Supply chain auxiliary decision-making method and system based on logistics big data

Abstract

The invention discloses a supply chain auxiliary decision-making method and system based on logistics big data, and relates to the technical field of supply chain decision-making management, and the method comprises the steps: obtaining user behavior data and user position data; obtaining a predicted demand quantity according to the user behavior data; according to the user position data, obtaining predicted distribution time; inputting the predicted demand quantity and the predicted distribution time into a distribution decision model to obtain a decision scheme; and managing the distribution links of the supply chain according to the decision scheme. According to the method, the demand quantity and the distribution time of the supply chain are predicted by obtaining the user behavior data and the position data, decision making is carried out on task distribution by considering the historical behavior data and the position information of the user, and the logistics service quality is improved while the cost of the logistics supply chain is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of supply chain decision management, and in particular to a supply chain decision-making assistance method and system based on logistics big data. Background Art

[0002] The current changes in consumer demand and the integration of technology in various scenarios require the improvement of the quality of the supply chain based on cost, efficiency and experience. Logistics services have always been an important link in the development of the supply chain. The basic structure of the logistics service supply chain is logistics service provider-logistics service integrator-user. Logistics service integrators establish logistics networks by integrating functional logistics companies, distribute user needs to multiple logistics service providers, and provide users with logistics service solutions by integrating, processing and managing basic logistics services.

[0003] As logistics gradually becomes the focus of competition, the decision-making scenarios of logistics service supply chain are becoming more and more diverse. Logistics big data improves the decision-making and execution capabilities of logistics systems through intelligent technology, and its role is reflected in all aspects of logistics services, such as optimizing warehousing logistics through the mining of massive data, improving transportation management, predicting sales, and providing decision support.

[0004] When a logistics service integrator receives an order from a user, it integrates and unifies the user's needs and then assigns the task to the logistics service provider, who then completes the logistics service. This process is called task allocation. In the logistics service supply chain, the user's needs and location, as well as the capabilities and resources of the logistics service provider, are important decision-making factors. The existing task allocation only considers the demand and provides services based on the resources and capabilities of the logistics service provider. It does not combine the user's historical data and consider whether the user has obtained satisfactory logistics services from the user's perspective, which leads to user dissatisfaction with the logistics service and low quality of logistics services. Summary of the invention

[0005] The purpose of the present invention is to provide a supply chain decision-making support method and system based on logistics big data. The present invention obtains user behavior data and location data, predicts the demand and delivery time of the supply chain, and makes decisions on task allocation based on the user's historical behavior data and location information, thereby reducing the cost of the logistics supply chain and improving the quality of logistics services.

[0006] The purpose of the present invention is achieved by the following technical means: In a first aspect, the present invention provides a supply chain decision-making assistance method based on logistics big data, comprising: Obtain user behavior data and user location data; Obtaining predicted demand based on the user behavior data; Obtaining a predicted delivery time based on the user location data; Inputting the predicted demand and the predicted delivery time into a distribution decision model to obtain a decision plan; According to the decision plan, the distribution link of the supply chain is managed.

[0007] Preferably, obtaining the predicted demand according to the user behavior data includes: Performing data cleaning on the user behavior data; The number of clicks, favorites, browsing time, shopping cart additions, purchases, and post-purchase ratings of users are extracted in different time ranges, and the maximum, minimum, average, and standard deviation are counted as features to obtain a feature data set. Normalizing the feature data set and calculating the correlation between features; Selecting a feature subset from the feature data set according to the correlation between the features; The feature subset is input into the demand forecasting neural network model to obtain the predicted demand.

[0008] Preferably, selecting a feature subset from the feature data set according to the correlation between the features comprises: Set reference thresholds; Extracting feature pairs whose correlation between the features is less than the reference threshold from the feature data set to obtain a feature pair data set, and calculating the correlation between the feature pairs and the target feature; Feature pairs whose correlation with the target feature is less than the reference threshold are removed from the feature pair data set to obtain the feature subset.

[0009] Preferably, before acquiring the user behavior data and the user location data, the method further includes: training the demand prediction neural network model, specifically including: Initialize the population size, number of iterations, learning factor, particle position and particle velocity of the particle swarm algorithm; Randomly generate population particles, and construct demand prediction neural network model parameters according to the population particles; Training the demand forecasting neural network model to obtain output values, and calculating population fitness according to the output values; Update the population particles according to the population fitness and modify the demand forecasting neural network model parameters; Repeat the steps until the number of iterations reaches a maximum, and obtain the demand forecasting neural network model.

[0010] Preferably, obtaining the predicted delivery time according to the user location data includes: Segmenting the user location data to obtain segmented location data; Performing clustering processing on the segmented position data to obtain clustered position data; Calculating the correlation degree of adjacent cluster position data; The predicted delivery time is obtained according to the correlation degree.

[0011] Preferably, the step of inputting the predicted demand and the predicted delivery time into a distribution decision model to obtain a decision solution includes: Taking the minimum cost of logistics supply chain as the goal, construct the minimum objective function; The minimum objective function is expressed as follows: in, is the number of suppliers, To predict the demand, For the Is the supplier Selected by a user, For the Suppliers to Quotes from users, For the Suppliers to The time penalty cost of each user, For the The delivery volume of each supplier, For the The prices of goods from suppliers; Determining constraint conditions according to the predicted demand and the predicted delivery time; The constraints are expressed as follows: in, For the Suppliers for The satisfaction level of each user, For the The acceptable price for each user, For the Suppliers to The service attitude of each user, For the The expected service attitude of each user, For the The weight coefficient of the price for each user, For the The weight coefficient of a user’s attitude towards the service, The actual delivery time is For the The predicted delivery time for each user, is the time penalty coefficient, For the The profit of a supplier, For the Suppliers to The service cost per user, For the The minimum profit of a supplier, is the service cost coefficient, The highest rating for the service attitude by users; According to the constraint conditions, a genetic algorithm is used to solve the minimum objective function to obtain the decision-making solution.

[0012] Preferably, the step of solving the minimum objective function by using a genetic algorithm according to the constraint conditions to obtain the decision-making solution comprises: Set the number of individuals, crossover probability, mutation probability and termination conditions; Calculating individual fitness according to the minimum objective function; An initial population is randomly generated, and intermediate individuals are formed according to the crossover probability. The intermediate individuals are mutated according to the mutation probability to obtain mutant individuals. Selecting the mutated individuals according to the individual fitness and the number of individuals to obtain a new population; Repeat the steps until the termination condition is reached, and output the mutated individual as the optimal solution to obtain the decision plan.

[0013] In a second aspect, the present invention provides a supply chain decision support system based on logistics big data, comprising: Data acquisition module: used to obtain user behavior data and user location data; Demand forecasting model: used to obtain predicted demand based on the user behavior data; Delivery time prediction module: used to obtain the predicted delivery time according to the user location data; Allocation decision module: used for inputting the predicted demand and the predicted delivery time into the allocation decision model to obtain a decision plan; Distribution management module: used to manage the distribution link of the supply chain according to the decision plan In a third aspect, the present invention provides an electronic device comprising a processor and a memory, wherein the memory is used to store computer program code, and the computer program code comprises computer instructions. When the processor executes the computer instructions, the electronic device executes the above-mentioned supply chain decision support method based on logistics big data.

[0014] In a fourth aspect, the present invention provides a computer-readable storage medium, in which a computer program is stored. The computer program includes program instructions, and when the program instructions are executed by a processor of an electronic device, the processor executes the above-mentioned supply chain decision-making support method based on logistics big data.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention obtains user behavior data and location data, predicts supply chain demand and delivery time, and makes decisions on task allocation based on user historical behavior data and location information, thereby reducing the cost of the logistics supply chain and improving the quality of logistics services.

[0016] The present invention obtains a feature set by performing data cleaning, feature construction and feature extraction on user behavior data, calculates the correlation between features to reduce the dimension of the feature set to obtain a feature subset, and inputs the feature subset into a demand forecasting neural network model to predict the delivery time, thereby reducing the prediction time of the demand forecasting neural network model.

[0017] The present invention optimizes the parameters of the demand forecasting neural network model by using an improved particle swarm optimization algorithm to obtain the optimal demand forecasting neural network model, thereby improving the prediction accuracy of the demand forecasting neural network model.

[0018] The present invention performs segmented clustering processing on user location data and then performs correlation calculation to mine the data, thereby obtaining user locations and predicted delivery times, and providing a data basis for task allocation decisions.

[0019] The present invention provides a method that takes into account the user's location, optimal delivery time and satisfaction with the logistics service provider, constructs a minimum objective function and constraints with the goal of minimizing the logistics supply chain cost, makes decisions on the distribution links of the supply chain based on the solution of the minimum objective function, ensures the user's logistics service experience, improves the quality of logistics services, and reduces the cost of the logistics supply chain.

[0020] The present invention solves the minimum objective function through a genetic algorithm, and obtains individuals of the final variant population as the optimal solution of the minimum objective function, thereby improving the accuracy of decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0023] Figure 1 A schematic diagram of a flow chart of a supply chain decision-making assistance method based on logistics big data provided in this embodiment; Figure 2 A schematic diagram of the process of step S1 provided in this embodiment; Figure 3 A schematic diagram of the process of step S14 provided in this embodiment; Figure 4 A schematic diagram of the process of step S0 provided in this embodiment; Figure 5 A schematic diagram of the process of step S3 provided in this embodiment; Figure 6 A schematic diagram of the process of step S4 provided in this embodiment; Figure 7 A schematic diagram of the process of step S43 provided in this embodiment; Figure 8 A schematic diagram of the structure of a supply chain decision support system based on logistics big data provided in this embodiment; Fig. 9 A schematic diagram of the structure of an electronic device provided in this embodiment. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative position relationship, movement status, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication will also change accordingly.

[0026] In addition, the descriptions of "first", "second", etc. in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0027] The present invention obtains user behavior data and location data, predicts supply chain demand and delivery time, and makes decisions on task allocation based on user historical behavior data and location information, thereby reducing the cost of the logistics supply chain and improving the quality of logistics services.

[0028] The present invention provides a supply chain decision-making assistance method based on logistics big data, such as Figure 1 As shown, the following steps are included: S1, obtain user behavior data and user location data; S2, based on user behavior data, obtains predicted demand; S3, obtain the predicted delivery time based on the user location data; S4, inputting the predicted demand and predicted delivery time into the allocation decision model to obtain a decision plan; S5, manages the distribution link of the supply chain based on the decision-making plan.

[0029] It should be noted that, in the context of big data, demand forecasting is crucial for the supply chain. Based on the real-time characteristics of big data, user behavior data can be used to predict demand. Logistics service providers can produce goods based on demand to ensure the profit of the entire supply chain. User behavior data comes from the big data platform of the supply chain, including browsing goods, collecting goods, adding goods to shopping carts, submitting orders and evaluating orders. These operations will be recorded by the big data platform to form user behavior data. These operations are related to demand, and demand can be predicted. For example, browsing behavior and collection behavior mean that users are more concerned about the goods and will perform further operations on the goods. Adding goods to shopping carts means that users have a strong desire to buy. Payment behavior is directly linked to sales, and evaluation behavior reflects users' repurchase and recommendation of goods. User location data includes the location information and time when the user performs the operation. The user's best delivery time can be predicted based on the user's location data.

[0030] In this embodiment, by obtaining user behavior data and location data, predicting supply chain demand and delivery time, and making decisions on task allocation based on the user's historical behavior data and location information, the cost of the logistics supply chain is reduced while improving the quality of logistics services.

[0031] In some embodiments, step S1, based on user behavior data, obtain predicted demand, such as Figure 2 As shown, the following steps are included: S11, cleaning the user behavior data; S12, extracting the number of clicks, number of favorites, browsing time, number of shopping cart additions, number of purchases, and post-purchase ratings of users in different time ranges, and counting the maximum value, minimum value, average value, and standard deviation as features to obtain a feature data set; S13, normalizing the feature data set and calculating the correlation between the features; S14, selecting a feature subset from the feature data set according to the correlation between the features; S15, inputting the feature subset into the demand prediction neural network model to obtain the predicted demand.

[0032] It should be noted that there are missing values ​​and unreasonable data in the original supply chain data. Data cleaning is to supplement the missing values, remove the unreasonable data or perform other operations. Missing data may occur when the machine fails during the recording process, and it can be supplemented by filling in zeros or taking the average of adjacent data. The unreasonable data situation is that the user behavior data obtained is very small, but the sales volume of the product is very good, or the user behavior data obtained is a lot, but the sales volume of the product is relatively low. For the former case, there may be fake orders, in which case the data needs to be removed. For the latter case, it may be that the price of the product is relatively high, resulting in low sales volume of the product, or there may be data crawlers, in which case the data needs to be selectively removed manually.

[0033] After cleaning the user behavior data, feature extraction is performed on the data, and new features are constructed based on the original features to provide more user information. User behavior data is extracted in different time ranges, and user behavior data of certain large events can be eliminated or predicted separately to ensure the stability of the data.

[0034] Before inputting the feature set into the demand forecasting neural network model for prediction, it is also necessary to remove redundant features, normalize the data, calculate the correlation between different features, and perform dimensionality reduction on the feature set, thereby reducing the prediction time of the demand forecasting neural network model.

[0035] The normalized conversion function is expressed as follows: in, is the normalized data, is the original data, is the data average, is the standard deviation of the data.

[0036] The correlation between features is expressed as follows: in, For the The first feature, is the average value of the first feature, For the The second feature, is the average value of the second feature.

[0037] In some embodiments, step S14 selects a feature subset from the feature data set based on the correlation between the features, such as Figure 3 As shown, the following steps are included: S141, setting a reference threshold; S142, extracting feature pairs whose correlation between features is less than a reference threshold from the feature data set to obtain a feature pair data set, and calculating the correlation between the feature pairs and the target feature; S143, removing feature pairs whose correlation with the target feature is less than a reference threshold from the feature pair data set to obtain a feature subset.

[0038] It should be noted that the selection of feature subsets uses the correlation between features as a selection index, sets a reference threshold as a threshold in the feature subset selection process, extracts feature pairs whose correlation between features is less than the reference threshold, and on the basis of the selected feature pairs, compares the correlation between the feature pairs and the target features, and eliminates the feature pairs with small correlation to obtain feature subsets. In some embodiments, the step size can be set, and after obtaining the feature subset, the value of the reference threshold can be reduced according to the step size, and a new feature subset can be obtained according to the new reference threshold. When the reference threshold is reduced to zero, each feature subset is evaluated to select the optimal feature subset.

[0039] In this embodiment, by calculating the correlation between different features, the feature set is reduced in dimension according to the correlation to obtain a feature subset, and the feature subset is input into the demand prediction neural network model for demand prediction, thereby reducing the prediction time of the demand prediction neural network model. The obtained optimal feature subset also improves the prediction accuracy of the demand prediction neural network model.

[0040] In some embodiments, before step S1, obtaining user behavior data and user location data, the following steps are also included: S0, training a demand prediction neural network model, such as Figure 4 As shown, the specific steps include: S01, initialize the population size, number of iterations, learning factor, particle position and particle velocity of the particle swarm algorithm; S02, randomly generate population particles, and build demand prediction neural network model parameters according to the population particles; S03, training the demand forecasting neural network model to obtain output values, and calculating the population fitness according to the output values; S04, updating the population particles according to the population fitness and correcting the demand forecasting neural network model parameters; S05, repeat the steps until the number of iterations reaches the maximum, and obtain the demand forecasting neural network model.

[0041] It should be noted that the specific form of population particles is as follows: in, For the A population of particles, is the number of iterations, is the number of neurons in the hidden layer of the demand forecasting neural network model, is the learning rate.

[0042] The demand forecasting neural network model parameters and modifications are expressed as follows: in, is the weight matrix of the input layer and the hidden layer, For the The particle position at the iteration number, For the The hidden layer output of the iteration number, is the output of the demand forecasting neural network model, is the weight matrix, is the learning factor.

[0043] The population fitness is expressed as follows: in, For the The fitness of a population particle, is the predicted value of the training set, is the output value of the training set, is the predicted value of the validation set, is the output value of the validation set.

[0044] The training set and the validation set can be obtained by dividing the historical user behavior data in proportion. The updating of the population particles is the same as that in the traditional particle swarm optimization algorithm, which compares the population fitness with the optimal individual fitness and the optimal global fitness, and the method of updating the particle position is the same, which will not be repeated here.

[0045] In this embodiment, the parameters of the demand forecasting neural network model are optimized by a particle swarm optimization algorithm to obtain an optimal demand forecasting neural network model, thereby improving the prediction accuracy of the demand forecasting neural network model.

[0046] In some embodiments, step S3 obtains a predicted delivery time based on the user location data, such as Figure 5 As shown, the following steps are included: S31, segmenting the user location data to obtain segmented location data; S32, clustering the segmented position data to obtain clustered position data; S33, calculating the correlation degree of adjacent cluster position data; S34, obtaining a predicted delivery time based on the correlation degree.

[0047] It should be noted that the data format of the user location data is as follows: in, is the horizontal coordinate information of the user's location data, is the vertical coordinate information of the user's location data, The time information of the user's location data.

[0048] Segment processing merges the data with the same position information in the position data. The segment processing is expressed as follows: in, For the to The user location data obtained by merging the user location data of For the and The horizontal coordinate information of the user's location data, For the and The vertical coordinate information of the user's location data, For the Time information of user location data, For the The time information of the user's location data.

[0049] Clustering processing sets a reference distance, clusters the location data in the segmented location data whose distance is not greater than the reference distance, and modifies the location data to be consistent with the location data of the cluster center. The clustering processing is expressed as follows: in, For the The horizontal coordinate of the user's location data, is the horizontal coordinate of the location data of the cluster center, For the The vertical coordinate of the user's location data, is the ordinate of the location data of the cluster center, is the reference distance.

[0050] The calculation of association degree can be achieved by calculating confidence and support. By calculating the confidence and support of all adjacent cluster location data, according to the size of confidence and support, the time information of cluster location data with large confidence and support is selected as the predicted delivery time.

[0051] In this embodiment, by performing segmented clustering processing on the user location data and then performing correlation calculation to mine the data, the predicted delivery time and user location are obtained, providing a data basis for task allocation decisions.

[0052] In some embodiments, in step S4, the predicted demand and predicted delivery time are input into the allocation decision model to obtain a decision solution, such as Figure 6 As shown, the following steps are included: S41, taking the minimum cost of logistics supply chain as the goal, construct the minimum objective function; The minimum objective function is expressed as follows: in, is the number of suppliers, To predict the demand, For the Is the supplier Selected by a user, For the Suppliers to Quotes from users, For the Suppliers to The time penalty cost of each user, For the The delivery volume of each supplier, For the The prices of goods from suppliers; S42, determining constraint conditions according to the predicted demand and predicted delivery time; The constraints are expressed as follows: in, For the Suppliers for The satisfaction level of each user, For the The acceptable price for each user, For the Suppliers to The service attitude of each user, For the The expected service attitude of each user, For the The weight coefficient of the price for each user, For the The weight coefficient of a user’s attitude towards the service, The actual delivery time is For the The predicted delivery time for each user, is the time penalty coefficient, For the The profit of a supplier, For the Suppliers to The service cost per user, For the The minimum profit of a supplier, is the service cost coefficient, The highest rating for the service attitude by users; S43, according to the constraints, a genetic algorithm is used to solve the minimum objective function to obtain a decision plan.

[0053] It should be noted that the minimum objective function and constraint function aim to minimize the cost of optimizing the logistics supply chain, while taking into account the user's location, delivery time, and satisfaction with the logistics service provider. The minimum objective function includes the service cost of the logistics service provider and the penalty cost for not delivering within the predicted delivery time. The service quotation of the logistics service provider can be adjusted according to the user's location data. In the constraint conditions, the user's satisfaction with the logistics service provider includes the user's quotation for the logistics service provider and the user's historical rating of the logistics service provider, that is, the logistics service provider's historical service attitude towards the user. The time penalty cost means that there is no penalty if the delivery is made before the delivery time predicted by big data, but a certain time cost is required for delivery after the predicted delivery time. At the same time, the minimum profit constraint of the logistics service provider needs to be met, and the total delivery volume needs to meet the predicted demand and the user's acceptable price is greater than the quotation of the logistics service provider. In this embodiment, the The weight coefficient of the price for each user , No. The weight coefficient of each user's attitude towards service and the service cost coefficient The value range is , No. Expected service attitude of each user The value range is .

[0054] In this embodiment, by considering the user's location, optimal delivery time and satisfaction with the logistics service provider, the minimum objective function and constraints are constructed with the goal of minimizing the logistics supply chain cost. Decisions are made on the distribution links of the supply chain based on the solution of the minimum objective function to ensure the user's logistics service experience, improve the quality of logistics services, and reduce the cost of the logistics supply chain.

[0055] In some embodiments, in step S43, according to the constraints, a genetic algorithm is used to solve the minimum objective function to obtain a decision solution, such as Figure 7 As shown, the following steps are included: S431, set the number of individuals, crossover probability, mutation probability and termination condition; S432, calculating the individual fitness according to the minimum objective function; S433, randomly generating an initial population, forming intermediate individuals according to the crossover probability, and performing mutation on the intermediate individuals according to the mutation probability to obtain mutant individuals; S433, select the mutant individuals according to their fitness and number to obtain a new population; S434, repeat the steps until the termination condition is reached, and output the mutated individual as the optimal solution to obtain a decision plan.

[0056] It should be noted that the genetic algorithm is used to solve the minimum objective function. When calculating the individual fitness, the inverse of the minimum objective function is selected as the individual fitness. After the initial population is randomly generated, several pairs of mothers are selected from the initial population to form intermediate individuals according to the probability of crossover. The number of mothers is greater than the number of individuals. After the mutant individuals are obtained, individuals with the same number of individuals are selected from the mutant individuals to form a new population.

[0057] In this embodiment, the minimum objective function is solved by a genetic algorithm, and the individuals of the final mutant population are used as the optimal solution of the minimum objective function, thereby improving the accuracy of decision-making.

[0058] A supply chain decision support system based on logistics big data, such as Figure 8 As shown, including: Data acquisition module: used to obtain user behavior data and user location data; Demand forecasting module: used to obtain forecasted demand based on user behavior data; Delivery time prediction module: used to obtain the predicted delivery time based on user location data; Allocation decision module: used to input the predicted demand and predicted delivery time into the allocation decision model to obtain a decision plan; Distribution management module: used to manage the distribution link of the supply chain according to the decision-making plan.

[0059] In this embodiment, by obtaining user behavior data and location data, predicting supply chain demand and delivery time, and making decisions on task allocation based on the user's historical behavior data and location information, the cost of the logistics supply chain is reduced while improving the quality of logistics services.

[0060] It should be understood that the disclosed system can be implemented in other ways. For example, the system embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, each functional module can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software functional modules.

[0061] An electronic device 2, such as Fig. 9As shown, a processor 21 and a memory 22, the memory 22 is used to store computer program codes, the computer program codes include computer instructions, when the processor 21 executes the computer instructions, the electronic device executes the above-mentioned supply chain decision-making assistance method based on logistics big data.

[0062] The electronic device 2 includes a processor 21, a memory 22, an output device 23, and an input device 24. The processor 21, the memory 22, the output device 23, and the input device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines, or buses, etc., which are not limited in the embodiments of the present invention. It should be understood that in various embodiments of the present invention, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, through various interfaces, transmission lines, buses, etc.

[0063] The processor 21 may be one or more graphics processing units (GPUs). When the processor 21 is a GPU, the GPU may be a single-core GPU or a multi-core GPU. Optionally, the processor 21 may be a processor group consisting of multiple GPUs, and the multiple processors are coupled to each other via one or more buses. Optionally, the processor 21 may also be other types of processors, etc., which are not limited in the embodiments of the present invention.

[0064] The memory 22 can be used to store computer program instructions and various computer program codes including program codes for executing the scheme of the present invention. Optionally, the memory 22 includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), and the memory 22 is used for related instructions and data.

[0065] The input device 24 is used to input data and / or signals, and the output device 23 is used to output data and / or signals. The output device 23 and the input device 24 can be independent devices or an integrated device.

[0066] A computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes the above-mentioned supply chain decision-making assistance method based on logistics big data.

[0067] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A supply chain decision-making assistance method based on logistics big data, characterized in that: include: Obtain user behavior data and user location data; Obtaining predicted demand based on the user behavior data; Obtaining a predicted delivery time based on the user location data; Inputting the predicted demand and the predicted delivery time into a distribution decision model to obtain a decision plan; According to the decision plan, the distribution link of the supply chain is managed.

2. The supply chain decision-making assistance method based on logistics big data according to claim 1 is characterized in that: The step of obtaining a predicted demand according to the user behavior data includes: Performing data cleaning on the user behavior data; The number of clicks, favorites, browsing time, shopping cart additions, purchases, and post-purchase ratings of users are extracted in different time ranges, and the maximum, minimum, average, and standard deviation are counted as features to obtain a feature data set. Normalizing the feature data set and calculating the correlation between features; Selecting a feature subset from the feature data set according to the correlation between the features; The feature subset is input into the demand forecasting neural network model to obtain the predicted demand.

3. The supply chain decision-making assistance method based on logistics big data according to claim 2 is characterized in that: The selecting a feature subset from the feature data set according to the correlation between the features comprises: Set reference thresholds; Extracting feature pairs whose correlation between the features is less than the reference threshold from the feature data set to obtain a feature pair data set, and calculating the correlation between the feature pairs and the target feature; Feature pairs whose correlation with the target feature is less than the reference threshold are removed from the feature pair data set to obtain the feature subset.

4. The supply chain decision-making assistance method based on logistics big data according to claim 2 is characterized in that: Before acquiring the user behavior data and the user location data, the method further includes: training the demand prediction neural network model, specifically including: Initialize the population size, number of iterations, learning factor, particle position and particle velocity of the particle swarm algorithm; Randomly generate population particles, and construct demand prediction neural network model parameters according to the population particles; Training the demand forecasting neural network model to obtain output values, and calculating population fitness according to the output values; Update the population particles according to the population fitness and modify the demand forecasting neural network model parameters; Repeat the steps until the number of iterations reaches a maximum, and obtain the demand forecasting neural network model.

5. The supply chain decision-making assistance method based on logistics big data according to claim 1 is characterized in that: The step of obtaining a predicted delivery time according to the user location data includes: Segmenting the user location data to obtain segmented location data; Performing clustering processing on the segmented position data to obtain clustered position data; Calculating the correlation degree of adjacent cluster position data; The predicted delivery time is obtained according to the correlation degree.

6. The supply chain decision-making assistance method based on logistics big data according to claim 1 is characterized in that: The step of inputting the predicted demand and the predicted delivery time into a distribution decision model to obtain a decision solution includes: Taking the minimum cost of logistics supply chain as the goal, construct the minimum objective function; The minimum objective function is expressed as follows: in, is the number of suppliers, To predict the demand, For the Is the supplier Selected by a user, For the Suppliers to Quotes from users, For the Suppliers to The time penalty cost of each user, For the The delivery volume of each supplier, For the The prices of goods from suppliers; Determining constraint conditions according to the predicted demand and the predicted delivery time; The constraints are expressed as follows: in, For the Suppliers for The satisfaction level of each user, For the The acceptable price for each user, For the Suppliers to The service attitude of each user, For the The expected service attitude of each user, For the The weight coefficient of the price for each user, For the The weight coefficient of a user’s attitude towards the service, The actual delivery time is For the The predicted delivery time for each user, is the time penalty coefficient, For the The profit of a supplier, For the Suppliers to The service cost per user, For the The minimum profit of a supplier, is the service cost coefficient, The highest rating for the service attitude by users; According to the constraint conditions, a genetic algorithm is used to solve the minimum objective function to obtain the decision-making solution.

7. The supply chain decision-making assistance method based on logistics big data according to claim 6 is characterized in that: The step of solving the minimum objective function according to the constraint conditions using a genetic algorithm to obtain the decision-making solution includes: Set the number of individuals, crossover probability, mutation probability and termination conditions; Calculating individual fitness according to the minimum objective function; An initial population is randomly generated, and intermediate individuals are formed according to the crossover probability. The intermediate individuals are mutated according to the mutation probability to obtain mutant individuals. Selecting the mutated individuals according to the individual fitness and the number of individuals to obtain a new population; Repeat the steps until the termination condition is reached, and output the mutated individual as the optimal solution to obtain the decision plan.

8. A supply chain decision support system based on logistics big data, characterized in that: include: Data acquisition module: used to obtain user behavior data and user location data; Demand forecasting model: used to obtain predicted demand based on the user behavior data; Delivery time prediction module: used to obtain the predicted delivery time according to the user location data; Allocation decision module: used for inputting the predicted demand and the predicted delivery time into the allocation decision model to obtain a decision plan; Distribution management module: used to manage the distribution link of the supply chain according to the decision plan.

9. An electronic device, characterized in that: It includes a processor and a memory, the memory is used to store computer program code, the computer program code includes computer instructions, and when the processor executes the computer instructions, the electronic device executes a supply chain decision-making support method based on logistics big data as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor of an electronic device, the processor executes a supply chain decision-making support method based on logistics big data as described in any one of claims 1 to 7.

Images