles are sequentially driven by the path of the originally generated town center TSP (travel business problem), and the vehicles are driven to the rural express points in the outer ring after the completion. And delivering the rural express points to be delivered, and after delivery is completed, driving to the town center express point closest to the last rural express point to be delivered. And judging whether the rural express delivery points are distributed, and if so, driving the vehicle to the starting point of the express delivery point of the town center. And finally merging the multi-loop path planning and outputting a result.
Step 2.3: the iteration is started.
Step 2.4: and calculating the fitness value of each chromosome, and selecting a parent population.
Step 2.5: offspring are generated using the improved neighborhood search strategy. And (3) for the path planning sequence of each town center express point of each chromosome, carrying out complete overturn by using the probability po_GA each time, and accepting the solution if the adaptability becomes large after the path reversal. And (3) adopting a partial reversal strategy for each rural express point, namely randomly generating two numbers for each rural express point, and reversing the sequence between the two random numbers. This result is accepted whether or not the fitness is getting large.
Step 2.6: offspring were generated using genetic strategies (roulette selection, single crossover, single parental variation).
Step 2.7: and 2.5, combining the offspring results generated in the steps 2.5 and 2.6, and selecting individuals with high fitness to generate a new population.
Step 2.8: and judging whether the maximum iteration times are reached, if so, turning to the step 2.7, and if not, turning to the step 2.4.
Step 2.9: and outputting the optimal path planning result and related information.
Step 3: the double-layer planning algorithm flow is specifically as follows:
step 3.1: the initialization of the algorithm and the information reading are carried out, and the method is concretely as follows:
step 3.2: and calculating the total express demand and the average express point capacity according to the building area, calculating the minimum number of express points according to the total express demand and the average express point capacity, and setting the maximum number of express points.
Step 3.3: and carrying out express point selection and building allocation iteration by using an immune algorithm, and outputting a corresponding optimal express point selection and building allocation scheme.
Step 3.4: and calculating the times of vehicle distribution and the requirement of each express point according to the result of the immune algorithm, and carrying out vehicle path planning operation iteration by using a hybrid genetic algorithm.
Step 3.5: and recording address selection, distribution and path planning solving result information under different express points.
Step 3.6: judging whether the number of the selected express points reaches the maximum number of the selected express points, if so, turning to the step 3.7, and if not, turning to the step 3.3.
Step 3.7: and finding out the number of the corresponding express points when the balance degree is optimal. By adoptingThe degree of equalization for each individual is calculated. />Respectively representing the maximum value and the minimum value of the distance between the building and the express delivery point from the viewpoint of residents;representing the maximum and minimum of cost and efficiency issues from the point of view of the rural express delivery,where β is the cost weight of the latency; d (D) max 、D min Respectively represent the arrangement of the common from the express enterprisesMaximum and minimum path costs to be paid for delivery.
Step 3.8: and (3) performing address selection operation of the express points according to the number of the express points, if the result fitness is higher than the fitness of the previous operation, turning to step 3.9, otherwise, continuing to turn to step 3.8 to perform iterative operation on the result obtained by the operation at the moment.
Step 3.9: and (5) carrying out path planning operation iteration by using a hybrid genetic algorithm and saving the result.
Step 3.10: and judging whether the iteration times are reached.
Step 3.11: the final LAP (addressing-allocation problem) and VRP (vehicle path problem) schemes in the record are output.
The invention has the beneficial effects that: the invention provides the method for carrying out the logistics distribution of the last kilometer in rural areas by adopting a common distribution-collection mode, simulating the actual express delivery demand by using the building distribution, crawling the actual rural road information and further researching the site selection-path planning problem of the last kilometer of the rural logistics. The research content of the rural 'last kilometer' problem is enriched, a new thought and a reference scheme are provided for solving the problem, and decision support can be provided for the problem of site selection and path planning of rural end logistics in practical application.
Drawings
FIG. 1 is a flow chart of a dual-layer planning provided by an embodiment of the invention for rural end logistics site selection-path planning;
FIG. 2 is a flowchart of an immune algorithm provided by an embodiment of the invention for rural end logistics site selection;
FIG. 3 is a flow chart of a hybrid genetic algorithm provided by an embodiment of the invention for rural end logistics path planning;
FIGS. 4 (a) -4 (d) are models describing the problem of rural end logistics path planning according to embodiments of the present invention; wherein fig. 4 (a) is a distribution of alternative express points in rural areas and express points in the center of the town, and fig. 4 (b), fig. 4 (c) and fig. 4 (d) are path plans between the alternative express points in rural areas and the express points in the center of the town when the distribution is performed for the first time, the second time and the third time, respectively.
FIG. 5 is a distribution diagram of a town center express point for a rural end logistics site selection-path planning problem according to an embodiment of the present invention;
FIG. 6 is an alternative rural courier point distribution diagram for the rural end logistics site selection-path planning problem in accordance with an embodiment of the present invention;
FIG. 7 is a diagram of a selection result of an express point according to an embodiment of the present invention;
FIG. 8 is a schematic diagram of a resulting vehicle delivery sequence for a rural end logistics path planning problem in accordance with an embodiment of the present invention;
FIG. 9 is a diagram of a result of selecting a rural delivery point according to a double-layer planning algorithm for a rural end logistics address-path planning problem in an embodiment of the present invention;
fig. 10 is a schematic diagram of a vehicle distribution sequence obtained by a double-layer planning algorithm for rural end logistics site selection-path planning according to an embodiment of the present invention.
Fig. 11 is a basic flow chart of the present invention.
Detailed Description
The following describes the embodiments of the present invention further with reference to the drawings and technical schemes.
As shown in fig. 11, the rural end logistics site selection-path planning method based on double-layer planning comprises the following steps:
step 1, data preprocessing
And crawling the information of a certain town POI from the POI interface of the Goldmap, obtaining data, and obtaining 700 POIs after de-duplication. Seven express points are found out from the crawled POIs, and all the express points are arranged in a densely populated area of the town center, as shown in fig. 4. Deleting the financial insurance service, the logistics and the like from the POIs of the crawling 7 as place names or major places and the like without the large class which becomes the condition of express points; and deleting the specific places such as wine, china, charging, pension and the like in the place names, removing POIs distributed in the town areas and nearby express delivery points, and finally obtaining 103 POIs of alternative express delivery points, as shown in fig. 5. Building information in the electronic map is extracted by ArcGIS, LOF anomaly detection, longitude and latitude conversion and the like are performed, then an administrative village center is found, and administrative region division is performed by using a tazion polygon, as shown in fig. 6. And finally, obtaining path information among the alternative express points.
Step 2, site selection modeling and solving
The overall objective function of the upper layer addressing model is as follows:
wherein x is m 、y m The positions of the alternative express points converted in longitude and latitude are respectively x j 、y j The positions of the buildings are converted in terms of longitude and latitude.
The constraint conditions are as follows:
a convenience store not selected as an express delivery point cannot cover a building:
each building can only be assigned a coverage area belonging to one express point:
the number of express points representing the selection needs to be within a certain number range:
express delivery of the range covered by each express delivery point has capacity constraint:
the distance from the building to the express point has a path length limitation:
each designated express delivery point needs to have coverage, at least more than one building:
0, 1 constraint of decision variables:
the addressing model is solved by an immune algorithm, and a flow chart of the immune algorithm is shown in fig. 2. The immune algorithm comprises the following steps:
step 2.1: initializing parameters and defining the encoding and decoding. The codes represent the numbers of the rural alternative express points and the buildings by natural numbers. The codes are divided into a plurality of rows, wherein the first row represents the number of the selected express points, and the rest is the codes allocated to the building. If the number of the express points in the first row is N, the operation is to select N express points. The (i+1) th row represents a building label collection obtained by distributing the selected (i) th express delivery point in the first row. The first row represents the length of the express point selection unchanged, the number of the row number codes allocated to the building remains unchanged, but the number of the representative buildings in each row of codes is variable.
Step 2.2: and generating an express point selection scheme of the initial solution. The method comprises the following steps: randomly extracting one point u from alternative express point set 1 The method comprises the steps of carrying out a first treatment on the surface of the For unreliability ofEach express point x calculationCalculating the distance between x and the selected express point; with weighted probability distance +.>Selecting the next express point for probability address; repeating the first two steps until the required number of express points is reached.
Step 2.3: the iteration number is defined, i.e. the iteration is started.
Step 2.4: performing antibody diversity evaluation calculations including affinity between antibody and antigen Affinity of antibody->Antibody concentrationDesired reproduction Rate->Wherein P represents a penalty factor for the distance constraint penalty function, Q represents a penalty factor for the capacity constraint inverse function, R v,s The number of the v-th antibody and the s-th antibody is the same; len represents the number of bits representing each antibody; t is a set threshold, commonly used t=0.7; m is the total number of antibodies set as described above; α is a diversity evaluation parameter, and α=0.95 is usually set.
Step 2.5: antibodies are preferentially selected from the structures as memory cells. Selecting by elite retention strategy, storing the antibody with higher affinity with antigen into memory cell, and storing the antibody with high desired proliferation rate of the rest antibody into memory cell.
Step 2.6: selecting parent population by the same method as selecting memory cells, and performing immune selection, cloning, mutation and clone inhibition operation.
Step 2.7: combining the memory cells with the new population of individuals to form a new structure.
Step 2.8: judging whether the iteration times are reached, if not, turning to the step 1.4, and if so, turning to the step 1.9.
Step 2.9: and outputting information such as optimal express point selection, resident allocation schemes, optimal fitness values, objective functions and the like. The result of the express point selection is shown in fig. 7.
Step 3, modeling and solving a path planning model
The underlying path planning model objective function is as follows:
wherein each time center express point o is opposite to town 2 Waiting time for collecting express deliveryThe method comprises the following steps:
for rural express delivery point i 2 Rural express delivery point i during delivery 2 Is to wait for a period of
Constraint conditions:
each rural express point can only be traversed by the vehicle once in total, including one entry and one exit:
each point can only be traversed once when the vehicle traverses the central express point of the town each time, including entering once and exiting once:
the load capacity constraint of the vehicle, namely the sum of the express delivery of the vehicle leaving the town center to the rural express delivery point at the time, cannot exceed the actual capacity constraint of the vehicle:
if the vehicle goes to rural areas, the vehicle must be delivered to more than one express delivery point:
0, 1 constraint of decision variables:
a flowchart for solving the underlying path planning model using a neighborhood-genetic algorithm is shown in fig. 3. The hybrid genetic algorithm comprises the following steps:
step 3.1: reading data and setting parameters of the hybrid genetic algorithm. The coding mode is represented by a plurality of lines of natural number codes. The first line is the path planning sequence of the rural express points, the path planning of the express points at the center of the town consists of a second line and a plurality of codes with the same length, and each code represents the sequence of collecting the express at the center of the town each time. Fitness is defined as/>
Step 3.2: the improved path shortest greedy strategy generates an initial solution for path planning. Calculating the number of times of distribution according to the total express delivery and the maximum load of the vehicle; the greedy strategy of finding out the express point closest to the current point generates a closed path formed by the center express point of the inner ring town and the rural express point of the outer ring respectively. Dividing the rural express points of the outer ring into a plurality of batches according to the limit of the vehicle-mounted capacity. When the vehicle is in the town center, the vehicles are driven in the path sequence of the initially generated town center TSP, and the vehicles are driven to the rural express points of the outer ring after the completion. And delivering the rural express points to be delivered, and after delivery is completed, driving to the town center express point closest to the last rural express point to be delivered. And judging whether the rural express delivery points are distributed, and if so, driving the vehicle to the starting point of the express delivery point of the town center. And finally merging the multi-loop path planning and outputting a result.
Step 3.3: the iteration is started.
Step 3.4: and calculating the fitness value of each chromosome, and selecting a parent population.
Step 3.5: offspring are generated using the improved neighborhood search strategy. And (3) for the path planning sequence of each town center express point of each chromosome, carrying out complete overturn by using the probability po_GA each time, and accepting the solution if the adaptability becomes large after the path reversal. And (3) adopting a partial reversal strategy for each rural express point, namely randomly generating two numbers for each rural express point, and reversing the sequence between the two random numbers. This result is accepted whether or not the fitness is getting large.
Step 3.6: offspring were generated using genetic strategies (roulette selection, single crossover, single parental variation).
Step 3.7: and 2.5, combining the offspring results generated in the steps 2.5 and 2.6, and selecting individuals with high fitness to generate a new population.
Step 3.8: and judging whether the maximum iteration times are reached, if so, turning to the step 2.7, and if not, turning to the step 2.4.
Step 3.9: and outputting the optimal path planning result and related information.
The path planning result is shown in fig. 8.
Step 4, double-layer planning solution
The main task of the double-layer planning algorithm is to firstly determine the proper quantity of the express points, and then select the express points in the rural area and solve the path planning, and the specific flow is shown in figure 1. The double-layer planning algorithm flow is specifically as follows:
step 4.1: the initialization of the algorithm and the information reading are carried out, and the method is concretely as follows:
step 4.2: and calculating the total express demand and the average express point capacity according to the building area, calculating the minimum number of express points according to the total express demand and the average express point capacity, and setting the maximum number of express points.
Step 4.3: and carrying out express point selection and building allocation iteration by using an immune algorithm, and outputting a corresponding optimal express point selection and building allocation scheme.
Step 4.4: and calculating the times of vehicle distribution and the requirement of each express point according to the result of the immune algorithm, and carrying out vehicle path planning operation iteration by using a hybrid genetic algorithm.
Step 4.5: and recording address selection, distribution and path planning solving result information under different express points.
Step 4.6: judging whether the number of the selected express points reaches the maximum number of the selected express points, if so, turning to the step 3.7, and if not, turning to the step 3.3.
Step 4.7: and finding out the number of the corresponding express points when the balance degree is optimal.
Step 4.8: and (3) performing address selection operation of the express points according to the number of the express points, if the result fitness is higher than the fitness of the previous operation, turning to step 3.9, otherwise, continuing to turn to step 3.8 to perform iterative operation on the result obtained by the operation at the moment.
Step 4.9: and (5) carrying out path planning operation iteration by using a hybrid genetic algorithm and saving the result.
Step 4.10: and judging whether the iteration times are reached.
Step 4.11: and outputting final LAP and VRP schemes in the record.
The result of selecting the rural express points according to the embodiment is shown in fig. 9, and the result of path planning is shown in fig. 10.
Claims (1)
1. A rural end logistics site selection-path planning method based on double-layer planning is characterized by comprising the following specific steps:
firstly, preprocessing data; the method comprises the steps of obtaining alternative express point information, namely crawling rural interest point information POIs, eliminating the large class which does not serve as express point conditions from the POIs, removing POIs distributed by existing express points in the town center area and the vicinity, and taking the rest POIs as rural alternative express points; for logistics demand distribution information, adopting building distribution to simulate the characteristics of rural logistics demand multipoint distribution, and simulating different logistics demands by using a density grading strategy based on building distribution;
secondly, establishing an operation study model for site selection and path planning problems; acquiring path information among rural POIs, and searching related rural express logistics distribution information from the path information; due to the specificity of rural roads, small-sized vehicles are adopted for carrying out multiple delivery until delivery of all rural express points is completed; during distribution, a common distribution center is not established at this time, vehicles directly go to express delivery points of different logistics companies to collect express delivery, and then the vehicles are distributed to rural express delivery points in batches; the rural residents directly go nearby to part of public places selected as express points in rural villages to take the express;
the method comprises the following steps:
(1) Upper layer site selection model establishment
The overall objective function is as follows:
wherein C is s The total cost of the rural express collection points is calculated; c (C) b The total cost of residents in rural areas; m is an alternative express point, and M e M, set m= { m|m=1, 2, …, f }; j is building, J e J, set j= { j|j=1, 2, …, g }; gamma is the unit express operation cost; c is the fixed cost of constructing an express delivery point; alpha is the weight of the express delivery cost; θ is the direct proportionality coefficient of the actual capacity and cost of the express point; μ is the weight of the express points biasing toward crowd density; u (u) mj The coverage of the express delivery point m comprises a building j which is 1, otherwise, the coverage of the express delivery point m is 0; u (u) m The express point m is selected as the express point, and is 1, otherwise, is 0;
d mj distance from building j to express point m; wherein x is m 、y m The positions of the alternative express points converted in longitude and latitude are respectively x j 、y j The positions of the buildings converted in longitude and latitude are respectively; v is administrative village, and V e V, set v= { v|v=1, 2, …, h }; j (J) v Representing a building J within the coverage area of an administrative village v, and J v ＝{1，2，…，g v }，g v ≤g；M v Alternative express point i in administrative village v coverage range and M v ＝{1，2，…，f v }，f v ≤f；
The constraint conditions are as follows:
a convenience store not selected as an express delivery point cannot cover a building:
each building can only be assigned a coverage area belonging to one express point:
the number of express points selected needs to be within a certain number range:
r represents the maximum applicable capacity of the express point m, s j The actual express delivery of the building j is provided with capacity constraint for the express delivery of the coverage range of each express delivery point:
r is the furthest distance between the express point and the building, and the distance from the building to the express point has path length limitation:
each designated express delivery point needs to have coverage, at least more than one building:
0, 1 constraint of decision variables:
u m ∈{0，1}，m∈M
u mj ∈{0，1}，m∈M，j∈J
(2) Lower path planning model establishment
Objective function F d The following are provided:
where O represents the town-center express point, O e O, o= { o|o=1, 2, …, p }; i is the selected express point in rural area, and I e I, i= { i|i=1, 2, …, q }; r is the number of vehicle path plans, and R e R, r= { r|r=1, 2, …, s }; w (w) i The actual demand of the rural express delivery point i is obtained; t (T) i The time when the vehicle arrives at the rural express point i;the time when the vehicle arrives at the express point o of the town center in the r-th delivery; />For vehicle r-th from rural express delivery collection point i 1 Travel to rural express delivery collection point i 2 1, otherwise 0;express point o from town center for vehicle r time 1 Travel to rural express delivery collection point o 2 1, otherwise 0; />The vehicle is driven from the town center express delivery point o to the rural express delivery collection point i for the r time, and if not, the vehicle is 0; />The vehicle is driven from the express delivery point i to the town center rural express delivery collection point o for the r time, and if not, the vehicle is 0; />Selected express point i for rural area 1 Travel to rural express delivery point i 2 Is a distance of (2); />Center express point o 1 Travel to town center express point o 2 Is a distance of (2); d, d oi The distance from the town center express delivery point o to the rural express delivery collection point i is set; d, d io The distance from the rural express delivery collection point i to the town center express delivery point o is travelled; ρ is a time satisfaction weight;
selected express point i for rural area 1 Travel to another selected express point i 2 Time of (2); />Express point o for town center 1 Travel to town center express point o 2 Time of (2); t is t oi The time for the town center express delivery point o to travel to the rural express delivery collection point i; t is t io The time for the rural express delivery collection point i to travel to the town center express delivery point o; express point o at center of town 2 Waiting time when collecting express delivery +.>The method comprises the following steps:
for rural express delivery point i 2 Rural express delivery point i during delivery 2 Is to wait for a period of
Constraint conditions:
each rural express point can only be traversed by the vehicle once in total, including one entry and one exit:
each point can only be traversed once when the vehicle traverses the central express point of the town each time, including entering once and exiting once:
the load capacity constraint of the vehicle is q, and the sum of the express delivery of the vehicle leaving the town center to the rural express delivery point can not exceed the actual capacity constraint of the vehicle:
if the vehicle goes to rural areas, the vehicle must be delivered to more than one express delivery point:
0, 1 constraint of decision variables:
thirdly, designing a nested double-layer planning algorithm based on an immune-hybrid genetic algorithm to solve the model;
step 1: the immune algorithm flow is specifically as follows:
step 1.1: initializing parameters; the codes represent the labels of alternative express points and buildings in rural areas by natural numbers; the codes are divided into a plurality of rows, wherein the first row represents the number of the selected express points, and the balance is the codes distributed to the building; if the number of the express points in the first row is N, the operation is to select N express points; the (i+1) th row represents a building label collection obtained by distributing the selected (i) th express delivery point in the first row; the first row represents the length of the express point selection unchanged, the number of the row number codes allocated to the rest representing the building is unchanged, but the number of the representative buildings in each row of codes is variable;
step 1.2: generating an express point selection scheme of an initial solution; the method comprises the following steps: randomly extracting one point u from alternative express point set 1 The method comprises the steps of carrying out a first treatment on the surface of the For each express point x which is not found, calculatingCalculating the distance between x and the selected express point; with weighted probability distance +.>Selecting the next express point for probability address; repeating the first two steps until the number of the required express points is reached;
step 1.3: defining iteration times, namely starting iteration;
step 1.4: performing antibody diversity evaluation calculations including affinity between antibody and antigen Affinity of antibody->Antibody concentrationDesired reproduction Rate->Wherein P represents a penalty factor for the distance constraint penalty function, Q represents a penalty factor for the capacity constraint inverse function, R v，s The number of the v-th antibody and the s-th antibody is the same; len represents the number of bits representing each antibody; t is a set threshold; m is the total number of antibodies; alpha is a diversity evaluation parameter;
step 1.5: selecting an antibody from the structure as a memory cell; selecting by elite retention strategy, firstly storing the antibody with higher affinity with antigen into memory cells, and then storing the antibody with high desired proliferation rate of the rest antibodies into memory cells;
step 1.6: selecting a parent population by the same method as selecting memory cells, and performing immune selection, cloning, mutation and clone inhibition operation;
step 1.7: combining the memory cells with the new population of individuals to form a new structure;
step 1.8: judging whether the iteration times are reached, if not, turning to the step 1.4, and if so, turning to the step 1.9;
step 1.9: outputting information such as optimal express point selection, resident allocation schemes, optimal fitness values, objective functions and the like;
step 2: the flow of the hybrid genetic algorithm is specifically as follows:
step 2.1: reading data and setting parameters of a hybrid genetic algorithm; the coding mode is represented by a plurality of lines of natural number codes; the first line is a path planning sequence of the rural express points, the path planning of the town center express points consists of a second line and a plurality of codes with the same length, and each code represents the sequence of collecting the express in the town center each time; fitness is defined as
Step 2.2: the improved path shortest greedy strategy generates an initial solution for path planning; calculating the number of times of distribution according to the total express delivery and the maximum load of the vehicle; the greedy strategy of finding out the express points closest to the current point generates a closed path formed by the center express point of the inner ring town and the rural express points of the outer ring respectively; dividing the outer ring rural express points into a plurality of batches according to the limit of vehicle-mounted capacity; when the vehicle is in the town center, the vehicles are sequentially driven by the original generated town center traveling business problem TSP path, and the vehicles are driven to the outer ring rural express points after the completion; the rural express points needing to be distributed at the time are distributed, and after the distribution is finished, the rural express points are driven to the town center express point closest to the last rural express point to be distributed at the time; judging whether the rural express delivery points are distributed completely, if so, driving the vehicle to the starting point of the express delivery point of the town center; finally merging the multi-loop path planning and outputting a result;
step 2.3: starting iteration;
step 2.4: calculating the adaptation value of each chromosome, and selecting a parent population;
step 2.5: generating offspring using the improved neighborhood search strategy; for the path planning sequence of each town center express point of each chromosome, the probability po_GA is used for completely overturning each time, and if the adaptability of the path is increased after the path is reversed, the solution is accepted; a partial reversal strategy is adopted for each rural express point, namely, two numbers are randomly generated for each rural express point, and the sequence between the two random numbers is reversed; accept this result whether or not the fitness is getting large;
step 2.6: generating offspring using genetic strategies;
step 2.7: combining the offspring results generated in the steps 2.5 and 2.6 and selecting individuals with high fitness to generate a new population;
step 2.8: judging whether the maximum iteration times are reached, if so, turning to step 2.7, and if not, turning to step 2.4;
step 2.9: outputting an optimal path planning result and related information;
step 3: the double-layer planning algorithm flow is specifically as follows:
step 3.1: the initialization of the algorithm and the information reading are carried out, and the method is concretely as follows:
step 3.2: calculating the total express demand and the average capacity of express points according to the building area, calculating the minimum number of express points according to the total express demand and the average capacity of the express points, and setting the maximum number of express points;
step 3.3: performing express point selection and building allocation iteration by using an immune algorithm, and outputting a corresponding optimal express point selection and building allocation scheme;
step 3.4: calculating the number of times of vehicle distribution and the requirement of each express point according to the result of the immune algorithm, and carrying out vehicle path planning operation iteration by using a hybrid genetic algorithm;
step 3.5: recording address selection, distribution and path planning solving result information under different express points;
step 3.6: judging whether the number of the selected express points reaches the maximum number of the selected express points, if so, turning to the step 3.7, and if not, turning to the step 3.3;
step 3.7: finding out the number of express points corresponding to the optimal balance degree; by adoptingCalculating the balance degree of each individual;/>respectively representing the maximum value and the minimum value of the distance between the building and the express delivery point from the viewpoint of residents;representing the maximum and minimum of cost and efficiency issues from the point of view of the rural express delivery,where β is the cost weight of the latency; d (D) max 、D min Respectively representing the maximum value and the minimum value of path cost required to be paid when the common distribution is arranged from the express enterprise;
step 3.8: performing address selection operation of the express points according to the number of the express points, if the result fitness is higher than the fitness of the previous operation, turning to step 3.9, otherwise, continuing to turn to step 3.8 to perform iterative operation on the result obtained by the operation at the moment;
step 3.9: carrying out path planning operation iteration by using a hybrid genetic algorithm and storing a result;
step 3.10: judging whether the iteration times are reached;
step 3.11: and outputting the final address-allocation problem and the scheme of the vehicle path problem in the record.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310542905.0A CN116797126A (en) | 2023-05-15 | 2023-05-15 | Rural terminal logistics site selection-path planning method based on double-layer planning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310542905.0A CN116797126A (en) | 2023-05-15 | 2023-05-15 | Rural terminal logistics site selection-path planning method based on double-layer planning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116797126A true CN116797126A (en) | 2023-09-22 |
Family
ID=88045472
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310542905.0A Pending CN116797126A (en) | 2023-05-15 | 2023-05-15 | Rural terminal logistics site selection-path planning method based on double-layer planning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116797126A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117689417A (en) * | 2023-12-20 | 2024-03-12 | 国网湖北省电力有限公司物资公司 | Optimization method and system applied to site selection of logistics distribution center |
-
2023
- 2023-05-15 CN CN202310542905.0A patent/CN116797126A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117689417A (en) * | 2023-12-20 | 2024-03-12 | 国网湖北省电力有限公司物资公司 | Optimization method and system applied to site selection of logistics distribution center |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Collaboration and transportation resource sharing in multiple centers vehicle routing optimization with delivery and pickup | |
| CN110888908B (en) | Charging station/pile recommendation system and method capable of achieving deep learning continuously | |
| Bai et al. | A bi-objective model for location planning of electric vehicle charging stations with GPS trajectory data | |
| Shrivastava et al. | A model for development of optimized feeder routes and coordinated schedules—A genetic algorithms approach | |
| CN101436345B (en) | System for forecasting harbor district road traffic requirement based on TransCAD macroscopic artificial platform | |
| Wei et al. | Research on construction of a cloud platform for tourism information intelligent service based on blockchain technology | |
| KR20190020852A (en) | Methods and systems for transport capacity scheduling | |
| CN106570595A (en) | Subway station position selection method and system based on space big data | |
| Ma et al. | A model for the stop planning and timetables of customized buses | |
| CN110909434B (en) | Urban public transport trunk line network design method under low-carbon guidance | |
| Huang et al. | An innovative approach to solve the carsharing demand-supply imbalance problem under demand uncertainty | |
| Asghari et al. | Environmental and social implications of incorporating carpooling service on a customized bus system | |
| CN108256969A (en) | A kind of public bicycles lease point dispatcher-controlled territory division methods | |
| CN116797126A (en) | Rural terminal logistics site selection-path planning method based on double-layer planning | |
| Huang et al. | A GIS-based framework for bus network optimization using genetic algorithm | |
| Baušys et al. | The residence plot selection model for family house in Vilnius by neutrosophic WASPAS method | |
| Akin et al. | Multicriteria analysis of planned intercity bus terminals in the metropolitan city of Istanbul, Turkey | |
| Kamel et al. | A modelling platform for optimizing time-dependent transit fares in large-scale multimodal networks | |
| Chen et al. | Reposition optimization in free-floating bike-sharing system: A case study in Shenzhen City | |
| CN111178580A (en) | Supermarket site selection method based on improved BP neural network | |
| Huang et al. | A bi-level model and hybrid heuristic algorithm for the optimal location of prefabricated building industrial park | |
| Tian et al. | A Long‐Term Shared Autonomous Vehicle System Design Problem considering Relocation and Pricing | |
| Ren et al. | Multi-objective optimization for multi-depot heterogeneous first-mile transportation system considering requests’ preference ranks for pick-up stops | |
| Qi et al. | Tourism Route Selection Model for Tourism Sustainable Development Based on Improved Genetic Algorithm | |
| Fan et al. | How does upzoning impact land use and transport: a case study of Seattle |
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 |