Intelligent logistics data processing method based on digitalization

Abstract

The present invention discloses a digital-based intelligent logistics data processing method, including periodically batch-acquiring multiple original order data in shopping software; extracting key logistics information from multiple original order data, the key logistics information at least includes recipient address information and item type information, parsing and classifying multiple original order data according to the key logistics information, and obtaining order classification data; based on the order classification data, determining the processing flow and distribution route of the logistics order, allocating logistics resources, and generating order execution data; uploading the order execution data to the corresponding logistics execution system, and monitoring the processing status of the order execution data by the logistics execution system. The present invention can realize intelligent and rapid classification of original order data, and can reduce the total time of picking and walking of loading and unloading equipment and improve delivery time by analyzing the picking path to optimize the objective function.

Description

Technical Field

[0001] The present invention relates to the field of logistics technology, and in particular to a digital-based intelligent logistics data processing method. Background Art

[0002] When e-commerce companies sell a certain product, they need to go through various links such as procurement, warehousing and distribution, and various types of data information are generated in this process. Since several orders are shipped to the same area during the same period, the current data processing technology generally records the data manually into the equipment and then classifies and summarizes it, which is prone to problems such as long data processing time and low efficiency. Summary of the invention

[0003] The purpose of the present invention is to provide a digital-based intelligent logistics data processing method to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] The digital-based intelligent logistics data processing method includes:

[0006] Periodically obtain multiple original order data in batches from shopping software;

[0007] Extracting key logistics information from multiple original order data, where the key logistics information at least includes recipient address information and item type information, and performing data parsing and classification on the multiple original order data according to the key logistics information to obtain order classification data;

[0008] Based on the order classification data, determine the processing flow and delivery route of the logistics order, allocate logistics resources, and generate order execution data. The logistics resources at least include transportation tools, delivery personnel, and logistics center resources;

[0009] Upload the order execution data to the corresponding logistics execution system and monitor the processing status of the order execution data by the logistics execution system.

[0010] Furthermore, the periodic batch acquisition of multiple original order data in the shopping software specifically includes:

[0011] Get the batch processing cycle set by the user;

[0012] According to the batch processing cycle, multiple original order data in the shopping software are periodically obtained in batches.

[0013] Furthermore, multiple original order data are analyzed and classified according to key logistics information, and the order classification data obtained specifically includes:

[0014] Adopt correlation model to extract features of multiple original order data according to correlation;

[0015] Cluster the extracted features according to their correlation and calculate the sample probability;

[0016] The sample probability is used as the input of the classification model, and the corresponding classification results are calculated.

[0017] Furthermore, the features by relevance include relevance between order titles and relevance between words.

[0018] Furthermore, the extracted features are clustered according to their correlation based on the Gaussian mixture model.

[0019] Furthermore, the correlation between order titles and word correlation is expressed based on title weights.

[0020] Furthermore, the logistics execution system issues order batch picking tasks to the logistics center resources based on the order execution data, and issues transportation and delivery tasks to the transportation vehicles and delivery personnel.

[0021] Furthermore, logistics center resources include warehousing and loading and unloading equipment to support centralized management and scheduling of orders.

[0022] Furthermore, the order classification data includes order data with the same destination area and order data with the same item type information.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] The present invention periodically obtains multiple original order data in the shopping software in batches according to the batch processing cycle, extracts the features of the multiple original order data according to the correlation through the correlation model, then clusters the extracted features according to the correlation and calculates the sample probability, uses the sample probability as the input of the classification model, and calculates the corresponding classification result, thereby realizing intelligent and rapid classification of the original order data and solving the problems of long data processing time and low efficiency.

[0025] In the present invention, according to the order data with the same item type information as the order classification data, the logistics execution system issues order batch picking tasks for the logistics center resources. The order batch picking tasks are to use loading and unloading equipment to perform sorting in the warehouse: by analyzing the picking path to optimize the objective function, the total picking walking time of the loading and unloading equipment can be reduced, thereby also reducing the picking walking distance of the loading and unloading equipment, thereby improving production capacity and delivery time. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the digital-based intelligent logistics data processing method of the present invention. DETAILED DESCRIPTION

[0027] 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.

[0028] See also Figure 1 , the present invention provides a technical solution:

[0029] The digital-based intelligent logistics data processing method includes:

[0030] Periodically obtain multiple original order data in batches from shopping software;

[0031] Extracting key logistics information from multiple original order data, where the key logistics information at least includes recipient address information and item type information, and performing data parsing and classification on the multiple original order data according to the key logistics information to obtain order classification data;

[0032] Based on the order classification data, determine the processing flow and delivery route of the logistics order, allocate logistics resources, and generate order execution data. The logistics resources at least include transportation tools, delivery personnel, and logistics center resources;

[0033] Upload the order execution data to the corresponding logistics execution system and monitor the processing status of the order execution data by the logistics execution system.

[0034] In the present invention, the periodic batch acquisition of multiple original order data in the shopping software specifically includes:

[0035] Get the batch processing cycle set by the user;

[0036] According to the batch processing cycle, multiple original order data in the shopping software are periodically obtained in batches.

[0037] At present, online shopping is becoming more and more popular. In a certain period of time, there are order data sent to the same area, such as the same city or county. Due to the large number, it is generally recorded by the client first and then summarized for delivery. At this stage, this recording and summarization method still needs further improvement.

[0038] In the present invention, multiple original order data are analyzed and classified according to key logistics information, and the order classification data obtained specifically includes:

[0039] The correlation model is used to extract the features of multiple original order data based on the correlation, that is, the correlation model can extract the features of the (original order data) document data;

[0040] Cluster the extracted features according to their correlation and calculate the sample probability ;

[0041] The sample probability is used as the input of the classification model, and the corresponding classification results are calculated to achieve intelligent and rapid classification of the original order data.

[0042] In the present invention, the features based on relevance include relevance between order titles and relevance between words.

[0043] In the present invention, the extracted features are clustered according to their correlation based on a Gaussian mixture model.

[0044] In the present invention, the correlation between order titles and word correlation is expressed based on title weights.

[0045] The features of multiple original order data extracted by using the correlation model include:

[0046] Correlation and word correlation between order titles of multiple original order data S as follows:

[0047] ;

[0048] in, , Respectively i The original order data m The weight of the word j The original order data m The weight of words, a total of n original order data, sim is the cosine similarity, which is used to express the similarity between words and obtain the title weight , whose expression is:

[0049] ;

[0050] Relevance of words S The mean of δ is a balancing parameter to adjust and S right The size of the effect. δ When taking 1, and The function curve gradually rises from left to right.

[0051] In the present invention, , ;

[0052] in:

[0053] ;

[0054] ;

[0055] ;

[0056] ;

[0057] ;

[0058] tf-idf (Term Frequency-Inverse Document Frequency) is term frequency-inverse document frequency, tf ( t m , l i ), tf ( t m , l j ) is the word frequency, respectively representing the m The word in i and j The frequency of occurrence in the original order data. The more frequent the word, the more t m The more important it is in the raw order data. idf ( t m ) is the inverse document frequency, N is the total number of documents, and n All the original order data contains the word t m The number of original order data, word t m The less the original order data is included, the more important it is. T represents a (feature) word sequence, T= { t 1 , t 2 … t m}.

[0059] In the present invention, the classification model adopts the Naive Bayesian model, and the posterior probability of the Naive Bayesian model is P ( y , x ), according to the Bayesian formula:

[0060] ;

[0061] in,P ( y | x ) is the event y In known events x The probability of occurrence, x = x 1, x 2… x i , P ( y ) is the event y The prior probability of occurrence, P ( x i | y ) is the event x i In Events y The probability of occurrence.

[0062] In the present invention, the user can set a batch processing cycle on the mobile terminal. By obtaining the batch processing cycle set by the user, at the beginning of each batch processing cycle, multiple original order data are imported in batches from the shopping software installed on the mobile terminal. By identifying key information of the multiple original order data imported in batches, multiple corresponding key logistics information are obtained, and the multiple key logistics information are analyzed and classified.

[0063] In this embodiment, the order classification data includes order data with the same destination area and order data with the same item type information. Order data with the same destination area is classified: such as orders sent to Province A, City B, and County C. Order data with the same item type information is classified: such as orders for hair dryers of Brand D and Type E.

[0064] In the present invention, the logistics execution system issues order batch picking tasks to the logistics center resources according to the order execution data, and issues transportation and delivery tasks to the transportation tools and delivery personnel.

[0065] In the present invention, the logistics center resources include warehousing and its facilities, loading and unloading equipment, etc., to support the centralized management and scheduling of orders.

[0066] In the present invention, according to the same order data as the item type information of the order classification data, the logistics execution system issues order batch picking tasks for the logistics center resources. The order batch picking tasks are to sort in the warehouse using loading and unloading equipment, which is specifically described below:

[0067] If any n 1 order O k ( k =1,2… k ) need to be picked up, each order contains several items,k The quantity of items in the order is q k . If the order is not split, n 1 order is picked up in batches, and each batch of orders is picked up once on one route, making the total walking distance for picking all orders D is Minimum, then the objective function is:

[0068] ;

[0069] The constraints are:

[0070] , constraint one;

[0071] , constraint 2;

[0072] , constraint three;

[0073] in, F is the capacity of the loading and unloading equipment, U is the set of all feasible order batches, u For each feasible order batch, d u To pick up u Total travel distance of all items in the batch order;

[0074] ;

[0075] Indicates u Order composition of batch orders;

[0076] ;

[0077] In the present invention, the objective function is solved to minimize the sum of picking routes of each batch of orders after dividing the orders into batches, and constraint one of the constraints indicates that the number of items contained in each batch of orders does not exceed the capacity of the loading and unloading equipment, and constraints two and three of the constraints indicate that each order can only be assigned to one batch.

[0078] In the present invention, the picking path is analyzed to optimize the objective function, thereby reducing the total picking walking time of the loading and unloading equipment, thereby reducing the picking walking distance of the loading and unloading equipment and thus improving production capacity and delivery time.

[0079] In the present invention, logistics planning software or algorithms are used to optimize the delivery routes to reduce the transportation cost and shorten the delivery time. The monitoring function or data analysis tool of the logistics management system is used to track and analyze the processing status of the order execution data. The processing status includes the progress of each stage such as order receiving, sorting, loading, and delivery.

[0080] The present invention, the undescribed parts are prior art.

[0081] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A digital-based intelligent logistics data processing method, characterized in that: include: Periodically obtain multiple original order data in batches from shopping software; Extracting key logistics information from multiple original order data, where the key logistics information at least includes recipient address information and item type information, and performing data parsing and classification on the multiple original order data according to the key logistics information to obtain order classification data; Based on the order classification data, determine the processing flow and delivery route of the logistics order, allocate logistics resources, and generate order execution data. The logistics resources at least include transportation tools, delivery personnel, and logistics center resources; Upload order execution data to the corresponding logistics execution system and monitor the processing status of order execution data by the logistics execution system; According to the key logistics information, multiple original order data are analyzed and classified, and the order classification data obtained includes: Adopt correlation model to extract features of multiple original order data according to correlation; Cluster the extracted features according to their correlation and calculate the sample probability; The sample probability is used as the input of the classification model, and the corresponding classification results are calculated; Features by relevance include relevance between order titles and word relevance; The correlation between order titles and word correlation is expressed based on title weights; Correlation and word correlation between order titles of multiple original order data S as follows: ； in, , Respectively i The original order data m The weight of the word j The original order data m The weight of words, a total of n original order data, sim is the cosine similarity, which is used to express the similarity between words and obtain the title weight , whose expression is: ； Relevance of words S The mean of δ is a balance parameter to adjust and S right The degree of influence of size, when δ When taking 1, and The function curve gradually rises from left to right.

2. The digital-based intelligent logistics data processing method according to claim 1, characterized in that: The periodic batch acquisition of multiple original order data in the shopping software specifically includes: Get the batch processing cycle set by the user; According to the batch processing cycle, multiple original order data in the shopping software are periodically obtained in batches.

3. The digital-based intelligent logistics data processing method according to claim 1, characterized in that: The extracted features are clustered according to their correlation based on the Gaussian mixture model.

4. The digital-based intelligent logistics data processing method according to claim 1, characterized in that: The logistics execution system issues order batch picking tasks to the logistics center resources based on the order execution data, and issues transportation and delivery tasks to the transportation vehicles and delivery personnel.

5. The digital-based intelligent logistics data processing method according to claim 1, characterized in that: Logistics center resources include warehousing and loading and unloading equipment to support centralized management and scheduling of orders.

6. According to the digital-based intelligent logistics data processing method as described in claim 1, the order classification data includes order data with the same destination area and order data with the same item type information.

Images