Path planning method considering carrier combinatorial optimization under uncertain requirements

Abstract

The invention discloses a route planning method considering carrier combinatorial optimization under an uncertain demand, and the method comprises the steps: firstly collecting and recording different demands of clients in cargo transportation, constructing a demand set, and building a matching mechanism of cargo attributes and the demand set; secondly, a fuzzy analytic hierarchy process is utilized to evaluate multiple capability indexes of the carrier, a carrier capability evaluation model is constructed, three-level priority search is carried out based on customer requirements, and a carrier combination scheme meeting the customer requirements is output; analyzing each cost, and establishing a cost model; and finally, solving the total cost through an improved adaptive genetic algorithm to obtain an optimized transportation path, and outputting an optimal carrier combination and transportation mode. According to the method, the transportation advantages of different carriers can be more comprehensively considered, the transportation and transfer difficulty faced by a single carrier is effectively reduced, the efficiency and service quality of multimodal transportation are improved, and the optimization requirements of a modern logistics system are met.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics route planning, and in particular to a route planning method that considers carrier combination optimization under uncertain demand. Background Art

[0002] With the rapid development of global trade and the increasing complexity of logistics networks, multimodal transport has attracted widespread attention as an efficient mode of cargo transportation. Multimodal transport achieves efficient and economical transportation of goods through the effective connection of multiple modes of transportation such as road, rail, and water transport. In the multimodal transport system, the selection of carriers and the planning of transportation routes directly affect the overall efficiency and service quality of the supply chain.

[0003] In the existing technology, traditional multimodal transport mostly adopts a fixed carrier combination for transportation. However, this method is often difficult to respond flexibly when faced with demand uncertainty, demand fluctuations and special cargo transportation requirements, and it is difficult to balance cost control and improvement of transportation quality. To overcome this limitation, companies often need to invest additional costs to make up for technical deficiencies or rely on the experience and familiarity of specific carriers, but this significantly increases operating costs and reduces overall efficiency. In this context, carrier combination optimization has become one of the important research issues in the field of multimodal transport route optimization. Its goal is to minimize transportation costs and improve transportation efficiency through reasonable carrier combination selection while meeting the timeliness and reliability requirements of goods.

[0004] Therefore, there is an urgent need for a route planning method that considers carrier portfolio optimization under uncertain demand. Summary of the invention

[0005] The present invention provides a route planning method considering carrier combination optimization under uncertain demand, which solves the problems of low response efficiency and high transportation cost caused by fixed carrier combination in traditional multimodal transport mode.

[0006] To achieve the above object, the present invention provides the following technical solution: a route planning method considering carrier combination optimization under uncertain demand, comprising the following steps:

[0007] Step 1: Collect and record different customer needs in cargo transportation, build a demand set, and match the needs with cargo attributes.

[0008] S11. Define U n Indicates the demand V u The demand set formed by collecting and recording the attributes of goods that all carriers are capable of transporting is expressed as follows:

[0009]

[0010] Among them U n represents the different requirements of customers in cargo transportation, w dn Indicates the high evaluation capability of the carriers selected at each stage, i Represents the weight of the corresponding deviation. Z represents the number of all available carriers in the entire system.

[0011] Step 2: Determine the priority of carriers based on customer needs, classify carrier capabilities into three categories: service quality, service cost, and comprehensive transportation capacity, and construct an evaluation matrix through fuzzy hierarchical analysis method.

[0012] S21. Match the concentrated demands with the attributes of the goods and index them, dividing the demands into three categories: service quality, where the indicators include order response rate, cargo damage rate, on-time delivery rate, and whether expedited delivery is supported, which are numbered 1, 2, 3, and 4 respectively; service cost, where the indicators include cargo damage cost and service price, which are numbered 5 and 6 respectively; comprehensive transportation capacity, where the indicators include the number of supported transportation modes, transportation vehicle capacity, and service complaint rate, which are numbered 7, 8, and 9 respectively.

[0013] S22. Fuzzy hierarchical analysis to calculate the scores of various indicators:

[0014] Definition mn The triple matrix A is used to fuzzily evaluate the carrier's various indicator capabilities, indicating the relative importance of each indicator. The specific representation of the triple matrix is ​​as follows:

[0015]

[0016] in For the lower bound, is the middle value, is the upper bound, m and n represent the index numbers, and represent the fuzzy number of the mth index relative to the nth index.

[0017] Step 3: Build a carrier combination selection mechanism, determine the three-level priority of carriers based on customer needs, search and generate the optimal carrier combination suitable for different transportation scenarios, and obtain the final carrier total score.

[0018] Step 3.1: Determine the three-level priorities and search for the right carrier combination:

[0019] According to the fuzzy indicator scores obtained in the above steps, three levels of priority are determined, among which direct customer demand is the first-level priority, indicators that can affect customer selection decisions within the transportation sector are the second-level priority, and other indicators with less or no influence are the third-level priority.

[0020] The fuzzy indicator scoring ranges for the three levels of priority are: 0.20-0.29 for the first level priority, 0.10-0.19 for the second level priority, and 0.0-0.9 for the third level priority.

[0021] 1) Define the fuzzy weight W d It is expressed as a triple of the weights of each indicator. The score of each indicator weight is calculated. According to the score of each indicator, the three-level priority search of the carrier indicator capability is performed. The specific expression of the fuzzy weight triple is as follows:

[0022]

[0023] Its fuzzy weight W d The specific calculation formula is:

[0024]

[0025] Here, k represents the kth level priority.

[0026] 2) Define the carrier's comprehensive score S d , which represents the total weighted score obtained by carrier d after searching for carrier combination solutions at three levels of priority according to customer needs. The specific calculation formula is:

[0027]

[0028] I is the indicator set, of which the service quality set is I1, the service cost set is I2, the comprehensive capability set is I3, and the small indicators of each classification set are represents the score of the d-th carrier on the y-th indicator in the k-th priority level.

[0029] Step 3.2: Eliminate the comprehensive score S according to the threshold set by the comprehensive score required by the actual customer or enterprise. d For the carriers that do not meet the threshold, output the combination plan of all carriers that meet the threshold and proceed to the next step of solution.

[0030] Step 4: Define the employment cost, transshipment cost, time window cost and penalty function in logistics transportation to cope with the impact of uncertain demand changes.

[0031] S41. Define carrier employment costs It means that the carrier will n The contract cost required to transport goods from node i to node j is specifically expressed as:

[0032]

[0033] Where D ij represents the distance between node i and node j, F nrepresents the n carriers hired, and C represents the collection of transportation methods required;

[0034] S42. Define the transshipment cost between different modes of transport as ω, which is in the following form:

[0035]

[0036] in Indicates that at node i, it is determined whether the mode of transportation has changed. If the mode of transportation has not changed, then If the mode of transportation changes, represents the cost of changing the mode of transportation at node i;

[0037] S43. Define the transportation cost as F T , which has the following form:

[0038]

[0039] in Represents the cost of keeping the goods fresh. represents the cargo damage rate of the d-th carrier at the k-th priority level, represents the cost of the corresponding order in the ij node, represents the response speed of the d-th carrier for the order of the k-th priority, Represents the unit cost of cargo transportation, D ij Represents the distance between nodes;

[0040] S44. Define the time window cost as F s , the time window adopts a soft time window. The early arrival of goods will incur a waiting cost θ1, while the late arrival of goods will incur a penalty cost θ2, which is as follows:

[0041]

[0042] in a t j The earliest delivery time requested by the customer. The time for goods to arrive. b t j The latest delivery time required by the customer. The on-time rate of carrier orders under different priorities, The service complaint rates of carriers at different priorities;

[0043] S45. Define the total cost calculation algorithm model F, which is as follows:

[0044]

[0045]

[0046] where x ij is the transportation decision variable from i to j, which takes a value of 0 or 1, indicating whether to select a carrier to carry out the transportation from i to j. Constraint 1 ensures that there is at least one carrier for each order; Constraint 2 ensures that the goods transported during the transportation process will not exceed the carrier's transportation capacity. max .

[0047] Step 5: Apply the improved adaptive genetic algorithm to solve the carrier combination and path optimization model in path planning, set the population size, crossover probability and mutation probability, and obtain the carrier combination and path plan that minimizes the total transportation cost. The specific steps are as follows:

[0048] S51, importing the attributes such as the carrier combination number and weight information, the coordinates of each transportation node, the cost and carrying capacity of different transportation modes, customer demand, and order time window data;

[0049] S52, set the population size of the improved GA algorithm to 50, the disaster threshold to 500 times, the initial crossover probability to 0.8, and the initial mutation probability to 0.05; encode each transportation path, and each path individual is composed of a node sequence, a carrier sequence, and the transportation mode adopted;

[0050] S53, calculating individual fitness, the fitness value is the sum of the total cost of each path individual, specifically:

[0051]

[0052] β[i]=(d,c): represents the carrier d and transportation method c selected for order i;

[0053] F[i,d,c]: represents the cost of carrier d and transportation method c selected for order i;

[0054] Z[i,d]: carrier selection scoring function that takes into account priority indicators.

[0055] Individuals with lower fitness have lower total costs and stronger adaptability. The individual with the lowest fitness is selected for the next step;

[0056] S54, iteratively calculate all carrier combinations, and output the carrier combination number, transportation mode, and path node number of the optimal solution if the iteration is successful. If the iteration exceeds the disaster threshold, the disaster factor is introduced to terminate the iteration to avoid falling into the local optimum, and then partially or completely reset the individuals with poor fitness, and generate new individuals by randomly selecting the node order or carrier combination to increase the population diversity, increase the mutation probability to accelerate the diversity search of the path solution, and enable the population to explore a better solution faster. After the disaster is over, the original crossover and mutation probabilities are restored.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. The present invention realizes the intelligent combination and dynamic path optimization of carriers in the multimodal transport system. On the premise of meeting the timeliness and reliability of cargo transportation, with the goal of minimizing transportation costs and maximizing transportation efficiency, it adapts to the fluctuations in demand in different transportation scenarios and realizes temporary alliances and resource sharing among carriers.

[0059] 2. In the present invention, in traditional multimodal transport, the present invention evaluates the carrier's capabilities based on the fuzzy hierarchical analysis method and performs a three-level priority search. This solution can flexibly combine multiple carriers to provide the best transportation method for different goods and needs, effectively reducing the limitations of a single carrier in the transportation and transshipment process.

[0060] 3. The present invention uses an improved adaptive genetic algorithm to perform path optimization under multi-dimensional requirements, so that the transportation chain has a higher risk resistance, especially maintaining strong fault tolerance in dealing with emergencies such as extreme weather and traffic congestion. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0062] In the attached picture:

[0063] Figure 1 It is a flow chart of the route planning method considering carrier combination optimization of the present invention. DETAILED DESCRIPTION

[0064] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0065] Example: Figure 1As shown in the figure, a path planning method considering carrier combination optimization under uncertain demand is used. Taking 50 carriers and 40 logistics nodes in a logistics network in a certain region of China as an example, the range roughly covers the western to eastern part of China. 5, 10, and 15 carriers are selected and accumulated to 50 selectable carriers respectively. The path planning scheme and the final cost are iteratively calculated, and compared with the other two traditional logistics models. The specific steps are as follows:

[0066] Step 1: Collect and record different customer needs in cargo transportation, build a demand set, and match the needs with cargo attributes.

[0067] S11. Define U n Indicates the demand V u The demand set formed by collecting and recording the attributes of goods that all carriers are capable of transporting is expressed as follows:

[0068]

[0069] Among them U n represents the different requirements of customers in cargo transportation, w dn Indicates the high evaluation capability of the carriers selected at each stage, i Represents the weight of the corresponding deviation. Z represents the number of all available carriers in the entire system.

[0070] Step 2: Determine the priority of carriers based on customer needs, classify carrier capabilities into three categories: service quality, service cost, and comprehensive transportation capacity, and construct an evaluation matrix through fuzzy hierarchical analysis method.

[0071] Step 2.1: Match the demands in the demand set with the attributes of the goods and index them, dividing the demands into three categories: service quality, where the indicators include order response rate, cargo damage rate, on-time delivery rate, and whether expedited delivery is supported, numbered 1, 2, 3, and 4 respectively; service cost, where the indicators include cargo damage cost and service price, numbered 5 and 6 respectively; comprehensive transportation capacity, where the indicators include the number of supported transportation modes, transportation vehicle capacity, and service complaint rate, numbered 7, 8, and 9 respectively.

[0072] Step 2.2: Fuzzy hierarchical analysis calculates the scores of each indicator:

[0073] Definition mn The triple matrix A is used to fuzzily evaluate the carrier's various indicator capabilities, indicating the relative importance of each indicator. The specific representation of the triple matrix is ​​as follows:

[0074]

[0075] in For the lower bound, is the middle value, is the upper bound, m and n represent the index numbers, and represent the fuzzy number of the mth index relative to the nth index.

[0076] Step 3: Build a carrier combination selection mechanism, determine the three-level priority of carriers based on customer needs, search and generate the optimal carrier combination suitable for different transportation scenarios, and obtain the final carrier total score.

[0077] Step 3.1: Determine the three-level priorities and search for the right carrier combination:

[0078] According to the fuzzy indicator scores obtained in the above steps, three levels of priority are determined, among which direct customer demand is the first-level priority, indicators that can influence customer selection decisions during transportation are the second-level priority, and other indicators with less or no influence are the third-level priority.

[0079] Among them, in step 2, the carrier's capabilities are quantified into nine indicators. For example, if the customer specifies that the demand is for a low cargo damage rate, then the cargo damage rate and cargo damage cost are the first-level priority for direct demand; when filling out an order, the customer feels that the service cost and on-time delivery rate are also very important, then these two items are the second-level priority; and the remaining five capabilities are the third-level priority.

[0080] Among them, the fuzzy indicator scoring range of the three-level priority is: the first-level priority is 0.20-0.29; the second-level priority is 0.10-0.19; and the third-level priority is 0.0-0.9.

[0081] 1) Define the fuzzy weight W d It is expressed as a triple of the weights of each indicator. The score of each indicator weight is calculated. According to the score of each indicator, the three-level priority search of the carrier indicator capability is performed. The specific expression of the fuzzy weight triple is as follows:

[0082]

[0083] Its fuzzy weight W d The specific calculation formula is:

[0084]

[0085] Here, k represents the kth level priority.

[0086] 2) Define the carrier's comprehensive score S d , which represents the total weighted score obtained by carrier d after searching for carrier combination solutions at three levels of priority according to customer needs. The specific calculation formula is:

[0087]

[0088] I is the indicator set, of which the service quality set is I1, the service cost set is I2, the comprehensive capability set is I3, and the small indicators of each classification set are represents the score of the d-th carrier on the y-th indicator in the k-th priority level.

[0089] Step 3.2: Eliminate the comprehensive score S according to the threshold set by the comprehensive score required by the actual customer or enterprise. d For the carriers that do not meet the threshold, output the combination plan of all carriers that meet the threshold and proceed to the next step of solution.

[0090] Step 4: Define the employment cost, transshipment cost, time window cost and penalty function in logistics transportation to cope with the impact of uncertain demand changes.

[0091] S41. Define carrier employment costs It means that the carrier will n The contract cost required to transport goods from node i to node j is specifically expressed as:

[0092]

[0093] Where D ij represents the distance between node i and node j, F n represents the n carriers hired, and C represents the collection of transportation methods required;

[0094] S42. Define the transshipment cost between different modes of transport as ω, which is in the following form:

[0095]

[0096] in Indicates that at node i, it is determined whether the mode of transportation has changed. If the mode of transportation has not changed, then If the mode of transportation changes, represents the cost of changing the mode of transportation at node i;

[0097] S43. Define the transportation cost as F T , which has the following form:

[0098]

[0099] in Represents the cost of keeping the goods fresh. represents the cargo damage rate of the d-th carrier at the k-th priority level, represents the cost of the corresponding order in the ij node, represents the response speed of the d-th carrier for the order of the k-th priority, Represents the unit cost of cargo transportation, D ij Represents the distance between nodes;

[0100] S44. Define the time window cost as F s , the time window adopts a soft time window. The early arrival of goods will incur a waiting cost θ1, while the late arrival of goods will incur a penalty cost θ2, which is as follows:

[0101]

[0102]

[0103] in a t j The earliest delivery time requested by the customer. The time for goods to arrive. b t j The latest delivery time required by the customer. The on-time rate of carrier orders under different priorities, The service complaint rates of carriers at different priorities;

[0104] S45. Define the total cost calculation algorithm model F, which is as follows:

[0105]

[0106] where x ij is the transportation decision variable from i to j, which takes a value of 0 or 1, indicating whether to select a carrier to carry out the transportation from i to j. Constraint 1 ensures that there is at least one carrier for each order; Constraint 2 ensures that the goods transported during the transportation process will not exceed the carrier's transportation capacity. max .

[0107] Step 5: Apply the improved adaptive genetic algorithm to solve the carrier combination and path optimization model in path planning, set the population size, crossover probability and mutation probability, and obtain the carrier combination and path plan that minimizes the total transportation cost. The specific steps are as follows:

[0108] S51, importing the attributes such as the carrier combination number and weight information, the coordinates of each transportation node, the cost and carrying capacity of different transportation modes, customer demand, and order time window data;

[0109] S52, set the population size of the improved GA algorithm to 50, the disaster threshold to 500 times, the initial crossover probability to 0.8, and the initial mutation probability to 0.05; encode each transportation path, and each path individual is composed of a node sequence, a carrier sequence, and the transportation mode adopted;

[0110] S53, calculating individual fitness, the fitness value is the sum of the total cost of each path individual, specifically:

[0111]

[0112] β[i]=(d,c): represents the carrier d and transportation method c selected for order i;

[0113] F[i,d,c]: represents the cost of carrier d and transportation method c selected for order i;

[0114] Z[i,d]: carrier selection scoring function that takes into account priority indicators.

[0115] Individuals with lower fitness have lower total costs and stronger adaptability. The individual with the lowest fitness is selected for the next step;

[0116] S54, iteratively calculate all carrier combinations, and output the carrier combination number, transportation mode, and path node number of the optimal solution if the iteration is successful. If the iteration exceeds the disaster threshold, the disaster factor is introduced to terminate the iteration to avoid falling into the local optimum, and then partially or completely reset the individuals with poor fitness, and generate new individuals by randomly selecting the node order or carrier combination to increase the population diversity, increase the mutation probability to accelerate the diversity search of the path solution, and enable the population to explore a better solution faster. After the disaster is over, the original crossover and mutation probabilities are restored.

[0117] The present invention also provides a storage medium, in which instructions are stored. When a computer reads the instructions, the computer executes the path planning method based on considering the optimization of carrier combinations.

[0118] The present invention also provides an electronic device, comprising a processor and the medium, wherein the processor executes instructions in the storage medium.

[0119] As described in the above embodiments, the cost comparison of the final output path planning scheme with the other two traditional logistics models can be concluded: the genetic iterative model that adds carbon emission parameters on the basis of the classic genetic iterative algorithm, and the genetic iterative model that considers carrier selection proposed in this article, can significantly reduce the total transportation cost. When the number of carriers is small, the cost of the carbon emission model and the carrier selection model is almost the same, the performance of the two models is similar, and the carbon emission model even performs better. However, as the number of carriers gradually increases and the available carrier combinations become more abundant, the method in this article shows a significant cost reduction trend.

[0120] Finally, it should be noted that the above description is only a preferred example of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A route planning method considering carrier combination optimization under uncertain demand, characterized in that: The following steps are involved: S1. Collect and record different needs of customers in cargo transportation, build a demand set, and match the needs with cargo attributes; S2. Determine the priority of carriers based on customer needs, classify carrier capabilities into three categories: service quality, service cost, and comprehensive transportation capacity, and construct an evaluation matrix through fuzzy hierarchical analysis method; S3. Build a carrier combination selection mechanism, determine the three-level priority of carriers according to customer needs, search and generate the optimal carrier combination suitable for different transportation scenarios, and obtain the final carrier total score; S4. Define the employment cost, transshipment cost, time window cost and penalty function in logistics transportation to cope with the impact of uncertain demand changes; S5. In route planning, the improved adaptive genetic algorithm is used to solve the carrier combination and route optimization model, and the population size, crossover probability and mutation probability are set to obtain the carrier combination and route plan that minimizes the total transportation cost.

2. The route planning method considering carrier combination optimization under uncertain demand according to claim 1 is characterized by: Step S1 specifically includes: S11. Define U n Indicates the demand V u The demand set formed by collecting and recording the attributes of goods that all carriers are capable of transporting is expressed as follows: Among them U n represents the different requirements of customers in cargo transportation, w dn Indicates the high evaluation capability of the carriers selected at each stage, i represents the weight of the corresponding deviation; Z represents the number of all available carriers in the entire system.

3. The route planning method considering carrier combination optimization under uncertain demand according to claim 1 is characterized by: Step S2 specifically includes: S21. Match the concentrated demands with the attributes of the goods and index them, and divide the demands into three categories: service quality, where the indicators include order response rate, cargo damage rate, on-time delivery rate, and whether expedited delivery is supported, numbered 1, 2, 3, and 4 respectively; service cost, where the indicators include cargo damage cost and service price, numbered 5 and 6 respectively; comprehensive transportation capacity, where the indicators include the number of supported transportation modes, transportation vehicle capacity, and service complaint rate, numbered 7, 8, and 9 respectively; S22, Fuzzy Hierarchy Analysis calculates the scores of each indicator, defining a mn The triple matrix A is used to fuzzily evaluate the carrier's various indicator capabilities, indicating the relative importance of each indicator. The specific representation of the triple matrix is ​​as follows: in For the lower bound, is the middle value, is the upper bound, m and n represent the index numbers, and represent the fuzzy number of the mth index relative to the nth index.

4. The route planning method considering carrier combination optimization under uncertain demand according to claim 1 is characterized by: Step S3 specifically includes: S31. Define fuzzy weight W d It is expressed as a triple of the weights of each indicator. The score of each indicator weight is calculated. According to the score of each indicator, the carrier indicator capability is searched in three levels of priority. The direct demand of customers is the first level priority, the indicators that can affect the customer's choice decision within the transportation are the second level priority, and the other indicators with little influence or no influence are the third level priority. The specific expression of the fuzzy weight triple is: Its fuzzy weight W d The specific calculation formula is: Among them, k represents the kth level priority; S32. Define the carrier's comprehensive score S d , which represents the total weighted score obtained by carrier d after searching for carrier combination solutions at three levels of priority according to customer needs. The specific calculation formula is: I is the indicator set, of which the service quality set is I1, the service cost set is I2, the comprehensive capability set is I3, and the small indicators of each classification set are represents the score of the d-th carrier on the y-th indicator in the k-th priority level.

5. The route planning method considering carrier combination optimization under uncertain demand according to claim 4 is characterized by: In step 3: After obtaining the total score of the carrier combination, the carriers with the comprehensive score S are eliminated according to the threshold set by the comprehensive score required by the actual customer or enterprise. d For the carriers that do not meet the threshold, output the combination plan of all carriers that meet the threshold and proceed to the next step of solution.

6. The route planning method considering carrier combination optimization under uncertain demand according to claim 1 is characterized by: Step S4 specifically includes: S41. Define carrier employment costs It means that the carrier will n The contract cost required to transport goods from node i to node j is specifically expressed as: Where D ij represents the distance between node i and node j, F n represents the n carriers hired, and C represents the collection of transportation methods required; S42. Define the transshipment cost between different modes of transport as ω, which is in the following form: in Indicates that at node i, it is determined whether the mode of transportation has changed. If the mode of transportation has not changed, then If the mode of transportation changes, represents the cost of changing the mode of transportation at node i; S43. Define the transportation cost as F T , which has the following form: in Represents the cost of keeping the goods fresh. represents the cargo damage rate of the d-th carrier at the k-th priority level, represents the cost of the corresponding order in the ij node, represents the response speed of the d-th carrier for the order of the k-th priority, Represents the unit cost of cargo transportation, D ij Represents the distance between nodes; S44. Define the time window cost as F s , the time window adopts a soft time window. The early arrival of goods will incur a waiting cost θ1, while the late arrival of goods will incur a penalty cost θ2, which is as follows: in The earliest delivery time requested by the customer. The time for goods to arrive. The latest delivery time required by the customer. The on-time rate of carrier orders under different priorities, The service complaint rates of carriers at different priorities; S45. Define the total cost calculation algorithm model F, which is in the following form: where x ij is the transportation decision variable from i to j, which takes a value of 0 or 1, indicating whether to select a carrier to carry out the transportation from i to j. Constraint 1 ensures that there is at least one carrier for each order; Constraint 2 ensures that the goods transported during the transportation process will not exceed the carrier's transportation capacity. max .

7. The route planning method considering carrier combination optimization under uncertain demand according to claim 1 is characterized by: In step S5, the specific steps include: S51, importing the attributes such as the carrier combination number and weight information, the coordinates of each transportation node, the cost and carrying capacity of different transportation modes, customer demand, and order time window data; S52, the adaptive genetic algorithm is an improved GA algorithm, the population size of the improved GA algorithm is set to 50, the disaster threshold is 500 times, the initial crossover probability is 0.8, and the initial mutation probability is 0.05; each transportation path is encoded, and each path individual is composed of a node sequence, a carrier sequence, and the adopted transportation mode; S53, calculating individual fitness, the fitness value is the sum of the total cost of each path individual, specifically: β[i]=(d,c): represents the carrier d and transportation method c selected for order i; F[i,d,c]: represents the cost of carrier d and transportation method c selected for order i; Z[i,d]: carrier selection scoring function that takes into account priority indicators; The individuals with lower fitness have lower total costs and stronger adaptability. The individuals with the lowest fitness are selected for the next step. S54. Iterate and calculate all carrier combinations. If the iteration is successful, output the carrier combination number, transportation method, and path node number of the optimal solution. If the iteration exceeds the disaster threshold, introduce the disaster factor and terminate the iteration to avoid falling into the local optimum. Then, perform a partial or complete reset on the individuals with poor fitness, generate new individuals by re-randomly selecting the node order or carrier combination to increase the population diversity, and restore the original crossover and mutation probabilities after the disaster.

Images