CN-121998696-A - Inventory constraint-oriented dynamic pricing collaborative optimization method and system

CN121998696ACN 121998696 ACN121998696 ACN 121998696ACN-121998696-A

Abstract

The application relates to the technical field of data processing, and discloses a dynamic pricing collaborative optimization method and a system for inventory constraint. The method comprises the steps of obtaining a historical sales data set and inventory parameters, calculating an inventory health index, fitting to obtain an initial price elastic coefficient, carrying out hyperbolic tangent transformation and modulation operation on the inventory health index to obtain a dynamic price elastic coefficient, carrying out demand prediction based on the dynamic price elastic coefficient, and establishing a multi-period yield optimization model for solving to obtain a pricing strategy and a replenishment scheme. The method solves the technical problem that the demand prediction model in the prior art cannot reflect the influence of the inventory state on the price sensitivity of the consumer, and improves the demand prediction precision and the controllability of the inventory evolution track.

Inventors

LIU MENGQIN
HUI MING
CHENG HONGFEI
ZHAO QINGTAO
LIU HONGXIA
HAI TAO

Assignees

南阳师范学院

Dates

Publication Date: 20260508
Application Date: 20260206

Claims (10)

1. An inventory constraint-oriented dynamic pricing collaborative optimization method, the method comprising: Step S1, acquiring a historical sales data set, a current stock quantity, a safety stock threshold value and a maximum stock capacity, and carrying out division operation by taking the difference value between the current stock quantity and the safety stock threshold value as a numerator and the difference value between the maximum stock capacity and the safety stock threshold value as a denominator to obtain an stock health index; S2, carrying out nonlinear regression fitting on the basis of price data and sales volume data in the historical sales data set to obtain an initial price elasticity coefficient, inputting the inventory health index into a hyperbolic tangent function to carry out nonlinear transformation to obtain a transformation value, multiplying the transformation value by an elasticity modulation intensity coefficient, adding 1 to obtain a modulation factor, and multiplying the modulation factor by the initial price elasticity coefficient to obtain a price elasticity coefficient which dynamically changes along with inventory health; S3, substituting the price elastic coefficient as a variable parameter into a demand calculation formula to predict the demand, and obtaining a demand prediction result associated with the stock state; And S4, establishing a multi-period gain optimization model according to the demand prediction result and solving to obtain a pricing strategy and a replenishment scheme.
2. The inventory constraint-oriented dynamic pricing collaborative optimization method according to claim 1, wherein step S1 comprises: extracting sales records in a preset historical time window from a sales system database, wherein the sales records comprise time sequence tags, price sequences at corresponding moments, actual sales volume sequences and inventory level sequences; Acquiring real-time inventory quantity at the current moment, a safety inventory threshold preset according to the storage physical space and the commodity shelf life and the maximum inventory capacity of the storage system from the inventory management system; Subtracting the real-time stock quantity from the safety stock threshold to obtain an inventory safety allowance, and subtracting the maximum inventory capacity from the safety stock threshold to obtain an inventory capacity interval; And dividing the stock safety allowance by taking the stock capacity interval as a denominator to obtain an index of the stock health degree.
3. The inventory constraint-oriented dynamic pricing collaborative optimization method according to claim 1, wherein performing a nonlinear regression fit based on price data and sales volume data in the historical sales data set in step S2 to obtain an initial price elasticity coefficient comprises: Extracting a price sequence and an actual sales volume sequence from the historical sales data set as training sample pairs; Constructing a basic demand expression in a power function form, wherein the basic demand expression is calculated by multiplying a price variable by a basic demand scale parameter after the price variable is raised to the negative power, and the exponent absolute value of the negative power is an initial price elasticity coefficient; Substituting the price sequence and the actual sales volume sequence into the basic demand expression, and calculating a square error between the predicted sales volume and the actual sales volume; And iteratively adjusting the basic demand scale parameter and the initial price elastic coefficient through a least square method to ensure that the sum of the square errors reaches the minimum value, thereby obtaining the initial price elastic coefficient.
4. The inventory constraint-oriented dynamic pricing collaborative optimization method according to claim 3, wherein in the step S2, the inventory health index is input into a hyperbolic tangent function to perform nonlinear transformation to obtain a transformation value, and the method comprises the steps of: multiplying the inventory health index with a preset inventory response sensitivity parameter to obtain an inventory state value with adjusted sensitivity; Inputting the stock state value into a hyperbolic tangent function to perform nonlinear mapping calculation, wherein the hyperbolic tangent function maps the input value to a numerical interval from minus 1 to plus 1; And carrying out numerical extraction on the output result of the hyperbolic tangent function to obtain a conversion value with the value range of minus 1 to plus 1.
5. The inventory constraint oriented dynamic pricing collaborative optimization method according to claim 4, wherein in the step S2, the modulation factor is obtained by multiplying the transformed value by an elastic modulation intensity coefficient and then adding 1, and the multiplying the modulation factor by the initial price elastic coefficient comprises: multiplying the transformation value with a preset elastic modulation intensity coefficient to obtain an elastic modulation quantity; adding the elastic modulation quantity with a value 1 to obtain a modulation factor; and multiplying the modulation factor by the initial price elasticity coefficient to obtain the price elasticity coefficient which dynamically changes along with the inventory health degree.
6. The inventory constraint-oriented dynamic pricing collaborative optimization method according to claim 1, wherein substituting the price elasticity coefficient as a variable parameter into a demand computing equation for demand prediction in step S3 comprises: Constructing a demand calculation formula, wherein the demand calculation formula calculates by multiplying a price variable by a basic demand scale parameter after taking a negative power of several, and the exponent absolute value of the negative power is a price elasticity coefficient; substituting the price elasticity coefficient which dynamically changes along with the inventory health degree into the demand calculation formula as the current value of the price elasticity coefficient; substituting the price variable to be optimized into the demand calculation formula, and calculating to obtain a demand predicted value through power operation and multiplication operation; And carrying out association marking on the demand quantity predicted value and the inventory health index to obtain an inventory state associated demand predicted result.
7. The inventory constraint-oriented dynamic pricing collaborative optimization method according to claim 6, wherein the establishing a multi-cycle revenue optimization model and solving in step S4 according to the demand prediction result comprises: Setting an optimization time range as a plurality of continuous decision periods, and constructing a discrete time inventory dynamic evolution equation based on the demand prediction result, wherein the inventory dynamic evolution equation obtains the inventory of the current period by subtracting the demand prediction value from the inventory of the previous period and adding the inventory supplement; Multiplying the price variable with the demand forecast value according to the demand forecast result to obtain sales income, and subtracting the purchase cost, the inventory holding cost, the backdrop penalty cost and the inventory overflow cost from the sales income to obtain single-period net income; accumulating and summing the single-period net benefits of all decision periods to construct an objective function of a multi-period benefit optimization model; And under the conditions of upper and lower limit constraint of stock quantity and upper limit constraint of replenishment quantity, solving the objective function through a mixed integer linear programming algorithm to obtain a pricing strategy and a replenishment scheme of each decision period.
8. An inventory constraint oriented dynamic pricing collaborative optimization system for implementing an inventory constraint oriented dynamic pricing collaborative optimization method in accordance with any of claims 1-7, the inventory constraint oriented dynamic pricing collaborative optimization system comprising: The acquisition module is used for acquiring a historical sales data set, a current stock quantity, a safety stock threshold value and a maximum stock capacity, taking the difference value between the current stock quantity and the safety stock threshold value as a numerator, and taking the difference value between the maximum stock capacity and the safety stock threshold value as a denominator to carry out division operation to obtain an stock health index; the fitting module is used for carrying out nonlinear regression fitting on the basis of the price data and sales volume data in the historical sales data set to obtain an initial price elastic coefficient, inputting the stock health index into a hyperbolic tangent function to carry out nonlinear transformation to obtain a transformation value, multiplying the transformation value by an elastic modulation intensity coefficient, adding 1 to obtain a modulation factor, and multiplying the modulation factor by the initial price elastic coefficient to obtain a price elastic coefficient which dynamically changes along with the stock health; The prediction module is used for substituting the price elastic coefficient as a variable parameter into a demand calculation formula to predict the demand so as to obtain a demand prediction result related to the inventory state; and the solving module is used for establishing a multi-period gain optimization model according to the demand prediction result and solving the multi-period gain optimization model to obtain a pricing strategy and a replenishment scheme.
9. The system of claim 8, wherein obtaining a historical sales data set, a current inventory amount, a safety inventory threshold value, and a maximum inventory capacity, dividing a difference between the current inventory amount and the safety inventory threshold value as a numerator, and a difference between the maximum inventory capacity and the safety inventory threshold value as a denominator, to obtain an inventory health index, comprising: extracting sales records in a preset historical time window from a sales system database, wherein the sales records comprise time sequence tags, price sequences at corresponding moments, actual sales volume sequences and inventory level sequences; Acquiring real-time inventory quantity at the current moment, a safety inventory threshold preset according to the storage physical space and the commodity shelf life and the maximum inventory capacity of the storage system from the inventory management system; Subtracting the real-time stock quantity from the safety stock threshold to obtain an inventory safety allowance, and subtracting the maximum inventory capacity from the safety stock threshold to obtain an inventory capacity interval; And dividing the stock safety allowance by taking the stock capacity interval as a denominator to obtain an index of the stock health degree.
10. The system of claim 9, wherein performing a non-linear regression fit based on the price data and sales volume data in the historical sales data set to obtain an initial price elasticity coefficient comprises: Extracting a price sequence and an actual sales volume sequence from the historical sales data set as training sample pairs; Constructing a basic demand expression in a power function form, wherein the basic demand expression is calculated by multiplying a price variable by a basic demand scale parameter after the price variable is raised to the negative power, and the exponent absolute value of the negative power is an initial price elasticity coefficient; Substituting the price sequence and the actual sales volume sequence into the basic demand expression, and calculating a square error between the predicted sales volume and the actual sales volume; And iteratively adjusting the basic demand scale parameter and the initial price elastic coefficient through a least square method to ensure that the sum of the square errors reaches the minimum value, thereby obtaining the initial price elastic coefficient.

Description

Inventory constraint-oriented dynamic pricing collaborative optimization method and system Technical Field The application relates to the technical field of data processing, in particular to a dynamic pricing collaborative optimization method and a system for inventory constraint. Background In the field of inventory management and pricing decision, in the prior art, a mode of separating pricing optimization and inventory management is generally adopted, namely, a demand prediction model with fixed parameters is firstly established according to historical sales data, then a pricing strategy and an inventory replenishment plan are respectively formulated based on the model, key parameters such as a price elastic coefficient in the demand prediction model are kept unchanged after fitting is completed, the pricing decision is mainly adjusted according to market demand fluctuation and a cost structure, inventory management is focused on setting of a safety inventory threshold value and a replenishment period, and a real-time information feedback and cooperative adjustment mechanism is lacked between the two. The main disadvantage of the prior art is that the static state of the demand prediction model cannot accurately reflect the influence of the stock state on the purchasing behavior of the consumer, when the stock level is changed significantly, the sensitivity of the consumer to the price is changed along with the change, the consumer has lower tolerance to price rising and tends to wait for price reduction promotion when the stock is abundant, the consumer has higher acceptance of the price and worries about shortage of the stock to accelerate purchasing decision when the stock is tension, but the fixed price elasticity coefficient cannot capture the modulating effect of the stock state to the price sensitivity, so that the demand prediction deviates from the actual situation, and the pricing strategy and the stock decision formulated based on the prediction are difficult to realize real collaborative optimization. Disclosure of Invention The application provides a dynamic pricing collaborative optimization method and a system for inventory constraint, which are used for solving the technical problem that a demand prediction model in the prior art cannot reflect the influence of an inventory state on price sensitivity of a consumer by constructing a dynamic price elastic coefficient modulation mechanism driven by an inventory health index, and improving the demand prediction precision and the controllability of an inventory evolution track. In a first aspect, the present application provides an inventory constraint-oriented dynamic pricing collaborative optimization method, where the inventory constraint-oriented dynamic pricing collaborative optimization method includes: Step S1, acquiring a historical sales data set, a current stock quantity, a safety stock threshold value and a maximum stock capacity, and carrying out division operation by taking the difference value between the current stock quantity and the safety stock threshold value as a numerator and the difference value between the maximum stock capacity and the safety stock threshold value as a denominator to obtain an stock health index; S2, carrying out nonlinear regression fitting on the basis of price data and sales volume data in the historical sales data set to obtain an initial price elasticity coefficient, inputting the inventory health index into a hyperbolic tangent function to carry out nonlinear transformation to obtain a transformation value, multiplying the transformation value by an elasticity modulation intensity coefficient, adding 1 to obtain a modulation factor, and multiplying the modulation factor by the initial price elasticity coefficient to obtain a price elasticity coefficient which dynamically changes along with inventory health; S3, substituting the price elastic coefficient as a variable parameter into a demand calculation formula to predict the demand, and obtaining a demand prediction result associated with the stock state; And S4, establishing a multi-period gain optimization model according to the demand prediction result and solving to obtain a pricing strategy and a replenishment scheme. The acquisition module is used for acquiring a historical sales data set, a current stock quantity, a safety stock threshold value and a maximum stock capacity, taking the difference value between the current stock quantity and the safety stock threshold value as a numerator, and taking the difference value between the maximum stock capacity and the safety stock threshold value as a denominator to carry out division operation to obtain an stock health index; the fitting module is used for carrying out nonlinear regression fitting on the basis of the price data and sales volume data in the historical sales data set to obtain an initial price elastic coefficient, inputting the stock health index into a hyperbolic tangent function