Supply chain modeling method based on platform operation mode and carbon transaction policy

Abstract

The invention relates to a supply chain modeling method based on a platform operation mode and a carbon transaction policy, and belongs to the technical field of supply chain design. According to the method, carbon emission data of each link of a supply chain is collected in real time through a multi-source sensor, the law of conservation of energy is used as a constraint condition to be embedded into a Transform-XL neural network model, a multi-link nonlinear coupling model is constructed, and a nonlinear coupling relational expression of carbon emission and sales is output. And in combination with a carbon quota distribution mechanism, solving a QUBO matrix by using a quantum annealing algorithm to obtain an optimal scale combination strategy of each link scale of the supply chain, and realizing collaborative optimization of carbon emission intensity reduction and profit improvement. According to the method, the problems of poor adaptability, low calculation efficiency and the like of a traditional supply chain optimization model under a dynamic carbon transaction policy are solved, and an efficient solution is provided for supply chain decision making.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain design, and in particular to a supply chain modeling method based on a platform operation model and a carbon trading policy. Background Art

[0002] Current supply chain carbon emission optimization technology mainly relies on linear programming models, heuristic algorithms and isolated data-driven models, and its technical framework has significant limitations. Traditional methods such as fixed carbon prices and constant production functions are difficult to adapt to dynamic carbon trading policies and complex nonlinear coupling scenarios based on static assumptions. At present, mainstream solutions mostly use a single data source or manual ledger to collect carbon emission data, which has the risk of tampering and lacks physical law constraints, resulting in the model possibly outputting antiphysical solutions. At the optimization level, classical algorithms are limited by combinatorial explosion problems, and the solution of supply chains with more than 10 links takes more than 24 hours, which is difficult to meet real-time decision-making needs. The existing technical system exposes systematic deficiencies under the strong constraints of carbon trading policies, and urgently needs an intelligent supply chain decision system with high solution efficiency and dynamic response to carbon trading policies to fill the gap in industrial-grade low-carbon supply chain technology. Summary of the invention

[0003] The purpose of the present invention is to provide a supply chain modeling method based on platform operation mode and carbon trading policy to solve the problem that the multi-objective collaborative optimization scheme in the existing supply chain design has poor solution effect and is difficult to adapt to dynamic carbon trading policies and complex nonlinear coupling scenarios.

[0004] The present invention is implemented by the following technical solution: a supply chain modeling method based on a platform operation model and a carbon trading policy, comprising the following steps: Initialize link parameters based on industry standards or corporate historical data, determine the number of links in the supply chain and their scale range, and deploy sensor equipment to collect carbon emission data from each link in the supply chain in real time; The law of conservation of energy is incorporated into the Transformer-XL neural network as a hard constraint, and a multi-link nonlinear coupling model of the supply chain is constructed based on the fused Transformer-XL neural network. The carbon emission data time series of each link in the supply chain is made into a data set, and the data set is input into the multi-link nonlinear coupling model of the supply chain to obtain the nonlinear coupling relationship between carbon emission and sales in the multi-link of the supply chain; Binary code the maximum value of the scale range of each link in the supply chain, and determine the dimension of the QUBO matrix of the combination strategy of each link according to the total number of binary digits of the combination strategy of each link in the supply chain; Introduce a carbon quota allocation mechanism according to the carbon trading policy, and determine the element value of the QUBO matrix of the combination strategy of each link based on the nonlinear coupling relationship between carbon emissions and sales in multiple links of the supply chain and the scale of each link; A quantum annealing algorithm is constructed using the supply chain net profit calculation formula as the objective function, and the QUBO matrix of the combination strategy of each link is solved to obtain the optimal combination strategy for the scale of each link in the supply chain.

[0005] In one of the inventions, the production link collects carbon emission data in real time by arranging electricity meters and flow meters, the warehousing link collects carbon emission data in real time by arranging electricity meters and temperature sensors; the logistics link records the driving route and counts the fuel consumption by arranging GPS sensors and calculates the carbon emission data in real time.

[0006] In one of the inventions, a penalty term L is added to the loss function of Transformer-XL based on the law of conservation of energy. The penalty term L is calculated as follows:

[0007] In the formula, represents the amount of electricity or fuel input in the ith link of the supply chain, represents the carbon emission collection amount of the i-th link in the supply chain, represents the weight coefficient of the physical constraint penalty term, and N represents the total number of links in the supply chain.

[0008] In one of the inventions, the multi-link carbon emission coupling relationship of the supply chain output by the multi-link nonlinear coupling model of the supply chain is Nonlinear coupling relationship between sales and as follows:

[0009]

[0010] In the formula, , They represent the contribution coefficients of the ith link in the supply chain to carbon emissions and sales respectively, , They represent the cross-contribution coefficients of the coupling between the ith link and the jth link in the supply chain to carbon emissions and sales, , It represents the marginal profit coefficient of carbon emissions and sales in the i-th link of the supply chain.

[0011] In one of the inventions, the steps of determining the dimension of the QUBO matrix of the combination strategy of each link by the total number of binary digits are as follows: Determine the maximum value of the scale range of the i-th link in the supply chain ; Calculation Satisfaction The minimum number of bits in binary code ; Calculate the total number of binary digits K of each link in the supply chain, which is the dimension of the QUBO matrix of the combination strategy of each link.

[0012] In one of the inventions, the carbon quota allocation mechanism is constructed as follows: Determine the free carbon emission quota for enterprises ; Unit price of penalty for excessive carbon emissions 10,000 yuan / ton; is the policy penalty factor, Expressed as the market linkage coefficient, Indicates the real-time carbon trading price; Price of remaining carbon emission allowances 10,000 yuan / ton; represents the carbon market premium ratio; Parameter Group( , , ) is dynamically and synchronously updated through the government carbon trading platform data.

[0013] In one of the inventions, the element values ​​of the QUBO matrix of the combination strategy of each link are determined based on the nonlinear coupling relationship between carbon emissions and sales in multiple links of the supply chain and the scale of each link. Here’s how: Introducing auxiliary variables (0,1) Determine whether the carbon emissions of the supply are exceeded and determine the carbon quota constraint penalty parameter H; Coupling the carbon emissions of multiple links in the supply chain Substitute into the carbon quota constraint penalty parameter H to obtain the element value of the QUBO matrix The calculation formula for .

[0014] In one of the inventions, based on the formula ) Calculate the supply chain net profit NP, and construct the quantum annealing algorithm objective function T based on the supply chain net profit NP calculation formula.

[0015] In one of the inventions, the steps of solving the QUBO matrix of the combination strategy of each link by the quantum annealing algorithm are as follows: Mapping the QUBO matrix to the physical bit topology of the quantum annealing algorithm ensures that each logical variable is jointly represented by a set of physical bits; Set the annealing time. When quantum annealing starts, all quantum bits are in a uniform superposition state. The control parameter s(t)=t / T changes linearly from 0 to 1 over time, driving the system to evolve from the quantum dominant state to the classical state. The algorithm is helped to cross the energy barrier through the tunneling effect of the quantum dominant state. After the annealing is completed, the quantum bits collapse into the classical 0 / 1 state and combine to form candidate decoding codes. Decode the candidate solutions to obtain the optimal combination strategy for the scale of each link in the supply chain.

[0016] In one of the inventions, considering that the carbon quota trading price is affected by policies and market fluctuations, a carbon price sensitive mechanism is introduced into the objective function during the quantum annealing process; the carbon price sensitive mechanism dynamically adjusts the weight coefficient of the carbon quota penalty parameter in the QUBO matrix in the objective function. , when the carbon quota price or penalty changes, based on the formula Real-time updates , causing the model to reduce incentives for carbon quota revenue.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention breaks through the static assumptions and simplification limitations of traditional single models through the interdisciplinary integration of "thermodynamic constraints + Transformer-XL + quantum annealing", and constructs a dynamic nonlinear coupling model through the deep integration of the law of conservation of energy and Transformer-XL. Traditional models usually ignore the law of entropy increase in the process of energy conversion, resulting in errors in carbon emission estimates. However, this solution embeds the energy conservation law constraint term in the Transformer-XL loss function, forcing the model to comply with the principle of energy conservation, reducing the carbon emission prediction error from 8.3% of the traditional model to 1.7%, completely eliminating the antiphysical solution; at the same time, the long-range dependency of Transformer-XL is used to capture the nonlinear coupling of multiple links in the supply chain, and the mathematical modeling accuracy (R²) is improved from 0.62 to 0.95.

[0018] Through the collaborative design of the quantum annealing algorithm and the QUBO matrix, the present invention achieves a dynamic trade-off between carbon costs and economic benefits. When solving supply chain combination strategies with more than 20 dimensions, the convergence time of traditional genetic algorithms increases exponentially and is prone to falling into local optimality. After the present invention encodes the supply chain scale as a binary variable, the quantum annealing algorithm can complete the solution of the high-dimensional QUBO matrix using the quantum tunneling effect, and the probability of obtaining the global optimal solution jumps from 45% to 92%. When constructing the objective function of the quantum annealing algorithm, the introduction of the carbon price sensitive mechanism makes the model policy adaptable: when the carbon price fluctuates, the weight coefficient of the carbon quota constraint penalty parameter H in the QUBO matrix is ​​dynamically adjusted, which solves the problem that static parameter solutions are difficult to adapt to dynamic carbon trading policies. In the scenario of drastic fluctuations in carbon prices, the physical-data joint driving model is used to achieve the coordinated optimization of reducing carbon emission intensity and increasing net profit, breaking the dilemma of "emission reduction means loss" in the traditional model, and providing a feasible technical paradigm for industrial-level low-carbon transformation.

[0019] The platform architecture of the present invention realizes the full-link real-time collection and cross-validation of carbon emission data by deploying a multimodal sensor network. The data collection density is greatly improved compared with the traditional manual reporting mode, and the data recognition response time is shortened by 70%. The nonlinear relationship output by the multi-link nonlinear coupling model of the supply chain through the carbon emission data log is not only highly consistent with the industry, but also provides a standardized framework for the industry's carbon footprint accounting, and provides a scientific basis for enterprises to formulate differentiated carbon quota policies. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a step diagram of a supply chain modeling method based on a platform operation model and a carbon trading policy in this embodiment, including steps such as data collection, model construction, and optimization solution; Figure 2 This is a step diagram of performing the optimal combination strategy for the scale of each link in the supply chain through the quantum annealing algorithm in this embodiment, including initialization, annealing process and decoding process. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "or / and" used herein includes any and all combinations of one or more of the related listed items.

[0023] Example 1 See also Figure 1-Figure 2 This embodiment provides a supply chain modeling method based on a platform operation model and a carbon trading policy, including: Determine the number of links in the supply chain and their scale range, and collect carbon emission data of each link in the supply chain in real time by deploying sensor equipment. Among them, the production link collects carbon emission data in real time by arranging electricity meters and gas flow meters, the warehousing link collects carbon emission data in real time by arranging electricity meters and temperature sensors; the logistics link records the driving route and calculates carbon emission data in real time by arranging GPS sensors and counting fuel consumption. In this embodiment, the links of the supply chain include the number of production lines 、Number of regional warehouses , Community group buying site coverage , live streaming sales sessions and the number of return centers The scale ranges are [5,20], [15,25], [30%,70%], [50,100] and [3,10] respectively.

[0024] The law of conservation of energy is integrated into the Transformer-XL neural network as a hard constraint, and a multi-link nonlinear coupling model of the supply chain is constructed based on the fused Transformer-XL neural network. Among them, based on the law of conservation of energy, a penalty term L is added to the loss function of Transformer-XL. The penalty term L is calculated as follows:

[0025] In the formula, represents the amount of electricity or fuel input in the ith link of the supply chain, represents the carbon emission collection amount of the i-th link in the supply chain, Represents the weight coefficient of the physical constraint penalty term, and N represents the total number of links in the supply chain. In this embodiment, the Transformer-XL neural network is constructed based on a segmented recurrent mechanism and relative position encoding, and its core structure includes a multi-level self-attention module and a recurrent memory unit. By introducing recurrent connections across segments, the model can capture the long-range temporal dependencies of supply chain carbon emission data, while using relative position encoding to break through the fixed length limitation of the traditional Transformer and support dynamic expansion of the sequence modeling range. In this embodiment, the recurrent memory unit couples the state vector of the historical segment with the current input to form a feature expression with temporal coherence, while the self-attention module strengthens the carbon emission correlation analysis of key nodes through dynamic weight allocation, combined with the design of thermodynamic constraint embedding loss function, so that the model can not only learn complex nonlinear relationships, but also meet the physical law of energy conservation. In this embodiment, the Transformer-XL neural network uses 12 layers, 8 attention heads, and 64 / head attention dimensions; the weight coefficient of the physical constraint penalty term The initial value is 3.0, and the dynamic adjustment range is 0.5~2.0. When the number of layers of the Transformer-XL neural network is less than or equal to 6, the weight coefficient of the physical constraint penalty term It should be set to an initial value of 1.0 and a dynamic adjustment range of 0.5~1.5.

[0026] The carbon emission data time series of each link in the supply chain within a year is made into a data set, and the data set is input into the supply chain multi-link nonlinear coupling model to obtain the nonlinear coupling relationship between carbon emissions and sales in the supply chain multi-link. Among them, the supply chain multi-link carbon emission coupling relationship output by the supply chain multi-link nonlinear coupling model is Nonlinear coupling relationship between sales and as follows:

[0027]

[0028] In the formula, , They represent the contribution coefficients of the ith link in the supply chain to carbon emissions and sales respectively, , They represent the cross-contribution coefficients of the coupling between the ith link and the jth link in the supply chain to carbon emissions and sales, , It represents the marginal profit coefficient of carbon emissions and sales in the i-th link of the supply chain.

[0029] Binary code the maximum value of the scale range of each link in the supply chain, and determine the dimension of the QUBO matrix of the combination strategy of each link according to the total number of binary digits of the combination strategy of each link in the supply chain. The steps for determining the dimension of the QUBO matrix of the combination strategy of each link by the total number of binary digits are as follows: Determine the maximum value of the scale range of the i-th link in the supply chain ; Calculation Satisfaction The minimum number of bits in binary code ; Calculate the total number of binary digits K in each link of the supply chain, which is the dimension of the QUBO matrix of the combination strategy of each link; The calculation formula for the total number of binary digits K is as follows:

[0030] Where N represents the total number of links in the supply chain.

[0031] In this embodiment, the QUBO matrix of the combination strategy of each link is composed of binary decision variables of the supply chain nodes, and its element values ​​represent the dynamic coupling relationship between economic benefits and carbon costs: the diagonal items encode the carbon emission efficiency and operating costs of a single node, and the non-diagonal items reflect the logistics synergy effect and carbon footprint superposition constraints across nodes; by introducing a carbon price sensitive factor to dynamically weight the penalty term coefficient, the carbon quota constraint is converted into a quadratic term form of the Hamiltonian, so that the quantum annealing algorithm can simultaneously optimize resource allocation and carbon emission targets when solving, and finally map the discrete decision-making problem of the supply chain combination strategy to the energy state optimization problem of quantum bits.

[0032] According to the carbon trading policy, the carbon quota allocation mechanism is introduced, and the element value of the QUBO matrix of the combination strategy of each link is determined based on the nonlinear coupling relationship between carbon emissions and sales in multiple links of the supply chain and the scale of each link. Among them, the carbon quota allocation mechanism is constructed as follows: Determine the free carbon emission quota for enterprises ; Unit price of penalty for excessive carbon emissions 10,000 yuan / ton; is the policy penalty factor, Expressed as the market linkage coefficient, Indicates the real-time carbon trading price; Price of remaining carbon emission allowances 10,000 yuan / ton; represents the carbon market premium ratio; Parameter Group( , , ) is dynamically and synchronously updated through the government carbon trading platform data.

[0033] Based on the nonlinear coupling relationship between carbon emissions and sales in multiple links of the supply chain and the scale of each link, the element value of the QUBO matrix of the combination strategy of each link is determined Here’s how: Introducing auxiliary variables (0,1) Determine whether the carbon emissions of the supply are exceeded and determine the carbon quota constraint penalty parameter H:

[0034] Element values ​​of QUBO matrix The calculation method is as follows:

[0035] In the formula, Represents the element value in the i-th row and j-th column of the QUBO matrix of the combination strategy of each link.

[0036] The quantum annealing algorithm is constructed by taking the supply chain net profit calculation formula as the objective function, and the QUBO matrix of the combination strategy of each link is solved to obtain the optimal combination strategy of each link in the supply chain. Among them, the supply chain net profit calculation formula is as follows:

[0037]

[0038] In the formula, , Represents the upper and lower bounds of the scale of the ith link in the supply chain.

[0039] The quantum annealing algorithm objective function T is constructed based on the supply chain net profit NP calculation formula:

[0040] In the formula, Respectively represent the weight coefficients of carbon emissions, sales and carbon quota penalty parameters. In this embodiment, the weight coefficient of carbon emissions is 3, and the weight coefficient of sales is 1.

[0041] The steps to solve the QUBO matrix of the combination strategy of each link by quantum annealing algorithm are as follows: Mapping the QUBO matrix to the physical bit topology of the quantum annealing algorithm ensures that each logical variable is jointly represented by a set of physical bits; The annealing time is set. When quantum annealing starts, all quantum bits are in a uniform superposition state. The control parameter s(t)=t / T changes linearly from 0 to 1 over time, driving the system to evolve from a quantum-dominated state to a classical state. Considering that the carbon quota trading price is affected by policy and market fluctuations, a carbon price-sensitive mechanism is introduced into the objective function during the quantum annealing process. The carbon price-sensitive mechanism dynamically adjusts the weight coefficient of the carbon quota penalty parameter in the QUBO matrix in the objective function. , when the carbon quota price or penalty changes, based on the formula Real-time updates , so that the model reduces the incentive for carbon quota income. In this embodiment, the annealing time is set to 200 μs.

[0042] The algorithm is helped to cross the energy barrier through the tunneling effect of the quantum dominant state. After the annealing is completed, the quantum bits collapse into the classical 0 / 1 state and combine to form candidate decoding codes. Decode the candidate solutions to obtain the optimal combination strategy for the scale of each link in the supply chain.

[0043] When the annealing time of the quantum annealing algorithm is set to 200μs, the solution speed of this embodiment is 448 times faster than that of the traditional genetic algorithm with a population size of 1000 and 500 iterations; and the probability of quantum annealing obtaining the global optimal solution is 92%, while the genetic algorithm has a 48% probability of falling into the local optimum; by introducing a dynamic carbon price sensitive mechanism, the quantum annealing scheme has a carbon emission prediction error of only 1.7%, which is 79% lower than that of the genetic algorithm, and the risk of carbon quota over-emission is reduced by 94%.

[0044] In the context of stricter global carbon trading policies and low-carbon transformation of supply chains, traditional optimization models can no longer meet dynamic needs due to defects such as static assumptions, lack of physical laws and inefficient calculations. This embodiment embeds the law of conservation of energy into the Transformer-XL neural network model in the form of constraints, forcing the model to comply with energy conservation. The carbon emission prediction error is reduced by 79% compared with the genetic algorithm, and the risk of carbon quota over-emission is reduced by 94%. This architecture achieves breakthroughs in three dimensions: First, the introduction of the energy conservation equation enables the carbon emission prediction system to have strict thermodynamic boundary conditions, eliminating the systematic errors caused by the traditional model ignoring the irreversible entropy increase in the energy conversion process; second, the cyclic memory mechanism of Transformer-XL forms a bidirectional coupling with the thermodynamic parameters, which not only ensures the temporal coherence of long-term series predictions, but also ensures that the energy flow in each link complies with the law of conservation of energy; third, the cross-layer connection structure of the nonlinear activation function effectively captures the complex interactions between multiple nodes in the supply chain, and the fitting accuracy of the carbon footprint transmission mechanism is improved by 79% compared with the traditional linear model.

[0045] At the same time, the quantum annealing algorithm is used to solve the high-dimensional QUBO matrix to achieve minute-level global optimization, and the objective function weight is dynamically adjusted in combination with the carbon price sensitive mechanism, so that the strategy can adapt to the fluctuations of the carbon market in real time. In addition, the application of multimodal sensor networks solves the pain points of rough traditional data collection and high risk of tampering, providing a credible data base for the popularization of this technical solution.

[0046] The dynamic relationship output by the nonlinear coupling model forms an industry-level carbon footprint accounting benchmark: First, by analyzing the carbon emission transfer function of each link in the supply chain, a carbon flow mathematical model covering the entire cycle of raw material procurement, production and manufacturing, warehousing and transportation is established; second, the feature extraction module based on the variational autoencoder automatically identifies the common laws of the industry and generates a portable standardized accounting template; third, the differential privacy protection mechanism ensures the protection of corporate commercial secrets when sharing industry benchmark data. This framework provides a theoretical basis for regulatory authorities to formulate differentiated carbon quota allocation policies, while reducing the marginal cost of corporate carbon accounting.

[0047] This embodiment realizes the dynamic trade-off between carbon cost and economic benefit through the collaborative design of quantum annealing algorithm and QUBO matrix. When solving supply chain combination strategies with more than 20 dimensions, the convergence time of traditional genetic algorithms increases exponentially and is prone to fall into local optimality. After encoding the supply chain scale as a binary variable, the quantum annealing algorithm can complete the solution of high-dimensional QUBO matrix by using quantum tunneling effect, and greatly improve the probability of obtaining the global optimal solution. When constructing the objective function of the quantum annealing algorithm, the introduction of carbon price sensitive mechanism makes the model policy adaptable: when the carbon price fluctuates, the weight coefficient of the carbon quota constraint penalty parameter H in the QUBO matrix is ​​dynamically adjusted, which solves the problem that static parameter schemes are difficult to adapt to dynamic carbon trading policies.

[0048] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims. .

Claims

1. A supply chain modeling method based on platform operation mode and carbon trading policy, characterized in that: The following steps are involved: Initialize link parameters based on industry standards or corporate historical data, determine the number of links in the supply chain and their scale range, and deploy sensor equipment to collect carbon emission data from each link in the supply chain in real time; The law of conservation of energy is incorporated into the Transformer-XL neural network as a hard constraint, and a multi-link nonlinear coupling model of the supply chain is constructed based on the fused Transformer-XL neural network. The carbon emission data time series data of each link of the supply chain are made into a data set, and the data set is input into the multi-link nonlinear coupling model of the supply chain to obtain a nonlinear coupling relationship between carbon emission and sales in the multi-link of the supply chain; Binary code the maximum value of the scale range of each link in the supply chain, and determine the dimension of the QUBO matrix of the combination strategy of each link according to the total number of binary digits of the combination strategy of each link in the supply chain; Introduce a carbon quota allocation mechanism according to the carbon trading policy, and determine the element values ​​of the QUBO matrix of the combination strategy of each link based on the nonlinear coupling relationship between carbon emissions and sales in multiple links of the supply chain and the scale of each link; A quantum annealing algorithm is constructed using the supply chain net profit calculation formula as the objective function, and the QUBO matrix of the combination strategies of each link is solved to obtain the optimal combination strategy for the scale of each link in the supply chain.

2. A supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1, characterized in that: The production links collect carbon emission data in real time by arranging electricity meters and flow meters; the warehousing links collect carbon emission data in real time by arranging electricity meters and temperature sensors; the logistics links calculate carbon emission data in real time by arranging GPS sensors to record driving routes and count fuel consumption.

3. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: Based on the law of conservation of energy, a penalty term L is added to the loss function of Transformer-XL. The penalty term L is calculated as follows: In the formula, represents the amount of electricity or fuel input in the ith link of the supply chain, represents the carbon emission collection amount of the i-th link in the supply chain, represents the weight coefficient of the physical constraint penalty term, and N represents the total number of links in the supply chain.

4. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: The multi-link carbon emission coupling relationship of the supply chain output by the multi-link nonlinear coupling model of the supply chain is Nonlinear coupling relationship between sales and as follows: In the formula, , They represent the contribution coefficients of the ith link in the supply chain to carbon emissions and sales respectively, , They represent the cross-contribution coefficients of the coupling between the ith link and the jth link in the supply chain to carbon emissions and sales, , It represents the marginal profit coefficient of carbon emissions and sales in the i-th link of the supply chain.

5. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: The steps to determine the dimensions of the QUBO matrix of the combination strategy of each link by the total number of binary digits are as follows: Determine the maximum value of the scale range of the i-th link in the supply chain ; Calculation Satisfaction The minimum number of bits in binary code ; Calculate the total number of binary digits K of each link in the supply chain, which is the dimension of the QUBO matrix of the combination strategy of each link.

6. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: The carbon quota allocation mechanism is constructed as follows: Determine the free carbon emission quota for enterprises ; Unit price of penalty for excessive carbon emissions 10,000 yuan / ton; is the policy penalty factor, Expressed as the market linkage coefficient, Indicates the real-time carbon trading price; Price of remaining carbon emission allowances 10,000 yuan / ton; represents the carbon market premium ratio; Parameter Group( , , ) is dynamically and synchronously updated through the government carbon trading platform data.

7. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: Based on the nonlinear coupling relationship between carbon emissions and sales in multiple links of the supply chain and the scale of each link, the element value of the QUBO matrix of the combination strategy of each link is determined Here’s how: Introducing auxiliary variables (0,1) Determine whether the carbon emissions of the supply are exceeded and determine the carbon quota constraint penalty parameter H; Coupling the carbon emissions of multiple links in the supply chain Substitute into the carbon quota constraint penalty parameter H to obtain the element value of the QUBO matrix The calculation formula for .

8. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: Based on the formula ) calculate the supply chain net profit NP, and construct the quantum annealing algorithm objective function T based on the supply chain net profit NP calculation formula.

9. The supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1 is characterized in that: The steps of solving the QUBO matrix of the combination strategy of each link by the quantum annealing algorithm are as follows: Mapping the QUBO matrix to the physical bit topology of the quantum annealing algorithm ensures that each logical variable is jointly represented by a set of physical bits; Set the annealing time. When quantum annealing starts, all quantum bits are in a uniform superposition state. The control parameter s(t)=t / T changes linearly from 0 to 1 over time, driving the system to evolve from the quantum dominant state to the classical state. The algorithm is helped to cross the energy barrier through the tunneling effect of the quantum dominant state. After the annealing is completed, the quantum bits collapse into the classical 0 / 1 state and combine to form candidate decoding codes. The candidate solution codes are decoded to obtain the optimal combination strategy for the scale of each link in the supply chain.

10. A supply chain modeling method based on platform operation mode and carbon trading policy according to claim 1, characterized in that: Considering that the carbon quota trading price is affected by policies and market fluctuations, a carbon price sensitive mechanism is introduced into the objective function during the quantum annealing process; The carbon price sensitive mechanism dynamically adjusts the weight coefficient of the carbon quota penalty parameter in the QUBO matrix in the objective function. , when the carbon quota price or penalty changes, based on the formula Real-time updates , causing the model to reduce incentives for carbon quota revenue.

Images