CN115689412A - Multi-period freight pricing and logistics network planning method - Google Patents
Multi-period freight pricing and logistics network planning method Download PDFInfo
- Publication number
- CN115689412A CN115689412A CN202211038505.8A CN202211038505A CN115689412A CN 115689412 A CN115689412 A CN 115689412A CN 202211038505 A CN202211038505 A CN 202211038505A CN 115689412 A CN115689412 A CN 115689412A
- Authority
- CN
- China
- Prior art keywords
- logistics
- algorithm
- node
- freight
- population
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a multi-period freight pricing and logistics network planning method, and relates to the technical field of logistics pricing and network planning. The method comprises the steps of obtaining related service information in a logistics planning database; preprocessing the service information to obtain model input data; determining a multi-period freight pricing and logistics network planning model according to the model input data; obtaining a hybrid element heuristic algorithm according to the particle swarm algorithm and the differential evolution algorithm; solving a multi-period freight pricing and logistics network planning model by using a mixed element heuristic algorithm to obtain a logistics planning scheme; and determining a final freight pricing and logistics network planning implementation scheme according to the logistics planning scheme. The invention can decide the freight price according to the behaviors of different types of customers in the logistics network planning, and also considers the selection of logistics carriers, thereby improving the practicability of the logistics network planning; and the planning efficiency can be improved on the premise of ensuring the solving quality, and the profit of enterprises is improved.
Description
Technical Field
The invention relates to the technical field of logistics pricing and network planning, in particular to a multi-period freight pricing and logistics network planning method.
Background
In today's business environment, revenue management issues have become a new dimension in logistics network planning. The freight rate not only determines the unit income of each commodity, but also influences the requirements of each client area, and can change the required logistics network planning such as logistics facilities, transportation routes and the like. However, in the real world, in addition to the logistics price affecting the customer demand, the customer satisfaction with the logistics service can also have a large impact on the demand. Different types of customers do not evaluate the logistics service level the same, so it is more realistic to make freight prices according to different types of customer behaviors. The existing logistics network planning usually only considers the selection of lines, and does not consider the selection of different logistics carriers on the same line, which results in the improvement of the cost of the whole logistics network and the reduction of the distribution efficiency. Therefore, a solution is needed to decide freight prices according to different types of customer behaviors in logistics network planning and to take account of the selection of logistics carriers. The logistics network planning problem under the complex structure generally has a mathematical model with nonlinear characteristics and exponentially increased complexity, and the existing algorithm is low in solving precision or consumes a large amount of solving time and computing resources, so that an algorithm capable of obtaining high precision in a short time is needed.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a multi-period freight pricing and logistics network planning method aiming at the defects of the prior art, the freight price can be decided according to different types of customer behaviors in the logistics network planning, the selection of logistics carriers is considered, and the practicability of the logistics network planning is improved.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows:
a multi-cycle freight pricing and logistics network planning method comprises the following steps:
acquiring related service information in a logistics planning database; the service information comprises: market potential freight price information, market potential customer node information, supply node information, transfer center node information, line information and logistics carrier information;
preprocessing the service information to obtain model input data;
determining a multi-period freight pricing and logistics network planning model according to the model input data;
obtaining a mixed element heuristic algorithm according to the particle swarm algorithm and the differential evolution algorithm;
solving a multi-period freight pricing and logistics network planning model by using a mixed element heuristic algorithm to obtain a logistics planning scheme; the logistics planning scheme comprises a freight rate pricing plan, a transfer center node plan, a line plan, a logistics carrier plan and a freight volume plan;
and determining a final freight pricing and logistics network planning implementation scheme according to the logistics planning scheme.
Adopt the produced beneficial effect of above-mentioned technical scheme to lie in: the multi-period freight pricing and logistics network planning method provided by the invention simultaneously considers different types of customer behaviors, logistics freight prices and logistics carriers in the multi-period logistics network planning, establishes a multi-period freight pricing and logistics network planning model, designs a mixed element heuristic algorithm based on a particle swarm algorithm and a differential evolution algorithm, improves planning efficiency on the premise of ensuring solving quality, and obtains multi-period network planning solutions of the complex network, such as freight pricing, transit center nodes, lines, logistics carriers, freight volume plans and the like. Compared with a basic network planning scheme, the freight price is decided according to different types of customer behaviors in the logistics network planning, the selection of logistics carriers is considered, the practicability of the logistics network planning is improved, the actual requirements are met, the planning efficiency can be improved on the premise of ensuring the solving quality, and the profit of enterprises is improved.
Drawings
Fig. 1 is a flowchart of a multi-cycle freight pricing and logistics network planning method provided in an embodiment of the present invention;
FIG. 2 is a diagram of a relationship between freight pricing and demand provided by an embodiment of the present invention;
fig. 3 is a schematic diagram of a multi-cycle freight pricing and logistics network planning model construction method according to an embodiment of the invention;
fig. 4 is a schematic diagram of a design process of a multi-cycle freight pricing and logistics network planning solving algorithm provided by the embodiment of the invention;
fig. 5 is a schematic diagram of a logistics network planning scheme provided by an embodiment of the invention;
fig. 6 is a flowchart of multi-cycle freight pricing and solution of logistics network planning according to an embodiment of the present invention.
Detailed Description
The following detailed description of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
As shown in fig. 1, the multi-cycle freight pricing and logistics network planning method provided by the present embodiment is as follows.
Step 1: and acquiring related service information in the logistics planning database. The service information includes: market potential freight price information, market potential customer node information, supply node information, transit center node information, route information, and logistics carrier information.
Step 2: and preprocessing the service information to obtain model input data. The method specifically comprises the following steps:
preprocessing the service information according to a period to obtain a period set; preprocessing the service information according to price grades to obtain a price grade set; preprocessing the service information according to a supply node to obtain a supply node set; preprocessing the service information according to a transit center node to obtain a transit center node set; preprocessing the service information according to a logistics carrier to obtain a logistics carrier set; preprocessing the service information according to client nodes to obtain a client node set; determining model input data according to the period set, the price level set, the supply node set, the transit center node set, the logistics carrier set and the customer node set.
As shown in fig. 2, different freight price levels correspond to different demands, different supply nodes have different supply capacities, different transit center nodes have different processing capacities and processing fees, and different logistics carriers have different transportation capacities and transportation fees.
And 3, step 3: and determining a multi-period freight pricing and logistics network planning model according to the model input data.
In this embodiment, the multi-period freight pricing and logistics network planning model is an objective function with the maximum profit, and restricts the capacities of the supply node, the transit center node and the logistics transport provider and the logistics network structure, as shown in fig. 3, the method for constructing the multi-period freight pricing and logistics network planning model of this embodiment includes the following steps:
step 301: acquiring the model input data;
step 302: determining a plurality of decision variables from the model input data.
In this embodiment, the decision variables include: logistics transport merchants among different nodes of the logistics network in different periods, construction conditions of transit center nodes in different periods, logistics freight price grades in different periods and commodity transportation volume among different nodes in different periods.
Decision variable 1:
in this embodiment, the decision variable 1 "logistics carrier between any two nodes of supply node and transit center node" passes
Is shown to be
u∈U,k∈K su T belongs to T, wherein T is a set of periods; s represents a collection of provisioning nodes; u represents a set of transit center nodes; k su Representing a collection of logistics forwarders between supply node s and transit centre node u.
The kth logistics carrier is selected between the supply node s and the transit centre node u in the period t, then
If not, then the mobile terminal can be switched to the normal mode,
decision variable 2:
in this embodiment, the decision variable 2 "transit center node and any two logistics carrier between the customer nodes" passes
Is shown to be
d∈D,k∈K ud T belongs to T, wherein T is a set of periods; u represents a set of transit center nodes; d represents a set of customer nodes; k ud Representing a collection of logistics carriers between transit center node u and customer node d.
During a period t, the kth logistics carrier is selected between the transit center node u and the customer node d, then
Otherwise
Decision variable 3:
in this embodiment, the decision variable 3 "transit center node" passes
Is shown to be
t∈T。
Transit center node u is selected at period t, then
Otherwise
Decision variable 4:
in this embodiment, the decision variable 4 "logistic freight price rating" passes
Is shown to be
L belongs to L, T belongs to T, wherein L is a set of price levels of the logistics freight.
At period tth the logistics freight price level is selected, then
Otherwise
Decision variable 5:
in this embodiment, the decision variable 5 "commodity traffic volume between any two nodes of the supply node and the transit center node" passes
Show that
u∈U,k∈K su T ∈ T, wherein
Decision variable 6:
in an embodiment, the decision variable 6 "commodity traffic volume between any two nodes of the transit center node and the customer node" passes
Show that
d∈D,k∈K ud T ∈ T, wherein
Step 303, constructing an objective function with the maximum profit as an objective.
In this embodiment, an objective function is constructed with a profit maximum objective based on decision variables. Specifically, total income is determined according to the logistics freight price and the actual commodity transportation volume, the total cost is determined according to the construction cost of the transit center node, the cooperation cost of the logistics transport provider, the processing cost of the transit center node, the transportation cost of the logistics transport provider and the shortage cost of the commodity, and finally the total profit is obtained:
wherein, the first and the second end of the pipe are connected with each other,
in order to be the price of the freight rate,
the construction cost is fixed for the transfer center,
for determining whether the transit center was selected before the current cycle,
the cooperation charge is fixed for the kth logistics transport provider between the supply node and the transit center node,
used for judging whether the kth logistics transport provider between the supply node and the transfer center node is selected before the current period,
the cooperation charge is fixed for the kth logistics carrier between the transit center node and the client node,
used for judging whether the kth logistics carrier between the transfer center node and the customer node is selected before the current period,
for the unit transportation cost of the kth logistics transport provider between the supply node and the transit center node,
for the unit transportation cost of the kth logistics transport provider between the transit center node and the customer node,
the unit treatment cost of the transfer center is saved,
the short-cut amount of the commodity is obtained,
is the unit cost of out-of-stock.
Step 304, determining a plurality of constraints of the objective function.
Constraint 1:
in this embodiment, the level of logistics service per cycle depends on the ratio of the actual delivery volume to the customer demand volume. Constraint 1 includes:
wherein the content of the first and second substances,
is the logistics service level.
Constraint 2:
in this embodiment, the customer demand per cycle depends not only on the price, but also on the customer satisfaction of the previous cycle. At the first period, the actual demand of the customer is equal to the market share of the customer; and from the second period, the actual demand of the current period client is determined by the freight pricing of the current period and the client satisfaction degree of the previous period. Since real world customers are often irrational, most people are not symmetrical in their sensitivity to loss and gain, and the pain facing loss greatly outweighs the pleasure facing gain. Therefore, the present embodiment describes attitudes of different types of customers with respect to the logistics service level, i.e., customer satisfaction, through the foreground theoretical cost function, so as to influence actual needs of the customers. The constraint 2 includes:
wherein the content of the first and second substances,
the level of service required for the customer is,
in order to be a market share for the customer,
for the actual demand of the customer, α d Risk attitude coefficient, beta, for different customers facing revenue d In the face of risk attitude coefficients at loss, λ d The loss aversion coefficient when facing loss.
Constraint 3:
in the embodiment, for the multi-period freight pricing and logistics network planning model, if a transit center node is constructed in a certain period, the construction cost of the transit center node does not need to be considered in the later period. The constraint 3 includes:
constraint 4:
in the embodiment, for the multi-period freight pricing and logistics network planning model, if cooperation with a logistics provider is achieved on a certain line between a certain period supply node and a transit center node, the cooperation cost of the logistics provider does not need to be considered on the line in the later period. The constraint 4 includes:
constraint 5:
in the embodiment, for the multi-cycle freight pricing and logistics network planning model, if cooperation with a logistics provider is achieved on a certain line between a certain cycle transit center node and a client node, the cooperation cost of the logistics provider does not need to be considered on the line in the later cycle. The constraint 5 includes:
constraint condition 6:
in the embodiment, the transportation capacity limit of the logistics provider between the supply node and the transfer center node is considered, and the quantity of the transported commodities cannot exceed the upper transportation capacity limit of the logistics provider. The constraint 6 includes:
wherein the content of the first and second substances,
for the kth logistics transportation between the supply node and the transit center nodeTransport capacity of the shipper.
Constraint condition 7:
in the embodiment, the transportation capacity limit of the logistics provider between the transit center node and the customer node is considered, and the number of transported commodities cannot exceed the upper limit of the transportation capacity of the logistics provider. The constraint 7 includes:
wherein, the first and the second end of the pipe are connected with each other,
the transportation capacity of the kth logistics transportation business between the central node and the customer node is transferred.
Constraint condition 8:
in this embodiment, the throughput of processing the commodity cannot exceed the upper limit of the processing capacity of the transit center node in consideration of the limit of the processing capacity of the transit center node. The constraint 8 includes:
wherein the content of the first and second substances,
is the handling capacity of the transit center.
Constraint 9:
in the present embodiment, the total amount of the commodity flowing out from the supply node may not exceed the maximum supply amount thereof in consideration of the supply capacity limit of the supply node. The constraint conditions 9 include:
wherein the content of the first and second substances,
is the provisioning capability of the provisioning node.
Constraint 10:
in this embodiment, the traffic balance limitation of the transit center node is considered, which means that the total amount of commodities entering a certain transit center node is equal to the total amount of commodities from the transit center node, that is, no commodity is retained by the transit center node. The constraint 10 includes:
constraint 11:
in this embodiment, it is considered that if a certain transit center node is not selected, the logistics carrier from the supply node to the transit center node cannot be selected. The constraint 11 includes:
constraint 12:
in this embodiment, it is considered that if a certain transit center node is not selected, the logistics transportation provider from the transit center node to the customer node cannot be selected. The constraints 12 include:
constraint condition 13:
in this embodiment, it is considered that if a certain transit center node is selected, at least one logistics carrier from the supply node to the transit center node provides services. The constraint conditions 13 include:
constraint 14:
in this embodiment, it is considered that if a certain transit center node is selected, at least one logistics carrier from the transit center node to the customer node provides services. The constraints 14 include:
constraint 15:
in the embodiment, the constraint of the customer demand node is considered, and the total amount of the actually arrived commodities cannot exceed the customer demand amount, namely the customer demand amount in the current period is equal to the sum of the shortage amount in the current period and the actually arrived amount in the current period. The constraints 15 include:
constraint 16:
in this embodiment, it is considered that only one logistics freight price level can be selected in each period. The constraints 16 include:
and 305, constructing a multi-period freight pricing and logistics network planning model according to the decision variables in the step 302, the objective function in the step 303 and the constraint conditions in the step 304.
And 4, step 4: and obtaining a mixed element heuristic algorithm according to the particle swarm algorithm and the differential evolution algorithm.
The design process of the multi-cycle freight pricing and logistics network planning solving algorithm is shown in fig. 4, and specifically includes:
step 401: the encoding process is that a one-dimensional integer encoding mode is adopted to solve the model, and the encoding process is divided into two parts;
a first part: firstly, numbering each node in the potential logistics network according to the sequence of the supply node, the transit center node and the client node, and converting the actual distribution of each node into a network withThe numbered potential logistics network planning diagram is converted from the figure 5 (a) to the figure 5 (b). Then, an adjacency matrix is established according to the number of logistics carriers between two nodes in the logistics network diagram, and the integer coding range of the particles at a certain position is determined according to the number of selectable logistics carriers between two nodes in the adjacency matrix. Assuming that there are n selectable logistics carriers between two nodes, for the particle swarm algorithm, the particle position encoding range is [0,2 ] n -1]The particle velocity encoding range is [ - (2) n -1),2 n -1](ii) a For the differential evolution algorithm, the particle encoding range is [0,2 ] n -1]. The encoding length of the first part of the particle depends on the number of node pairs in the potential logistics network with selectable logistics carriers, i.e. the number of non-0 elements in the adjacency matrix, set to z.
A second part: freight price ratings for different customer nodes are encoded. Assuming that the number of client nodes in a potential logistics network is d, the freight price grade is l, and for the particle swarm algorithm, the particle position coding range is [0, l-1], and the particle speed coding range is [ - (l-1), l-1]; for the differential evolution algorithm, the particle encoding range is [0, l-1]. The encoding length of the second part of the particle depends on the number of client nodes in the potential logistics network, i.e. d.
In summary, for the multi-cycle freight pricing and logistics network planning model, the encoding length of the particle in each cycle is (z + d), and assuming T cycles in total, the total encoding length of the particle is T x (z + d).
Step 402: a decoding process, which is divided into two parts according to the encoding process, and the specific decoding process comprises the following steps:
a first part: the first part in the coding process is decoded in a binary mode, and the selection condition of the logistics transport provider between any two nodes can be obtained only by converting the integer code into the binary 0-1. Specifically, the logistics carriers between two nodes (i, j) are numbered. Then, converting the integer code into binary 0-1 code, judging the selection condition of the logistics transport quotient between two nodes according to the sequence from low order to high order, if the k-th order is 1, indicating the kth order between two nodes (i, j)k logistics carriers are selected, i.e. decision variables
The value of (b) is 1; if the k bit is 0, it means that the k logistics carrier between two nodes (i, j) is not selected, i.e. the decision variable
The value of (d) is 0. Therefore, the selection condition of the decision variable transit center node can be determined.
Assume that there are three alternative logistics forwarders between node 1 and node 3, i.e. n =3. The binary 0-1 decoding corresponding to the integer code and the selected logistics carrier conditions are shown in table 1.
Table 1 decoding first partial example
A second part: decoding a second part of the encoding process, the particle integer encoding range being [0, l-1]]Then the freight price range of the client node in the potential logistics network is [1, l ]]. When the encoded value of the particle is m (m ≦ l), then the decision variable is made
Has a value of 1, i.e. the freight price level m is selected by the client node j e D within the period T e T. Assuming that the highest level of freight price, l =4, the encoding range of the particle at each customer node is [0,3]Then the freight price level range of the client node in the potential logistics network is [1,4 ]]The specific correspondence is shown in table 2.
Table 2 decoding second partial example
Step 403: and in the initialization process, a random initialization mode in an encoding range is adopted. For the particle swarm algorithm, the initialization comprises the random initialization of two factors of speed and position; for the differential evolution algorithm, each dimension of each particle is randomly initialized.
Step 404: the updating process of the particle swarm algorithm and the differential evolution algorithm is slightly different, and specifically comprises the following steps:
for the particle swarm optimization, assume that there are m particles in the population, and the dimension of the solution space is N, x i 、v i 、p i And p g Respectively representing the current position, the current speed, the best solution of history and the best solution of the whole situation that all the particles in the population pass through, wherein the updating formula of the speed of the particles in the population is as follows:
the update formula of the particle position in the population is as follows:
wherein the inertia weight is w, the learning factor is c 1 、c 2 Are all non-negative real numbers; xi, eta are equal to U0, 1]. The method specifically comprises the following steps:
TT represents the current iteration number of the algorithm, and TT represents the maximum iteration number of the algorithm.
The particles in the population are rounded first after updating the speed and position, but may exceed their own range. The treatment is carried out by the following method:
if the value of a position code of a certain bit in the particle exceeds the upper code bound, the upper code bound value is taken as the position code value of the bit; and if the value of the position code of a certain bit in the particle is lower than the lower code bound, taking the lower code bound value as the position code value of the bit. If the value of a certain bit velocity code in the particle exceeds the value range, the same method is adopted for processing.
For a differential evolution algorithm, an updating process comprises mutation operation, crossover operation and a selection strategy, a DE/best/2/bin method is adopted, wherein best represents an optimal value as a base number, 2 represents the number of differential vectors, and bin represents a binomial crossover strategy, and the method specifically comprises the following steps:
each individual in the population generates a corresponding variant individual according to the following formula:
V i (tt+1)＝X best (tt)+F(tt)[(X r1 (tt)-X r2 (tt))+(X r3 (tt)-X r4 (tt))]
wherein i represents the index number of the target individual, r 1 、r 2 、r 3 、r 4 Respectively represent index numbers of different individuals in the population, and i and r 1 、r 2 、r 3 、r 4 Are different from each other, X best (tt) represents the best individuals in the tt generation population,
the difference vector is generated by randomly selecting 4 different individuals and then for X best (tt) carrying out mutation operation. Since integer coding is adopted in the coding process, the result is rounded after the calculation by the above formula. Since the rounding of the variant may be done beyond the coding range, each individual in the population is treated as follows:
wherein, V ij (tt + 1) represents the value of the ith individual in the (tt + 1) th generation population in the jth dimension,
representing the maximum value of the individual in the j-th dimension.
Then the current individual X i (tt) and variant individuals V i (tt + 1) is subjected to a crossover operation, thereby producingNew test individual U i (tt)：
Where j denotes the j dimension of the ith individual, rand is a random number between [0,1], CR is the probability of crossover, and CR ∈ [0,1].
Finally, a selection strategy of a greedy strategy is adopted to carry out U on the test individuals i (tt) and Current subject X i (tt) evaluation was performed, and the more excellent individuals between the two were selected as the individuals of the offspring.
Step 405: after freight pricing and basic structure of the logistics network are determined, the minimum-cost maximum flow algorithm needs to distribute transportation volume for each logistics carrier in the logistics network according to customer requirements and limiting factors such as capacity and cost in the logistics network. The embodiment evaluates the selected network by using a minimum cost maximum flow algorithm and calculates the total cost of the network. Because the minimum cost maximum flow algorithm has certain limitation when solving the network planning problem, the method is suitable for solving the network problem which only has a single source point and a single sink point and has no capacity limitation of an intermediate node. Therefore, the potential logistics network is converted, the original network is converted into a single source point and a single sink point, and the intermediate node has no network problem of capability limitation. The method specifically comprises the following steps:
and adding a virtual source point as a starting point of the whole network and adding a virtual sink point as a termination point of the whole network. The capacity of virtual arcs from the virtual source point to the supply node and from the client node to the virtual sink is set to be infinite, and the cooperation cost and the unit transportation cost are both 0. In addition, because the transit center node also has the processing capacity constraint and the cost attribute, each transit center node is divided into two points, and the processing capacity constraint and the cost attribute of the transit center node are transferred to a virtual arc between the two nodes. The processing capacity constraint and the unit processing cost attribute of the transit center node are converted into the transport capacity constraint and the unit transport cost attribute of the corresponding virtual transport arc, and the construction cost of the transit center node is converted into the construction cost of the corresponding virtual transport arc.
Step 406: in the solution repairing process, since the obtained solution may not meet the practical constraint condition limit, that is, a network disconnection situation may occur, it is necessary to repair the infeasible solution.
The network disconnection is divided into three cases:
for the condition that the client node is disconnected with the network, one or more arcs are randomly added in the network to be connected with the client node, the position of the disconnected node pair in the particle is found, and finally an integer is randomly selected in a value range to replace a 0 value of an original position;
for the situation that the supply node is not communicated with the network, one or more arcs are randomly added in the network to be connected with the supply node, and the other processing methods are the same as the situations;
for the situation that the transit center node is not communicated with the network, if the situation that an arc enters and an arc does not exit exists, one or more arcs are randomly added in the network from the transit center node; if the arc is not in the arc-out condition, one or more arcs reaching the transit center node are randomly added in the network from the supply node, and the rest processing methods are the same as the above conditions.
New infeasible solutions may be generated in the repair process, so that the logistics network needs to be repaired for multiple times, and the logistics network needs to be detected before and after each repair, so that the commodities cannot smoothly arrive at the client side until the whole network is completely communicated.
Step 407: and evaluating the solution, wherein the maximum value of the total profit in the logistics network is used as an adaptive value function, namely an objective function. It is intended that of all particles that satisfy the constraints, the particle that maximizes the total profit in the network can be found. The particles are the obtained target logistics network.
Step 408: and an algorithm selection mechanism, wherein in order to improve planning efficiency on the premise of ensuring the solving quality, an optimal algorithm selection mechanism is designed, and which algorithm is used in each iteration process is determined according to the variation condition of the individual adaptation value in the population and the variation condition of the global optimal value of the population. The method specifically comprises the following steps:
firstly, performing particle swarm algorithm operation on an initial randomly generated population X (0) according to the change situation of the individual adaptive value in the population, wherein the individuals in the original population are X (0) = (X) 1 (0),x 2 (0),...,x NP (0) And the updated population is u = (u) in the updated population 1 ,u 2 ,...,u NP ) Introducing a new variable lbp in the process of updating the optimal value and the optimal solution of the current individual, wherein the lbp is the number of the individuals in the original population X (0) substituted in the updated population u; the differential evolution algorithm operation is also carried out on the initial randomly generated population X (0), and the individuals in the population after the updating are
Introducing a new variable lbd in the process of updating the current individual optimal value and the optimal solution, wherein the lbd is the updated population u * The number of individuals in the substitute original population X (0); the variation of the individual fitness value in the population is represented by the variable lb, where
Introducing a variable gb in the global search process according to the global optimum value change condition of the population, and if a particle swarm algorithm is used in the tt-th generation iteration process and the current global optimum value is not inferior to the previous generation global optimum value, gb =1; otherwise gb =2. If a differential evolution algorithm is used in the tth generation iteration process, and the current global optimal value is not inferior to the previous generation global optimal value, gb =2; else gb =1.
And an algorithm selection mechanism determines which algorithm is used in the next iteration process according to the values of each generation of lb and gb. When lb is larger than or equal to 0.5 and gb =1, it indicates that the number of individual adaptive values updated in the population after the operation of the particle swarm algorithm is larger, and the global optimal solution is updated compared with the previous generation, and the next generation uses the particle swarm algorithm. When lb is less than 0.5 and gb =2, it indicates that the number of individual fitness value updates in the population after the differential evolution algorithm operation is more, and the global optimal solution is updated compared with the previous generation, and the next generation uses the differential evolution algorithm. When lb is more than or equal to 0.5 and gb =2, it is stated that the number of individual fitness value updates in the population after the operation of the particle swarm algorithm is more, but the global optimal solution is not updated, a random number rand is generated according to the roulette rule, and if rand is less than or equal to lb, the next generation uses a differential evolution algorithm; if rand > lb, the next generation uses the particle swarm algorithm. When lb is less than 0.5 and gb =1, it indicates that the number of individual fitness value updates in the population after the operation of the differential evolution algorithm is more, but the global optimal solution is not updated, and according to the roulette rule, a random number rand is generated, and if rand is less than lb, the next generation uses the differential evolution algorithm; if rand is larger than or equal to lb, the next generation uses a particle swarm algorithm. The specific selection is shown in table 3:
TABLE 3 Algorithm selection mechanism
| Individual adaptive value judgment condition | Global optimum judgment condition | Judgment conditions for roulette | Algorithm used by next generation |
| lb≥0.5 | gb＝1 | - | Particle swarm algorithm |
| lb≥0.5 | gb＝2 | rand≤lb | Differential evolution algorithm |
| lb≥0.5 | gb＝2 | rand＞lb | Particle swarm algorithm |
| lb＜0.5 | gb＝2 | - | Differential evolution algorithm |
| lb＜0.5 | gb＝1 | rand＜lb | Differential evolution algorithm |
| lb＜0.5 | gb＝1 | rand≥lb | Particle swarm algorithm |
Step 409: and (5) ending the algorithm according to a condition, and ending the algorithm if the maximum iteration times are reached.
And 5: and solving the multi-period freight pricing and logistics network planning model to obtain a logistics planning scheme.
The specific flow of multi-cycle freight pricing and solution of logistics network planning is shown in fig. 6, and includes:
step 501: initializing parameters in a population scale NP, a dimension D, a maximum iteration number NG, an initial population X, a particle swarm algorithm and a differential evolution algorithm.
Step 502: setting initial iteration times, namely tt =0, simultaneously solving a multi-cycle freight pricing and logistics network planning model by using a particle swarm algorithm and a differential evolution algorithm according to the steps 401-407, calculating initial values of lb and gb, and selecting an algorithm used for the first iteration according to an algorithm selection mechanism of the step 408.
Step 503: if the particle swarm algorithm is judged to be selected, solving a multi-period freight pricing and logistics network planning model by using the particle swarm algorithm according to the steps 401 to 407; if the differential evolution algorithm is judged to be selected, the differential evolution algorithm is used for solving the multi-period freight pricing and logistics network planning model according to the steps 401-407.
Step 504: updating the values of lb and gb according to the above solution results, and selecting the algorithm to be used in the next iteration according to the algorithm selection mechanism in step 408.
Step 505: and updating the current individual optimal value, the current optimal solution, the current global optimal value and the current global optimal solution.
Step 506: judging whether a termination condition is met, and if the termination condition is met, ending the solution; if not, return to step 503.
And obtaining a logistics planning scheme according to the solving algorithm process.
Step 6: determining a final freight pricing and logistics network planning implementation according to the logistics planning scheme, as shown in fig. 5 (c).
In the embodiment, different types of customer behaviors, logistics freight prices and logistics carriers are considered in the multi-cycle logistics network planning, a multi-cycle freight pricing and logistics network planning model is established, a mixed meta-heuristic algorithm is designed based on a particle swarm algorithm and a differential evolution algorithm, the planning efficiency is improved on the premise of ensuring the solving quality, and multi-cycle network planning solutions such as freight pricing, transit center nodes, lines, logistics carriers and freight volume plans of the complex network are obtained. The scheme can formulate the logistics freight price according to different types of customer behaviors in the multi-period logistics network planning, and considers the selection of logistics carriers, thereby improving the practicability of the logistics network planning, meeting the actual requirements and improving the profits of enterprises.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.
Claims (8)
1. A multi-period freight pricing and logistics network planning method is characterized in that: the method comprises the following steps:
acquiring related service information in a logistics planning database;
preprocessing the service information to obtain model input data;
determining a multi-period freight pricing and logistics network planning model according to the model input data;
obtaining a hybrid element heuristic algorithm according to the particle swarm algorithm and the differential evolution algorithm;
solving a multi-period freight pricing and logistics network planning model by using a mixed element heuristic algorithm to obtain a logistics planning scheme;
and determining a final freight pricing and logistics network planning implementation scheme according to the logistics planning scheme.
2. The multi-cycle freight pricing and logistics network planning method of claim 1, wherein: the service information includes: market potential freight price information, market potential customer node information, supply node information, transit center node information, route information, and logistics carrier information.
3. The multi-cycle freight pricing and logistics network planning method of claim 2, wherein: preprocessing the service information to obtain model input data, which specifically comprises the following steps:
preprocessing the service information according to a period to obtain a period set; preprocessing the service information according to price grades to obtain a price grade set; preprocessing the service information according to a supply node to obtain a supply node set; preprocessing the service information according to a transit center node to obtain a transit center node set; preprocessing the service information according to a logistics carrier to obtain a logistics carrier set; preprocessing the service information according to client nodes to obtain a client node set;
determining model input data according to the period set, the price level set, the supply node set, the transit center node set, the logistics carrier set and the customer node set.
4. The multi-cycle freight pricing and logistics network planning method of claim 3, wherein: the multi-period freight pricing and logistics network planning model takes the maximum profit as an objective function, and restricts the capacity of a supply node, a transit center node and a logistics carrier and the structure of a logistics network;
determining a decision variable, and constructing an objective function by taking the maximum profit as an objective; the total income is determined according to the logistics freight price and the actual commodity transportation volume, the total cost is determined according to the construction cost of the transfer center node, the cooperation cost of the logistics transport provider, the processing cost of the transfer center node, the transportation cost of the logistics transport provider and the shortage cost of the commodities, and finally the total profit is obtained, namely the objective function, the expression of which is as follows:
wherein T is a period set, L is a price level set, S is a supply node set, U is a transfer center node set, D is a client node set, and K su Is a logistics carrier set from a supply node to a transit center node, K ud Is a logistics carrier set between a transit center node and a client node,
in order to be the price of the freight rate,
to divert the traffic for the kth logistics carrier between the central node to the customer node,
for the case of selection of a freight price level,
the construction cost is fixed for the transfer center,
in order to select the case of the transit center,
for determining whether a transit center was selected before the current cycle,
the cooperation cost is fixed for the kth logistics transport business between the supply node and the transit center node,
for the selection of the kth logistics carrier from the supply node to the transit center node,
used for judging whether the kth logistics transport provider between the supply node and the transfer center node is selected before the current period,
the cooperation charge is fixed for the kth logistics carrier between the transit center node and the client node,
to turn toThe selection of the k-th logistics transportation business from the transportation center node to the customer node,
used for judging whether the kth logistics carrier between the transfer center node and the customer node is selected before the current period,
for the unit transportation cost of the kth logistics transport provider between the supply node and the transit center node,
for the traffic of the kth logistics carrier between the supply node and the transit center node,
for the unit transportation cost of the kth logistics transport provider between the transit center node and the customer node,
in order to transfer the unit treatment cost of the center,
the short-cut amount of the commodity is obtained,
cost per unit of out-of-stock;
the expression of the constraint condition is as follows:
wherein, the first and the second end of the pipe are connected with each other,
in order to meet the level of the logistics service,
the level of service required for the customer is,
in order to be a market share for the customer,
is the actual demand of the customer, alpha d Risk attitude coefficient, beta, for different customers facing revenue d In the face of risk attitude coefficients at loss, λ d In the face of the loss aversion coefficient at the time of loss,
for the transport capacity of the kth logistics carrier between the supply node and the transit center node,
to divert the transport capacity of the kth logistics carrier between the central node to the customer node,
in order to be able to handle the capacity of the transit centre,
is the provisioning capability of the provisioning node.
5. The multi-cycle freight pricing and logistics network planning method of claim 4, wherein: the hybrid meta-heuristic algorithm specifically comprises the following steps:
solving the model by adopting a one-dimensional integer coding mode, and dividing a coding process into two parts;
dividing the decoding process into two parts according to the encoding process, wherein the first part corresponds to logistics network planning, and the second part corresponds to freight pricing;
a random initialization mode is adopted in the population initialization process;
updating the particles in the population according to the updating formulas of the particle swarm algorithm and the differential evolution algorithm;
according to the encoding, decoding, initializing and updating processes, freight rate pricing and a basic structure of a logistics network can be determined, and the problem of multi-period freight rate pricing and logistics network planning is converted into a problem of distributing transportation volume to each logistics carrier in the logistics network;
solving a new problem according to a minimum cost maximum flow algorithm;
repairing the infeasible solution by using a repairing strategy according to the restriction condition limit existing in reality;
taking the maximum value of the total profit in the logistics network, namely an objective function, as an adaptive value function;
determining an algorithm selection mechanism according to the variation condition of the individual adaptive value in the population and the variation condition of the global optimal value of the population to decide which algorithm is used in each iteration process, namely which algorithm is used for solving the multi-period freight pricing and logistics network planning model;
and taking the maximum iteration number as a termination condition of the algorithm, namely finishing the algorithm when the maximum iteration number is reached.
6. The multi-cycle freight pricing and logistics network planning method of claim 5, wherein: the specific method of the algorithm selection mechanism comprises the following steps:
firstly, performing particle swarm algorithm operation on an initial randomly generated population X (0) according to the change situation of the individual adaptive value in the population, wherein the individuals in the original population are X (0) = (X) 1 (0),x 2 (0),...,x NP (0) U = (u) for individuals in the population after update 1 ,u 2 ,...,u NP ) Introducing a new variable lbp in the process of updating the optimal value and the optimal solution of the current individual by using NP as the population scale, wherein the lbp is the number of the individuals in the original population X (0) substituted in the updated population u; the differential evolution algorithm operation is also carried out on the initial randomly generated population X (0), and the individuals in the population after the updating are
Introducing a new variable lbd in the process of updating the current individual optimal value and the optimal solution, wherein the lbd is the updated population u * The number of individuals in the original population X (0) is replaced; the variation of the individual fitness value in the population is represented by the variable lb, where
Introducing a variable gb in the global search process according to the change condition of the global optimum value of the population, and if the particle swarm algorithm is used in the tt-th generation iteration process and the current global optimum value is not different from the previous generation global optimum value, gb =1; otherwise gb =2; if a differential evolution algorithm is used in the tt-th iteration process, and the current global optimum value is not inferior to the previous global optimum value, gb =2; otherwise gb =1;
an algorithm selection mechanism determines which algorithm is used in the next iteration process according to the values of each generation of lb and gb, and the specific steps are as follows: when lb is more than or equal to 0.5 and gb =1, it indicates that the number of individual fitness value updates in the population after the operation of the particle swarm algorithm is more, and the global optimal solution is updated compared with the previous generation, and the next generation uses the particle swarm algorithm; when lb is less than 0.5 and gb =2, it is shown that the number of individual fitness value updates in the population after the operation of the differential evolution algorithm is greater, and the global optimal solution is updated compared with the previous generation, and the next generation uses the differential evolution algorithm; when lb is more than or equal to 0.5 and gb =2, it is stated that the number of individual fitness value updates in the population after the operation of the particle swarm algorithm is more, but the global optimal solution is not updated, according to the roulette rule, a random number rand is generated, if rand is less than or equal to lb, the next generation uses the differential evolution algorithm, if rand is greater than lb, the next generation uses the particle swarm algorithm; when lb is less than 0.5 and gb =1, it indicates that the number of individual fitness value updates in the population after the operation of the differential evolution algorithm is more, but the global optimal solution is not updated, and according to the roulette rule, a random number rand is generated, if rand is less than lb, the next generation uses the differential evolution algorithm, and if rand is greater than or equal to lb, the next generation uses the particle swarm algorithm.
7. The multi-cycle freight pricing and logistics network planning method of claim 6, wherein: the specific process for solving the multi-period freight pricing and logistics network planning model comprises the following steps:
step 501: initializing parameters in a population scale NP, a dimension D, a maximum iteration number NG, an initial population X, a particle swarm algorithm and a differential evolution algorithm;
step 502: setting initial iteration times, namely tt =0, simultaneously solving a multi-cycle freight pricing and logistics network planning model by using a particle swarm algorithm and a differential evolution algorithm, calculating initial values of lb and gb, and selecting an algorithm used in the first iteration according to an algorithm selection mechanism;
step 503: if the particle swarm algorithm is judged to be selected, the particle swarm algorithm is used for solving a multi-period freight pricing and logistics network planning model; if the differential evolution algorithm is judged and selected, the differential evolution algorithm is used for solving a multi-period freight pricing and logistics network planning model;
step 504: updating the values of lb and gb according to the above-mentioned solution result, and selecting the algorithm used in the next iteration according to the algorithm selection mechanism;
step 505: updating the current individual optimal value, the optimal solution, the current global optimal value and the optimal solution;
step 506: judging whether a termination condition is met, if so, ending the solution to obtain a logistics planning scheme; if not, return to step 503.
8. The multi-cycle freight pricing and logistics network planning method of claim 7, wherein: the logistics planning scheme comprises a freight pricing plan, a transfer center node plan, a line plan, a logistics carrier plan and a freight volume plan.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211038505.8A CN115689412A (en) | 2022-08-29 | 2022-08-29 | Multi-period freight pricing and logistics network planning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211038505.8A CN115689412A (en) | 2022-08-29 | 2022-08-29 | Multi-period freight pricing and logistics network planning method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115689412A true CN115689412A (en) | 2023-02-03 |
Family
ID=85061394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211038505.8A Pending CN115689412A (en) | 2022-08-29 | 2022-08-29 | Multi-period freight pricing and logistics network planning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115689412A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115829183A (en) * | 2023-02-22 | 2023-03-21 | 四川港投新通道物流产业投资集团有限公司 | Cold-chain logistics path planning method, device, equipment and readable storage medium |
| CN117113608A (en) * | 2023-10-23 | 2023-11-24 | 四川港投新通道物流产业投资集团有限公司 | Cold-chain logistics network node layout method and equipment |
-
2022
- 2022-08-29 CN CN202211038505.8A patent/CN115689412A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115829183A (en) * | 2023-02-22 | 2023-03-21 | 四川港投新通道物流产业投资集团有限公司 | Cold-chain logistics path planning method, device, equipment and readable storage medium |
| CN117113608A (en) * | 2023-10-23 | 2023-11-24 | 四川港投新通道物流产业投资集团有限公司 | Cold-chain logistics network node layout method and equipment |
| CN117113608B (en) * | 2023-10-23 | 2024-02-13 | 四川港投新通道物流产业投资集团有限公司 | Cold-chain logistics network node layout method and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111626477B (en) | Multi-type joint transport path optimization method considering uncertain conditions | |
| CN115689412A (en) | Multi-period freight pricing and logistics network planning method | |
| Yin | Multiobjective bilevel optimization for transportation planning and management problems | |
| Chen et al. | Quantifying the value of leadtime information in a single-location inventory system | |
| CN111311158B (en) | Electric logistics vehicle path planning method under limited charging facility condition | |
| CN109886490B (en) | Matching optimization method for combined vehicle combined transportation | |
| CN109919532B (en) | Logistics node determining method and device | |
| CN111553507A (en) | Multi-commodity-flow-based optimization method for Chinese and European container transportation scheme | |
| CN110705741B (en) | Multi-distribution center vehicle path optimization method based on improved ant colony algorithm | |
| CN110619441A (en) | Leader-based GA-PSO (genetic algorithm-particle swarm optimization) soft time window vehicle path optimization method | |
| CN114897217A (en) | Fresh food electricity merchant front-end bin site selection-path planning method | |
| CN115759917A (en) | Logistics path planning method based on improved mixed ant colony algorithm | |
| CN113822461A (en) | Track traffic cross-line operation optimization method, system, equipment and storage medium | |
| Esmaeili et al. | Designing a hub location and pricing network in a competitive environment. | |
| CN114239915A (en) | Platform centralized supply and demand scheme optimization method based on supply and demand matching optimization model | |
| Chang et al. | Assess economic and environmental trade-off for inland port location | |
| CN116187889A (en) | Full-chain molten iron intermodal transportation scheme system and evaluation method | |
| WO2023082315A1 (en) | Intelligent parsing method and system for whole electronic solid waste recycling process | |
| CN115689081A (en) | Dynamic optimization method for kitchen garbage collection and transportation path based on adaptive cooperative mechanism | |
| Ye et al. | A fuzzy vehicle routing assignment model with connection network based on priority-based genetic algorithm | |
| Tao et al. | A differential evolution with reinforcement learning for multi-objective assembly line feeding problem | |
| Pan | The optimization model of the vendor selection for the joint procurement from a total cost of ownership perspective | |
| Hasanzadeh et al. | DESIGNING A THREE LAYER SUPPLY CHAIN CONSIDERING DIFFERENT TRANSPORTATION CHANNELS AND DELIVERY TIME DEPENDENT DEMAND. | |
| Khademi et al. | Addressing a sustainable production-distribution supply chain network design problem considering carbon emissions policies by a hybrid whale optimization algorithm | |
| Gugat et al. | Optimising of'Sea Toll'Network Design on the Maritime Freight Transportation Network |
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 |