A fruit and vegetable prediction pricing and replenishment planning method based on an ARIMA time series model

Abstract

The application belongs to the technical field of supermarket planning, and provides a fruit and vegetable prediction pricing and replenishment planning method based on an ARIMA time series model, which comprises data collection, data preprocessing, model establishment, model training, prediction pricing, pricing strategy formulation, replenishment planning, monitoring and adjustment, and comprehensive optimization; the application can accurately predict the sales volume of the next week through time series analysis of the historical sales data of fruit and vegetable commodities by using the ARIMA model, can provide reliable basis for replenishment and pricing decisions, can calculate the total replenishment quantity according to the predicted sales volume and the loss rate, can ensure sufficient commodity supply and avoid excessive inventory, can reduce loss and waste, and can construct a cost-plus pricing model based on the price demand elasticity theory, can dynamically adjust the price in combination with the predicted sales volume, and can realize supply-demand balance and maximum profit.

Description

Technical Field

[0001] This invention belongs to the field of supermarket planning technology, specifically a method for predictive pricing and replenishment planning of fruits and vegetables based on the ARIMA time series model. Background Technology

[0002] In fresh food supermarkets, the replenishment and pricing strategies for fruits and vegetables are crucial to the operational efficiency and profitability of the supermarkets due to their short shelf life and sensitive sales time. The ARIMA model, as a common time series forecasting model, has the advantages of simple structure, fast forecasting speed, and convenient operation. It is suitable for capturing the trend and cyclical changes in the sales volume of fruits and vegetables. At the same time, by combining the theory of price elasticity of demand and nonlinear programming methods, replenishment and pricing strategies for fruits and vegetables can be formulated more scientifically, thereby maximizing the daily net revenue of supermarkets.

[0003] However, current traditional replenishment and pricing methods often rely on experience-based judgment, making it difficult to accurately predict market demand and dynamically adjust prices.

[0004] To address these issues, those skilled in the art have proposed a method for predictive pricing and replenishment planning of fruits and vegetables based on the ARIMA time series model. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a method for predictive pricing and replenishment planning of fruits and vegetables based on the ARIMA time series model. This method solves the problem that traditional replenishment and pricing methods in the prior art often rely on experience-based judgments, making it difficult to accurately predict market demand and dynamically adjust prices.

[0006] A method for predictive pricing and replenishment planning of fruits and vegetables based on the ARIMA time series model includes the following steps:

[0007] S1. Data Collection: Collect historical sales data, including daily sales volume, daily sales price, and cost information;

[0008] S2. Data Preprocessing: Cleaning, preprocessing, and transforming the collected data to ensure data quality and consistency;

[0009] S3. Model Building: Select an ARIMA model suitable for fruit and vegetable sales data, and determine the model parameters based on the characteristics of the data;

[0010] S4. Model Training: Train the ARIMA model using the preprocessed data;

[0011] S5. Predictive pricing: Use a trained model to predict the demand for fruits and vegetables in the future.

[0012] S6. Pricing Strategy Formulation: Based on forecasted demand and market conditions, formulate a reasonable pricing strategy;

[0013] S7. Replenishment Planning: Determine the quantity and timing of replenishment based on forecasted demand and current inventory levels;

[0014] S8. Monitoring and Adjustment: Regularly monitor the difference between actual sales data and forecast results, and adjust and optimize the model as needed;

[0015] S9. Comprehensive Optimization: Integrate forecast pricing and replenishment planning with other factors (such as cost, profit targets, etc.) to achieve overall optimization.

[0016] Preferably, in step S2, the cleaning, preprocessing, and transformation of the collected data includes the following steps:

[0017] S2.1 First, identify and correct errors and inconsistencies in the collected data, including handling missing values, handling duplicate values, and standardization; the formula for standardization is as follows:

[0018]

[0019] Where, x new x represents standardized data; x represents raw data; x represents... max x represents the maximum value of the sample data. min The minimum value of the sample data;

[0020] S2.2 Then, the cleaned data is preprocessed, including dimensionality reduction. The steps of the dimensionality reduction process are as follows:

[0021] First, assume λ1, λ2, ..., λ n Let R be the eigenvalues ​​of the correlation coefficient matrix. Then, the variance contribution rate of the k-th principal component is:

[0022]

[0023] The formula for calculating the cumulative variance contribution rate is:

[0024]

[0025] S2.3 Then, the preprocessed data is transformed to meet the requirements of the ARIMA model.

[0026] Preferably, in step S2, the establishment of the ARIMA model includes determining the model parameters p, d, and q, performing differencing on the non-stationary time series, and estimating and validating the model parameters. The model validation uses a verification algorithm to verify that the ARIMA model is correctly configured and can operate normally. Specifically, this involves setting outliers for the ARIMA model configuration and then using a Gaussian mixture model algorithm to test the running data of the configured ARIMA model. The formula for the Gaussian mixture model algorithm is as follows:

[0027]

[0028] Where x represents the amount of ARIMA model data; π i Represents the weights of the Gaussian mixture model; u i This represents the mean of the i-th Gaussian component;

[0029] This Gaussian mixture model algorithm can help outlier detection algorithms find anomalous data targets in the configured ARIMA model. Its function formula is:

[0030] L(λ|x)=P(x|λ)=ΠP(x i |λ).

[0031] Preferably, in step S4, the step of training the ARIMA model is as follows:

[0032] Step 1: Global Model Initialization and Distribution: The central server initializes the model and then distributes the global model parameters ω for this round. g Distribute to each node;

[0033] Step 2, Local Node Training: Participating nodes receive global model parameters, use local datasets for local training, and update the parameters of the local model. The calculation method is shown in Equation (2). In each training iteration, stochastic gradient descent is used as the local optimizer. After a set number of rounds of SGD calculation, the local node model uploads the new model parameters to the central server.

[0034]

[0035] Where t represents the current training round, η represents the learning rate, ▽ is the gradient calculation symbol, and f i (ω k The input parameter for the training instance is ω. k The resulting losses;

[0036] Step 3: The central server receives and aggregates the node model parameters: The penalty regularization term is calculated using equation (1), and the global model parameter ω is obtained by aggregating the local models. gThe aggregation method is shown in equation (3). When the penalty regularization term u is 1, the corresponding algorithm process is as follows:

[0037]

[0038] Step 4: Global Model Distribution: The central server distributes the global model parameters to each node;

[0039] Step 5: Repeat steps 2-4 until convergence or the required number of training rounds is reached.

[0040] Preferably, a fruit and vegetable forecasting, pricing, and replenishment planning system based on an ARI MA time series model includes:

[0041] The demand analysis module clarifies the specific needs of fruit and vegetable retailers;

[0042] The data collection and preprocessing module collects relevant data and performs preprocessing according to requirements;

[0043] The model building and prediction module builds the ARI MA model and predicts future sales volume.

[0044] The strategy formulation module develops pricing and replenishment strategies based on forecast results.

[0045] The implementation and monitoring module is used to put strategies into practice and monitor market changes in real time, making dynamic adjustments as needed.

[0046] A processor is configured to execute the above-described method for predictive pricing and replenishment planning of fruits and vegetables based on an ARI MA time series model.

[0047] A machine-readable storage medium storing instructions that, when executed by a processor, configure the processor to perform the aforementioned fruit and vegetable forecasting, pricing, and replenishment planning method based on an ARI MA time series model.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] 1. Improve forecast accuracy: By using the ARIMA model to perform time series analysis on historical sales data of fruits and vegetables, it is possible to accurately predict sales volume for the next week, providing a reliable basis for replenishment and pricing decisions;

[0050] 2. Optimize replenishment strategy: Calculate the total replenishment volume based on the predicted sales volume and loss rate to ensure sufficient supply of goods and avoid excessive inventory, thereby reducing losses and waste;

[0051] 3. Dynamic pricing adjustment: Based on the theory of price elasticity of demand, a cost-plus pricing model is constructed, and the price is dynamically adjusted in combination with the predicted sales volume to achieve supply and demand balance and maximize profits;

[0052] 4. Enhance the scientific nature of decision-making: By comprehensively considering the total replenishment volume, pricing strategy, and daily net profit of supermarkets through nonlinear programming models, the optimal replenishment and pricing plan is formulated, thereby improving the scientific nature and rationality of decision-making;

[0053] 5. Enhance risk resistance: This method can flexibly respond to market changes, adjust replenishment and pricing strategies in a timely manner, and enhance the risk resistance of supermarkets in complex market environments. Attached Figure Description

[0054] Figure 1 This is a flowchart of the fruit and vegetable forecasting pricing and replenishment planning method based on the ARIMA time series model of the present invention.

[0055] Figure 2 This is a framework diagram of the fruit and vegetable forecasting pricing and replenishment planning system based on the ARIMA time series model of the present invention. Detailed Implementation

[0056] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0057] like Figure 1 As shown:

[0058] Example 1: This invention provides a method for predictive pricing and replenishment planning of fruits and vegetables based on an ARIMA time series model, comprising the following steps:

[0059] S1. Data Collection: Collect historical sales data, including daily sales volume, daily sales price, and cost information;

[0060] S2. Data Preprocessing: Cleaning, preprocessing, and transforming the collected data to ensure data quality and consistency;

[0061] S3. Model Building: Select an ARIMA model suitable for fruit and vegetable sales data, and determine the model parameters based on the characteristics of the data;

[0062] S4. Model Training: Train the ARIMA model using the preprocessed data;

[0063] S5. Predictive pricing: Use a trained model to predict the demand for fruits and vegetables in the future.

[0064] S6. Pricing Strategy Formulation: Based on forecasted demand and market conditions, formulate a reasonable pricing strategy;

[0065] S7. Replenishment Planning: Determine the quantity and timing of replenishment based on forecasted demand and current inventory levels;

[0066] S8. Monitoring and Adjustment: Regularly monitor the difference between actual sales data and forecast results, and adjust and optimize the model as needed;

[0067] S9. Comprehensive Optimization: Integrate forecast pricing and replenishment planning with other factors.

[0068] As shown above, this invention uses the ARIMA model to perform time-series analysis on historical sales data of fruits and vegetables, accurately predicting sales volume for the coming week and providing a reliable basis for replenishment and pricing decisions. It calculates the total replenishment volume based on the predicted sales volume and loss rate, ensuring sufficient supply and avoiding excessive inventory, thus reducing losses and waste. A cost-plus pricing model is constructed based on the price elasticity of demand theory, dynamically adjusting prices according to the predicted sales volume to achieve supply-demand balance and maximize profits. A nonlinear programming model comprehensively considers the total replenishment volume, pricing strategy, and daily net profit of supermarkets to formulate the optimal replenishment and pricing plan, improving the scientific and rational nature of decision-making. This method can flexibly respond to market changes, adjust replenishment and pricing strategies in a timely manner, and enhance the resilience of supermarkets in complex market environments.

[0069] In one embodiment of the present invention, the cleaning, preprocessing and transformation of the collected data includes the following steps:

[0070] S2.1 First, identify and correct errors and inconsistencies in the collected data, including handling missing values, handling duplicate values, and standardization; the formula for standardization is as follows:

[0071]

[0072] Where, x new x represents standardized data; x represents raw data; x represents... max x represents the maximum value of the sample data. min The minimum value of the sample data;

[0073] S2.2 Then, the cleaned data is preprocessed, including dimensionality reduction. The steps of the dimensionality reduction process are as follows:

[0074] First, assume λ1, λ2, ..., λ n Let R be the eigenvalues ​​of the correlation coefficient matrix. Then, the variance contribution rate of the k-th principal component is:

[0075]

[0076] The formula for calculating the cumulative variance contribution rate is:

[0077]

[0078] S2.3 Then, the preprocessed data is transformed to meet the requirements of the ARIMA model.

[0079] As can be seen from the above, standardization formulas can ensure the consistency and comparability of data, enabling data of different dimensions to be analyzed and predicted in the same model. At the same time, dimensionality reduction algorithms can extract the main features of the data, reduce data dimensionality, and improve computational efficiency and the accuracy of model predictions.

[0080] In one embodiment of the present invention, the establishment of the ARIMA model includes determining the model parameters p, d, and q; performing differencing on the non-stationary time series to stabilize it for time series analysis; and estimating and validating the model parameters. The model validation employs a verification algorithm to verify that the ARIMA model is correctly configured and can operate normally. Specifically, this involves setting outliers for the ARIMA model configuration and then using a Gaussian mixture model algorithm to test the running data of the configured ARIMA model. The formula for the Gaussian mixture model algorithm is as follows:

[0081]

[0082] Where x represents the amount of ARIMA model data; π i Represents the weights of the Gaussian mixture model; u i This represents the mean of the i-th Gaussian component;

[0083] This Gaussian mixture model algorithm can help outlier detection algorithms find anomalous data targets in the configured ARIMA model. Its function formula is:

[0084] L(λ|x)=P(x|λ)=ΠP(x i |λ).

[0085] As shown above, the Gaussian mixture model algorithm can help detect abnormal data targets in the operation of the ARIMA model, thereby ensuring the correct configuration and normal operation of the model.

[0086] In one embodiment of the present invention, the steps for training the ARIMA model are as follows:

[0087] Step 1: Global Model Initialization and Distribution: The central server initializes the model and then distributes the global model parameters ω for this round. g Distribute to each node;

[0088] Step 2, Local Node Training: Participating nodes receive global model parameters, use local datasets for local training, and update the parameters of the local model. The calculation method is shown in Equation (2). In each training iteration, stochastic gradient descent is used as the local optimizer. After a set number of rounds of SGD calculation, the local node model uploads the new model parameters to the central server.

[0089]

[0090] Where t represents the current training round, η represents the learning rate, ▽ is the gradient calculation symbol, and f i (ω k ) is the input parameter ω for the training instance. k The resulting losses;

[0091] Step 3: The central server receives and aggregates the node model parameters: The penalty regularization term is calculated using equation (1), and the global model parameter ω is obtained by aggregating the local models. g The aggregation method is shown in equation (3). When the penalty regularization term u is 1, the corresponding algorithm process is as follows:

[0092]

[0093] Step 4: Global Model Distribution: The central server distributes the global model parameters to each node;

[0094] Step 5: Repeat steps 2-4 until convergence or the required number of training rounds is reached.

[0095] As can be seen from the above, stochastic gradient descent and aggregation formulas in model training can efficiently train models in a distributed environment, improving the model's generalization ability and prediction performance.

[0096] like Figure 2 As shown:

[0097] Example 2: This invention provides a fruit and vegetable forecasting pricing and replenishment planning system based on the ARI MA time series model, including a demand analysis module, a data collection and preprocessing module, a model building and prediction module, a strategy formulation module, and an implementation and monitoring module, wherein the demand analysis module, data collection and preprocessing module, model building and prediction module, strategy formulation module, and implementation and monitoring module are electrically connected in sequence.

[0098] The demand analysis module clarifies the specific needs of fruit and vegetable retailers;

[0099] The data collection and preprocessing module collects relevant data and performs preprocessing according to requirements;

[0100] The model building and prediction module builds the ARI MA model and predicts future sales volume.

[0101] The strategy formulation module develops pricing and replenishment strategies based on forecast results.

[0102] The implementation and monitoring module is used to put strategies into practice and monitor market changes in real time, making dynamic adjustments as needed.

[0103] Those skilled in the art will understand that embodiments of this application can be provided as a system or a computer program product. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0104] This application is described with reference to flowchart illustrations and / or block diagrams of devices (systems) and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0105] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0106] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0107] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0108] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0109] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0110] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, article, or apparatus that includes that element.

[0111] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for predictive pricing and replenishment planning of fruits and vegetables based on an ARIMA time series model, characterized in that: Includes the following steps: S1. Data Collection: Collect historical sales data, including daily sales volume, daily sales price, and cost information; S2. Data Preprocessing: Cleaning, preprocessing, and transforming the collected data to ensure data quality and consistency; S3. Model Building: Select an ARIMA model suitable for fruit and vegetable sales data, and determine the model parameters based on the characteristics of the data; S4. Model Training: Train the ARIMA model using the preprocessed data; S5. Predictive pricing: Use a trained model to predict the demand for fruits and vegetables in the future. S6. Pricing Strategy Formulation: Based on forecasted demand and market conditions, formulate a reasonable pricing strategy; S7. Replenishment Planning: Determine the quantity and timing of replenishment based on forecasted demand and current inventory levels; S8. Monitoring and Adjustment: Regularly monitor the difference between actual sales data and forecast results, and adjust and optimize the model as needed; S9. Comprehensive Optimization: Integrate forecast pricing and replenishment planning with other factors.

2. The method for predictive pricing and replenishment planning of fruits and vegetables based on an ARIMA time series model as described in claim 1, characterized in that: In step S2, the cleaning, preprocessing, and transformation of the collected data includes the following steps: S2.1 First, identify and correct errors and inconsistencies in the collected data, including handling missing values, handling duplicate values, and standardization; the formula for standardization is as follows: Where, x new x represents standardized data; x represents raw data; x represents... max x represents the maximum value of the sample data. min The minimum value of the sample data; S2.2 Then, the cleaned data is preprocessed, including dimensionality reduction. The steps of the dimensionality reduction process are as follows: First, assume λ1, λ2, ..., λ n Let R be the eigenvalues ​​of the correlation coefficient matrix. Then, the variance contribution rate of the k-th principal component is: The formula for calculating the cumulative variance contribution rate is: S2.3 Then, the preprocessed data is transformed to meet the requirements of the ARIMA model.

3. The method for predictive pricing and replenishment planning of fruits and vegetables based on an ARIMA time series model as described in claim 1, characterized in that: In step S2, the establishment of the ARIMA model includes determining the model parameters p, d, and q, performing differencing on the non-stationary time series, and estimating and validating the model parameters. The model validation uses a verification algorithm to verify that the ARIMA model is correctly configured and can operate normally. Specifically, this involves setting outliers for the ARIMA model configuration and then using a Gaussian mixture model algorithm to test the running data of the configured ARIMA model. The formula for the Gaussian mixture model algorithm is as follows: Where x represents the amount of ARIMA model data; π i Represents the weights of the Gaussian mixture model; u i This represents the mean of the i-th Gaussian component; This Gaussian mixture model algorithm can help outlier detection algorithms find anomalous data targets in the configured ARIMA model. Its function formula is: L(λx)=P(xλ)=ΠP(x i l)。 4. The method for predictive pricing and replenishment planning of fruits and vegetables based on an ARIMA time series model as described in claim 1, characterized in that: In step S4, the steps for training the ARIMA model are as follows: Step 1: Global Model Initialization and Distribution: The central server initializes the model and then distributes the global model parameters ω for this round. g Distribute to each node; Step 2, Local Node Training: Participating nodes receive global model parameters, use local datasets for local training, and update the parameters of the local model. The calculation method is shown in Equation (2). In each training iteration, stochastic gradient descent is used as the local optimizer. After a set number of rounds of SGD calculation, the local node model uploads the new model parameters to the central server. Where t represents the current training round, and η represents the learning rate. f is the gradient calculation symbol. i (ω k The input parameter for the training instance is ω. k The resulting losses; Step 3: The central server receives and aggregates the node model parameters: The penalty regularization term is calculated using equation (1), and the global model parameter ω is obtained by aggregating the local models. g The aggregation method is shown in equation (3). When the penalty regularization term u is 1, the corresponding algorithm process is as follows: Step 4: Global Model Distribution: The central server distributes the global model parameters to each node; Step 5: Repeat steps 2-4 until convergence or the required number of training rounds is reached.

5. The fruit and vegetable forecasting, pricing, and replenishment planning system based on the ARIMA time series model as described in claim 1, characterized in that: include: The demand analysis module clarifies the specific needs of fruit and vegetable retailers; The data collection and preprocessing module collects relevant data and performs preprocessing according to requirements; The model building and prediction module builds an ARIMA model and predicts future sales volume. The strategy formulation module develops pricing and replenishment strategies based on forecast results. The implementation and monitoring module is used to put strategies into practice and monitor market changes in real time, making dynamic adjustments as needed.

6. A processor, characterized in that, It is configured to perform the fruit and vegetable forecasting pricing and replenishment planning method based on the ARIMA time series model according to any one of claims 1 to 4.

7. A machine-readable storage medium storing instructions thereon, characterized in that, When executed by the processor, the instruction causes the processor to be configured to perform the fruit and vegetable forecasting pricing and replenishment planning method based on the ARIMA time series model according to any one of claims 1 to 4.

Images