Supply chain cost optimization method and device based on big data, equipment and medium

Abstract

The invention discloses a supply chain cost optimization method and device based on big data, equipment and a medium. The method comprises the steps that supply chain real-time data and supply chain historical data of a target supply chain phase are collected through Internet of Things equipment; establishing a supply chain data view based on the supply chain real-time data and the supply chain historical data; processing the supply chain data view according to a machine learning model to determine prediction data of the target supply chain; and performing linear programming processing on the prediction data of the target supply chain to determine an optimization strategy corresponding to the target supply chain. Based on the above technical scheme, demand prediction and supply chain bottleneck identification are carried out through the processed data set and the machine learning model, market demand changes can be captured more accurately, inventory cost, transportation cost and supply chain risks are optimized through a linear programming algorithm, and more efficient cost control and risk management are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of supply chain management, and particularly relates to a supply chain cost optimization method and device based on big data, equipment and medium. BACKGROUND

[0002] In modern supply chain management, enterprises face multiple complex challenges, such as high inventory costs, frequent transportation cost fluctuations, and potential operational risks in various links of the supply chain. To address these issues, enterprises urgently need more accurate data support and intelligent decision-making models.

[0003] However, traditional supply chain management usually relies on simple statistical models for demand forecasting, such as moving average method, linear regression, etc. The above methods have insufficient prediction accuracy when dealing with complex and variable market environments, and cannot effectively capture the nonlinear changes and seasonal fluctuations of market demand, resulting in frequent inventory accumulation or shortage. SUMMARY

[0004] The present application provides a supply chain cost optimization method and device based on big data, equipment and medium, which can more accurately capture market demand changes by applying machine learning models for demand forecasting and supply chain bottleneck identification, and optimize inventory costs, transportation costs and supply chain risks through linear programming algorithms, achieving more efficient cost control and risk management.

[0005] According to one aspect of the present application, a supply chain cost optimization method based on big data is provided, the method comprising:

[0006] collecting real-time supply chain data and historical supply chain data of the target supply chain through Internet of Things devices;

[0007] establishing a supply chain data view based on the real-time supply chain data and the historical supply chain data;

[0008] processing the supply chain data view according to a machine learning model to determine predicted data of the target supply chain;

[0009] linearly programming the predicted data of the target supply chain to determine an optimization strategy corresponding to the target supply chain.

[0010] According to another aspect of the present application, a supply chain cost optimization device based on big data is provided, the device comprising:

[0011] a data collection module for collecting real-time supply chain data and historical supply chain data of the target supply chain through Internet of Things devices;

[0012] The data processing module is configured to establish a supply chain data view based on the supply chain real-time data and the supply chain historical data.

[0013] The data prediction module is configured to process the supply chain data view according to a machine learning model to determine prediction data of the target supply chain.

[0014] The optimization strategy determination module is configured to perform linear programming processing on the prediction data of the target supply chain to determine an optimization strategy corresponding to the target supply chain.

[0015] According to another aspect of the present application, an electronic device is provided, which comprises:

[0016] at least one processor; and

[0017] a memory in communication with the at least one processor; wherein

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the big data-based supply chain cost optimization method of any of the embodiments of the present application.

[0019] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to perform the big data-based supply chain cost optimization method of any of the embodiments of the present application when executed by the processor.

[0020] The technical solution of the embodiments of the present application collects supply chain real-time data and supply chain historical data of a target supply chain through an Internet of Things device, establishes a supply chain data view based on the supply chain real-time data and the supply chain historical data, processes the supply chain data view according to a machine learning model to determine prediction data of the target supply chain, and performs linear programming processing on the prediction data of the target supply chain to determine an optimization strategy corresponding to the target supply chain. Based on the above technical solution, demand prediction and supply chain bottleneck identification are performed by applying a machine learning model to the processed data set, which can more accurately capture market demand changes, and inventory cost, transportation cost and supply chain risk are optimized through a linear programming algorithm, thereby achieving more efficient cost control and risk management.

[0021] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor to limit the scope of the present application. Other features of the present application will become apparent through the following description. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to make the technical solution in the embodiments of the present application clearer, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0023] Figure 1 is a flow chart of a supply chain cost optimization method based on big data provided by an embodiment of the present application;

[0024] Figure 2 is a flow chart of a supply chain cost optimization method based on big data provided by an embodiment of the present application;

[0025] Figure 3 is a structural schematic diagram of a supply chain cost optimization device based on big data provided by an embodiment of the present application;

[0026] Figure 4 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to make the technical solution in the embodiments of the present application clearer, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0028] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] Figure 1A flowchart of a supply chain cost optimization method based on big data is provided for an embodiment of the present application. The embodiment can be applicable to the case of optimizing the cost of a supply chain. The method can be executed by a supply chain cost optimization device based on big data. The device can be realized in the form of hardware and / or software and can be configured in an electronic device. As shown in Figure 1 The method specifically includes the following steps:

[0030] S110, collecting supply chain real-time data and supply chain historical data of a target supply chain through an Internet of Things device.

[0031] The Internet of Things device can be a sensing device deployed in the supply chain. The target supply chain can be understood as a supply chain that needs to be optimized in cost. The supply chain real-time data can be real-time data collected through the Internet of Things device. The supply chain historical data can be data generated in the process of historical operation of the supply chain. The real-time data includes inventory level, transportation status and environmental conditions. The historical supply chain data includes procurement records, sales data and supplier performance

[0032] Specifically, real-time data in the warehouse and transportation is collected through the IoT device, including inventory level, transportation status and environmental conditions. Historical supply chain data, including procurement records, sales data and supplier performance, is integrated. Exemplarily, IoT devices can be deployed at key nodes of the supply chain, such as production equipment, storage shelves, transportation vehicles, etc., to collect real-time data and historical data.

[0033] S120, establishing a supply chain data view based on the supply chain real-time data and the supply chain historical data.

[0034] The supply chain data view can be a visualized data view obtained after integrating and processing various data in the whole process of the supply chain.

[0035] Specifically, ETL tools are used to process missing values, abnormal values and duplicate data. The processed multi-source data is centralized through a data lake to establish a unified data view. Exemplarily, data integration tools, such as ETL (extraction, transformation, loading) processes, can be used to uniformly clean, convert and integrate real-time and historical data scattered in different systems and formats, to ensure the consistency and accuracy of the data. Data warehouse technology is used to build a multi-dimensional data model. The integrated data is organized and stored according to dimensions such as supply chain links, time series and product categories. Then, according to big data visualization technology, such as BI (business intelligence) tools or customized visualization platforms, the data is presented in the form of intuitive charts, dashboards, etc., to form a supply chain data view.

[0036] On the basis of the above technical solutions, the supply chain data view is established based on the supply chain real-time data and the supply chain historical data, including: data preprocessing of the supply chain real-time data and the supply chain historical data according to a data processing tool; establishing a supply chain data lake corresponding to the target supply chain based on the preprocessed supply chain real-time data and the supply chain historical data; and obtaining the supply chain data view by processing the supply chain data in the supply chain data lake according to a preset processing mode.

[0037] The preset processing mode includes data modeling, defining data structure, formulating data standard, and creating data dictionary. The data processing tool can be a tool for preprocessing data, which can be an ETL tool. The supply chain data lake is a centralized data storage system designed for supply chain management, used for storing structured, semi-structured and unstructured data.

[0038] Specifically, the ETL tool is used to preprocess the two types of data, including data cleaning to remove noise and outliers, data conversion to unify format and coding, data integration to merge multi-source data, and based on the preprocessed data, a supply chain data lake corresponding to the target supply chain is constructed, and distributed storage technology such as Hadoop HDFS is used to process the data in the data lake according to the preset processing mode. Through data modeling, the association between data is clear; defining data structure, standardizing data storage format; formulating data standards to ensure data accuracy; creating data dictionary to provide data explanation and description, and generating supply chain data view.

[0039] For example, the ETL (Extract, Transform, Load) tool is used to process the data. First, the original data is extracted from multiple data sources, including warehouse management system, transportation management system and historical records, etc. Through data conversion, missing values, outliers and duplicate data in the data are processed. Missing values can be filled by interpolation method or substitution method, outliers are corrected by statistical analysis or rule correction, and duplicate data is cleaned by de-duplication algorithm. The cleaned data is stored in the data lake, which allows storage of structured, semi-structured and unstructured data, and unifies management to ensure that data from different sources can be integrated on a unified platform. In the data lake, a unified data view is created using data modeling techniques, including defining data structure, formulating data standard and creating data dictionary. The data view can include data warehouse models (such as star model or snowflake model) or data virtualization layer to provide a clear data basis for analysis and optimization.

[0040] S130, processing the supply chain data view according to the machine learning model to determine the predicted data of the target supply chain.

[0041] The machine learning model can be a pre-trained machine learning model, and can include a long short-term memory network model and a random forest regression model. The long short-term memory network model is used to determine demand prediction data, and the random forest regression model is used to determine bottleneck prediction data.

[0042] Specifically, feature data closely related to the prediction target is filtered from the supply chain data view, such as historical sales volume, inventory turnover rate, order delivery cycle, etc. According to the type of prediction problem, such as time series prediction or classification prediction, a suitable machine learning algorithm is selected, such as an LSTM neural network for time series sales prediction and a random forest algorithm for supply chain risk classification prediction. The model is trained using historical data, and the model performance is optimized through cross-validation and parameter tuning to ensure that the model has high accuracy and generalization ability. The trained model is applied to the real-time supply chain data view to generate prediction data for the target supply chain, such as sales prediction for a future period of time, inventory demand prediction, etc.

[0043] On the basis of the above technical solution, the prediction data includes demand prediction data; the machine learning model is used to process the supply chain data view to determine the prediction data of the target supply chain, including: extracting demand data of the target supply chain from the supply chain data view based on a preset time step; and processing the demand data based on a long short-term memory network model to determine demand prediction data of the target supply chain.

[0044] The preset time step can be a pre-set time step for extracting data. The long short-term memory network model is provided with at least two layers of LSTM units; the first LSTM unit is used to output the hidden state of all time steps; the second LSTM unit is used to output the hidden state of the last time step; and an anti-overfitting layer is added after the second layer of LSTM units. The demand prediction data is a set of information used to estimate future product or service demand.

[0045] Specifically, according to a preset time step, such as daily, weekly or monthly, historical demand data of the target supply chain is extracted from the supply chain data view, and a long short-term memory network model is constructed, which includes at least two layers of LSTM units. The first layer of LSTM units is responsible for processing the input demand data sequence and outputting the hidden state of all time steps. The second layer of LSTM units further processes the output of the first layer, focusing on extracting long-term dependencies of the sequence, and only outputs the hidden state of the last time step. It should be noted that, in order to prevent model overfitting, an anti-overfitting layer, such as a Dropout layer or a regularization layer, is added after the second layer of LSTM units. The trained model is used to process the demand data to determine the demand prediction data of the target supply chain.

[0046] For example, the demand change is predicted by using an LSTM model, and the specific structure of the LSTM model is as follows: an input layer with an input shape of (batch_size, time_steps, features), where batch_size represents the number of samples in each training batch; time_steps represents the time step; and features represent the number of features at each time step;

[0047] The first LSTM layer contains 64 LSTM units, uses the tanh activation function by default, and uses the sigmoid activation function for internal state, where the return sequence is set to True to output the hidden state of all time steps for the next LSTM layer; the second LSTM layer contains 32 LSTM units, and the return sequence is set to False to output the hidden state of the last time step; a Droport layer is added after the second LSTM layer; the fully connected layer contains 16 neurons for nonlinear transformation of the output of the second LSTM layer, and the ReLU activation function is used; and the output layer contains 1 neuron using the linear activation function.

[0048] The loss function of the LSTM model is as follows: ; wherein, represents the total loss function of the LSTM model; represents the weighted mean square error; represents the regularization term; ; wherein, represents the weight of the i-th sample or time step; represents the true value of the i-th sample of the LSTM model; represents the predicted value of the i-th sample of the LSTM model; ; wherein, represents the regularization coefficient; p represents the total number of LSTM model parameters; represents the j-th LSTM model parameter.

[0049] In this embodiment, the model can effectively capture long-term dependencies and complex patterns in time series through a two-layer LSTM structure. The first LSTM layer returns the hidden states of all time steps, preserving the complete time series information, allowing the second LSTM to further process and refine this information. This hierarchical processing approach helps improve the model's prediction accuracy. A Dropout layer is added after the second LSTM, which randomly drops a portion of neurons to prevent the model from over-relying on specific neurons, effectively reducing the risk of overfitting. This is particularly important for LSTM models that handle high-dimensional features. The loss function uses Weighted MSE, which allows different weights to be set for different samples or time steps, allowing the model to better focus on key time steps or important samples and improve prediction accuracy at important times. At the same time, the loss function includes an L2 regularization term, which helps control the complexity of the model, reduce overfitting of parameters, and improve the model's generalization ability.

[0050] On the basis of the above technical solution, the prediction data includes bottleneck prediction data; the prediction data of the target supply chain is determined by processing the supply chain data view according to the machine learning model, including: determining the supply chain link corresponding to the target supply chain, and extracting the link data corresponding to each supply chain link from the supply chain data view; processing the link data based on a random forest regression model to determine the bottleneck prediction data corresponding to each supply chain link.

[0051] Among them, the supply chain link can be each flow link in the supply chain, which can include raw material procurement, production processing, logistics distribution, inventory management, etc. The bottleneck prediction data is a set of key information used to identify, evaluate and prevent potential bottleneck links in supply chain management.

[0052] Specifically, the supply chain links of the target supply chain are identified, and from the comprehensive supply chain data view, the link data corresponding to each supply chain link is extracted according to the business logic and data characteristics of each link, including but not limited to processing time, resource utilization rate, cost expenditure, quality pass rate, etc. A random forest regression model is constructed to predict bottlenecks for each supply chain link to determine bottleneck prediction data.

[0053] For example, the loss function of the random forest regression model is as follows: ; wherein, L represents the total loss function of the random forest regression model; L represents the mean square error; L represents the difference loss between trees; L represents the model complexity penalty term; α, β, γ represent weight coefficients respectively; ; wherein, n represents the number of samples; L represents the true value of the i-th sample; This represents the predicted value of the i-th sample, output by the random forest regression model; ;in, This represents the total number of trees in the random forest; This represents the set of predicted values ​​for the k-th tree; This represents the variance of the prediction results for the k-th tree; Where λ represents the regularization coefficient; This represents the depth of the k-th tree.

[0054] S140. Perform linear programming on the forecast data of the target supply chain to determine the optimization strategy corresponding to the target supply chain.

[0055] Among them, optimization strategies can be strategic recommendations for optimizing the target supply chain.

[0056] Specifically, the target supply chain forecast data obtained through machine learning models, such as demand forecasts, resource availability, and cost budgets, are used as input parameters for linear programming. A linear programming model is constructed, defining the decision variables (e.g., production volume, inventory levels, transportation volume), the objective function (e.g., minimizing total cost, maximizing service level), and constraints (e.g., resource limitations, time windows, capacity constraints). The model is then solved using a linear programming solver, such as CPLEX or Gurobi, to obtain the optimal values ​​of the decision variables for achieving the objective under given constraints. Based on the solution results, optimization strategies corresponding to the target supply chain are determined, such as adjusting production plans, optimizing inventory allocation, and improving logistics routes.

[0057] The technical solution of this invention collects real-time and historical supply chain data of the target supply chain through IoT devices; establishes a supply chain data view based on the real-time and historical data; processes the supply chain data view using a machine learning model to determine the predicted data for the target supply chain; and performs linear programming on the predicted data to determine the corresponding optimization strategy for the target supply chain. Based on the above technical solution, by applying a machine learning model to demand forecasting and supply chain bottleneck identification using the processed dataset, it is possible to more accurately capture changes in market demand and optimize inventory costs, transportation costs, and supply chain risks through linear programming algorithms, thereby achieving more efficient cost control and risk management.

[0058] In one possible implementation of the present invention Figure 2 A flowchart illustrating a supply chain cost optimization method based on big data, as provided in this embodiment of the invention, is shown below. Figure 2 As shown, the method for determining the optimization strategy corresponding to the target supply chain based on linear programming processing of the forecast data of the target supply chain in this embodiment further includes the following steps:

[0059] S210. Establish the linear programming objective function corresponding to the target supply chain, and determine the constraints corresponding to the target supply chain.

[0060] The constraints include inventory constraints and transportation capacity constraints. The objective function of the linear programming problem can be a function used for linear solution.

[0061] Specifically, a linear programming objective function corresponding to the target supply chain is established, and constraints corresponding to the target supply chain are determined. For example, when minimizing total cost, variables such as procurement cost, production cost, transportation cost, and inventory holding cost need to be comprehensively considered. By integrating the objective function and constraints, a complete linear programming model is formed.

[0062] Based on the above technical solutions, a linear programming objective function corresponding to the target supply chain is established, including: constructing a linear programming objective function with the goal of minimizing the total cost of the supply chain.

[0063] The total supply chain cost is the weighted sum of inventory costs, transportation costs, and supply chain risks. Inventory costs are calculated by summing the unit holding cost of each inventory category multiplied by the inventory quantity; transportation costs are calculated by summing the unit transportation cost of each transportation route multiplied by the transportation volume; and supply chain risks are calculated by summing the risk weight of each link multiplied by the bottleneck prediction value.

[0064] Specifically, the objective function of the linear programming algorithm is as follows: The optimization objective is to reduce inventory costs, transportation costs, and supply chain risks; inventory costs: In the formula: m represents the quantity of each inventory category; This represents the unit holding cost of the i-th type of inventory; Indicates the holding quantity of the i-th type of inventory; transportation cost: In the formula: n represents the number of transportation routes; This represents the unit transportation cost of the j-th transportation route; Indicates the transport volume of the j-th transport route; Supply chain risk: Where p represents the number of links in the supply chain; This represents the risk weight of the k-th supply chain link; This represents the predicted bottleneck value for the k-th link in the supply chain.

[0065] The linear programming algorithm and constraints are integrated into a single linear programming problem for optimization, as follows: ;in, The overall objective function is the weighted sum of inventory costs, transportation costs, and supply chain risks.

[0066] Constraints include: Inventory constraints: ;in, This represents the safety stock level of the i-th type of inventory; This represents the holding quantity of the i-th type of inventory; This represents the predicted demand value at time step t;

[0067] Transportation capacity constraints: ;in, This represents the maximum transport capacity of the j-th transport route; This represents the transport volume of the j-th transport route; ;in, This represents the predicted bottleneck value for the k-th link in the supply chain. This represents the bottleneck threshold at the k-th link in the supply chain. A linear programming solver (such as CPLEX, Gurobi, or an open-source linear programming library) is used to solve the above model to obtain the optimal inventory levels, transportation volumes, and risk control solutions.

[0068] S220. Solve the linear programming objective function based on constraints and predicted data to determine the optimization strategy.

[0069] The optimization strategies include optimal inventory levels, optimal transportation volumes, and risk control measures.

[0070] Specifically, the previously constructed linear programming model is imported into a professional solving tool, such as CPLEX or Gurobi. Input data includes predicted demand and resource availability at each stage, as well as constraints such as resource limitations, logistics capacity, and inventory capacity. During the solution process, the tool automatically searches the optimal solution space that satisfies all constraints based on the objective function (minimizing the total supply chain cost) and the constraints. After solving, key optimization strategies are extracted from the results: first, determining the optimal inventory level, i.e., the minimum holding level for each inventory category while meeting demand and service levels; second, identifying the optimal transportation volume, including the optimal transportation batch size and frequency for each transportation route; and third, generating risk control plans, developing targeted risk mitigation measures based on the supply chain risk assessment results, such as setting safety stock and diversifying supplier selection.

[0071] The technical solution of this invention collects data from warehouses and transportation in real time through IoT devices and integrates historical data, thus solving the problems of information silos and incomplete data. ETL tools are used to handle missing values, outliers, and duplicate data, and a unified data view is established through a data lake, ensuring data accuracy and consistency. Based on the processed dataset, machine learning models are applied for demand forecasting and supply chain bottleneck identification. Compared to traditional statistical models, this method can more accurately capture changes in market demand. Finally, linear programming algorithms are used to optimize inventory costs, transportation costs, and supply chain risks, achieving more efficient cost control and risk management compared to experience-based judgment. Overall, this invention significantly improves the accuracy and responsiveness of supply chain management, effectively addressing the shortcomings of traditional methods.

[0072] Figure 3 This is a schematic diagram of a supply chain cost optimization device based on big data, provided as an embodiment of the present invention. Figure 3 As shown, the device includes: a data acquisition module 310, a data processing module 320, a data prediction module 330, and an optimization strategy determination module 340.

[0073] Data acquisition module 310 is used to collect real-time and historical supply chain data of the target supply chain through IoT devices;

[0074] Data processing module 320 is used to build a supply chain data view based on real-time supply chain data and historical supply chain data;

[0075] The data prediction module 330 is used to process the supply chain data view based on the machine learning model to determine the predicted data for the target supply chain.

[0076] The optimization strategy determination module 340 is used to perform linear programming processing on the forecast data of the target supply chain to determine the optimization strategy corresponding to the target supply chain.

[0077] Based on the above technical solution, the data processing module is used to preprocess real-time and historical supply chain data using data processing tools; establish a supply chain data lake corresponding to the target supply chain based on the preprocessed real-time and historical supply chain data; and process the supply chain data in the supply chain data lake according to preset processing methods to obtain a supply chain data view. The preset processing methods include data modeling, defining data structures, formulating data standards, and creating a data dictionary.

[0078] Based on the above technical solution, the forecast data includes demand forecast data and a data forecast module, which is used to extract the demand data of the target supply chain from the supply chain data view based on a preset time step; the demand data is processed according to the Long Short-Term Memory Network model to determine the demand forecast data of the target supply chain; wherein, the Long Short-Term Memory Network model is set with at least two layers of LSTM units; the first LSTM unit is used to output the hidden state of all time steps; the second LSTM unit is used to output the hidden state of the last time step; and an anti-overfitting layer is added after the second LSTM unit.

[0079] Based on the above technical solution, the prediction data includes bottleneck prediction data and a data prediction module, which is used to determine the supply chain links corresponding to the target supply chain and extract the link data corresponding to each supply chain link from the supply chain data view; the link data is processed based on the random forest regression model to determine the bottleneck prediction data corresponding to each supply chain link.

[0080] Based on the above technical solution, the optimization strategy determination module is used to establish a linear programming objective function corresponding to the target supply chain and determine the constraints corresponding to the target supply chain; based on the constraints and forecast data, the linear programming objective function is solved to determine the optimization strategy, which includes the optimal inventory level, the optimal transportation volume, and the risk control scheme.

[0081] Based on the above technical solution, an optimization strategy determination module is used to construct a linear programming objective function with the goal of minimizing the total supply chain cost; where the total supply chain cost is the weighted sum of inventory cost, transportation cost, and supply chain risk.

[0082] Based on the above technical solutions, the constraints include inventory constraints and transportation capacity constraints.

[0083] The technical solution of this invention collects real-time and historical supply chain data of the target supply chain through IoT devices; establishes a supply chain data view based on the real-time and historical data; processes the supply chain data view using a machine learning model to determine the predicted data for the target supply chain; and performs linear programming on the predicted data to determine the corresponding optimization strategy for the target supply chain. Based on the above technical solution, by applying a machine learning model to demand forecasting and supply chain bottleneck identification using the processed dataset, it is possible to more accurately capture changes in market demand and optimize inventory costs, transportation costs, and supply chain risks through linear programming algorithms, thereby achieving more efficient cost control and risk management.

[0084] The supply chain cost optimization device based on big data provided in the embodiments of the present invention can execute the supply chain cost optimization method based on big data provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method.

[0085] Figure 4 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0086] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0087] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0088] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as big data-based supply chain cost optimization methods.

[0089] In some embodiments, the big data-based supply chain cost optimization method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the big data-based supply chain cost optimization method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the big data-based supply chain cost optimization method by any other suitable means (e.g., by means of firmware).

[0090] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0091] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0092] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0093] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0094] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0095] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0096] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0097] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A supply chain cost optimization method based on big data, characterized in that, include: Collect real-time and historical supply chain data of the target supply chain through IoT devices; A supply chain data view is established based on the real-time supply chain data and the historical supply chain data. The supply chain data view is processed using a machine learning model to determine the predicted data for the target supply chain; Linear programming is used to process the forecast data of the target supply chain to determine the optimization strategy corresponding to the target supply chain.

2. The method according to claim 1, characterized in that, The process of establishing a supply chain data view based on the real-time supply chain data and the historical supply chain data includes: The real-time and historical data of the supply chain are preprocessed using data processing tools. A supply chain data lake corresponding to the target supply chain is established based on preprocessed real-time supply chain data and historical supply chain data. The supply chain data in the supply chain data lake is processed according to a preset processing method to obtain the supply chain data view. The preset processing method includes data modeling, defining data structure, formulating data standards, and creating a data dictionary.

3. The method according to claim 1, characterized in that, The forecast data includes demand forecast data; The step of processing the supply chain data view based on a machine learning model to determine the predicted data for the target supply chain includes: Demand data for the target supply chain is extracted from the supply chain data view based on a preset time step. The demand data is processed using a long short-term memory network model to determine the demand forecast data for the target supply chain. The Long Short-Term Memory (LSTM) network model has at least two LSTM units; the first LSTM unit is used to output the hidden state of all time steps; the second LSTM unit is used to output the hidden state of the last time step; and an anti-overfitting layer is added after the second LSTM unit.

4. The method according to claim 1, characterized in that, The predicted data includes bottleneck predicted data; the step of processing the supply chain data view based on a machine learning model to determine the predicted data for the target supply chain includes: Identify the supply chain links corresponding to the target supply chain, and extract the link data corresponding to each supply chain link from the supply chain data view; The data from each stage of the supply chain is processed using a random forest regression model to determine the bottleneck prediction data corresponding to each stage.

5. The method according to claim 1, characterized in that, The step of performing linear programming processing on the forecast data of the target supply chain to determine the optimization strategy corresponding to the target supply chain includes: Establish a linear programming objective function corresponding to the target supply chain, and determine the constraints corresponding to the target supply chain; The optimization strategy is determined by solving the linear programming objective function based on the constraints and the predicted data, wherein the optimization strategy includes the optimal inventory level, the optimal transportation volume, and the risk control scheme.

6. The method according to claim 5, characterized in that, The establishment of the linear programming objective function corresponding to the target supply chain includes: The objective function of the linear programming is constructed with the goal of minimizing the total cost of the supply chain; wherein the total cost of the supply chain is the weighted sum of inventory cost, transportation cost and supply chain risk.

7. The method according to claim 5, wherein the constraints include inventory constraints and transportation capacity constraints.

8. A supply chain cost optimization device based on big data, characterized in that, include: The data acquisition module is used to collect real-time and historical supply chain data of the target supply chain through IoT devices; The data processing module is used to establish a supply chain data view based on the real-time supply chain data and the historical supply chain data; The data prediction module is used to process the supply chain data view based on the machine learning model to determine the predicted data for the target supply chain; The optimization strategy determination module is used to perform linear programming processing on the forecast data of the target supply chain to determine the optimization strategy corresponding to the target supply chain.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the big data-based supply chain cost optimization method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the big data-based supply chain cost optimization method according to any one of claims 1-7.

Images