A cross-border logistics data information processing method and system

Abstract

The present invention relates to the field of cross-border logistics data processing technology, and in particular, to a cross-border logistics data information processing method and system, the steps of which include: obtaining cross-border logistics data; extracting features from the cross-border logistics data, setting the target variable as the transportation time, solving the correlation coefficient between the extracted features and the target variable, and constructing a composite risk score feature vector based on the correlation coefficient; inputting the composite risk score feature vector into a trained LSTM model for calculation to obtain the minimum transportation time of cross-border logistics. The application of the present invention can not only accurately predict logistics time, but also provide decision support for logistics optimization, which is helpful to improve the efficiency and reliability of cross-border logistics.

Description

Technical Field

[0001] The present invention relates to the technical field of cross-border logistics data processing, and in particular to a method and system for cross-border logistics data information processing. Background Art

[0002] Cross-border logistics time prediction is a key challenge in e-commerce and international trade. With the acceleration of globalization, accurately predicting logistics time is becoming more and more important. At present, traditional cross-border logistics time prediction methods are usually manual predictions, which have many significant defects. On the one hand, manual predictions are highly dependent on personal experience and are easily affected by subjective factors, making it difficult to ensure the consistency and accuracy of predictions; on the other hand, faced with a large amount of complex logistics data, manual processing is inefficient and cannot respond to rapidly changing market demands in a timely manner. In addition, it lacks scalability and is difficult to cope with the challenges brought about by the expansion of business scale. Based on this, in response to the above problems, we designed a cross-border logistics data information processing method and system. Summary of the invention

[0003] The purpose of the present invention is to provide a cross-border logistics data information processing method and system, which innovatively combines data-driven and machine learning methods. Through feature extraction and correlation analysis, it can identify key factors from complex cross-border logistics data, improve the interpretability of the model, and take minimizing transportation time as the objective function of the LSTM model. It can not only accurately predict logistics time, but also provide decision support for logistics optimization, which helps to improve the efficiency and reliability of cross-border logistics.

[0004] The embodiments of the present invention are implemented by the following technical solutions:

[0005] A method for processing cross-border logistics data information, the method comprising the steps of:

[0006] Obtain cross-border logistics data;

[0007] Extract features from cross-border logistics data, set the target variable as transportation time, calculate the correlation coefficient between the extracted features and the target variable, and construct a composite risk score feature vector based on the correlation coefficient;

[0008] The composite risk score feature vector is input into the trained LSTM model for calculation to obtain the minimum transportation time for cross-border logistics.

[0009] Optionally, the cross-border logistics data is subjected to feature extraction to obtain the transportation time features of the cross-border logistics data. , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics , wherein the specific calculation formula of the transportation time characteristic is:

[0010]

[0011] The specific calculation formula of the transportation distance characteristic is:

[0012]

[0013] in, is the latitude, is the longitude, is the radius of the Earth, For the end time, is the start time.

[0014] Optionally, the target variable is set as the transportation time, and the correlation coefficient between the extracted features and the target variable is solved as follows:

[0015] Set the target variable to transportation time , through the correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The correlation coefficient with the target variable;

[0016] Select the transport distance feature in descending order of correlation coefficients. , Characteristics of customs clearance times and product weight characteristics as output.

[0017] Optionally, the correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The specific calculation formula for the correlation coefficient between the target variable and the target variable is:

[0018]

[0019]

[0020] Among them, the correlation matrix The number 1 in , the correlation matrix The number 2 in the , the correlation matrix The number 3 in the , the correlation matrix The number 4 in , the correlation matrix The number 5 in , is the correlation coefficient between features x and y, , are the i-th observation values ​​of features x and y, respectively. , are the average values ​​of features x and y respectively, and N is the total number of observations.

[0021] Optionally, the composite risk score feature vector is constructed based on the correlation coefficient, which is specifically:

[0022] Selection based transport distance feature , Characteristics of customs clearance times and product weight characteristics , construct geographical location feature G and transportation mode feature M;

[0023] Combine the selected features and constructed features to construct the composite risk score feature vector R.

[0024] Optionally, the combination of the selected features and the constructed features is used to construct a composite risk score feature vector, and the specific calculation formula is:

[0025]

[0026]

[0027] in, , , , are weight coefficients, is the maximum transport distance, is the maximum number of customs clearances, is the maximum product weight, is a function based on geographic location and mode of transportation, R is a composite risk score vector, and X is the input vector of the trained LSTM model.

[0028] Optionally, the composite risk score feature vector is input into a trained LSTM model for calculation, and the objective function calculation formula of the LSTM model is:

[0029]

[0030] in, is the number of training samples, is the transportation time of the i-th training sample, are LSTM model parameters;

[0031] The constraints are as follows:

[0032]

[0033]

[0034]

[0035]

[0036]

[0037] in, is the commodity weight of the i-th sample, is the maximum allowable product weight, is the transportation cost of the i-th sample, is the average maximum allowable cost, is the output of the LSTM model, is the error term, , , are the unit costs of distance, weight and time respectively, is the transport distance.

[0038] A cross-border logistics data information processing system, comprising:

[0039] Cross-border logistics data acquisition module, to obtain cross-border logistics data;

[0040] The cross-border logistics data extraction module extracts features from cross-border logistics data, sets the target variable as transportation time, calculates the correlation coefficient between the extracted features and the target variable, and constructs a composite risk score feature vector based on the correlation coefficient;

[0041] The cross-border logistics data processing module inputs the composite risk score feature vector into the trained LSTM model for calculation to obtain the minimum transportation time for cross-border logistics.

[0042] The technical solution of the embodiment of the present invention has at least the following advantages and beneficial effects:

[0043] The embodiment of the present invention innovatively combines data-driven and machine learning methods. Through feature extraction and correlation analysis, it can identify key factors from complex cross-border logistics data, improve the interpretability of the model, and use minimizing transportation time as the objective function of the LSTM model. It can not only accurately predict logistics time, but also provide decision support for logistics optimization, which helps to improve the efficiency and reliability of cross-border logistics. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flow chart of a method for processing cross-border logistics data information provided in accordance with an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0046] like Figure 1 As shown, the present invention provides one embodiment: a method for processing cross-border logistics data information, the steps of the method comprising:

[0047] Obtain cross-border logistics data;

[0048] Extract features from cross-border logistics data, set the target variable as transportation time, calculate the correlation coefficient between the extracted features and the target variable, and construct a composite risk score feature vector based on the correlation coefficient;

[0049] The composite risk score feature vector is input into the trained LSTM model for calculation to obtain the minimum transportation time for cross-border logistics.

[0050] In this embodiment, the feature extraction of cross-border logistics data is performed to obtain the transportation time feature of the cross-border logistics data. , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics , wherein the specific calculation formula of the transportation time characteristic is:

[0051]

[0052] The specific calculation formula of the transportation distance characteristic is:

[0053]

[0054] in, is the latitude, is the longitude, is the radius of the Earth, For the end time, is the start time.

[0055] In the implementation, the target variable is set as the transportation time, and the correlation coefficient between the extracted features and the target variable is solved, which is specifically:

[0056] Set the target variable to transportation time , through the correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The correlation coefficient with the target variable;

[0057] Select the transport distance feature in descending order of correlation coefficients. , Characteristics of customs clearance times and product weight characteristics as output.

[0058] The correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The specific calculation formula for the correlation coefficient between the target variable and the target variable is:

[0059]

[0060]

[0061] Among them, the correlation matrix The number 1 in , the correlation matrix The number 2 in the , the correlation matrix The number 3 in the , the correlation matrix The number 4 in , the correlation matrix The number 5 in , is the correlation coefficient between features x and y, , are the i-th observation values ​​of features x and y, respectively. , are the average values ​​of features x and y respectively, and N is the total number of observations.

[0062] In this embodiment, the composite risk score feature vector is constructed based on the correlation coefficient, which is specifically:

[0063] Selection based transport distance feature , Characteristics of customs clearance times and product weight characteristics , construct geographical location feature G and transportation mode feature M;

[0064] Combine the selected features and constructed features to construct the composite risk score feature vector R.

[0065] Specifically, the geographical location feature G is constructed by coding the origin and destination, and the mode of transportation feature M is constructed by coding different modes of transportation, which are divided into sea transportation, air transportation, land transportation and railway transportation.

[0066] More specifically, the composite risk score feature vector is constructed by combining the selected features and the constructed features, and the specific calculation formula is as follows:

[0067]

[0068]

[0069] in, , , , are weight coefficients, is the maximum transport distance, is the maximum number of customs clearances, is the maximum product weight, is a function based on geographic location and mode of transportation, which can be designed according to historical data. R is the composite risk score vector, and X is the input vector of the trained LSTM model.

[0070] In this embodiment, the composite risk score feature vector is input into the trained LSTM model for calculation. The objective function calculation formula of the LSTM model is:

[0071]

[0072] in, is the number of training samples, is the transportation time of the i-th training sample, are LSTM model parameters;

[0073] The constraints are as follows:

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] in, is the commodity weight of the i-th sample, is the maximum allowable product weight, is the transportation cost of the i-th sample, is the average maximum allowable cost, is the output of the LSTM model, is the error term, , , are the unit costs of distance, weight and time respectively, is the transport distance.

[0080] In implementation, the objective function designed in this embodiment specifically takes the minimum transportation time as the objective function, and takes into account the non-negative constraints of transportation time, transportation capacity constraints, budget constraints, logistics time calculation constraints and cost calculation constraints. The objective function aims to make the average logistics transportation time reach the preset target, reflecting the pursuit of efficiency. The non-negative constraint of transportation time ensures the practical feasibility of the prediction results. The transportation capacity constraint takes into account the actual constraints of logistics resources. The budget constraint balances time efficiency and cost control. The logistics time calculation constraint introduces the LSTM model, which improves the accuracy of the prediction and the ability to capture complex time series relationships. The cost calculation constraint fully considers the impact of distance, weight and time on cost. Based on the objective function and constraints of this embodiment, not only can the prediction accuracy be improved, but also a more comprehensive and operational guidance can be provided for logistics decision-making. It can be understood that the specific construction and training process of the LSTM model is an existing process, and this embodiment adopts this process, and only optimizes the objective function and constraints.

[0081] In combination with the above embodiments, the present invention further provides: a cross-border logistics data information processing system, comprising:

[0082] Cross-border logistics data acquisition module, to obtain cross-border logistics data;

[0083] The cross-border logistics data extraction module extracts features from cross-border logistics data, sets the target variable as transportation time, calculates the correlation coefficient between the extracted features and the target variable, and constructs a composite risk score feature vector based on the correlation coefficient;

[0084] The cross-border logistics data processing module inputs the composite risk score feature vector into the trained LSTM model for calculation to obtain the minimum transportation time for cross-border logistics.

[0085] It can be understood that the cross-border logistics data information processing system provided in this embodiment and the cross-border logistics data information processing method provided in the above embodiment are based on the same inventive concept. For more specific working principles of each module in the embodiment of the present invention, please refer to the above embodiment and will not be repeated in the embodiment of the present invention.

[0086] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A cross-border logistics data information processing method, characterized in that: The steps of the method include: Obtain cross-border logistics data; Extract features from cross-border logistics data, set the target variable as transportation time, calculate the correlation coefficient between the extracted features and the target variable, and construct a composite risk score feature vector based on the correlation coefficient; The composite risk score feature vector is input into the trained LSTM model for calculation to obtain the minimum transportation time for cross-border logistics; The feature extraction of cross-border logistics data is performed to obtain the transportation time features of the cross-border logistics data. , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics , wherein the specific calculation formula of the transportation time characteristic is: The specific calculation formula of the transportation distance characteristic is: in, is the latitude, is the longitude, is the radius of the Earth, For the end time, is the starting time; The target variable is set as the transportation time, and the correlation coefficient between the extracted features and the target variable is solved, which is specifically: Set the target variable to transportation time , through the correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The correlation coefficient with the target variable; Select the transport distance feature in descending order of correlation coefficients. , Characteristics of customs clearance times and product weight characteristics As output; The correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The specific calculation formula for the correlation coefficient between the target variable and the target variable is: Among them, the correlation matrix The number 1 in , the correlation matrix The number 2 in the , the correlation matrix The number 3 in the , the correlation matrix The number 4 in , the correlation matrix The number 5 in , is the correlation coefficient between features x and y, , are the i-th observation values ​​of features x and y, respectively. , are the average values ​​of features x and y respectively, and N is the total number of observations; The composite risk score feature vector is constructed based on the correlation coefficient, which is specifically: Selection based transport distance feature , Characteristics of customs clearance times and product weight characteristics , construct geographical location features G and transportation mode features M; Combining the selected features with the constructed features, a composite risk score feature vector R is constructed; The composite risk score feature vector is constructed by combining the selected features and the constructed features. The specific calculation formula is: in, , , , are weight coefficients, is the maximum transport distance, is the maximum number of customs clearances, is the maximum product weight, is a function based on geographic location and mode of transportation, R is a composite risk score vector, and X is the input vector of the trained LSTM model.

2. The cross-border logistics data information processing method according to claim 1 is characterized in that: The composite risk score feature vector is input into the trained LSTM model for calculation. The objective function calculation formula of the LSTM model is: in, is the number of training samples, is the transportation time of the i-th training sample, are LSTM model parameters; The constraints are as follows: in, is the commodity weight of the i-th sample, is the maximum allowable product weight, is the transportation cost of the i-th sample, is the average maximum allowable cost, is the output of the LSTM model, is the error term, , , are the unit costs of distance, weight and time respectively, is the transport distance.

3. A cross-border logistics data information processing system, characterized in that: include: Cross-border logistics data acquisition module, to obtain cross-border logistics data; The cross-border logistics data extraction module extracts features from cross-border logistics data, sets the target variable as transportation time, calculates the correlation coefficient between the extracted features and the target variable, and constructs a composite risk score feature vector based on the correlation coefficient; The cross-border logistics data processing module inputs the composite risk score feature vector into the trained LSTM model for calculation to obtain the minimum transportation time for cross-border logistics; The feature extraction of cross-border logistics data is performed to obtain the transportation time features of the cross-border logistics data. , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics , wherein the specific calculation formula of the transportation time characteristic is: The specific calculation formula of the transportation distance characteristic is: in, is the latitude, is the longitude, is the radius of the Earth, For the end time, is the starting time; The target variable is set as the transportation time, and the correlation coefficient between the extracted features and the target variable is solved, which is specifically: Set the target variable to transportation time , through the correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The correlation coefficient with the target variable; Select the transport distance feature in descending order of correlation coefficients. , Characteristics of customs clearance times and product weight characteristics As output; The correlation matrix Solving the transportation time characteristics of cross-border logistics data , Transportation distance characteristics , Characteristics of customs clearance times , Product quantity characteristics and product weight characteristics The specific calculation formula for the correlation coefficient between the target variable and the target variable is: Among them, the correlation matrix The number 1 in , the correlation matrix The number 2 in the , the correlation matrix The number 3 in the , the correlation matrix The number 4 in , the correlation matrix The number 5 in , is the correlation coefficient between features x and y, , are the i-th observation values ​​of features x and y, respectively. , are the average values ​​of features x and y respectively, and N is the total number of observations; The composite risk score feature vector is constructed based on the correlation coefficient, which is specifically: Selection-based transport distance feature , Characteristics of customs clearance times and product weight characteristics , construct geographical location features G and transportation mode features M; Combining the selected features with the constructed features, a composite risk score feature vector R is constructed; The composite risk score feature vector is constructed by combining the selected features and the constructed features. The specific calculation formula is: in, , , , are weight coefficients, is the maximum transport distance, is the maximum number of customs clearances, is the maximum product weight, is a function based on geographic location and mode of transportation, R is a composite risk score vector, and X is the input vector of the trained LSTM model.

Images