Product pricing optimization method and system considering carbon quota

Abstract

The invention discloses a product pricing optimization method and system considering a carbon quota, relates to the technical field of carbon emission analysis, and solves the problem that the prediction accuracy is limited due to the fact that most carbon transaction price prediction models in the prior art depend on a single historical carbon price time sequence. According to the method, the carbon emission of a production unit product is obtained from a database, and the product wholesale price given by a manufacturer is marked as a decision variable; respectively constructing a demand function of a retailer and a profit function of a manufacturer based on the decision variables; obtaining the optimal demanded quantity of the retailer according to the decision variable; constructing a decision optimization model of the manufacturer according to the optimal demand of the retailer; obtaining a profit value of the manufacturer through a decision optimization model of the manufacturer; calling a quantum behavior particle swarm algorithm, and taking the profit value as a fitness value of particles; obtaining the number of iterations and an optimal pricing decision based on a quantum behavior particle swarm algorithm; according to the invention, the accuracy of predicting the carbon transaction price is improved.

Description

Technical Field

[0001] The present invention belongs to the field of carbon emission analysis and relates to carbon emission optimization technology, specifically a method and system for optimizing product pricing by considering carbon quotas. Background Art

[0002] By incorporating carbon quotas into the product pricing system, companies will pay more attention to carbon emissions and take a series of measures to reduce carbon emissions, which will also help protect the environment. Manufacturers restricted by carbon quotas have significant differences in production decisions from traditional models. Manufacturers regard carbon quotas as key production factors, consider the benefits or costs caused by quota trading, and the trade-off between product pricing and carbon reduction costs. Incorporating carbon quotas into the product pricing system can strengthen the guiding role of the policy. By adjusting the price and distribution method of carbon quotas, companies can be guided to adjust their production methods and product structure to achieve the set target strategies.

[0003] Existing technologies determine the processing plan before producing mechanical products, and the calculation of carbon reduction costs depends on the carbon price in the future production stage. Effectively predicting carbon trading prices becomes the key to decision-making. Existing technologies mostly use models to predict carbon trading prices that rely on a single historical carbon price time series and fail to fully consider a variety of other influencing factors, thereby limiting the accuracy of the prediction and increasing the decision-making risk of manufacturers.

[0004] The present invention provides a method and system for optimizing product pricing taking carbon quota into consideration to solve the above technical problems. Summary of the invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a method and system for optimizing product pricing taking into account carbon quotas, which is used to solve the technical problem that most of the models for predicting carbon trading prices in the prior art rely on a single historical carbon price time series and fail to fully consider a variety of other influencing factors, thereby limiting the accuracy of the prediction and increasing the decision-making risk of manufacturers.

[0006] To achieve the above object, a first aspect of the present invention provides a method for optimizing product pricing taking carbon quotas into consideration, comprising:

[0007] Obtain the carbon emissions per unit of production from the database and mark it as decision variable one; obtain the wholesale price of the product given by the manufacturer and mark it as decision variable two; construct the retailer's demand function based on decision variable one, and calculate the manufacturer's total cost based on the demand function;

[0008] Retrieve the total cost of the manufacturer, construct the profit function based on the decision variable 2 and propose the function constraints;

[0009] Retrieving decision variables to obtain the optimal demand of retailers; building a decision optimization model for manufacturers based on the optimal demand of retailers; wherein the decision variables include: decision variable 1 and decision variable 2;

[0010] Obtain the manufacturer's profit value through the manufacturer's decision optimization model;

[0011] The quantum behavior particle swarm algorithm is called and the profit value is used as the fitness value of the particle. The number of iterations and the optimal pricing decision are obtained based on the quantum behavior particle swarm algorithm.

[0012] Preferably, constructing the retailer's demand function based on decision variable 1 includes:

[0013] The initial carbon quota is obtained from the database and marked as ω, the carbon emission per unit of production is marked as e, and the retailer product demand parameter is marked as N i ; Where i is the number of retailers, i = 1, 2, 3, ..., n; n is a positive integer;

[0014] By formula Calculate the difference between the initial carbon quota and the actual carbon emissions generated;

[0015] Determine whether σ is greater than or equal to 0; if yes, then σ is the carbon trading income of the manufacturing enterprise; if no, then σ is the carbon trading expenditure of the manufacturing enterprise;

[0016] Get the retailer's price demand elasticity from the database and label it as a i , the retailer’s carbon footprint demand elasticity is denoted by b i ; The retail price of the product given by the retailer is marked as P retail ;

[0017] By formula N i (P retail ,e)=N i -a i ·P retail -b i ·eCalculate the retailer's demand function.

[0018] Preferably, the calculating the total cost of the manufacturer according to the demand function comprises:

[0019] The carbon footprint per unit product before emission reduction is marked as C tr The carbon footprint of a unit product after emission reduction is marked as C low , through the formula The emission reduction cost is calculated to be a quadratic function of the emission reduction rate; k is the R&D cost coefficient invested by the enterprise for emission reduction, and k>0; the total cost of the manufacturer includes production cost, carbon emission reduction cost generated by adopting low-carbon technology, transportation cost and carbon trading cost;

[0020] Get the wholesale price of the product given by the manufacturer and mark it as P wholesale , obtain the carbon trading price through the carbon trading price prediction model, and mark the carbon trading price as P EUA The manufacturer's cost per unit of product is denoted by P cost , the retailer’s price mark is P retail ; The manufacturer's cost of shipping a unit of product is denoted by P transport ;

[0021] Among them, through the formula Calculate the manufacturer's production cost;

[0022] By formula Calculate the manufacturer’s carbon reduction costs;

[0023] By formula Calculate the manufacturer's shipping cost;

[0024] By formula Get the manufacturer's carbon trading costs;

[0025] The total cost of the manufacturer is calculated by the formula ZZC=SXC+TJC+YSC-TJC.

[0026] It should be noted that the larger the value of k, the greater the cost invested by the manufacturer.

[0027] Preferably, obtaining the carbon trading price through the carbon trading price prediction model includes:

[0028] Historical carbon trading price data, EU carbon emission quota closing price, crude oil spot closing price and Shanghai and Shenzhen 300 Index are obtained from the database and integrated into a price series as the first factor, and the first factor is input into the carbon trading price prediction model; the carbon trading price is output through the carbon trading price prediction model, and the error value between the current carbon trading price and the historical carbon trading price is calculated, and the error value is input into the carbon trading price prediction model as the second factor, and the carbon trading price prediction model is trained to re-output the carbon trading price; wherein, the carbon trading price prediction model is constructed according to an artificial intelligence model; the artificial intelligence model includes a convolutional neural network model or a long short-term memory neural network model.

[0029] Preferably, the profit function is constructed based on the second decision variable and function constraints are proposed, including:

[0030] By formula Calculate the manufacturer's profit function;

[0031] Retailers based on the manufacturer's P wholesale and e optimizes P retail , to maximize profits;

[0032] Get the retailer's demand function N i (P retail , e), calculated by the formula Derive the retailer’s profit function;

[0033] Construction constraint 1: P retail -P wholesale >0, indicating that the retail price is greater than the wholesale price;

[0034] Construction constraint 2: N i (P retail , e)>0, indicating that the retailer’s product demand function is greater than zero;

[0035] Construction constraint three: C tr ≥C low ≥0, indicating that the carbon footprint per unit product after emission reduction measures are taken is smaller than the carbon footprint before emission reduction.

[0036] Preferably, obtaining the optimal demand of the retailer includes:

[0037] Retrieve Pwholesale and e, and determine N i (P retail ,e) and (P retail -P wholesale ) are both greater than zero; if so, then N i (P retail ,e)·(P retail -P wholesale ) is maximized; otherwise, continue to judge;

[0038] By formula P * retail (P wholesale ,e)=(N i -b i ·e+a i ·P wholesale ) / 2a i Calculate the retailer's optimal retail price;

[0039] By formula N i * (P retail ,e)=N i * (P * retail (P wholesale ,e))=(N i -3b i ·ea i ·P wholesale ) / 2 calculates the optimal demand of the corresponding retailer.

[0040] Preferably, the decision optimization model of the manufacturer constructed according to the optimal demand of the retailer includes:

[0041] Retrieve the optimal retail price of the retailer and the corresponding optimal demand of the retailer and import them into each cost function of the manufacturer to obtain:

[0042]

[0043] Retrieve the total cost of the manufacturer ZZC = SXC + TJC + YSC - TJC;

[0044] Then through the formula Calculate the profit function of the manufacturer;

[0045] Through the formula Establish the decision optimization model of the manufacturer.

[0046] It should be noted that the manufacturer optimizes its own decisions through the dynamic decision results of the retailer. Therefore, the optimal decisions of the retailer can be fully utilized when analyzing the decisions of the manufacturer.

[0047] Preferably, the number of iterations obtained based on the quantum-behaved particle swarm optimization algorithm includes:

[0048] S110: Through the formula Calculate the distribution probability of particles in the search space described by the quantum wave function form; x c Is the central position of the normal distribution, and x is the position of the particle;

[0049] S120: Mark the upper limit of the search space as x max , Mark the lower limit of the search space as x min , And mark the random number uniformly distributed in the [0, 1] interval as r i ;

[0050] Through the formula Calculate the initial position of the particle ;

[0051] Take the initial position as the initial individual optimum And the global optimum g 0 ;

[0052] S130: Update the particle positions in the QPSO through the quantum wave function and random variables;

[0053] S140: Through the formula Calculate the individual optimum position of the corresponding particle after update;

[0054] Through the formula Calculate the global optimal position of the corresponding particle after update; where f() is the objective function, representing the fitness value of the particle at the updated position ;

[0055] S150: Update the contraction factor by constructing a linearly decreasing strategy; where the linearly decreasing strategy is: β max and β min are the initial value and the final value of β respectively; t max is the maximum number of iterations; t is the current number of iterations;

[0056] S160: Retrieve the maximum number of iterations t max ; Determine whether the current number of iterations has reached the maximum number of iterations t max ; If yes, stop the iteration and output the number of iterations; if no, continue the iteration.

[0057] It should be noted that the global optimal position changes very little in multiple iterations, indicating that the algorithm has converged; the initial position formula of the particle is based on a one-dimensional particle space, and the calculation method is the same for any dimension; the contraction factor β is used to control the search range and usually gradually decreases with the increase of the generation number.

[0058] Preferably, updating the position of the particle in QPSO through the quantum wave function and random variables includes:

[0059] S131: Calculate the guiding position for the i-th particle in the t-th generation through the formula ;

[0060] S132: Calculate the updated position of the particle through the formula ;

[0061] Where is the guiding position of the i-th particle; is the position of the i-th particle at the t-th generation; u is a random number uniformly distributed in the interval (0, 1); the "±" sign is determined by the random method to determine the search direction of the particle.

[0062] The present invention selects the positive or negative sign by the random method so that the particle can search on both sides of the guiding point; this randomness helps to increase the diversity of the search and prevent the particle from falling into the local optimal solution prematurely.

[0063] Preferably, obtaining the optimal pricing decision based on the quantum-behaved particle swarm algorithm includes:

[0064] S210: Initialize the parameters, where the parameters include: the particle swarm size, the contraction factor β, and the initial particle position;

[0065] S220: Retrieve the carbon trading price P EUA , and calculate the fitness values of all particles;

[0066] S230: Compare the fitness value of the current particle with its historical optimal fitness value, and determine whether the fitness value of the current particle is higher than the historical optimal fitness value of the corresponding particle; if yes, update the fitness value of the current particle to the optimal fitness value of the current particle; if no, use the historical optimal fitness value of the corresponding particle as the optimal position fitness value of the current particle;

[0067] S240: Compare the optimal position fitness value of each current particle with the global optimal fitness value, and determine whether the global optimal fitness value is higher than the optimal position fitness value of the corresponding particle; if yes, the global optimal fitness value remains unchanged; if no, use the optimal position fitness value of the corresponding particle as the global optimal fitness value;

[0068] S250: Calculate the historical best position and the global best position of each particle;

[0069] S260: Determine whether the iterative operation satisfies the termination condition; if yes, terminate the operation and output the corresponding optimal pricing decision; if no, return to step S220; where the termination condition is that the number of iterations reaches the maximum number of iterations.

[0070] It should be noted that the maximum number of iterations is determined by weighing the convergence performance and the calculation time based on the actual work experience of the staff, and the specific value is adjusted through experiments to adapt to specific problems. Generally, it starts with an assignment of tmax = 200 for testing.

[0071] To achieve the above object, the second aspect of the present invention provides a system for optimizing product pricing considering carbon quotas, including: a profit function construction module and a decision optimization module;

[0072] Profit function construction module: Obtain the carbon emissions per unit product from the database and label it as decision variable one; obtain the product wholesale price given by the manufacturer and label it as decision variable two; construct the demand function of the retailer based on decision variable one, and calculate the total cost of the manufacturer according to the demand function;

[0073] Retrieve the total cost of the manufacturer, construct a profit function based on decision variable two, and propose function constraint conditions;

[0074] Decision optimization module: Retrieve the decision variables, and obtain the optimal demand quantity of the retailer; construct a decision optimization model of the manufacturer according to the optimal demand quantity of the retailer; where the decision variables include: decision variable one and decision variable two;

[0075] Obtain the profit value of the manufacturer through the decision optimization model of the manufacturer;

[0076] Invoke the quantum-behaved particle swarm optimization algorithm, and use the profit value as the fitness value of the particle; obtain the number of iterations and the optimal pricing decision based on the quantum-behaved particle swarm optimization algorithm.

[0077] Compared with the prior art, the beneficial effects of the present invention are as follows: Considering carbon quotas as an important factor affecting decisions, which is not available in traditional prediction models; based on the quantum-behaved particle swarm optimization algorithm (QPSO), compared with the traditional particle swarm optimization algorithm, the movement of QPSO particles is described by probability distribution, avoiding the risk of particles falling into local optimal solutions, making QPSO have stronger global search ability; QPSO introduces the concept of quantum behavior, and its algorithm implementation is relatively simple, easy to program and apply, and has strong adaptability, performing well in high-dimensional and complex search spaces. And by constructing the retailer's demand function through decision variable one, it can reflect consumers' preferences for environmentally friendly products, and then calculate the total cost of the manufacturer according to the demand function, which helps the manufacturer understand the relationship between production costs and market demand; by constructing the profit function through decision variable two and proposing corresponding function constraint conditions, it can ensure that while the manufacturer pursues profit maximization, it also takes into account actual factors such as market demand and production costs; by invoking decision variables to construct the retailer's optimal demand quantity model, it helps the retailer determine the best purchase quantity, thereby reducing costs and increasing profits; according to the retailer's optimal demand quantity to construct the manufacturer's decision optimization model, it can comprehensively consider multiple factors such as market demand, production costs, and wholesale prices, and provide the manufacturer with the optimal pricing strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0079] Figure 1 It is a schematic flowchart of the pricing optimization of the present invention;

[0080] Figure 2 It is a schematic diagram of the specific steps for obtaining the carbon trading price of the present invention;

[0081] Figure 3 It is a schematic diagram of the specific steps for constructing the decision optimization model of the present invention;

[0082] Figure 4 It is a schematic diagram of the specific steps for the optimal pricing decision of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0083] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0084] Please refer to Figure 1 , an embodiment of the first aspect of the present invention provides a method for optimizing product pricing considering carbon quotas, including:

[0085] Obtain the carbon emissions per unit product from the database and label it as decision variable one; obtain the wholesale price of the product given by the manufacturer and label it as decision variable two; construct the demand function of the retailer based on decision variable one, and calculate the total cost of the manufacturer according to the demand function;

[0086] Retrieve the total cost of the manufacturer, construct a profit function based on decision variable two and propose function constraint conditions;

[0087] Retrieve the decision variables, obtain the optimal demand quantity of the retailer; construct a decision optimization model for the manufacturer according to the optimal demand quantity of the retailer; wherein, the decision variables include: decision variable one, decision variable two;

[0088] Obtain the profit value of the manufacturer through the decision optimization model of the manufacturer;

[0089] Retrieve the quantum-behaved particle swarm algorithm, and use the profit value as the fitness value of the particle; obtain the number of iterations and the optimal pricing decision based on the quantum-behaved particle swarm algorithm.

[0090] Please refer to Figure 2 , the specific steps to obtain the carbon trading price. Obtain historical carbon trading price data, the closing price of EU carbon emission allowances, the closing price of crude oil spot, and the CSI 300 index from the database and integrate them into a price series as the first factor, and input the first factor into the carbon trading price prediction model; output the carbon trading price through the carbon trading price prediction model, calculate the error value between the current carbon trading price and the historical carbon trading price, and input the error value as the second factor into the carbon trading price prediction model to train the carbon trading price prediction model to re-output the carbon trading price; wherein, the carbon trading price prediction model is constructed according to an artificial intelligence model; the artificial intelligence model includes a convolutional neural network model or a long short-term memory neural network model.

[0091] The carbon trading price P EUA can be obtained from the carbon trading prediction model, and then the approximate optimal solutions of C low and P wholesale are obtained through the quantum-behaved particle swarm algorithm, thus solving the difficulties brought by the non-convexity of the model and avoiding getting stuck in local optima during the solution process.

[0092] See also Figure 3 ,The specific steps of building the decision optimization model are to obtain the initial carbon quota from the database and mark it as ω, the carbon emission per unit of production is marked as e, and the retailer product demand parameter is marked as N i ; Wherein, i is the number of retailers, i = 1, 2, 3, ..., n; n is a positive integer;

[0093] By formula Calculate the difference between the initial carbon quota and the actual carbon emissions generated;

[0094] Determine whether σ is greater than or equal to 0; if yes, then σ is the carbon trading income of the manufacturing enterprise; if no, then σ is the carbon trading expenditure of the manufacturing enterprise;

[0095] Get the retailer's price demand elasticity from the database and label it as a i , the retailer’s carbon footprint demand elasticity is denoted by b i ; The retail price of the product given by the retailer is marked as P retail ;

[0096] By formula N i (P retail ,e)=N i -a i ·P retail -b i eCalculate the retailer’s demand function;

[0097] By formula Calculate the manufacturer's profit function;

[0098] Retailers based on the manufacturer's P wholesale and e optimizes P retail , to maximize profits;

[0099] Get the retailer's demand function N i (P retail , e), calculated by the formula Derive the retailer’s profit function;

[0100] Construction constraint 1: P retail -P wholesale >0, indicating that the retail price is greater than the wholesale price;

[0101] Construction constraint 2: N i (P retail , e)>0, indicating that the retailer’s product demand function is greater than zero;

[0102] Construction constraint three: C tr ≥C low≥0, indicating that the carbon footprint per unit product after emission reduction measures are taken is smaller than the carbon footprint before emission reduction.

[0103] For example, the decision-making process of manufacturers is not only limited by the total carbon emission cap set by the government, but also affected by the price fluctuations in the carbon trading market. Therefore, the profit structure of manufacturers mainly includes two aspects: on the one hand, it comes from product sales revenue, and on the other hand, it depends on the profit and loss of carbon quota trading. Based on the above analysis, It represents the difference between the initial carbon quota and the actual carbon emissions. If σ>0, it is the carbon trading income of the manufacturing enterprise; if σ<0, it is the carbon trading expenditure of the manufacturing enterprise. Assuming that the product carbon footprint and retail price are negatively correlated with market demand, the demand function of retailer i is expressed as a linear function: N i (P retail ,e)=N i -a i ·P retail -b i e;

[0104] By calling the established profit function and constraints, we can see that the manufacturer's wholesale price P wholesale , Carbon emissions per unit product after carbon reduction C low and the retailer’s price P retail , is the main factor affecting the profits and carbon emissions of manufacturers and retailers; since the retailer's decision needs to be based on the manufacturer's decision P wholesale and e, so the following analysis of the retailer's optimal decision is based on the given P wholesale and e, in N i (P retail ,e) and (P retail -P wholesale ) are greater than zero, N i (P retail ,e)·(P retail -P wholesale )maximize.

[0105] because It's about P retail convex function, so the retailer's optimal retail price is: P * retail (P wholesale ,e)=(N i -b i ·e+a i ·P wholesale ) / 2a i ;

[0106] The optimal demand of the retailer is:

[0107] See also Figure 4 , the specific steps of the optimal pricing decision,

[0108] S110: By formula The calculated quantum wave function describes the distribution probability of particles in the search space; x c is the center position of the normal distribution, and x is the position of the particle;

[0109] S120: Mark the upper limit of the search space as x max , the lower limit of the search space is marked as x min , a random number uniformly distributed in the interval [0, 1] is marked as r i ;

[0110] By formula Calculated particles The initial position of

[0111] The initial position is taken as the initial individual optimal and the global optimal g 0 ;

[0112] S130: Update the particle position in QPSO through quantum wave functions and random variables;

[0113] S131: By formula The guiding position of the i-th particle for the t-th generation is calculated;

[0114] S132: By formula Calculate the updated particle position;

[0115] in, is the guiding position of the ith particle; is the position of the ith particle in the tth generation; u is a random number uniformly distributed in the interval (0, 1); the “±” sign is determined by a random method to determine the search direction of the particle;

[0116] S140: By formula Calculate the optimal individual position of the corresponding particle after update;

[0117] By formula The global optimal position of the corresponding particle after the update is calculated; where f() is the objective function, which indicates the particle's updated position The fitness value at ;

[0118] S150: Update the shrinkage factor by constructing a linear decreasing strategy; wherein the linear decreasing strategy is: β max and β min are the initial and final values ​​of β respectively; tmax is the maximum number of iterations; t is the current number of iterations;

[0119] S160: Retrieve the maximum number of iterations t max ; Determine whether the current number of iterations has reached the maximum number of iterations t max ; If yes, stop the iteration and output the number of iterations; if no, continue the iteration;

[0120] S210: Initializing parameters, where the parameters include: particle swarm size, contraction factor β, and initialization particle position;

[0121] S220: Get the carbon trading price P EUA , calculate the fitness values ​​of all particles;

[0122] S230: Compare the fitness value of the current particle with its historical optimal fitness value to determine whether the fitness value of the current particle is higher than the historical optimal fitness value of the corresponding particle; if yes, update the fitness value of the current particle to the optimal fitness value of the current particle; if no, use the historical optimal fitness value of the corresponding particle as the optimal position fitness value of the current particle;

[0123] S240: Compare the current optimal position fitness value of each particle with the global optimal fitness value to determine whether the global optimal fitness value is higher than the optimal position fitness value of the corresponding particle; if yes, the global optimal fitness value remains unchanged; if no, the optimal position fitness value of the corresponding particle is used as the global optimal fitness value;

[0124] S250: Calculate the historical best position and the global best position of each particle;

[0125] S260: Determine whether the iterative operation meets the termination condition; if yes, the operation is terminated and the corresponding optimal pricing decision is output; if no, return to step S220; wherein the termination condition is that the number of iterations reaches the maximum number of iterations.

[0126] For example, the quantum-behaved particle swarm optimization algorithm (QPSO) is an improved particle swarm optimization algorithm based on the principles of quantum mechanics. QPSO introduces the concept of quantum mechanics to make the behavior of particles different from that of classical PSO, thereby improving the global search ability and convergence of the algorithm; the quantum-behaved particle swarm algorithm draws on the uncertainty principle and wave function description in quantum mechanics, abandons the speed update method of classical PSO, and instead uses the position probability distribution of particles to describe the state of particles; the state of particles is no longer described by a certain speed and position, but by a wave function to describe its distribution probability in the search space, using the quantum wave function form as

[0127] Initialize the particle swarm, including the initial position and initial fitness value of the particles. The positions of the particles are randomly generated in the search space.

[0128] Particle position initialization: particle The initial position of is usually randomly generated in the search space;

[0129] By formula Calculate the initial position of the particle;

[0130] Use quantum wave functions and random variables to update the particle position. The calculation formula is:

[0131] in: is the guiding position of the ith particle; is the position of the i-th particle in the t-th generation; β is a contraction factor used to control the search range, which usually decreases with the increase of generations to ensure the convergence of the algorithm; the setting of the contraction factor β has an important influence on the performance of QPSO; generally speaking, β decreases with the increase of generations to enhance the local development ability of the later search; u is a random number uniformly distributed in the interval (0, 1); the "±" sign indicates the random selection of positive and negative signs to determine the search direction of the particle.

[0132] The above steps are based on the quantum-behaved particle swarm optimization algorithm (QPSO), which is an improved particle swarm optimization algorithm based on the principles of quantum mechanics. QPSO introduces the concept of quantum mechanics to make the behavior of particles different from that of classical PSO, thereby improving the global search ability and convergence of the algorithm. The positive or negative sign is selected by random method so that particles can search on both sides of the guide point. This randomness helps to increase the diversity of the search and prevent particles from falling into the local optimal solution too early.

[0133] The second aspect of the present invention provides a system for optimizing product pricing considering carbon quotas, including: a profit function construction module, a decision optimization module;

[0134] Profit function construction module: obtain the carbon emissions per unit of production from the database and mark it as decision variable one; obtain the wholesale price of the product given by the manufacturer and mark it as decision variable two; construct the retailer's demand function based on decision variable one, and calculate the manufacturer's total cost based on the demand function;

[0135] Retrieve the total cost of the manufacturer, construct the profit function based on the decision variable 2 and propose the function constraints;

[0136] Decision optimization module: retrieve decision variables to obtain the optimal demand of retailers; build a decision optimization model for manufacturers based on the optimal demand of retailers; the decision variables include: decision variable 1 and decision variable 2;

[0137] Obtain the manufacturer's profit value through the manufacturer's decision optimization model;

[0138] The quantum behavior particle swarm algorithm is called and the profit value is used as the fitness value of the particle. The number of iterations and the optimal pricing decision are obtained based on the quantum behavior particle swarm algorithm.

[0139] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0140] The working principle of the present invention is as follows: the present invention obtains the carbon emission per unit of production from a database and marks it as decision variable one; obtains the wholesale price of the product given by the manufacturer and marks it as decision variable two; constructs the demand function of the retailer based on the decision variable one, and calculates the total cost of the manufacturer according to the demand function; constructs the profit function based on the decision variable two and proposes function constraints; retrieves the decision variables to obtain the optimal demand of the retailer; constructs the decision optimization model of the manufacturer according to the optimal demand of the retailer; obtains the profit value of the manufacturer through the decision optimization model of the manufacturer; retrieves the quantum behavior particle swarm algorithm, and uses the profit value as the fitness value of the particle; obtains the number of iterations and the optimal pricing decision based on the quantum behavior particle swarm algorithm.

[0141] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for optimizing product pricing considering carbon quotas, characterized in that: include: Obtain the carbon emissions per unit of production from the database and mark it as decision variable one; obtain the wholesale price of the product given by the manufacturer and mark it as decision variable two; Construct the retailer's demand function based on decision variable one, and calculate the manufacturer's total cost based on the demand function; Retrieve the total cost of the manufacturer, construct the profit function based on the decision variable 2 and propose the function constraints; Retrieving decision variables to obtain the optimal demand of retailers; building a decision optimization model for manufacturers based on the optimal demand of retailers; wherein the decision variables include: decision variable 1 and decision variable 2; Obtain the manufacturer's profit value through the manufacturer's decision optimization model; The quantum behavior particle swarm algorithm is called and the profit value is used as the fitness value of the particle. The number of iterations and the optimal pricing decision are obtained based on the quantum behavior particle swarm algorithm.

2. A method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The step of constructing a retailer's demand function based on decision variable 1 includes: The initial carbon quota is obtained from the database and marked as ω, the carbon emission per unit of production is marked as e, and the retailer product demand parameter is marked as N i ; Where i is the number of retailers, i = 1, 2, 3, ..., n; n is a positive integer; By formula Calculate the difference between the initial carbon quota and the actual carbon emissions generated; Determine whether σ is greater than or equal to 0; if yes, then σ is the carbon trading income of the manufacturing enterprise; if no, then σ is the carbon trading expenditure of the manufacturing enterprise; Get the retailer's price demand elasticity from the database and label it as a i , the retailer’s carbon footprint demand elasticity is denoted by b i ; The retail price of the product given by the retailer is marked as P retail ; By formula N i (P retail ,e)=N i -a i ·P retail -b i ·eCalculate the retailer's demand function.

3. A method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The total cost of the manufacturer is calculated according to the demand function, including: The carbon footprint per unit product before emission reduction is marked as C tr The carbon footprint of a unit product after emission reduction is marked as C low , through the formula The emission reduction cost is calculated to be a quadratic function of the emission reduction rate; k is the R&D cost coefficient invested by the enterprise for emission reduction, and k>0; the total cost of the manufacturer includes production cost, carbon emission reduction cost generated by adopting low-carbon technology, transportation cost and carbon trading cost; Get the wholesale price of the product given by the manufacturer and mark it as P wholesale , obtain the carbon trading price through the carbon trading price prediction model, and mark the carbon trading price as P EUA The manufacturer's cost per unit of product is denoted by P cost , the retailer’s price mark is P retail The manufacturer's cost of shipping a unit of product is denoted by P transport ; Among them, through the formula Calculate the manufacturer's production cost; By formula Calculate the manufacturer’s carbon reduction costs; By formula Calculate the manufacturer's shipping cost; By formula Get the manufacturer's carbon trading costs; The total cost of the manufacturer is calculated by the formula ZZC=SXC+TJC+YSC-TJC.

4. A method for optimizing product pricing considering carbon quotas according to claim 3, characterized in that: The obtaining of the carbon trading price through the carbon trading price prediction model includes: Historical carbon trading price data, carbon emission quota closing price, crude oil spot closing price and stock market index are obtained from the database and integrated into a price series as the first factor, and the first factor is input into the carbon trading price prediction model; the carbon trading price is output through the carbon trading price prediction model, and the error value between the current carbon trading price and the historical carbon trading price is calculated, and the error value is input into the carbon trading price prediction model as the second factor, and the carbon trading price prediction model is trained to re-output the carbon trading price; wherein, the carbon trading price prediction model is constructed according to an artificial intelligence model; the artificial intelligence model includes a convolutional neural network model or a long short-term memory neural network model.

5. The method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The profit function is constructed based on the second decision variable and the function constraint conditions are proposed, including: By formula Calculate the manufacturer's profit function; Retailers based on the manufacturer's P wholesale and e optimizes P retail , to maximize profits; Get the retailer's demand function N i (P retail , e), calculated by the formula Derive the retailer’s profit function; Construction constraint 1: P retail -P wholesale >0, indicating that the retail price is greater than the wholesale price; Construction constraint 2: N i (P retail , e)>0, indicating that the retailer’s product demand function is greater than zero; Construction constraint three: C tr ≥C low ≥0, indicating that the carbon footprint per unit product after emission reduction measures are taken is smaller than the carbon footprint before emission reduction.

6. A method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The obtaining of the optimal demand of the retailer comprises: Retrieve Pwholesale and e, and determine N i (P retail ,e) and (P retail -P wholesale ) are both greater than zero; if so, then N i (P retail ,e)·(P retail -P wholesale ) is maximized; otherwise, continue to judge; By formula P * retail (P wholesale ,e)=(N i -b i ·e+a i ·P wholesale ) / 2a i Calculate the retailer's optimal retail price; By formula N i * (P retail ,e)=N i * (P * retail (P wholesale ,e))=(N i -3b i ·ea i ·P wholesale ) / 2 calculates the optimal demand of the corresponding retailer.

7. A method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The decision optimization model for the manufacturer is constructed according to the optimal demand of the retailer, including: The optimal retail price of the retailer and the optimal demand of the retailer are imported into the cost function of each manufacturer to obtain: Retrieve the manufacturer’s total cost ZZC = SXC + TJC + YSC - TJC; Then through the formula Calculate the manufacturer's profit function; By formula Build a decision optimization model for manufacturers.

8. The method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The method of obtaining the number of iterations based on the quantum behavior particle swarm algorithm includes: S110: By formula The calculated quantum wave function describes the distribution probability of particles in the search space; x c is the center position of the normal distribution, and x is the position of the particle; S120: Mark the upper limit of the search space as x max , the lower limit of the search space is marked as x min , a random number uniformly distributed in the interval [0, 1] is marked as r i ; By formula Calculated particles The initial position of The initial position is taken as the initial individual optimal and the global optimal g 0 ; S130: Update the particle position in QPSO through quantum wave functions and random variables; S140: By formula Calculate the optimal individual position of the corresponding particle after update; By formula The global optimal position of the corresponding particle after the update is calculated; where f() is the objective function, which indicates the particle's updated position The fitness value at ; S150: Update the shrinkage factor by constructing a linear decreasing strategy; wherein the linear decreasing strategy is: β max and β min are the initial and final values ​​of β respectively; t max is the maximum number of iterations; t is the current number of iterations; S160: Retrieve the maximum number of iterations t max ; Determine whether the current number of iterations has reached the maximum number of iterations t max ; If yes, stop the iteration and output the number of iterations; if no, continue the iteration.

9. A method for optimizing product pricing considering carbon quotas according to claim 8, characterized in that: The updating of the particle positions in the QPSO by using quantum wave functions and random variables includes: S131: By formula The guiding position of the i-th particle for the t-th generation is calculated; S132: By formula Calculate the updated particle position; in, is the guiding position of the ith particle; is the position of the ith particle in the tth generation; u is a random number uniformly distributed in the interval (0, 1); the "±" sign is determined by a random method to determine the search direction of the particle.

10. The method for optimizing product pricing considering carbon quotas according to claim 1, characterized in that: The method of obtaining the optimal pricing decision based on the quantum behavior particle swarm algorithm includes: S210: Initializing parameters, where the parameters include: particle swarm size, contraction factor β, and initialization particle position; S220: Get the carbon trading price P EUA , calculate the fitness values ​​of all particles; S230: Compare the fitness value of the current particle with its historical optimal fitness value to determine whether the fitness value of the current particle is higher than the historical optimal fitness value of the corresponding particle; if yes, update the fitness value of the current particle to the optimal fitness value of the current particle; if no, use the historical optimal fitness value of the corresponding particle as the optimal position fitness value of the current particle; S240: Compare the current optimal position fitness value of each particle with the global optimal fitness value to determine whether the global optimal fitness value is higher than the optimal position fitness value of the corresponding particle; if yes, the global optimal fitness value remains unchanged; if no, the optimal position fitness value of the corresponding particle is used as the global optimal fitness value; S250: Calculate the historical best position and the global best position of each particle; S260: Determine whether the iterative operation meets the termination condition; if yes, the operation is terminated and the corresponding optimal pricing decision is output; if no, return to step S220; wherein the termination condition is that the number of iterations reaches the maximum number of iterations.

11. A system for optimizing product pricing considering carbon quotas, applied to a method for optimizing product pricing considering carbon quotas as claimed in any one of claims 1 to 10, characterized in that: include: Profit function building module, decision optimization module; Profit function construction module: obtain the carbon emissions per unit of production from the database and mark it as decision variable one; obtain the wholesale price of the product given by the manufacturer and mark it as decision variable two; construct the retailer's demand function based on decision variable one, and calculate the manufacturer's total cost based on the demand function; Retrieve the total cost of the manufacturer, construct the profit function based on the decision variable 2 and propose the function constraints; Decision optimization module: retrieve decision variables to obtain the optimal demand of retailers; build a decision optimization model for manufacturers based on the optimal demand of retailers; the decision variables include: decision variable 1 and decision variable 2; Obtain the manufacturer's profit value through the manufacturer's decision optimization model; The quantum behavior particle swarm algorithm is called and the profit value is used as the fitness value of the particle. The number of iterations and the optimal pricing decision are obtained based on the quantum behavior particle swarm algorithm.

Images