Reverse express state identification and tracking management method and system based on big data

Abstract

The invention provides a reverse express state recognition and tracking management method and system based on big data, and relates to the technical field of reverse express tracking management. The tracking management method comprises the following steps: S1, obtaining commodity information, position information and a commodity state of reverse express; s2, obtaining the return process and cost of reverse express return; s3, carrying out non-return risk assessment and comprehensive value assessment on the damaged commodity, and making a processing decision; s4, acquiring return state data of each return link in the return process; s5, reverse express loss risk prediction is carried out, and the reverse express loss risk probability is obtained; and S6, setting a corresponding tracking management scheme according to the loss risk probability. According to the invention, through carrying out non-return risk assessment and comprehensive value assessment on reverse express, a more reasonable damaged commodity disposal decision is made, and the overall loss is reduced to the greatest extent. The loss risk is effectively reduced by predicting the loss risk probability of the reverse express and adopting differentiated tracking management measures.

Description

Technical Field

[0001] The present invention relates to the technical field of reverse express status identification and tracking management, and in particular to a reverse express status identification and tracking management method based on big data, and a reverse express status identification and tracking management system based on big data. Background Art

[0002] As an important part of the modern service industry, the express delivery industry plays an increasingly important role in economic and social development. The popularization of the Internet and the rapid development of information technology have enabled logistics information to be transmitted in real time and accurately. Customers can easily place orders and check the logistics status of express delivery through the Internet, and enterprises can also achieve efficient tracking and management of goods. The continuous development and optimization of transportation modes such as roads, railways, and aviation have built a huge and efficient transportation network. This provides a solid foundation for the rapid transportation of express delivery and greatly shortens the transportation time of goods. The application of automated warehousing systems and intelligent sorting equipment has improved the efficiency of cargo storage and processing. Through the use of technologies such as barcodes and RFID, goods can be quickly identified and classified, reducing errors and time costs of manual operations. With the vigorous development of e-commerce, consumers' demand for the distribution of goods has increased dramatically. The convenience of online shopping has prompted the rapid expansion of the express delivery industry to meet consumers' expectations for fast and accurate delivery. The government has introduced a series of policies to support the development of the logistics industry, standardize the market order, and promote the healthy development of express delivery companies. At the same time, the continuous improvement of relevant laws and regulations also protects the legitimate rights and interests of consumers and enterprises.

[0003] Returns are an inevitable part of express delivery business. When consumers shop online, they may be dissatisfied with the product after purchase, or the size or color may not match the product because they cannot directly touch and feel the product. Some products may be damaged during production, transportation or storage, causing consumers to request a return after receiving them. In order to increase consumers' purchasing confidence, some merchants offer unconditional returns, which to a certain extent increases the number of returns. For example, incorrect address filling, inaccurate recipient information, express delivery delays and other reasons may result in the recipient not being able to receive the goods in time, and thus choosing to return the goods.

[0004] As the express delivery industry continues to develop, the tracking and management of express returns has become increasingly important. In reverse logistics management, many companies mainly rely on manual intervention and lack a systematic data collection and analysis mechanism. It is impossible to fully grasp the specific dynamic information of returned items at each link, making it difficult to discover and solve problems in a timely manner. For damaged goods, a simple disposal strategy of direct return or direct scrapping is usually adopted. The lack of comprehensive consideration of factors such as the degree of damage and residual value makes it difficult to make the best decision and minimize losses. At the same time, reverse express delivery is more likely to be lost in the transportation link, but many companies have failed to establish an effective loss risk prediction mechanism. The lack of differentiated control measures for different risk levels makes it difficult to reduce the loss risk level in a targeted manner. Summary of the invention

[0005] The present invention provides a method and system for reverse express status identification and tracking management based on big data, so as to solve the defects of the prior art that it is difficult to minimize the loss and the delivery is easily lost.

[0006] On the one hand, the present invention provides a method for identifying and tracking reverse express delivery status based on big data, which is used to identify the status of reverse express delivery and perform tracking management, including: S1. Obtain the product information, current location information and product status of the reverse express delivery. The product information includes product type, volume and weight. The product status includes damaged products and intact products.

[0007] S2. Obtain the return process and return cost of reverse express return through current location information and product information.

[0008] S3. For reverse express deliveries of damaged goods, a non-return risk assessment and a comprehensive value assessment are conducted, and a handling decision is made based on the results of the comprehensive value assessment. The handling decisions include direct abandonment, discounted sales, and return via the original route.

[0009] S4. Track the reverse express delivery according to the preset cycle to obtain the return status data of each return link in the return process. The return status data includes the actual time data, transportation route information and weight change information of each return link.

[0010] S5. Construct a loss prediction model based on multi-layer perceptron to predict the loss risk of reverse express delivery, and optimize the hyperparameters of the loss prediction model through the Hippo algorithm, input the return status data, and output the loss risk probability of reverse express delivery.

[0011] S6. Classify the loss risk level of reverse express delivery according to the loss risk probability, and set up a corresponding tracking management plan for each loss risk level.

[0012] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, in step S2, the return cost includes the return logistics fee and the management time cost, and the calculation formula of the return cost is:

[0013] In the formula, g represents the return logistics cost, represents the unit time cost coefficient, and H represents the management time.

[0014] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, in step S3, the process of conducting non-return risk assessment includes: Collect historical relevant data, including product information, damage conditions and corresponding non-return risks.

[0015] The relevant data are coded and normalized to obtain standardized data.

[0016] Build an evaluation model based on the AdaBoost algorithm, input product information and damage conditions, and output the corresponding non-return risk.

[0017] The evaluation model is trained and the model parameters that meet the test accuracy are retained.

[0018] The product information and damage conditions of reverse express delivery are coded and standardized, and input into the trained evaluation model to output the non-return risk of reverse express delivery, which includes no risk and risk.

[0019] Conduct a comprehensive value assessment on risk-free damaged goods, and return risky damaged goods along the original route.

[0020] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, in step S3, the process of performing comprehensive value assessment includes: S31. Obtain a damaged position image, and evaluate the damaged goods in combination with customer feedback information to obtain a damage ratio.

[0021] S32. Calculate the residual utilization value based on the damage ratio.

[0022] S33. Comprehensive value assessment of reverse express delivery is performed, and the calculation formula is expressed as:

[0023] In the formula, It represents the inventory pressure conversion coefficient, Y represents the inventory quantity of the commodity, S represents the cost required to process the commodity, and B represents the residual utilization value.

[0024] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, the process of making a processing decision includes: If C≤0, choose to give up directly.

[0025] If P>C>0, choose discount sales, where P represents the expected value that can be obtained from discount sales, and the formula is:

[0026] In the formula, Indicates the discount percentage.

[0027] If C ≥ P, the product will be marked as damaged and returned along the original route.

[0028] According to a reverse express status identification and tracking management method based on big data provided by the present invention, when a discount sale is selected, it is determined whether the customer can purchase at the expected value. If so, a discount sale is made; otherwise, it is marked as a damaged product and returned along the original route.

[0029] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, the process of constructing a loss prediction model based on a multi-layer perceptron includes: S51. Collect historical return status data of reverse express and mark whether it is lost.

[0030] S52, preprocessing the historical return status data, the preprocessing includes supplementing missing values, cleaning outliers and standardizing data to obtain preprocessed data, and dividing the preprocessed data into a training set, a test set and a validation set.

[0031] S53, construct a multi-layer perceptron model, including an input layer for receiving return status data, multiple hidden layers, each hidden layer including a preset number of neurons, and an output layer for outputting the loss risk probability of reverse express delivery.

[0032] S54. Select an activation function and use the Sigmoid function as the activation function of the multilayer perceptron model. The expression formula of the Sigmoid function is:

[0033] In it, the input value x is between 0 and 1, and the output value f(x) varies smoothly between 0 and 1.

[0034] S55. Train the multilayer perceptron model on the training set, use the test set to evaluate the trained multilayer perceptron model, calculate the accuracy, retain the multilayer perceptron model parameters with the highest accuracy, and obtain the loss prediction model.

[0035] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, in step S5, the process of optimizing the hyperparameters of the loss prediction model by the Hippo algorithm includes: The hyperparameter combination is composed of the learning rate, the number of hidden layer neurons and the regularization parameter.

[0036] A population of hippos is randomly generated, and each individual hippo represents a set of hyperparameters.

[0037] The validation set data is used to evaluate the performance of the loss prediction model under the hyperparameter combination represented by each hippo individual as the fitness value.

[0038] According to the fitness value, the individual optimal position of each hippopotamus and the global optimal position of the entire hippopotamus are updated. The hippopotamus position update formula is:

[0039] in, and are the positions of the i-th hippopotamus individual in the j-th dimension at the t+1th and tth iterations, respectively. is the weight that controls the approach to the optimal individual, is the weight controlling random exploration. is the position of the best hippopotamus in the current iteration in the jth dimension, is the position of a hippopotamus randomly selected from the population in the jth dimension.

[0040] According to a method for reverse express status identification and tracking management based on big data provided by the present invention, in step S6, the loss risk level includes a low loss risk level, a medium loss risk level and a high loss risk level. Setting a corresponding tracking management plan includes: For reverse express with low loss risk level, tracking will be carried out according to the preset cycle.

[0041] For reverse express with medium loss risk level, increase tracking frequency.

[0042] For reverse express with high loss risk level, real-time tracking is carried out and an emergency response mechanism is established.

[0043] On the other hand, the present invention also provides a reverse express status identification and tracking management system based on big data, comprising: The data collection module is used to obtain the product information, current location information and product status of reverse express delivery.

[0044] The return path and cost analysis module is used to obtain the return process and return cost of reverse express returns.

[0045] The damaged goods processing module is used to conduct non-return risk assessment and comprehensive value assessment on the reverse express delivery of damaged goods, and make processing decisions based on the comprehensive value assessment results.

[0046] The reverse express loss prediction module is used to track reverse express according to the preset cycle, obtain the return status data of each link, build a loss prediction model based on the multi-layer perceptron to predict the loss risk of reverse express, and optimize the hyperparameters of the loss prediction model through the Hippo algorithm, input the return status data, and output the loss risk probability of reverse express.

[0047] The hierarchical management and control module is used to classify the loss risk level of reverse express delivery according to the loss risk probability, and set up a corresponding tracking management plan for each loss risk level.

[0048] The reverse express status identification and tracking management method and system based on big data provided by the present invention can make more reasonable decisions on the disposal of damaged goods by conducting risk assessment and comprehensive value assessment of reverse express delivery for non-return items, taking into account both the value of the goods and the disposal cost, and can also prevent subsequent losses caused by the failure to return damaged goods, thereby minimizing overall losses. By constructing a loss prediction model, the probability of loss risk of reverse express delivery can be accurately predicted, and differentiated tracking management measures can be taken for different loss risk levels to effectively reduce the risk of loss. The method based on big data can effectively solve the key pain points in reverse logistics management, has significant advantages in improving operational efficiency, reducing costs and risks, and provides strong support for the intelligent transformation of reverse logistics of enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 It is a flow chart of a method for reverse express delivery status identification and tracking management based on big data provided by an embodiment of the present invention; Figure 2 It is a structural diagram of a reverse express status identification and tracking management system based on big data provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0052] Combine the following Figure 1-Figure 2 The present invention describes the reverse express status identification and tracking management method and system based on big data.

[0053] Figure 1 It is a flow chart of a method for reverse express status identification and tracking management based on big data provided by an embodiment of the present invention.

[0054] like Figure 1 As shown, the reverse express status identification and tracking management method and system based on big data provided by the embodiment of the present invention, the execution subject can be the reverse express status identification and tracking management method based on big data, the method includes: S1. Obtain the order number, product information, current location information and product status of the reverse express delivery. The product information includes product type, volume and weight. The product status includes damaged products and intact products.

[0055] In this embodiment, the real-time reverse express order number, commodity information, and current location information data are obtained from the system interface or APP of the express company, and a unified commodity information coding system is established to standardize the commodity type, size and weight attributes. At the same time, a complete data verification mechanism is established to identify and correct possible errors or omissions in the data, and the data source is regularly reviewed and evaluated to ensure the accuracy and completeness of the data.

[0056] By classifying the status of goods, we can determine the reason for reverse delivery. If the customer returns the goods because they are damaged, we can consider selling them at a discount.

[0057] S2. Obtain the return process and return cost of reverse express return through current location information and product information.

[0058] The return cost includes the return logistics fee and management time cost. The calculation formula of the return cost is:

[0059] In the formula, g represents the return logistics cost, represents the unit time cost coefficient, and H represents the management time.

[0060] The specific return logistics costs are calculated based on the weight and volume of the goods, combined with the return freight price system of different express companies. At the same time, the cost differences of different return methods need to be considered, including self-pickup, door-to-door collection, etc. For some remote areas or special products, additional surcharges need to be considered.

[0061] Management time includes the time required for a series of processes such as receiving returns, inspecting goods, recording, and refunding. The unit time cost coefficient is calculated based on the company's labor cost budget and reflects the value of labor input. For different return processing processes, the time cost of each link is calculated separately.

[0062] S3. For reverse express deliveries of damaged goods, a non-return risk assessment and a comprehensive value assessment are conducted, and a handling decision is made based on the results of the comprehensive value assessment. The handling decisions include direct abandonment, discounted sales, and return via the original route.

[0063] The process of conducting a non-return risk assessment includes: Collect historical relevant data, including product information, damage conditions and corresponding non-return risks.

[0064] The relevant data are coded and normalized to obtain standardized data.

[0065] Build an evaluation model based on the AdaBoost algorithm, input product information and damage conditions, and output the corresponding non-return risk.

[0066] The evaluation model is trained and the model parameters that meet the test accuracy are retained.

[0067] The product information and damage conditions of reverse express delivery are coded and standardized, and input into the trained evaluation model to output the non-return risk of reverse express delivery, which includes no risk and risk.

[0068] Conduct a comprehensive value assessment on risk-free damaged goods, and return risky damaged goods along the original route.

[0069] In this embodiment, the risk of non-return is assessed based on the type and damage of the goods, which can prevent subsequent losses caused by damaged goods not being returned. For example, special goods such as chemicals and batteries may pose safety hazards and environmental pollution risks even if they are slightly damaged.

[0070] The process of conducting a comprehensive value assessment includes: S31. Obtain the damaged position image, and evaluate the damaged goods in combination with customer feedback information to obtain the damage ratio, which is calculated as follows:

[0071] In the formula, A represents the damage ratio, N represents the value of the damaged part, and M represents the value of the complete product.

[0072] S32. Calculate the residual utilization value based on the damage ratio. The calculation formula is:

[0073] Where B represents the residual utilization value.

[0074] S33. Comprehensive value assessment of reverse express delivery is performed, and the calculation formula is expressed as:

[0075] In the formula, It represents the inventory pressure conversion coefficient, Y represents the inventory quantity of the commodity, S represents the cost required to process the commodity, and B represents the residual utilization value.

[0076] In this embodiment, the inventory pressure conversion coefficient can be dynamically adjusted according to the enterprise's inventory management strategy and cost accounting system. When the damaged goods are large in size and heavy in weight, S represents the cost of transporting the damaged goods to the shipping point or recycling point, including labor costs, transportation costs, etc.

[0077] The process for making treatment decisions includes: If C≤0, choose to give up directly.

[0078] If P>C>0, choose discount sales, where P represents the expected value that can be obtained from discount sales, and the formula is:

[0079] In the formula, Indicates the discount percentage.

[0080] If C ≥ P, the product will be marked as damaged and returned along the original route.

[0081] When choosing a discount sale, determine whether the customer can purchase at the expected value. If so, offer a discount sale. Otherwise, mark the item as damaged and return it along the original path.

[0082] S4. Track the reverse express delivery according to the preset cycle to obtain the return status data of each return link in the return process. The return status data includes the actual time data, transportation route information and weight change information of each return link.

[0083] S5. Construct a loss prediction model based on multi-layer perceptron to predict the loss risk of reverse express delivery, and optimize the hyperparameters of the loss prediction model through the Hippo algorithm, input the return status data, and output the loss risk probability of reverse express delivery.

[0084] The process of building a loss prediction model based on a multi-layer perceptron includes: S51. Collect historical return status data of reverse express and mark whether it is lost.

[0085] S52, preprocessing the historical return status data, the preprocessing includes supplementing missing values, cleaning outliers and standardizing data to obtain preprocessed data, and dividing the preprocessed data into a training set, a test set and a validation set.

[0086] Use z-score to check the outliers in the historical return status data, remove data points that are beyond the reasonable range and obvious erroneous data, use the average value to fill in missing values, and eliminate duplicates in the data to reduce redundancy. The z-score formula is expressed as:

[0087] Where d is the data point, is the sample mean, is the sample standard deviation, when When it is greater than the set threshold, the data point is considered an outlier.

[0088] The mean value filling formula is:

[0089] Here, m is the number of non-missing values ​​in the dataset.

[0090] S53, construct a multi-layer perceptron model, including an input layer for receiving return status data, multiple hidden layers, each hidden layer including a preset number of neurons, and an output layer for outputting the loss risk probability of reverse express delivery.

[0091] S54. Select an activation function and use the Sigmoid function as the activation function of the multilayer perceptron model. The expression formula of the Sigmoid function is:

[0092] In it, the input value x is between 0 and 1, and the output value f(x) varies smoothly between 0 and 1.

[0093] S55. Train the multilayer perceptron model on the training set, use the test set to evaluate the trained multilayer perceptron model, calculate the accuracy, retain the multilayer perceptron model parameters with the highest accuracy, and obtain the loss prediction model.

[0094] The process of optimizing the hyperparameters of the loss prediction model through the Hippo algorithm includes: The hyperparameter combination is composed of the learning rate, the number of hidden layer neurons and the regularization parameter.

[0095] A population of hippos is randomly generated, and each individual hippo represents a set of hyperparameters.

[0096] The validation set data is used to evaluate the performance of the loss prediction model under the hyperparameter combination represented by each hippo individual as the fitness value.

[0097] According to the fitness value, the individual optimal position of each hippopotamus and the global optimal position of the entire hippopotamus are updated. The hippopotamus position update formula is:

[0098] in, and are the positions of the i-th hippopotamus individual in the j-th dimension at the t+1th and tth iterations, respectively. is the weight that controls the approach to the optimal individual, is the weight controlling random exploration. is the position of the best hippopotamus in the current iteration in the jth dimension, is the position of a hippopotamus randomly selected from the population in the jth dimension.

[0099] S6. Classify the loss risk level of reverse express delivery according to the loss risk probability, and set up a corresponding tracking management plan for each loss risk level.

[0100] The loss risk levels include low loss risk level, medium loss risk level and high loss risk level. Setting up corresponding tracking management plans includes: For reverse express with low loss risk level, tracking will be carried out according to the preset cycle.

[0101] For reverse express with medium loss risk level, increase tracking frequency.

[0102] For reverse express with high loss risk level, real-time tracking is carried out and an emergency response mechanism is established.

[0103] In this embodiment, for reverse express delivery with low loss risk level, regular tracking is performed according to a certain preset period (such as every 3-5 days). The tracking method may include telephone inquiries, checking logistics information, etc. The transportation status is timely grasped and problems are dealt with in time.

[0104] Increase the tracking frequency for reverse express delivery with medium loss risk level, such as active tracking every 1-2 days. Strengthen cooperation with logistics service providers and pay close attention to the dynamics of logistics nodes. Establish an early warning mechanism and initiate the disposal process in time if an abnormality occurs.

[0105] We implement real-time full-process tracking of reverse express deliveries with high loss risk levels and monitor the dynamics of goods throughout the process. We maintain close communication with logistics service providers and formulate emergency response plans. When necessary, we adopt special measures such as manual tracking to ensure the safety of goods.

[0106] In summary, this embodiment provides a reverse express status identification and tracking management method based on big data. By conducting non-return risk assessment and comprehensive value assessment on reverse express, a more reasonable decision on the disposal of damaged goods is made, which takes into account both the value of the goods and the disposal cost. It can also prevent subsequent losses caused by the failure to return damaged goods, and minimize overall losses. By constructing a loss prediction model, the loss risk probability of reverse express can be accurately predicted, and differentiated tracking management measures can be taken for different loss risk levels to effectively reduce the risk of loss. The method based on big data can effectively solve the key pain points in reverse logistics management, and has significant advantages in improving operational efficiency, reducing costs and risks, and provides strong support for the intelligent transformation of enterprise reverse logistics.

[0107] Based on the same general inventive concept, the present invention also protects a reverse express status identification and tracking management system based on big data. The reverse express status identification and tracking management system based on big data provided by the present invention is described below. The reverse express status identification and tracking management system based on big data described below and the reverse express status identification and tracking management method based on big data described above can be referenced to each other.

[0108] Figure 2 It is a structural diagram of a reverse express status identification and tracking management system based on big data provided by an embodiment of the present invention.

[0109] like Figure 2 As shown, the reverse express status identification and tracking management system based on big data includes a data collection module, a return path and cost analysis module, a damaged goods processing module, a reverse express loss prediction module and a hierarchical management and control module.

[0110] The data collection module is used to obtain the order number, product information, current location information and product status of reverse express delivery.

[0111] The return path and cost analysis module is used to obtain the return process and return cost of reverse express returns.

[0112] The damaged goods processing module is used to conduct non-return risk assessment and comprehensive value assessment on the reverse express delivery of damaged goods, and make processing decisions based on the comprehensive value assessment results.

[0113] The reverse express loss prediction module is used to track reverse express according to the preset cycle, obtain the return status data of each link, build a loss prediction model based on the multi-layer perceptron to predict the loss risk of reverse express, and optimize the hyperparameters of the loss prediction model through the Hippo algorithm, input the return status data, and output the loss risk probability of reverse express.

[0114] The hierarchical management and control module is used to classify the loss risk level of reverse express delivery according to the loss risk probability, and set up a corresponding tracking management plan for each loss risk level.

[0115] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying and tracking reverse express delivery status based on big data, which is used to identify the status of reverse express delivery and conduct tracking management, characterized in that: include: S1. Obtaining the commodity information, current location information and commodity status of the reverse express delivery, wherein the commodity information includes commodity type, volume and weight, and the commodity status includes damaged commodities and intact commodities; S2. Obtaining the return process and return cost of the reverse express return through the current location information and the commodity information; S3. Conduct a non-return risk assessment and a comprehensive value assessment on the reverse express delivery of the damaged goods, and make a handling decision based on the comprehensive value assessment result, wherein the handling decision includes direct abandonment, discounted sales, and return according to the original route; S4. Track the reverse express delivery according to a preset cycle, and obtain the return status data of each return link in the return process, wherein the return status data includes the actual time data, transportation path information and weight change information of each return link; S5. Construct a loss prediction model based on a multi-layer perceptron to predict the loss risk of reverse express delivery, optimize the hyperparameters of the loss prediction model through the Hippo algorithm, input the return status data, and output the loss risk probability of the reverse express delivery; S6. Classify the reverse express delivery into loss risk levels according to the loss risk probability, and set a corresponding tracking management plan for each loss risk level.

2. The method for reverse express status identification and tracking management based on big data according to claim 1 is characterized in that: In step S2, the return cost includes return logistics fee and management time cost, and the calculation formula of the return cost is: ; In the formula, g represents the return logistics cost, represents the unit time cost coefficient, and H represents the management time.

3. The method for reverse express status identification and tracking management based on big data according to claim 1 is characterized in that: In step S3, the process of conducting non-return risk assessment includes: Collect historical relevant data, including product information, damage conditions, and corresponding non-return risks; Performing code conversion and numerical standardization processing on the relevant data to obtain standardized data; Construct an evaluation model based on the AdaBoost algorithm, input the product information and damage condition, and output the corresponding non-return risk; Training the evaluation model and retaining model parameters that meet the test accuracy; The commodity information and damage conditions of the reverse express are coded and standardized, and input into the trained evaluation model to output the non-return risk of the reverse express, where the non-return risk includes no risk and risk; A comprehensive value assessment is conducted on the risk-free damaged goods, and the risky damaged goods are returned along the original route.

4. The method for reverse express status identification and tracking management based on big data according to claim 1 is characterized in that: In step S3, the process of performing the comprehensive value assessment includes: S31, obtaining a damaged position image, and evaluating the damaged product in combination with customer feedback information to obtain a damage ratio; S32, calculating the residual utilization value according to the damage ratio; S33, conduct a comprehensive value assessment on the reverse express delivery, and the calculation formula is expressed as: ; In the formula, It represents the inventory pressure conversion coefficient, Y represents the inventory quantity of the commodity, S represents the cost required to process the commodity, and B represents the residual utilization value.

5. The method for reverse express status identification and tracking management based on big data according to claim 1, characterized in that: The process for making the processing decision includes: If C≤0, choose to give up directly; If P>C>0, choose discount sales, where P represents the expected value that can be obtained from discount sales, and the formula is: ; In the formula, Indicates the discount ratio; If C ≥ P, the product will be marked as damaged and returned along the original route.

6. The method for reverse express status identification and tracking management based on big data according to claim 5 is characterized in that: When the discount sale is selected, it is determined whether the customer can purchase at the expected value. If so, a discount sale is made; otherwise, the product is marked as damaged and returned along the original path.

7. The method for reverse express status identification and tracking management based on big data according to claim 1, characterized in that: In step S5, the process of constructing a loss prediction model based on a multi-layer perceptron includes: S51. Collect historical return status data of reverse express delivery and mark whether it is lost; S52, preprocessing the historical return status data, wherein the preprocessing includes supplementing missing values, cleaning abnormal values ​​and data standardization to obtain preprocessed data, and dividing the preprocessed data into a training set, a test set and a validation set; S53, constructing a multi-layer perceptron model, including an input layer for receiving the return status data; multiple hidden layers, each hidden layer including a preset number of neurons; and an output layer for outputting the loss risk probability of the reverse express delivery; S54, select an activation function, and use the Sigmoid function as the activation function of the multilayer perceptron model. The expression formula of the Sigmoid function is: ; Among them, the input value x is between 0 and 1, and the output value f(x) changes smoothly between 0 and 1; S55, training the multilayer perceptron model on the training set, using the test set to evaluate the trained multilayer perceptron model, calculating the accuracy, retaining the multilayer perceptron model parameters with the highest accuracy, and obtaining a loss prediction model.

8. The method for reverse express status identification and tracking management based on big data according to claim 7 is characterized in that: In step S5, the process of optimizing the hyperparameters of the loss prediction model by the Hippo algorithm includes: The hyperparameter combination is formed by the learning rate, the number of hidden layer neurons and the regularization parameter; Randomly generate a hippo population, each hippo individual represents a set of hyperparameters; Using the validation set data to evaluate the performance of the loss prediction model under the hyperparameter combination represented by each hippopotamus individual as a fitness value; The individual optimal position of each hippopotamus and the global optimal position of the entire hippopotamus are updated according to the fitness value. The hippopotamus position update formula is: ; in, and are the positions of the i-th hippopotamus individual in the j-th dimension at the t+1th and tth iterations respectively; is the weight that controls the approach to the optimal individual, is the weight controlling random exploration; is the position of the best hippopotamus in the current iteration in the jth dimension, is the position of a hippopotamus randomly selected from the population in the jth dimension.

9. The method for reverse express status identification and tracking management based on big data according to claim 1, characterized in that: In step S6, the loss risk level includes a low loss risk level, a medium loss risk level and a high loss risk level; Setting the corresponding tracking management scheme includes: For reverse express with low loss risk level, tracking is performed according to the preset period; For reverse express with medium loss risk level, increase tracking frequency; For reverse express with high loss risk level, real-time tracking will be carried out and an emergency response mechanism will be established.

10. A reverse express status identification and tracking management system based on big data, characterized in that: The reverse express status identification and tracking management system based on big data includes: The data collection module is used to obtain the product information, current location information and product status of reverse express delivery; The return path and cost analysis module is used to obtain the return process and return cost of reverse express returns; The damaged goods processing module is used to conduct non-return risk assessment and comprehensive value assessment on the reverse express delivery of damaged goods, and make processing decisions based on the comprehensive value assessment results; The reverse express loss prediction module is used to track reverse express according to the preset cycle, obtain the return status data of each link, build a loss prediction model based on the multi-layer perceptron to predict the loss risk of reverse express, and optimize the hyperparameters of the loss prediction model through the Hippo algorithm, input the return status data, and output the loss risk probability of reverse express; The hierarchical management and control module is used to classify the loss risk level of reverse express delivery according to the loss risk probability, and set up a corresponding tracking management plan for each loss risk level.

Images