A multi-modal transport multi-objective decision-making method for quantifying carbon emission impact

Abstract

The application discloses a kind of multimodal transport multi-objective decision research methods of quantifying carbon emission influence, comprising the following steps: step 1, according to transport node, route and freight parameter, establish multimodal transport multi-objective path decision model under uncertain condition;Step 2, the optimization target and constraint condition in multimodal transport multi-objective path decision problem under uncertain condition are constructed by processing uncertain condition;Step 3, path decision model under different carbon emission policy is established, and path optimization under the influence of carbon emission policy is carried out.The present application makes up the method deficiency of current large space scale regional carbon emission policy on traffic decision influence, solves the quantification problem of the nature and ability of path decision influence under uncertain condition and carbon emission policy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of traffic planning, and in particular to a multi-modal transport multi-objective decision-making research method for quantifying the influence of carbon emissions. BACKGROUND

[0002] Implementing reasonable carbon emission policies helps to better reduce carbon emissions and transportation costs, alleviate road congestion, and promote the popularization of multi-modal transport, forming a virtuous cycle of comprehensive transport development.

[0003] Currently, in the process of multi-modal transport, more and more scholars carry out path optimization research under carbon emission policies based on the consideration of transport environmental protection, and existing research has proved that carbon emission policies have a significant impact on path decision-making. However, the line network characteristics and regional economic volume of the real multi-modal transport network will affect the implementation of carbon emission policies. At this time, the significance of the influence of carbon emission policies on path decision-making will also change. At present, most of the research adopts a virtual network set by man, and whether the experimental results and analysis conclusions can be applied to the actual transport process still needs to be further verified. Therefore, how to construct a reasonable multi-modal transport multi-objective decision-making method that can quantify the influence of carbon emission policies has become a problem to be solved. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a multi-modal transport multi-objective decision-making research method for quantifying the influence of carbon emissions, which can solve the problem of quantifying the influence of uncertain conditions and carbon emission policies on path decision-making.

[0005] To solve the above technical problems, the present application provides a multi-modal transport multi-objective decision-making research method for quantifying the influence of carbon emissions, comprising the following steps:

[0006] Step 1: establishing a multi-modal transport multi-objective path decision-making model under uncertain conditions according to the transportation nodes, routes and freight parameters;

[0007] Step 2: processing the uncertain conditions to construct the optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision-making problem under uncertain conditions;

[0008] Step 3: establishing a path decision-making model under different carbon emission policies to optimize the path under the influence of carbon emission policies.

[0009] Preferably, in step 1, the freight parameters include transportation speed, node transfer time, delivery time and transportation cost.

[0010] Preferably, in step 1, the multi-modal transport multi-objective path decision model under uncertain conditions is specifically as follows: two optimization objectives of expected total transportation time and expected total transportation cost are involved, and a multi-objective random expected value model as shown in equation (1) is established:

[0011]

[0012] wherein x is an n-dimensional decision vector, ξ is a random vector, f i (x,ξ) is an objective function, i=1,2,…,m, g j (x,ξ) is a set of random constraint functions, j=1,2,…,p, and E is an expected value operator.

[0013] Preferably, in step 2, the optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision problem under uncertain conditions are constructed, which are specifically as follows: the optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision problem under uncertain conditions are constructed, wherein the total transportation time T0 includes the transportation time T1 between cargo nodes and the transfer time T2 at the nodes, both of which are uncertain values, and the expected total transportation time E(T0) in the multi-objective random expected value model is as follows:

[0014]

[0015]

[0016] E(T0)=E(T1)+E(T2) (4)

[0017] In equations (2), (3) and (4), h, i, j represent transportation nodes, h, i, j∈N; (h, i) represents a transportation path; k, l represent cargo transportation modes, k, l∈M; (Km) represents the distance of path (h, i) under mode k; M is a set of transportation modes; M (h,i) (Km) represents the distance of path (h, i) under mode k; M is a set of transportation modes; M N is a set of transportation nodes; N i (Km) represents the distance of path (h, i) under mode k; M is a set of transportation modes; M (h) represents the transfer time of mode k to mode l at node i; (Km) represents the distance of path (h, i) under mode k; (Km / h) represents the transportation speed of path (h, i) under mode k; is a decision variable, if the cargo is transported through (h, i) by mode k, otherwise is a decision variable, if the transportation mode is converted at i Otherwise

[0018] The total transportation cost C0 includes the sum of transportation cost and transshipment cost C1, and the delay penalty cost C2; wherein the transportation cost and transshipment cost are irrelevant to the transportation time, the delay penalty cost is relevant to the total transportation time, and belongs to uncertain value, the expected total transportation cost E(C0) is expressed as follows:

[0019]

[0020] C2 = max(μ, 0)C od (T0-T od ) (6)

[0021] E(C0) = C1 + E[C2] (7)

[0022] In the formula (5) (6) (7), represents the transportation cost (yuan) of the transportation path (h, i) under the kth transportation mode; represents the transshipment cost (yuan) of the node i from the kth mode to the lth mode; Q0 is the total weight of the goods (TEU); T0 is the total transportation time of the goods (h); Tod is the agreed delivery time (h); Cod is the unit delay penalty cost (yuan / h); max(μ, 0) is the max function, which finds the maximum value of μ and 0;

[0023] Under the deterministic condition, whether the goods can be delivered to the destination on time only exists two cases of yes and no; but under the uncertain condition, whether the goods can be delivered on time belongs to the random variable subject to 0-1 distribution, the probability and expectation of on-time delivery are equal, so the variable h(λ r ) is defined, if the goods are delivered on time, i.e. T0-T od < 0, h(λ r ) = 1; otherwise h(λ r ) = 0, the expected constraint of on-time delivery E[h(λ r )] is expressed as:

[0024] E[h(λ r )] = P(T0-T od < 0) ≥ α (8)

[0025] Wherein, T0 is the total transportation time of the goods (h); T od is the agreed delivery time (h); α is the expected threshold of on-time delivery of goods, α ∈ (0, 1), the setting of the threshold eliminates part of the path scheme which performs poorly in transportation time, reduces the calculation complexity, and the constraint allows the total transportation time of part of the path to exceed the transportation time limit T od , and a penalty cost based on the total transportation cost must be paid;

[0026] Estimate the expected total transportation time and the expected constraint of on-time delivery, random simulation generates r groups of samples Where 1 < r < R, when R→∞,

[0027]

[0028]

[0029] When the total number of random simulation R is high enough, the left side of formula (9), (10) is the estimated value of the expected total transportation time, the expected delay penalty cost and the expected constraint of on-time delivery.

[0030] Preferably, in step 3, the path decision model under different carbon emission policies is established, and the path optimization under the influence of carbon emission policies specifically includes the following steps:

[0031] Step S31, without considering the transportation carbon emission:

[0032] Under this condition, without considering the carbon emissions generated in the multimodal transport process, the minimum total transportation time T0 and total transportation cost C0 are taken as the target, and the expected estimated value is taken as the constraint, which is specifically as follows:

[0033]

[0034] Step S32, considering the case of mandatory carbon emission policy;

[0035] The mandatory carbon emission policy refers to that the regulatory agency allocates a fixed emission limit to the transportation enterprise, and this policy requires that the total carbon emission Em0 in the process of transporting goods should be less than or equal to the emission threshold Em proposed by the regulatory department u , Em0 is the sum of transportation carbon emission Em1 and transfer carbon emission Em2, which is a determined value, and the calculation method is as follows:

[0036]

[0037]

[0038] Em0=Em1+Em2 (14)

[0039] For different transportation modes, the calculation methods of carbon emissions are different; if k∈M s , this type of transportation mode has a fixed route and operation plan, and the running speed is stable, and the carbon emission is only related to the transportation distance, such as railway transportation; if k∈M r , this type of transportation mode has no fixed route and operation plan, and the transportation speed may be affected by uncertain factors, such as highway transportation; therefore, under the mandatory carbon emission policy, the optimization target and constraint condition of the model are as follows:

[0040]

[0041] In particular, if k∈M r ,but:

[0042]

[0043] Road transport speed is inversely proportional to carbon emissions; the slower the vehicle speed, the more carbon dioxide is produced, and the greater the environmental impact. The formula for calculating carbon emissions from road transport is as follows:

[0044]

[0045] in, Indicates van type The carbon emissions per unit of speed at which the truck operates are expressed in kg / km; {ε0,ε1,ε2,ε3,ε4,ε5,ε6} represent correlation coefficients, which are only related to the size of the truck. For large vans, the coefficient values ​​are {1576,-17.6,0,0.00117,0,36067,0}; Q0 is the total weight of the cargo, H k It is the load capacity of the van, and the quotient of the two is the upper limit of the number of vans in the fleet;

[0046] Step S33: Consider the carbon tax policy.

[0047] A carbon tax is levied by regulatory agencies on businesses, charging a fee for each unit of carbon dioxide emitted. The purpose of a carbon tax is to encourage businesses to adopt green logistics technologies to reduce carbon emissions. The tax rate is E on the total carbon emissions generated during transportation. w carbon tax C Em And included in the total transportation cost, under the carbon tax policy, the optimization objective and constraints of the model are as follows:

[0048]

[0049] Step S34: Consider the situation regarding carbon cap-and-trade policies;

[0050] Carbon trading policy refers to a market regulated by a third party where transportation companies have a certain amount of carbon emission credits that can be traded. If carbon emissions are below the limit, the remaining credits can be sold for profit; if carbon emissions exceed the limit, the excess credits must be purchased from the market. Under this policy, transportation companies possess a fixed carbon emission credit limit. l If there is a remaining quota, it will be sold, and the profit will offset part of the transportation costs. If the quota is insufficient, additional quota needs to be purchased, and the purchase expenditure will be included in the total transportation cost. The optimization objective and constraints of the model are as follows:

[0051]

[0052] Wherein, delta = g(x), x = Em0-Em l If x < 0, delta = E s , E s is the carbon quota selling price (Yuan / Kg) ; if x >= 0, delta = E b ; the above model is subject to the following constraint conditions:

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060] Constraint (20) represents adjacent transport path cargo flow conservation constraints; constraint (21) ensures that the same logistics demand adopts no more than one transport mode on a transport path; constraint (22) ensures that the cargo transfer link carried out at a node is no more than once; constraint (23) ensures that the cargo amount on any path is less than or equal to the path passing service capacity; constraint (24) ensures that the cargo amount of any transfer node is less than or equal to its transfer capacity; constraint (25) represents that cargo does not occur transfer at the starting point or the terminal; constraint (26) represents the value of the decision variable.

[0061] The beneficial effects of the present application are: the present application establishes a multimodal transport multi-objective path decision model under uncertain conditions, then processes the uncertain conditions, constructs the optimization target and constraint condition in the multimodal transport multi-objective path decision problem under uncertain conditions, and finally establishes a path decision model under different carbon emission policies, carries out path optimization under the influence of carbon emission policies, makes up for the insufficient method of current large-scale regional carbon emission policy influence on traffic decision, and solves the quantization problem of the influence nature and ability of uncertain conditions and carbon emission policies on path decision. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 It is a method flowchart of the present application.

[0063] Figure 2 It is a multimodal transport test network and transport mileage schematic diagram of the present application.

[0064] Figure 3 A multi-modal transport path decision result diagram under different tax rates of the present application. DETAILED DESCRIPTION

[0065] As Figure 1 shown, a multi-modal transport multi-objective decision research method for quantifying carbon emission impact includes the following steps:

[0066] Step S1, establishing a multi-modal transport multi-objective path decision model under uncertain conditions.

[0067] Since under the determined condition, the transport speed and the transfer time of the goods both belong to uncertain values, if modeling and solving according to the determined condition, the decision result often does not match the actual transport characteristics, and the reliability is greatly reduced. Therefore, the present application takes the average value, i.e. the mathematical expectation, of the function corresponding to the uncertain variable to process such problems. While reducing the complexity of solving, the decision result is more in line with the actual transport characteristics.

[0068] Involving two optimization objectives of expected total transport time and expected total transport cost, a multi-objective random expectation model as formula (1) needs to be established:

[0069]

[0070] Wherein, x is an n-dimensional decision vector, ξ is a random vector, the probability density function of which is φ(ξ), f i (x,ξ) is an objective function, i=1,2,…,m. g j (x,ξ) is a set of random constraint functions, j=1,2,…,p, and E is an expectation operator.

[0071] Step S2, processing the uncertain conditions to construct the optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision problem under uncertain conditions.

[0072] The optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision problem under uncertain conditions are constructed. Among them, the total transport time T0 includes the transport time T1 between the nodes of goods and the transfer time T2 on the nodes, both of which belong to uncertain values. In the multi-objective random expectation model, the expected total transport time can be expressed in the following form:

[0073]

[0074]

[0075] E(T0)=E(T1)+E(T2) (4)

[0076] The total transportation cost Co includes the sum of transportation cost and transshipment cost C1 and the delay penalty cost C2. The transportation cost and transshipment cost are not related to the transportation time, while the delay penalty cost is related to the total transportation time and is an uncertain value. The expected total transportation cost can be expressed as follows:

[0077]

[0078] C2 = max(μ, 0)C od (T0-T od ) (6)

[0079] E(C0) = C1 + E[C2] (7)

[0080] Under the deterministic condition, whether the goods can be delivered on time to the destination only exists two cases of yes and no. But under the uncertain condition, whether the goods can be delivered on time belongs to a random variable subject to 0-1 distribution, and the probability and expectation of on-time delivery are equal. Therefore, the variable h(λ r ) is defined, if the goods are delivered on time, i.e. T0-T od < 0, h(λ r ) = 1; otherwise h(λ r ) = 0. Thus, the expected on-time delivery constraint can be expressed as:

[0081] E[h(λ r )] = P(T0-T od < 0) ≥ α (8)

[0082] Wherein, α is the expected threshold of on-time delivery, α ∈ (0, 1). The setting of the threshold eliminates some path schemes that perform poorly in transportation time, reducing the computational complexity. The constraint allows the total transportation time of some paths to exceed the transportation time limit T od , but a penalty cost based on the total transportation cost must be paid.

[0083] The expected total transportation time and the expected on-time delivery constraint can be estimated by using the above method. A random simulation generates r groups of samples where 1 < r < R. When R → ∞,

[0084]

[0085]

[0086] When the total number of random simulations R is high enough, the left side of formulas (9) and (10) is the estimated value of the expected total transportation time, the expected delay penalty cost and the expected on-time delivery constraint.

[0087] Step S3, establish a path decision model under different carbon emission policies, and optimize the path under the influence of carbon emission policies.

[0088] The low-carbon transportation is realized, and the implementation of carbon tax policy is effective. In order to explore the influence on path decision after implementing corresponding carbon emission policy under uncertain conditions, four multi-objective path planning models are established, including not considering transportation carbon emission, considering forced carbon emission policy, considering carbon tax policy and considering carbon total quantity control and transaction policy. After estimating the expected value, the optimization objectives and constraint conditions of the model under different carbon emission policies are as follows:

[0089] S31, without considering the carbon emission in the process of multimodal transport:

[0090] Under this condition, the carbon emission in the process of multimodal transport is not considered. The minimum total transportation time T0 and total transportation cost C0 are taken as the target, and the expected estimated value is taken as the constraint, and the specific conditions are as follows:

[0091]

[0092] S32, considering the forced carbon emission policy.

[0093] The forced carbon emission policy refers to that the regulatory agency allocates fixed emission limit to the transportation enterprise, and the policy requires that the total carbon emission Em0 in the process of goods transportation should be less than or equal to the emission threshold Em proposed by the regulatory department. u Em0 is the sum of transportation carbon emission Em1 and transfer carbon emission Em2, which is a determined value, and the calculation method is as follows:

[0094]

[0095]

[0096] Em0=Em1+Em2 (14)

[0097] For different transportation modes, the calculation methods of carbon emission are different. If k∈M s , the transportation mode has fixed route and opening plan, the running speed is stable, and the carbon emission is only related to the transportation distance, such as railway transportation; if k∈M r , the transportation mode has no fixed route and opening plan, and the transportation speed may be affected by uncertain factors, such as highway transportation. Therefore, under the forced carbon emission policy, the optimization objectives and constraint conditions of the model are as follows:

[0098]

[0099] In particular, if k∈M r , then:

[0100]

[0101] The speed of road transportation is inversely proportional to the carbon emissions, the slower the vehicle transport speed, the more carbon dioxide generated, the greater the impact on the environment. According to the standard proposed by the European Commission at the MEET conference, the formula for calculating the carbon emissions of road transportation is as follows:

[0102]

[0103] wherein, represents the unit carbon emissions generated by the van traveling at a speed of , the unit is Kg / km;{ε0,ε1,ε2,ε3,ε4,ε5,ε6} represents the relevant coefficients, which are only related to the size of the van, for large vans, the values of the coefficients are{1576,-17.6,0,0.00117,0,36067,0} respectively;Q0 is the total weight of the goods, H k is the load capacity of the van, and the quotient of the two is the upper limit of the number of vans in the fleet.

[0104] S33, considering the case of carbon tax policy.

[0105] Carbon tax is collected by regulatory agencies from enterprises, and the cost of carbon dioxide emissions per unit is charged. The purpose of collecting carbon tax is to guide enterprises to adopt green logistics technology to reduce carbon emissions. The total amount of carbon emissions during transportation is taxed at a rate of E w carbon tax C Em , which is included in the total transportation cost. Under the carbon tax policy, the optimization objective and constraint conditions of the model are as follows:

[0106]

[0107] S34, considering the case of carbon total amount control and trading policy.

[0108] Carbon trading policy means that transportation enterprises have a certain amount of carbon emission quota in a third-party supervised market, and the quota is allowed to trade. If the carbon emissions are lower than the limit, the remaining quota can be sold for profit; if the carbon emissions are higher than the limit, the excess quota needs to be purchased from the market [8] . Under this policy, transportation enterprises have a fixed carbon emission quota Em l . If there is a surplus, the surplus will be sold, and the profit will offset part of the transportation cost; if the quota is insufficient, additional quota needs to be purchased, and the purchase cost will be included in the total transportation cost. The optimization objective and constraint conditions of the model are as follows:

[0109]

[0110] wherein, δ=g(x), x=Em0-Em l . If x<0, δ=E s; if x ≥ 0, δ = E b The above models are subject to the following constraints:

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118] Constraint (20) represents the adjacent transportation path cargo flow conservation constraint; constraint (21) ensures that the same logistics demand adopts no more than one transportation mode on a transportation path; constraint (22) ensures that the number of cargo transfer links at a node is no more than one; constraint (23) ensures that the amount of cargo on any path is less than or equal to the path's passing service capacity; constraint (24) ensures that the amount of cargo at any transfer node is less than or equal to its transfer capacity; constraint (25) represents that cargo does not transfer at the origin or the destination; and constraint (26) represents the value of the decision variable.

[0119] As shown in a small "public-rail-water" intermodal transport network Figure 2 , for two different groups of transportation scenarios, an exact solution strategy is used for solving. The transportation origin and destination in the figure have been labeled, and other nodes have transfer functions.

[0120] Under the carbon tax policy, two groups of uncertain scenarios are set for multi-objective path decision analysis, and the two groups of scenarios are completely the same except for the carbon tax rate. Among them, the carbon tax rate in scenario 1 is 0.05 yuan / KgCO 2e , and the carbon tax rate in scenario 2 is 32 yuan / KgCO 2e . The highway transportation speed and transfer time of scenario 1 and scenario 2 obey the normal distribution, and the transportation speed of other modes is shown in Table 2.

[0121] Table 2 Uncertain highway transportation speed and transfer time random distribution

[0122]

[0123] The cargo transportation and transfer unit price is shown in Table 3, where x represents the transportation mileage.

[0124] Table 3 Cargo transportation and transfer unit price

[0125]

[0126]

[0127] Carbon emissions from road transport are calculated using formula (17). The unit carbon emissions from rail and water transport and the unit carbon emissions from freight transshipment are shown in Table 4.

[0128] Table 4 Carbon Emissions per Unit of Freight Transport and Transshipment

[0129]

[0130] according to Figure 3 As can be seen from the path decision results, the reduction in carbon tax rate lowers the total transportation cost, thus significantly improving the non-dominated solution set of the path scheme. Secondly, the carbon tax rate is 32 yuan / kg CO2. 2e In this case, the probability of finding an optimal solution in the non-dominated solution set is higher, the total transportation time increases slightly, and the total transportation cost will be significantly improved; while the carbon tax rate is 0.05 yuan / kg CO. 2e In this case, the increase in transportation time does not significantly improve transportation costs. Furthermore, the frontier range formed by the non-dominated solution set shows that increasing the carbon tax rate does not significantly improve the total transportation time of route options, because these route options primarily use rail transport. The numerical examples demonstrate that using an exact solution strategy can intuitively analyze the impact of changes in the carbon tax rate on route decision-making.

Claims

1. A method for multi-modal transport multi-objective decision-making research quantifying the impact of carbon emissions, characterized in that, Comprising the following steps: Step 1, according to the transport node, route and freight parameters, the multi-modal transport multi-objective path decision model under uncertain conditions is established; two optimization objectives of expected total transportation time and expected total transportation cost are involved, and a multi-objective random expected value model as shown in formula (1) is established: where x is an n-dimensional decision vector, ξ is a random vector with probability density function φ(ξ), f i (x,ξ) is an objective function, i = 1, 2, …, m, g j (x,ξ) is a set of random constraint functions, j = 1, 2, …, p, E is an expectation operator; Step 2, the optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision problem under uncertain conditions are constructed; the optimization objectives and constraint conditions in the multi-modal transport multi-objective path decision problem under uncertain conditions are constructed, wherein the total transportation time T0 includes the transportation time T1 between the freight nodes and the transfer time T2 at the nodes, both of which belong to uncertain values, and in the multi-objective random expected value model, the expected total transportation time E(T0) is in the following form: E(T0)=E(T1)+E(T2) (4) In the formulas (2), (3), (4), h, i, j represent the transportation nodes, h, i, j ∈ N; (h, i) represents the transportation path; k, l represent the transportation mode, k, l ∈ M; represents the distance of the path (h, i) under the k transportation mode, with the unit of Km; M is the transportation mode set; M (h,i) represents the transportation mode set of the path (h, i), N is the transportation node set; N i represents the adjacent node set of the point i, T i k,l represents the transfer time of the mode k to the mode l at the node i, with the unit of h; represents the distance of the path (h, i) under the k transportation mode, with the unit of Km; represents the transportation speed of the path (h, i) under the k transportation mode, with the unit of Km / h; is the decision variable, if the goods are transported through the (h, i) transportation arc with the k transportation mode, otherwise is the decision variable, if the transportation mode conversion occurs at the point i, otherwise The total transportation cost C0 includes the sum of transportation cost and transfer cost C1, and the delay penalty cost C2; wherein the expected total transportation cost E(C0) is in the following form: C2 = max(μ,0)C od (T0-T od )(6)E(C0) = C1 + E[C2] (7) In formulas (5) (6) (7), represents the transportation cost of transportation path (h, i) under mode k, with unit of yuan; represents the transfer cost of mode k to mode l at node i, with unit of yuan; Q0is the total weight of goods TEU; T0is the total transportation time of goods, with unit of h; Todis the agreed delivery time, with unit of h; Codis the unit delay penalty cost, with unit of yuan / h; max(μ, 0) is the max function, which finds the maximum value between μ and 0; Under deterministic conditions, the goods can arrive at the destination on time or not, there are only two cases; but in the case of uncertainty, the goods on time delivery or not belong to the 0-1 distribution of random variables, the probability of on time delivery and the expectation are equal, so the variable h(λ r ) is defined, if the goods arrive on time, that is, T0-T od <0, h(λ r ) = 1; otherwise h(λ r ) = 0, the expected value of the goods on time delivery constraint E[h(λ r )] is represented as: E[h(λ r )] = P(T0-T od < 0) ≥ α (8) where T0is the total time of goods transportation, in h; T od is the agreed delivery time, in h; a is the threshold of the expectation of on-time delivery, a ∈ (0, 1), which allows the total time of transportation of part of the path to exceed the time limit T od , a penalty fee based on the total cost of transportation must be paid. Estimate the expected total transportation time and on-time delivery expectation constraints, stochastic simulation generates r sets of samples where 1 < r < R, when R→∞, When the total number of random simulations R is high enough, the left side of formula (9), (10) is the estimated value of the expected total transportation time, the expected delay penalty cost and the on-time delivery expectation constraint; Step 3, the path decision model under different carbon emission policies is established, and the path optimization under the influence of carbon emission policies is carried out.

2. The method of claim 1, wherein the method further comprises: In step 1, the freight parameters include transportation speed, node transfer time, delivery time and transportation cost.

3. The method of claim 1, wherein the method further comprises: In step 3, the path decision model under different carbon emission policies is established, and the path optimization under the influence of carbon emission policies is carried out, which includes the following steps: Step S31, without considering the transportation carbon emission: Under this condition, without considering the carbon emission generated in the process of multimodal transport, the minimum of total transportation time T0 and total transportation cost C0 is taken as the objective, and the on-time delivery expectation constraint is taken as the constraint, which is specifically as follows: Step S32, considering the case of forced carbon emission; The total carbon emission amount Em0 during the transportation of the goods shall be less than or equal to the emission threshold Em u Em0 is the sum of the transportation carbon emission Em1 and the transit carbon emission Em2, which is a determined value, and the calculation method is as follows: Em0=Em1+Em2(14) For different transport modes, the calculation of carbon emissions is different; if k ∈ M s , the transport mode has a fixed route and operation plan, the running speed is stable, and the carbon emissions are only related to the transport distance. If k ∈ M r , the transportation mode has no fixed route and operation plan, and the transportation speed is affected by uncertain factors; therefore, under the mandatory carbon emission policy, the optimization objectives and constraints of the model are as follows: In particular, if k e M r then: The calculation formula of highway transportation carbon emission is as follows: in, Indicates van type The carbon emissions per unit of speed at which the truck operates are expressed in kg / km; {ε0,ε1,ε2,ε3,ε4,ε5,ε6} represent correlation coefficients, which are only related to the size of the truck. For large vans, the coefficient values ​​are {1576,-17.6,0,0.00117,0,36067,0}; Q0 is the total weight of the cargo, H k It is the load capacity of the van, and the quotient of the two is the upper limit of the number of vans in the fleet; Step S33, considering the case of carbon tax; The optimization objectives and constraint conditions of the model are as follows: Step S34, considering the case of carbon total amount control and transaction; Transportation companies have fixed carbon emission allowances Em l If there is surplus, the surplus will be sold, and the profit will be deducted from part of the transportation cost; if the allowance is insufficient, additional allowances need to be purchased, and the purchase cost will be included in the total transportation cost. The optimization objective and constraint conditions of the model are as follows: Wherein, δ = g(x), x = Em0-Em l , if x < 0, δ = E s , E s is the carbon credit sell-off price, unit: yuan / Kg; if x ≥ 0, δ = E b ; the above models are subject to the following constraints: Constraint (20) represents the adjacent transportation path freight flow conservation constraint; constraint (21) ensures that the same logistics demand adopts no more than one transportation mode on a transportation path; constraint (22) ensures that the freight transfer link at a node does not exceed once; constraint (23) ensures that the freight volume on any path is less than or equal to the path passing service capacity; constraint (24) ensures that the freight volume at any transfer node is less than or equal to its transfer capacity; constraint (25) represents that the freight does not transfer at the starting point or the terminal; constraint (26) represents the value of the decision variable.

Images