A joint optimization method for grain loading mode and transportation path

Abstract

The present invention discloses a method for jointly optimizing a grain loading mode and a transportation path, comprising: obtaining normal transportation time, reasonable transportation loss, transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters of grain transportation from a departure point to a destination, constructing transportation decision variables according to the transportation parameters, and constructing a transportation objective function and constraint conditions with the minimum actual transportation total cost according to the four parameters and the transportation decision variables; integrating the four parameters, the transportation decision variables, the objective function and the constraint conditions to obtain a grain transportation mathematical model, and solving the mathematical model using an improved genetic algorithm according to the constraint conditions to obtain an optimal transportation path; the present invention is more accurate in responding to differentiated demands in terms of transportation time, transportation quality, transportation cost and the like, and taking into account different loading modes, transportation modes and more loss costs, the results obtained when determining the grain transportation path are better and more resources are saved.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-mode transportation path optimization, and in particular to a method for jointly optimizing a grain loading mode and a transportation path. Background Art

[0002] As the most basic guarantee for human survival and development, ensuring food security has long been a primary concern for countries around the world. At present, the geographical separation between food production and consumption is a common phenomenon. Building a high-quality food circulation system has become an important guarantee for ensuring the balance of food supply and demand and food security. At the same time, as a special type of cargo, food is prone to loss. In the practice of food transportation, the upper limit of the acceptable loss rate is usually stipulated in the transportation contract. This provision has increased the hidden costs of the cargo owner in disguise, affecting the cargo owner's choice of transportation organization method, and has also become a difficulty in reducing logistics costs.

[0003] Focusing on grain transportation, Han Jianjun et al. constructed a grain transshipment hub location and path optimization model from the perspective of reducing carbon emissions and transportation costs. Ndembe et al. took grain transportation in the Great Plains of the United States as the research object and found that the impact of railway grain transportation service quality on the consignor's consignment selection behavior is greater than the impact of simple transportation prices. Zhang et al. proposed a hybrid algorithm for optimizing the dispatch of emergency vehicles for grain based on artificial immunity and ant colony algorithm, with the maximum satisfaction of emergency point demand, total distribution cost and distribution time as optimization objectives. Prajapati et al. constructed a grain logistics optimization model under the background of e-commerce that considers the one-kilometer transportation cost at both ends, carbon emission cost, inventory cost, accident and loss cost, and time penalty cost. Sharma et al. constructed a grain supply chain network optimization model under the background of sudden epidemics, including personal distance cost, carbon emission cost, fixed cost and transportation cost. Maiyar et al. constructed a two-stage optimization model combining grain transportation tactics and operations between state-level grain depots and central-level grain depots in India, and designed a composite particle swarm algorithm with efficient solution capabilities. Maiyar et al. constructed a path optimization model with the dual objectives of minimizing the cost of the grain supply network and greenhouse gas emissions based on factors such as grain loss, hub location, and capacity constraints.

[0004] At present, for the commodity grain, relevant research has not fully considered the perishability of grain transportation and the diversity of loading methods and transportation methods in the transportation organization mode. The existing path optimization method lacks quantitative parameters for the multi-dimensional losses in the whole process of grain transportation, resulting in the problem of hidden costs that cannot be measured, resulting in the transportation plan formed is not the most ideal plan for the shipper. The existing path optimization method lacks quantitative parameters for the differences in loading speed and loss rate of different loading methods such as bagged, bulk and containerized loading. The essential characteristics of "scattered transportation" and "bulk to container transportation" are not accurately described, which is not conducive to the combined optimization of transportation organization methods. In addition, the existing path optimization method optimizes a single transportation path, and lacks technical methods for jointly optimizing loading methods and transportation paths. Summary of the invention

[0005] The present invention provides a method for jointly optimizing grain loading methods and transportation routes, so as to overcome the technical problem that the diversity of transportation methods, loading methods and losses is not taken into account in the grain route planning process, resulting in not only a large amount of time and resources being spent in the grain transportation process, but also the loss cannot be reduced.

[0006] In order to achieve the above object, the technical solution of the present invention is:

[0007] A method for jointly optimizing grain loading mode and transportation path, comprising:

[0008] S1: Obtain the normal transportation time and reasonable transportation loss of grain from the origin to the destination, where the reasonable transportation loss is the natural loss of grain during the transportation process;

[0009] S2: Obtaining transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters in the grain transportation process, constructing transportation decision variables according to the transportation parameters, and constructing a grain transportation objective function and constraint conditions with the minimum actual total transportation cost according to the four parameters and the transportation decision variables; the transportation parameters include transportation mode and loading mode, and the loss cost parameters include loss rates under different loading modes and different transportation modes;

[0010] S3: Integrate the four parameters, transportation decision variables, grain transportation objective function and constraint conditions to obtain a grain transportation mathematical model, wherein the grain transportation mathematical model is used to determine an optimal loading method and an optimal transportation route for grain;

[0011] S4: Based on the constraints, an improved genetic algorithm is used to solve the grain transportation mathematical model to obtain an optimal transportation path that combines the loading method and the transportation method. The optimal transportation path is a transportation path with the lowest actual total transportation cost within normal transportation time and reasonable transportation losses.

[0012] Further, S2 obtains the transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters in the grain transportation process, constructs the transportation decision variables according to the transportation parameters, and constructs the grain transportation objective function and constraint conditions with the minimum actual transportation total cost according to the four parameters and the transportation decision variables; the transportation parameters include the transportation mode and the loading mode, and the loss cost parameters include the loss rates under different loading modes and different transportation modes, including:

[0013] S21. Obtaining transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters during grain transportation:

[0014] The transport parameters include but are not limited to:

[0015] G = N, A, K, represents the grain transportation network; N is the set of transportation nodes, N = 1, 2, 3, ...i, ...j; A is the set of transportation arcs, representing the connection between different modes of transportation at the same location; K is the set of transportation modes, K = {1, 2, 3}, 1, 2, 3 represent road, rail and water transportation respectively; M is the set of loading modes, M = {1, 2, 3}, 1, 2, 3 represent bag, bulk and container loading respectively;

[0016] Q is the total weight of the transported goods, in tons;

[0017] ——Transport distance of the kth transport mode between nodes i and j (km);

[0018] ——The unit loading speed (t / h) of converting the transportation mode k between nodes i and j to g at node j when the mth loading mode is selected;

[0019] ——Transport speed of the kth mode of transport between nodes i and j (km / h);

[0020] ——The maximum transport capacity of the kth transport mode between nodes i and j when the mth loading mode is selected, m∈M, M={1,2,3};

[0021] ——The maximum reloading capacity at node j when the mth loading mode is selected;

[0022] T AO ——Final arrival time of goods (h);

[0023] T——latest arrival time required for goods (h);

[0024] The transportation cost parameters include but are not limited to:

[0025] C – total transportation cost;

[0026] ——Unit transportation cost (yuan / (t·km)) of transporting goods between nodes i and j using the mth loading method and the kth transportation method, g∈K, K={1,2,3};

[0027] ——Under the mth loading mode, the unit transshipment cost (yuan / t) of converting the transport mode k between nodes i and j to g at node j;

[0028] The carbon emission cost parameters include but are not limited to:

[0029] c e ——Unit carbon emission cost (yuan / kg);

[0030] ——Specific carbon emissions of the kth mode of transportation between nodes i and j (kg / (t·km));

[0031] ——the unit carbon emissions (kg / t) of converting the transport mode k between nodes i and j to g at node j;

[0032] The loss cost parameters include but are not limited to:

[0033] ——Under the mth loading mode, the unit loss rate (%) of the kth transportation mode between nodes i and j;

[0034] ——Under the mth loading mode, the unit loss rate (%) of converting the transportation mode k between nodes i and j to g at node j, g∈K;

[0035] p——unit value of grain (yuan / t);

[0036] S22. Constructing transportation decision variables according to the transportation mode and loading mode, wherein the transportation decision variables include but are not limited to:

[0037]

[0038] S23, constructing a grain transportation objective function with the minimum actual transportation total cost according to the four parameters and the transportation decision variables, as shown in formula (1),

[0039]

[0040] S24, constructing constraint conditions according to the four parameters and the transportation decision variables, as shown in formula (2) to formula (8),

[0041]

[0042]

[0043] in, represents the decision variable for selecting mode k between transport nodes j and i; formula (2) indicates that only one mode of transport can be selected between any nodes; formula (3) indicates that at most one mode of transport can be changed at a transport node; formula (4) indicates that only one loading method can be selected between any nodes; formula (5) ensures transport continuity and avoids loops; formula (6) ensures that there is sufficient transport capacity between nodes; formula (7) ensures that there is sufficient transshipment capacity at the node; and formula (8) ensures that the total transport time is less than the required arrival time.

[0044] Furthermore, according to the constraint conditions, the improved genetic algorithm is used to solve the grain transportation mathematical model to obtain the optimal transportation path combining the loading mode and the transportation mode. The optimal transportation path is the transportation path with the minimum actual total transportation cost within the normal transportation time and reasonable transportation loss, including:

[0045] S41, using an improved genetic algorithm to perform chromosome encoding and decoding on the loading mode and transportation node;

[0046] S42, generating an initial population of encoded and decoded chromosomes that meet the transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints, wherein the initial population is a transportation scheme formed by combining different transportation modes, different loading modes and different nodes;

[0047] S43, using the elite retention strategy to process the initial population, and directly retaining E chromosomes in the initial population to the offspring according to the degree of excellence;

[0048] S44, calculating the merits of the chromosomes of the initial population, and processing the merits using a roulette wheel method, and selecting two parent chromosomes from the initial population;

[0049] S45, performing crossover and mutation operations on the two parent chromosomes to obtain two new daughter chromosomes;

[0050] S46, repeating S44-S45, selecting different parent chromosomes to perform crossover and mutation operations, until the number of new offspring chromosomes reaches a preset value, obtaining an offspring chromosome population, and forming a new offspring population equal to the initial population size together with the E chromosomes directly retained in the offspring;

[0051] S47. Repeat steps S42-S46, select transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints of different transportation plans, construct multi-generation chromosome populations, and stop calculating until the improved genetic algorithm solution converges, and output the relevant solution information of the optimal chromosome in the latest generation population, so as to obtain the optimal grain transportation plan.

[0052] Furthermore, the loading mode is encoded with a real value, and the loading mode is set to 1, 2, and 3; 1 represents bagging, 2 represents bulk, and 3 represents container loading;

[0053] The transport node is coded in binary, 1 means the node is selected, and 0 means the node is not selected.

[0054] Furthermore, the generation process of the initial population is:

[0055] S421, constructing a node sequence table, wherein the node sequence table adopts a combination of real-value coding and binary coding, wherein the real values ​​are 1, 2, and 3, respectively representing bagged, bulk, and container loading, and numbers 1, 2, 3, ..., R represent nodes in the transportation network, wherein number 1 is the initial node, and number R is the terminal node;

[0056] S422, randomly select a loading mode, and randomly select the next node connected to the initial node 1;

[0057] S423, repeat the node selection operation until the set end node is reached, and a chromosome with a length of n+1 is obtained, where n is the number of nodes in the node set N;

[0058] S424. Repeat steps S422-S423, select different nodes, until the number of chromosomes reaches a preset value, and obtain an initial population that satisfies the transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints.

[0059] Furthermore, the initial population is processed using an elite retention strategy, and E chromosomes in the initial population are directly retained in the offspring according to the magnitude of the merits, including:

[0060] Sort the chromosomes in the population in descending order of merit, and keep the top E chromosomes as elite individuals directly in the offspring. The elite retention strategy is shown in formula (9):

[0061] E = round(r × N) (9)

[0062] Among them, E is the number of elite chromosomes, r is the elite retention rate, N is the population size, and round is the rounding operation.

[0063] Further, the merits of the chromosomes of the initial population are calculated, and the merits are processed using a roulette wheel method, and two parent chromosomes are selected from the initial population, including:

[0064] S441, Calculate the degree of superiority

[0065] Assume f(a) is the total cost C corresponding to the a-th chromosome that satisfies the constraint condition, and calculate the merit value according to the total cost, as shown in formula (10):

[0066]

[0067] S442. According to the merit value, two parent chromosomes are selected from the initial population by using a roulette wheel method, as shown in formula (11).

[0068]

[0069] Where N is the population size, F a is the merit value of the ath chromosome, and F is the sum of the merits of all chromosomes;

[0070] The probability of a chromosome being selected is calculated based on the sum of the merit value and the chromosome merit, as shown in formula (12):

[0071]

[0072] Among them, p a represents the probability of the ath chromosome being selected.

[0073] Further, the two parent chromosomes are subjected to crossover and mutation operations to obtain two new daughter chromosomes, including:

[0074] Find out the number of identical transport nodes passed by the two parent chromosomes, marked as L, and randomly select a value S between 1 and L in the order of the number of nodes from small to large, and use the Sth identical transport node passed as the crossover starting point, and the gene between the starting point and the arrival point as the crossover fragment;

[0075] Perform a crossover operation on the two parent chromosomes at the starting gene point of the crossover segment, that is, exchange the gene crossover segments of the two parent chromosomes to obtain two new daughter chromosomes;

[0076] A mutation operator is introduced to mutate the carrier mode fund point in the parent chromosome, and the mutated chromosome is transferred to the new daughter chromosome to obtain two mutated daughter chromosomes.

[0077] The present invention discloses a method for jointly optimizing grain loading modes and transportation routes, selects parameters such as the loading speed and loss rate difference of different loading modes such as bagging, bulk and container loading, and constructs a mathematical model including transportation cost, carbon emission cost, loss cost and loading modes of different transportation modes, and solves the mathematical model using an improved genetic algorithm. Compared with the prior art that weakens the influence of loading modes on transportation organization, the method makes the differentiated demand response of consignors in terms of transportation time, transportation quality, transportation cost, etc. more accurate, and takes into account different loading modes, transportation modes and more loss costs, so that the results obtained in determining the grain transportation route are better and more resources are saved. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0079] Figure 1 A method flow chart of a combined optimization method of grain loading mode and transportation path according to the present invention;

[0080] Figure 2 This is a schematic diagram of the transport node conversion of the present invention;

[0081] Figure 3 The total cost and cost composition of the optimal transportation path for different loading modes of the present invention;

[0082] Figure 4 It is a schematic diagram of the proportion of road, rail and water transportation structure of the optimal transportation scheme under different time constraints of the present invention;

[0083] Figure 5 Schematic diagram of genetic algorithm coding;

[0084] Figure 6 This is a schematic diagram of chromosome crossover and mutation in genetic algorithm;

[0085] Figure 7 The figure is a schematic diagram of a transportation network according to an embodiment of the present invention. DETAILED DESCRIPTION

[0086] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0087] This embodiment provides a method for jointly optimizing grain loading mode and transportation path. Figure 1 As shown, including;

[0088] S1: Obtain the normal transportation time and reasonable transportation loss of grain from the origin to the destination, where the reasonable transportation loss is the natural loss of grain during the transportation process;

[0089] S2: Obtaining transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters in the grain transportation process, constructing transportation decision variables according to the transportation parameters, and constructing a grain transportation objective function and constraint conditions with the minimum actual total transportation cost according to the four parameters and the transportation decision variables; the transportation parameters include transportation mode and loading mode, and the loss cost parameters include loss rates under different loading modes and different transportation modes;

[0090] S3: Integrate the four parameters, transportation decision variables, grain transportation objective function and constraint conditions to obtain a grain transportation mathematical model, wherein the grain transportation mathematical model is used to determine an optimal loading method and an optimal transportation route for grain;

[0091] S4: Based on the constraints, an improved genetic algorithm is used to solve the grain transportation mathematical model to obtain an optimal transportation path that combines the loading method and the transportation method. The optimal transportation path is a transportation path with the lowest actual total transportation cost within normal transportation time and reasonable transportation losses.

[0092] Specifically, firstly, the normal transportation time and reasonable transportation loss of grain from the origin to the destination are obtained, and the reasonable transportation loss is the natural loss of grain during the transportation process. Then, the transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters of the grain transportation process are obtained according to the preset origin and destination. According to the transportation parameters, transportation decision variables are constructed, and according to the four parameters and the transportation decision variables, the objective function and constraint conditions of grain transportation with the minimum actual total transportation cost are constructed. The transportation parameters include transportation mode and loading mode, and the loss cost parameters include loss rates under different loading modes and different transportation modes. The transportation cost, carbon emission cost and loss cost are taken into account, revealing the hidden cost part in the total cost structure. , which provides a more accurate basis for total cost estimation and can propose the optimal transportation organization plan; integrates the four parameters, transportation decision variables, grain transportation objective function and constraints to obtain a grain transportation mathematical model, which is used to determine the optimal loading method and optimal transportation path for grain, forming a mathematical model to facilitate actual calculations; finally, according to the constraints, uses an improved genetic algorithm to solve the grain transportation mathematical model, and obtains the optimal transportation path combining the loading method and the transportation method. The optimal transportation path is the transportation path with the minimum actual total transportation cost within normal transportation time and reasonable transportation loss. The use of an improved genetic algorithm can jump out of the local optimal solution and obtain the global optimal solution, that is, the optimal grain transportation plan.

[0093] In a specific embodiment, the scheme for obtaining the normal transportation time and reasonable transportation loss of grain from the origin to the destination, where the reasonable transportation loss is the natural loss of grain during the transportation process, is:

[0094] Determine the relative distance between the grain transportation origin and destination based on actual conditions, determine the normal transportation time, and determine the natural loss of grain under different transportation methods, loading methods, and transshipment methods during the transportation time.

[0095] In a specific embodiment, the transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters in the grain transportation process are obtained, and the transportation decision variables are constructed according to the transportation parameters, and the grain transportation objective function and constraint conditions with the minimum actual transportation total cost are constructed according to the four parameters and the transportation decision variables; the transportation parameters include the transportation mode and the loading mode, and the loss cost parameters include the loss rate under different loading modes and different transportation modes.

[0096] S21. Obtaining transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters during grain transportation:

[0097] The transport parameters include but are not limited to:

[0098] G = N, A, K, represents the grain transportation network; N is the set of transportation nodes, N = 1, 2, 3, ...i, ...j; A is the set of transportation arcs, representing the connection between different modes of transportation at the same location; K is the set of transportation modes, K = {1, 2, 3}, 1, 2, 3 represent road, rail and water transportation respectively; M is the set of loading modes, M = {1, 2, 3}, 1, 2, 3 represent bag, bulk and container loading respectively;

[0099] Q is the total weight of the transported goods, in tons;

[0100] ——Transport distance of the kth transport mode between nodes i and j (km); Table 1 shows the transport distance of each transport arc segment,

[0101] Table 1 Transport distance of transport arc

[0102]

[0103] ——The unit loading speed (t / h) of converting the transportation mode k between nodes i and j to g at node j when the mth loading mode is selected;

[0104] ——Transport speed of the kth mode of transport between nodes i and j (km / h);

[0105] ——The maximum transport capacity of the kth transport mode between nodes i and j when the mth loading mode is selected, m∈M, M={1,2,3};

[0106] ——The maximum reloading capacity at node j when the mth loading mode is selected;

[0107] T AO ——Final arrival time of goods (h);

[0108] T——latest arrival time required for goods (h);

[0109] The transportation cost parameters include but are not limited to:

[0110] C – total transportation cost;

[0111] ——Unit transportation cost (yuan / (t·km)) of transporting goods between nodes i and j using the mth loading method and the kth transportation method, g∈K, K={1,2,3};

[0112] ——Under the mth loading mode, the unit transshipment cost (yuan / t) of converting the transport mode k between nodes i and j to g at node j;

[0113] The carbon emission cost parameters include but are not limited to:

[0114] c e ——Unit carbon emission cost (yuan / kg);

[0115] ——Specific carbon emissions of the kth mode of transportation between nodes i and j (kg / (t·km));

[0116] ——the unit carbon emissions (kg / t) of converting the transport mode k between nodes i and j to g at node j;

[0117] The loss cost parameters include but are not limited to:

[0118] ——Under the mth loading mode, the unit loss rate (%) of the kth transportation mode between nodes i and j;

[0119] ——Under the mth loading mode, the unit loss rate (%) of converting the transportation mode k between nodes i and j to g at node j, g∈K;

[0120] p——unit value of grain (yuan / t);

[0121] The transportation costs, unit carbon emissions, transportation loss rates and transportation speeds of different transportation and loading methods are shown in Table 2. The transshipment costs, carbon emissions, transshipment loss rates and transshipment speeds of transshipment between different transportation methods are shown in Table 3.

[0122] Table 2 Transportation costs, unit carbon emissions, transportation loss rate and transportation speed of different transportation and loading methods

[0123]

[0124]

[0125] Table 3 Transshipment costs, carbon emissions, transshipment loss rate and transshipment speed between different modes of transport

[0126]

[0127] S22. Constructing transportation decision variables according to the transportation mode and loading mode, wherein the transportation decision variables include but are not limited to:

[0128]

[0129] S23. Constructing a grain transportation objective function with the minimum actual transportation total cost according to the four parameters and the transportation decision variables, as shown in formula (13):

[0130]

[0131] S14, constructing constraint conditions according to the four parameters and the transportation decision variables, as shown in formulas (14) to (20),

[0132]

[0133] in, represents the decision variable for selecting mode k between transport nodes j and i; formula (14) indicates that only one mode of transport can be selected between any nodes; formula (15) indicates that at most one mode of transport can be changed at a transport node; formula (16) indicates that only one loading method can be selected between any nodes; formula (17) ensures transport continuity and avoids loops; formula (18) ensures that there is sufficient transport capacity between nodes; formula (19) ensures that there is sufficient transshipment capacity at the node; and formula (20) ensures that the total transport time is less than the required arrival time.

[0134] In this plan, transportation costs, carbon emission costs and loss costs are taken into account, revealing the hidden cost part of the total cost structure, making the total cost calculation more accurate and able to propose the optimal transportation organization plan. The composition of various costs is as follows: Figure 3 As shown in the figure, it can be seen that the loss cost considered in this plan accounts for a large proportion and is a key factor affecting the design of the transportation organization plan. Therefore, considering the loss cost can obtain a better transportation route; the loading speed and loss rate differences of different loading methods such as bagging, bulk and container loading are selected and defined, such as Figure 4 As shown in the figure, the proportion of road, rail and water transportation structure of the optimal transportation plan under different time constraints is demonstrated, revealing the difference characteristics of the optimal transportation plan under different time demand scenarios of consignors. The proportion of transportation modes selected at different times is different. Compared with the existing technology that weakens the impact of loading methods on transportation organization, the differentiated demand response of consignors in terms of transportation time, transportation quality, transportation cost, etc. is more accurate.

[0135] In a specific embodiment, the four parameters, transportation decision variables, grain transportation objective function and constraint conditions are integrated to obtain a grain transportation mathematical model, and the grain transportation mathematical model is used to determine the optimal loading method and optimal transportation path of grain:

[0136] The four parameters, transportation decision variables, grain transportation objective function and constraints are integrated to form a grain transportation mathematical model including a variety of cost parameters, transportation mode related parameters, loading mode related parameters, transportation decision variables, grain transportation objective function and constraints. That is, the transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters, decision variables and formulas (13) to (20) are integrated to form a grain transportation mathematical model.

[0137] In this solution, a mathematical model is formed to provide a data basis for path planning.

[0138] In a specific embodiment, according to the constraint conditions, an improved genetic algorithm is used to solve the grain transportation mathematical model to obtain an optimal transportation path combining the loading method and the transportation method. The optimal transportation path is a transportation path with the minimum actual total transportation cost within normal transportation time and reasonable transportation loss. The solution is:

[0139] S41, using an improved genetic algorithm to perform chromosome encoding and decoding on the loading mode and transportation node;

[0140] The loading mode is encoded with a real value, and the loading mode is set to 1, 2, and 3; 1 represents bagging, 2 represents bulk, and 3 represents container loading, and the decoding process is the use process;

[0141] The transport node is coded using binary code, 1 is set to select the node, and 0 is set to not select the node;

[0142] by Figure 2 For example, the transportation network Figure 5 The code is shown to indicate that the bulk loading method is selected, and the transportation route is 1-2-4-6-8-10-15-16-17-18, that is, the grain is transported from node 1 to node 2 by road in bulk loading, and the road is changed to the railway of node 4 at node 2, and the grain is transported from node 4 to node 6 by railway, and the railway is changed to the waterway of node 8 at node 6, and the grain is transported from node 8 to node 10 by waterway, and the grain is transported from node 8 to node 10 by waterway, and the grain is transported from node 15 to node 16 by railway, and the grain is transported from node 16 to node 17 by road at node 17, and the grain is transported from node 17 to terminal 18 by road.

[0143] S42, generating an initial population of encoded and decoded chromosomes that meet the transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints, wherein the initial population is a transportation scheme formed by combining different transportation modes, different loading modes and different nodes:

[0144] S421, constructing a node sequence table, wherein the node sequence table adopts a combination of real-value coding and binary coding, wherein the real values ​​are 1, 2, and 3, respectively representing bagged, bulk, and container loading, and numbers 1, 2, 3, ..., R represent nodes in the transportation network, wherein number 1 is the initial node, and number R is the terminal node;

[0145] S422, randomly select a loading mode, and randomly select the next node connected to the initial node 1;

[0146] S423, repeat the node selection operation until the set end node is reached, and a chromosome with a length of n+1 is obtained, where n is the number of nodes in the node set N;

[0147] S424, repeating steps S422-S423, selecting different nodes, until the number of chromosomes reaches a preset value, and obtaining an initial population that satisfies the transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints. In this embodiment, the number of chromosomes of the selected initial population is 200;

[0148] S43. S43. Use the elite retention strategy to process the initial population, and directly retain E chromosomes in the initial population to the offspring according to the size of the merits:

[0149] Sort the chromosomes in the population in descending order of merit, and keep the top E chromosomes as elite individuals directly in the offspring. The elite retention strategy is shown in formula (21):

[0150] E = round(r × N) (21)

[0151] Where E is the number of elite chromosomes, r is the elite retention rate, the value of r is set to [0.1, 0.15], N is the population size, and round is a rounding operation;

[0152] The reason for adopting the elite retention strategy is that after the crossover and mutation operations, the quality of the offspring solution individuals cannot be guaranteed, which will cause the loss of excellent genes of the parent solution individuals, and unnecessary redundant operations will be generated in the solution process, which will affect the optimization ability and solution speed of the solution. Therefore, the elite retention strategy is introduced to directly inherit the excellent parent individuals to the offspring individual population at a certain random ratio. The solution individuals are the chromosomes that meet the constraints.

[0153] S44, calculating the merits of the chromosomes of the initial population, and processing the merits using a roulette wheel method, and selecting two parent chromosomes from the initial population:

[0154] S441, calculate fitness

[0155] Let f(a) be the total cost C corresponding to the a-th chromosome that satisfies the constraint condition, and calculate the merit value according to the total cost, as shown in formula (22),

[0156]

[0157] The larger the f(a) value is, the worse the solution effect of the chromosome that satisfies the constraint condition is. Therefore, the inverse of the f(a) value is marked as the quality F of the ath chromosome that satisfies the constraint condition. a , the larger the merit value is, the better the chromosome that satisfies the constraint condition is;

[0158] S442. According to the merit value, two parent chromosomes are selected from the initial population by using a roulette wheel method, as shown in formula (23).

[0159]

[0160] Map the merit value to the area of ​​the roulette wheel. The higher the merit value, the larger the corresponding area, and the greater the probability of being selected. N is the population size, and F is the a is the merit value of the ath chromosome, and F is the sum of the merits of all chromosomes;

[0161] The probability of a chromosome being selected is calculated based on the sum of the merit value and the chromosome merit, as shown in formula (24):

[0162]

[0163] Among them, p a represents the probability of the ath chromosome being selected;

[0164] S45, performing crossover and mutation operations on the two parent chromosomes to obtain two new daughter chromosomes:

[0165] The operation process of the crossover operator is to find out the number of nodes of the same transport nodes passed by the two parent chromosomes, marked as L, and randomly select a value S between 1 and L in the order of the number of nodes from small to large, and use the Sth same transport node passed as the crossover starting point, and the genes between the starting point and the arrival point as the crossover fragment;

[0166] like Figure 5 As shown, at the starting gene point of the crossover segment, i.e., point b in the figure, a crossover operation is performed on the two parent chromosomes, that is, the gene crossover segments of the two parent chromosomes are exchanged to obtain two new offspring;

[0167] At the same time, a mutation operator is introduced to mutate the carrier mode fund point in the parent chromosome with a certain probability, such as Figure 6As shown at the midpoint c, the transport mode in the parent generation 1 is changed from bulk represented by 2 to container loading represented by 3, and is inherited to the offspring 1 to obtain the mutated offspring;

[0168] S46, repeating S44-S45, selecting different parent chromosomes to perform crossover and mutation operations, until the number of new offspring chromosomes reaches a preset value, i.e., 200-E, to obtain an offspring chromosome population, and together with the E chromosomes directly retained in the offspring, form a new offspring population equal to the initial population size;

[0169] S47. Repeat steps S42-S46, select transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints of different transportation plans, construct multi-generation chromosome populations, and stop calculating until the improved genetic algorithm solution converges, and output the relevant solution information of the optimal chromosome in the latest generation population, so as to obtain the optimal grain transportation plan; in this embodiment, the number of iterations selected according to actual conditions and experience is 150 times.

[0170] In this scheme, an improved genetic algorithm is selected to solve the optimal path. The improved genetic algorithm has global search capabilities, can jump out of the local optimal solution and obtain the global optimal solution. It can evaluate individuals in the population at the same time, and can also perform operations such as selection, crossover, and mutation in parallel, which can improve the calculation speed of the algorithm; using the elite retention strategy, the chromosomes with better genes can be inherited as a whole to the next generation of the population, which is conducive to improving the convergence speed of the algorithm. Crossover and mutation operations on chromosomes can enrich the solution space, so that the obtained solution is the global optimal solution and avoid falling into the local optimal solution.

[0171] The present invention takes the transportation of 1500 tons of grain in a certain place as an example:

[0172] A batch of grain weighing 1,500 tons is to be transported from a grain depot in Qiqihar City (indicated by the letter A) to an industrial park in Ganzhou City (indicated by the letter O). During the transportation, it passes through 13 cities represented by letters B to N, including Harbin and Shenyang. The transportation network is as follows: Figure 7 As shown;

[0173] By adding road, rail and water conversion nodes, the various transportation routes between cities are converted into a single transportation arc between any two nodes. The converted nodes are numbered from 1 to 32, and the transportation arcs are numbered from 1 to 36. Considering the three different loading methods of bagging, bulk and container loading, the goal is to minimize the total cost including transportation cost, carbon emission cost and loss cost, and determine the reasonable transportation route and loading method;

[0174] The genetic algorithm is used to solve the optimal path, transportation arc-transportation method, total cost and transportation time under different loading methods as shown in Table 1.

[0175] Table 1 Optimal path solutions under different loading modes

[0176]

[0177]

[0178] A comparison shows that the optimal routes for the three loading methods are the same, and they adopt a road-rail-water transport organizational model. At the same time, the total cost of container loading is significantly better than the total cost of bulk and bagged loading. Under the optimal total cost condition, the transportation time of the optimal container loading route is 130.6 hours, which is better than the 139.6 hours under the optimal bulk transportation route and the 135.6 hours under the optimal bagged transportation route. Analyzing the proportion of different transportation modes in the total transportation route, the road transportation distance accounts for 1.06% of the optimal route, the railway transportation distance accounts for 39.16%, and the water transportation distance accounts for 59.78%, reflecting the advantages of waterways and railways in long-distance transportation and the role of roads in distribution at both ends.

[0179] Explanation of the optimal solution: Grain is loaded in containers, starting from the initial node 1, passing through nodes 2, 3, 5, 10, 14, 20, 23, 27, 30, and 31 to reach the terminal node 32. Road transportation is used between nodes 1 and 2, road transportation is transferred to railway operation between nodes 2 and 3, railway transportation is used between nodes 3 and 5, railway transportation is used between nodes 5 and 10, railway transportation is used between nodes 10 and 14, railway transportation is transferred to water transportation between nodes 14 and 20, water transportation is used between nodes 20 and 23, water transportation is used between nodes 23 and 27, railway transportation is used between nodes 27 and 30, railway transportation is transferred to road operation between nodes 30 and 31, and road transportation is used between nodes 31 and 32.

[0180] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for joint optimization of grain loading mode and transportation path, characterized in that: include: S1: Obtain the normal transportation time and reasonable transportation loss of grain from the origin to the destination, where the reasonable transportation loss is the natural loss of grain during the transportation process; S2: Obtain transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters in the grain transportation process, construct transportation decision variables according to the transportation parameters, and construct the grain transportation objective function and constraint conditions with the minimum actual transportation total cost according to the four parameters and the transportation decision variables; the transportation parameters include transportation mode and loading mode, and the loss cost parameters include loss rates under different loading modes and different transportation modes; the specific steps of S2 are as follows: S21. Obtaining transportation parameters, transportation cost parameters, carbon emission cost parameters and loss cost parameters during grain transportation: The transport parameters include: G = N, A, K, G represents the grain transportation network; N is the set of transportation nodes, N = 1, 2, 3, ... i, ... j; A is the set of transportation arcs, which represents the connection between different modes of transportation at the same location; K is the set of transportation modes, K = {1, 2, 3}, 1, 2, 3 represent road, rail and water transportation respectively; M is the set of loading modes, M = {1, 2, 3}, 1, 2, 3 represent bagged, bulk and container loading respectively; Q is the total weight of the transported goods, in tons; ——Transport distance of the kth transport mode between transport nodes i and j (km); ——The unit loading speed (t / h) of converting the transport mode k between transport nodes i and j to g at transport node j when the mth loading mode is selected; ——Transport speed of the kth transport mode between transport nodes i and j (km / h); ——The maximum transport capacity of the kth transport mode between transport nodes i and j under the mth loading mode, m∈M, M={1,2,3}; ——The maximum transshipment capacity at transport node j when the mth loading mode is selected; T AO ——Final arrival time of goods (h); T——latest arrival time required for goods (h); The transportation cost parameters include: C – total transportation cost; ——Unit transportation cost (yuan / (t·km)) of transporting goods between transport nodes i and j using the mth loading method via the kth transport method, g∈K, K={1,2,3}; ——Under the mth loading mode, the unit transshipment cost (yuan / t) of converting the transport mode k between transport nodes i and j to g at transport node j; The carbon emission cost parameters include: c e ——Unit carbon emission cost (yuan / kg); ——Specific carbon emissions of the kth mode of transport between transport nodes i and j (kg / (t·km)); ——the unit carbon emissions (kg / t) of transport mode k between transport nodes i and j converted to g at transport node j; The loss cost parameters include: ——Under the mth loading mode, the unit loss rate (%) of the kth transportation mode between transportation nodes i and j; ——Under the mth loading mode, the unit loss rate (%) of converting the transport mode k between transport nodes i and j to g at transport node j, g∈K; p——unit value of grain (yuan / t); S22. Constructing transportation decision variables according to the transportation mode and loading mode, wherein the transportation decision variables include: S23, constructing a grain transportation objective function with the minimum actual transportation total cost according to the four parameters and the transportation decision variables, as shown in formula (1), S24, constructing constraint conditions according to the four parameters and the transportation decision variables, as shown in formula (2) to formula (8), in, represents the decision variable for selecting transport mode k between transport nodes j and i; formula (2) indicates that only one transport mode can be selected between any transport nodes, formula (3) indicates that at most one transport mode change occurs at a transport node, formula (4) indicates that only one loading method can be selected between any transport nodes, formula (5) ensures transport continuity and avoids loop formation, formula (6) ensures that there is sufficient transport capacity between transport nodes, formula (7) ensures that there is sufficient transshipment capacity at the transport node, and formula (8) ensures that the total transport time is less than the required delivery time; S3: Integrate the four parameters, transportation decision variables, grain transportation objective function and constraint conditions to obtain a grain transportation mathematical model, wherein the grain transportation mathematical model is used to determine an optimal loading method and an optimal transportation route for grain; S4: Based on the constraints, an improved genetic algorithm is used to solve the grain transportation mathematical model to obtain an optimal transportation path that combines the loading method and the transportation method. The optimal transportation path is a transportation path with the lowest actual total transportation cost within normal transportation time and reasonable transportation losses.

2. A method for joint optimization of grain loading mode and transportation path according to claim 1, characterized in that: S4 solves the grain transportation mathematical model using an improved genetic algorithm according to the constraint conditions to obtain an optimal transportation path combining the loading method and the transportation method. The optimal transportation path is a transportation path with the minimum actual total transportation cost within normal transportation time and reasonable transportation loss, including: S41, using an improved genetic algorithm to perform chromosome encoding and decoding on the loading mode and transportation node; S42, generating an initial population of encoded and decoded chromosomes that meet the transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints, wherein the initial population is a transportation scheme formed by combining different transportation modes, different loading modes and different transportation nodes; S43, using the elite retention strategy to process the initial population, and directly retaining E chromosomes in the initial population to the offspring according to the degree of excellence; S44, calculating the merits of the chromosomes of the initial population, and processing the merits using a roulette wheel method, and selecting two parent chromosomes from the initial population; S45, performing crossover and mutation operations on the two parent chromosomes to obtain two new daughter chromosomes; S46, repeating S44-S45, selecting different parent chromosomes to perform crossover and mutation operations, until the number of new offspring chromosomes reaches a preset value, obtaining an offspring chromosome population, and forming a new offspring population equal to the initial population size together with the E chromosomes directly retained in the offspring; S47. Repeat steps S42-S46, select transportation parameters, transportation cost parameters, carbon emission cost parameters, loss cost parameters and constraints of different transportation plans, construct multi-generation chromosome populations, and stop calculating until the improved genetic algorithm solution converges, and output the relevant solution information of the optimal chromosome in the latest generation population, so as to obtain the optimal grain transportation plan.

3. A method for joint optimization of grain loading mode and transportation path according to claim 2, characterized in that: The loading mode is encoded with a real value, and the loading mode is set to 1, 2, or 3; 1 represents bagging, 2 represents bulk, and 3 represents container loading; The transport node is coded in binary, 1 means the transport node is selected, and 0 means the transport node is not selected.

4. A method for joint optimization of grain loading mode and transportation path according to claim 2, characterized in that: The generation process of the initial population is: S421, constructing a transportation node sequence table, wherein the transportation node sequence table adopts a combination of real-valued coding and binary coding, wherein the real values ​​are 1, 2, and 3, respectively representing bagged, bulk, and container loading, and numbers 1, 2, 3, ..., R represent transportation nodes in the transportation network, wherein number 1 is the initial node, and number R is the terminal node; S422, randomly select a loading method, and randomly select the next transportation node connected to the initial node 1; S423, repeating the operation of transport node selection until reaching the set end node, and obtaining a chromosome with a length of n+1, where n is the number of transport nodes in the transport node set N; S424. Repeat steps S422-S423, select different transport nodes, until the number of chromosomes reaches a preset value, and obtain an initial population that meets the transport parameters, transport cost parameters, carbon emission cost parameters, loss cost parameters and constraints.

5. The method for joint optimization of grain loading mode and transportation path according to claim 2, characterized in that: The initial population is processed using an elite retention strategy, and E chromosomes in the initial population are directly retained in the offspring according to the magnitude of the merits, including: Sort the chromosomes in the population in descending order of merit, and keep the top E chromosomes as elite individuals directly in the offspring. The elite retention strategy is shown in formula (9): E = round(r × N) (9) Among them, E is the number of elite chromosomes, r is the elite retention rate, N is the population size, and round is the rounding operation.

6. A method for joint optimization of grain loading mode and transportation path according to claim 2, characterized in that: Calculating the merits of the chromosomes of the initial population, and processing the merits using a roulette wheel method, and selecting two parent chromosomes from the initial population, including: S441, Calculate the degree of superiority Assume f(a) is the total cost C corresponding to the a-th chromosome that satisfies the constraint condition, and calculate the merit value according to the total cost, as shown in formula (10): S442. According to the merit value, two parent chromosomes are selected from the initial population by using a roulette wheel method, as shown in formula (11). Where N is the population size, F a is the merit value of the ath chromosome, and F is the sum of the merits of all chromosomes; The probability of a chromosome being selected is calculated based on the sum of the merit value and the chromosome merit, as shown in formula (12): Among them, p a represents the probability of the ath chromosome being selected.

7. A method for joint optimization of grain loading mode and transportation path according to claim 2, characterized in that: Performing crossover and mutation operations on the two parent chromosomes to obtain two new daughter chromosomes, including: Find out the number of transport nodes that the two parent chromosomes pass through, marked as L, and randomly select a value S between 1 and L in the order of the number of transport nodes from small to large, and use the Sth identical transport node as the crossover starting point, and the gene between the starting point and the arrival point as the crossover fragment; Perform a crossover operation on the two parent chromosomes at the starting gene point of the crossover segment, that is, exchange the gene crossover segments of the two parent chromosomes to obtain two new daughter chromosomes; The mutation operator is introduced to mutate the carrier mode gene point in the parent chromosome, and the mutated chromosome is transferred to the new daughter chromosome to obtain two mutated daughter chromosomes.

Images