Method, device and equipment for determining logistics route generation strategy

Abstract

The present application relates to the field of logistics, and specifically to a method, device and equipment for determining a logistics route generation strategy. The present application provides a method for determining a logistics route generation strategy, comprising: obtaining transaction volume data of a logistics route and business feature data of the logistics route; clustering the transaction volume data and business feature data as samples to obtain a plurality of cluster cluster data; constructing a logistics route generation strategy based on the plurality of cluster cluster data, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

Description

Technical Field

[0001] The present application relates to the field of logistics, and specifically to a method, device and equipment for determining a logistics route generation strategy. Background Art

[0002] In the traditional logistics route planning work, price adjusters manually formulate each logistics route based on the cost and market of each logistics route, and these logistics routes can be subsequently pushed to the market and to consumers.

[0003] With the continuous development of computer technology, the existing logistics work can be realized through online logistics platforms. After the client side determines the logistics itinerary, the server side will generate the logistics route corresponding to this logistics itinerary based on the logistics itinerary reported by the current client side and the relevant data of the original logistics route.

[0004] In the process of studying the existing related technologies, the inventors found that some of the data related to the original logistics routes had low accuracy, which easily led to low accuracy of the logistics routes generated based on these data. Summary of the invention

[0005] The present application provides a method, device and equipment for determining a logistics route generation strategy, which are used to improve the accuracy of the logistics route generation strategy to a certain extent in the process of determining the logistics route generation strategy.

[0006] In a first aspect, the present application provides a method for determining a logistics route generation strategy, the method comprising:

[0007] Obtaining transaction volume data and business characteristic data of logistics routes;

[0008] The transaction volume data and the business characteristic data are clustered as samples to obtain multiple cluster data;

[0009] Based on multiple cluster data, a logistics route generation strategy is constructed, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

[0010] In combination with the first aspect of the present application, in a first possible implementation of the first aspect of the present application, the transaction volume data and the business feature data are clustered as samples to obtain a plurality of cluster data including:

[0011] Perform data cleaning on transaction volume data and business feature data respectively;

[0012] The cleaned transaction volume data and business feature data are used as samples for clustering processing to obtain multiple cluster data.

[0013] In combination with the first possible implementation of the first aspect of the present application, in the second possible implementation of the first aspect of the present application, the service characteristic data includes at least one of line characteristic data, service scenario characteristic data, date characteristic data and quotation characteristic data.

[0014] In combination with the second possible implementation of the first aspect of the present application, in a third possible implementation of the first aspect of the present application, the transaction volume data and the business feature data after data cleaning are used as samples for clustering processing, and the obtained multiple cluster data include:

[0015] The transaction volume data after data cleaning, the business scenario feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain first cluster cluster data;

[0016] When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the path feature data after data cleaning, the quotation feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain the second cluster data;

[0017] When the second cluster data passes the unimodal check and the single check, the second cluster data is confirmed as a plurality of cluster data.

[0018] In combination with the third possible implementation manner of the first aspect of the present application, in a fourth possible implementation manner of the first aspect of the present application, when the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line feature data after data cleaning, the quotation feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing, and the second cluster data includes:

[0019] When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain the first sub-cluster data, wherein the line feature data identifies the line direction;

[0020] The first cluster data, the cleaned trading volume data, the cleaned quotation feature data and the cleaned date feature data are clustered as samples to obtain the second sub-cluster data, wherein the quotation feature data indicates an upper line trend;

[0021] The first sub-cluster cluster data and the second sub-cluster cluster data are confirmed as the second cluster cluster data.

[0022] In combination with the third or fourth possible implementation of the first aspect of the present application, in a fifth possible implementation of the first aspect of the present application, building a logistics route generation strategy based on multiple cluster data includes:

[0023] From multiple cluster data, the distribution values ​​of logistics routes with different route quotations in logistics routes with the same business scenarios, the same dates and the same route lengths are extracted as the corresponding user acceptance values;

[0024] Based on the business scenario, date, route length, route quotation and user acceptance value, a logistics route set is constructed as a logistics route generation strategy, wherein the logistics route set includes candidate logistics routes with different business scenarios, different dates, different route lengths, different route quotations and different user acceptance values.

[0025] In combination with the fifth possible implementation of the first aspect of the present application, in a sixth possible implementation of the first aspect of the present application, after building a logistics route set based on business scenarios, dates, route lengths, route quotations, and user acceptance values ​​as a logistics route generation strategy, the method further includes:

[0026] In the logistics route set, multiple candidate logistics routes are sorted according to the size of the corresponding expected benefits.

[0027] In a second aspect, the present application provides a device for determining a logistics route generation strategy, the device comprising:

[0028] A transceiver unit, used to obtain transaction volume data and business characteristic data of the logistics route;

[0029] The processing unit is used to cluster the transaction volume data and the business characteristic data as samples to obtain a plurality of cluster data; based on the plurality of cluster data, a logistics route generation strategy is constructed, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

[0030] In conjunction with the second aspect of the present application, in a first possible implementation manner of the second aspect of the present application, the processing unit is specifically configured to:

[0031] Perform data cleaning on transaction volume data and business feature data respectively;

[0032] The cleaned transaction volume data and business feature data are used as samples for clustering processing to obtain multiple cluster data.

[0033] In combination with the first possible implementation of the second aspect of the present application, in the second possible implementation of the second aspect of the present application, the service characteristic data includes at least one of line characteristic data, service scenario characteristic data, date characteristic data and quotation characteristic data.

[0034] In combination with the second possible implementation of the first aspect of the present application, in a third possible implementation of the first aspect of the present application, the processing unit is specifically configured to:

[0035] The transaction volume data after data cleaning, the business scenario feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain first cluster cluster data;

[0036] When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the path feature data after data cleaning, the quotation feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain the second cluster data;

[0037] When the second cluster data passes the unimodal check and the single check, the second cluster data is confirmed as a plurality of cluster data.

[0038] In combination with the third possible implementation of the second aspect of the present application, in a fourth possible implementation of the second aspect of the present application, the processing unit is specifically configured to:

[0039] When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain the first sub-cluster data, wherein the line feature data identifies the line direction;

[0040] The first cluster data, the cleaned trading volume data, the cleaned quotation feature data and the cleaned date feature data are clustered as samples to obtain the second sub-cluster data, wherein the quotation feature data indicates an upper line trend;

[0041] The first sub-cluster cluster data and the second sub-cluster cluster data are confirmed as the second cluster cluster data.

[0042] In combination with the third or fourth possible implementation of the second aspect of the present application, in a fifth possible implementation of the second aspect of the present application, the processing unit is specifically configured to:

[0043] From multiple cluster data, the distribution values ​​of logistics routes with different route quotations in logistics routes with the same business scenarios, the same dates and the same route lengths are extracted as the corresponding user acceptance values;

[0044] Based on the business scenario, date, route length, route quotation and user acceptance value, a logistics route set is constructed as a logistics route generation strategy, wherein the logistics route set includes candidate logistics routes with different business scenarios, different dates, different route lengths, different route quotations and different user acceptance values.

[0045] In combination with the fifth possible implementation manner of the second aspect of the present application, in a sixth possible implementation manner of the second aspect of the present application, the processing unit is further configured to:

[0046] In the logistics route set, multiple candidate logistics routes are sorted according to the size of the corresponding expected benefits.

[0047] In a third aspect, the present application also provides a device for determining a logistics route generation strategy, including a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps of any method provided in the embodiments of the present application are executed.

[0048] In a fourth aspect, the present application also provides a computer-readable storage medium, which stores a plurality of instructions, and the instructions are suitable for a processor to load to execute the steps of any method provided in the embodiments of the present application.

[0049] From the above content, it can be concluded that the present application has the following beneficial effects:

[0050] In the process of constructing a logistics route generation strategy, obtaining the transaction volume data and business feature data of the logistics route and taking both of them as samples can effectively avoid clustering errors or clustering anomalies that are easily caused by using a single data as a sample. Then, clustering processing is performed on both the transaction volume data and the business feature data. Through clustering processing, different common features of the transaction volume data and the business feature data are extracted, and samples containing common features are classified into one category. In this way, cluster cluster data with a larger data volume than the original data volume can be obtained, avoiding uneven sample distribution and limited sample size. Then, based on different cluster cluster data corresponding to different common features, a logistics route generation strategy is constructed. When the logistics route generation strategy is used to generate the target logistics route corresponding to the input target logistics itinerary, the target logistics route can be matched according to the common features corresponding to the target logistics itinerary, thereby improving the accuracy of the target logistics route to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0052] Figure 1 A flow chart of a method for determining a logistics route generation strategy for this application;

[0053] Figure 2 A schematic diagram of a process of clustering processing for this application;

[0054] Figure 3 A schematic diagram of a scenario for single peak verification of this application;

[0055] Figure 4 A schematic diagram of a scenario for clustering processing in this application;

[0056] Figure 5 Another schematic diagram of the clustering process for this application;

[0057] Figure 6 Another schematic diagram of the clustering process for this application;

[0058] Figure 7 A flow chart of a logistics route generation strategy for this application;

[0059] Figure 8 A schematic diagram of a structure of a device for determining a logistics route generation strategy of the present application;

[0060] Fig. 9 A structural schematic diagram of a device for determining the logistics route generation strategy of this application. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0062] In the following description, the specific embodiments of the present application will be described with reference to the steps and symbols performed by one or more computers, unless otherwise stated. Therefore, these steps and operations will be mentioned several times as being performed by a computer, and the computer execution referred to herein includes the operation of a computer processing unit by an electronic signal representing data in a structured form. This operation converts the data or maintains it at a location in the memory system of the computer, which can be reconfigured or otherwise change the operation of the computer in a manner familiar to testers in the field. The data structure maintained by the data is a physical location in the memory, which has specific characteristics defined by the data format. However, the principles of the present application are described in the above text, which does not represent a limitation, and testers in the field will understand that the various steps and operations described below can also be implemented in hardware.

[0063] The principles of the present application may be operated using many other general or specific purpose computing, communication environments or configurations. Examples of well-known computing systems, environments and configurations suitable for use with the present application may include, but are not limited to, handheld phones, personal computers, servers, multiprocessor systems, microcomputer-based systems, mainframe computers, and distributed computing environments, including any of the above systems or devices.

[0064] The terms "first", "second", and "third" in this application are used to distinguish different objects rather than to describe a specific order. In addition, the terms "include", "have", and any variations thereof are intended to cover non-exclusive inclusions.

[0065] First of all, before introducing this application, let me first introduce the relevant content of this application regarding the application background.

[0066] The method, device and computer-readable storage medium for determining the logistics route generation strategy provided in the present application can be applied to the device for determining the logistics route generation strategy to improve the accuracy of the logistics route generation strategy to a certain extent.

[0067] In the existing related technologies, the generation strategy of logistics routes is generally obtained by relevant staff based on the analysis of historical logistics routes, that is, it is manually configured, and its accuracy is often low, which is not suitable for today's volatile market environment and user needs.

[0068] Based on the above-mentioned defects of the existing related technologies, the present application provides a method for determining a logistics route generation strategy. In the process of determining the logistics route generation strategy, both transaction volume data and business feature data are introduced as samples. The two are clustered and constructed to obtain a logistics route generation strategy, which improves the accuracy of the logistics route generation strategy to a certain extent.

[0069] Among them, in the present application, the device for determining the logistics route generation strategy can be understood as a server device, a physical host or a user equipment (UE) and other hardware devices with data processing capabilities. The UE can specifically be a terminal device such as a smart phone, a tablet computer, a laptop computer, a PDA, a desktop computer or a personal digital assistant (PDA).

[0070] A logistics route can be understood as the transportation path used when providing logistics transportation services, which is used to transport logistics items. Specifically, it can be an express route in the express delivery industry, for example.

[0071] Next, we will introduce the method for determining the logistics route generation strategy provided by this application.

[0072] First, see Figure 1 , Figure 1 A flow chart of a method for determining a logistics route generation strategy of the present application is shown. In the present application, the method for determining a logistics route generation strategy may specifically include:

[0073] Step S101, obtaining transaction volume data of the logistics route and business characteristic data of the logistics route;

[0074] Step S102, clustering the transaction volume data and the business feature data as samples to obtain a plurality of cluster data;

[0075] Step S103, constructing a logistics route generation strategy based on the multiple cluster data, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to the input target logistics itinerary.

[0076] From the above Figure 1 It can be seen from the illustrated embodiment that, in the process of constructing a logistics route generation strategy, the present application obtains the transaction volume data and business feature data of the logistics route and uses the two together as samples, which can effectively avoid clustering errors or clustering anomalies that are easily caused by using a single data as a sample. Then, the transaction volume data and the business feature data are clustered, and the different common features of the transaction volume data and the business feature data are extracted through clustering. The samples containing the common features are classified into one category, so that cluster cluster data with a larger data volume than the original data volume can be obtained, avoiding uneven sample distribution and limited sample size. Then, based on different cluster cluster data corresponding to different common features, a logistics route generation strategy is constructed. When the logistics route generation strategy is used to generate a target logistics route corresponding to the input target logistics itinerary, the target logistics route can be matched according to the common features corresponding to the target logistics itinerary, thereby improving the accuracy of the target logistics route to a certain extent.

[0077] The following continues Figure 1 The specific implementation methods of each step of the illustrated embodiment are described in detail:

[0078] In this application, the logistics route generation strategy can be determined in two ways: active and passive.

[0079] Active determination method. Taking the server as an example, the device for determining the logistics route generation strategy can establish a connection with the business platform of the logistics company, so as to obtain the transaction volume data and business characteristic data of the logistics routes put into the market from the business platform on a regular or real-time basis, and further retrieve the transaction volume data and business characteristic data of other logistics routes configured by relevant staff.

[0080] When the logistics routes and their transaction volume data and business characteristic data are updated, the method for determining the logistics route generation strategy of the present application can be triggered to update the logistics route generation strategy to obtain a logistics route generation strategy that meets the latest actual situation.

[0081] Passive determination method. Taking the case where the device for determining the logistics route generation strategy is a laptop computer, the relevant application installed on the laptop computer can receive the transaction volume data and business characteristic data of a series of logistics routes input by the staff, and then trigger the method for determining the logistics route generation strategy of this application based on the confirmation of the staff, and start to determine the corresponding logistics route generation strategy based on the transaction volume data and business characteristic data of this series of logistics routes.

[0082] Among them, the transaction volume data can be understood as the sales volume of the logistics route. The transaction volume data can be the transaction volume data corresponding to the actual business behavior after the logistics route is put into the market, or it can also be the configured transaction volume data, which is the transaction volume data configured for some logistics routes that are under preparation, have not been put into the market or have just been put into the market and have no actual transaction volume or lack of actual transaction volume;

[0083] Business feature data is used to indicate the business features matched by the logistics route in the preset different business feature sets. The business feature refers to a concept abstracted from the common characteristics of the logistics business of the logistics route. Similar to the transaction volume, the business feature data can be the business feature data corresponding to the actual business behavior of the logistics route after it is put on the market, or it can also be the configured business feature data, for some logistics routes that are under preparation, have not been put on the market or have just been put on the market, and the business feature data configured for these logistics routes that have no actual transaction volume or lack actual transaction volume.

[0084] In an exemplary implementation, the service characteristic data may specifically include at least one of line characteristic data, service scenario characteristic data, date characteristic data, and quotation characteristic data.

[0085] Among them, the route characteristic data is used to indicate the route characteristics of the logistics route itself, such as the starting logistics nodes and their respective regions, the target logistics nodes and their respective regions, the logistics nodes along the way and their respective regions, the range of longitude and latitude changes, the range of altitude changes, the route length and other characteristics.

[0086] Business scenario feature data is used to indicate the business scenario features of the logistics route, such as different levels of starting logistics nodes, heavy rain weather, sandstorm weather, haze weather, heavy cargo logistics routes, small piece logistics routes, document logistics routes, expedited logistics routes and other features.

[0087] Date feature data is used to indicate the date features of the logistics business of the logistics route during working hours, such as year, month, day, hour, minute, second, week, solar term, statutory holidays, e-commerce holidays and other date features.

[0088] The quotation characteristic data is used to indicate the feasible price quoted by the logistics company to the user who enjoys the service after taking into account the logistics cost of the logistics business of the logistics route. The user can be an ordinary consumer such as a sender or an e-commerce user, or the node of the logistics node itself.

[0089] In this application, when constructing logistics routes, both transaction volume data and business feature data are used as samples as the original data for constructing logistics route generation strategies, which can effectively avoid clustering errors or clustering anomalies that are easily caused by using a single data as a sample.

[0090] After obtaining the transaction volume data and business feature data of the logistics route, they can be clustered. During the clustering process, the two are classified according to preset common features or unknown common features to obtain multiple sets. Each common feature can correspond to a cluster data, among which a specific data can be classified not only in one cluster data, but also in another cluster at the same time. In this way, cluster data with a larger amount of data than the original data can be obtained, avoiding uneven sample distribution and limited sample size.

[0091] In the clustering process, in order to improve the validity of the data, data cleaning and verification may also be involved. After data cleaning and data verification, specific clustering processing can be carried out.

[0092] In another exemplary implementation, Figure 2A schematic diagram of a clustering process of the present application is shown. The clustering process involved in the present application may include the following steps S201 and S202:

[0093] Step S201, performing data cleaning on the transaction volume data and the business characteristic data respectively;

[0094] Exemplarily, data cleaning in this application refers to the preprocessing of data, deleting duplicate invalid data, discovering and correcting identifiable errors in the data, including checking data consistency, invalid values, missing values, avoiding erroneous data or abnormal data that conflicts with each other.

[0095] Data consistency check refers to checking whether the data meets the requirements based on the reasonable value range and mutual relationship of each variable, and discovering data that exceeds the normal range, is logically unreasonable or contradictory. For example, if a variable measured with a 1-7 scale has a value of 0, or a negative number appears in the quotation, it should be considered as exceeding the normal value range. Data consistency can usually be achieved by adjusting the data type, adjusting the date, adjusting the uppercase and lowercase, removing spaces, etc.

[0096] Common processing methods for invalid values ​​and missing values ​​are: estimation, whole case deletion, variable deletion, and pairwise deletion.

[0097] Estimation. The simplest way is to use the sample mean, median or mode of a variable to replace invalid values ​​and missing values. This method is simple, but it does not fully consider the existing information in the data and the error may be large. Another way is to estimate through correlation analysis or logical inference between variables based on the respondents' answers to other questions. For example, the business scenario of a logistics route may be related to logistics nodes of certain levels, and the possibility of having heavy cargo logistics business can be inferred based on the node level of the combined logistics nodes.

[0098] The whole case deletion is to remove samples with missing values. Since many questionnaires may have missing values, this approach may result in a significant reduction in the effective sample size and an inability to fully utilize the collected data. Therefore, it is only suitable for situations where key variables are missing or the proportion of samples with invalid or missing values ​​is very small.

[0099] Variable deletion. If a variable has many invalid and missing values ​​and is not particularly important for the research question, you can consider deleting the variable. This approach reduces the number of variables for analysis but does not change the sample size.

[0100] Paired deletion uses a special code (usually 9, 99, 999, etc.) to represent invalid and missing values, while retaining all variables and samples in the data set. However, only samples with complete answers are used in specific calculations, so different analyses will have different effective sample sizes due to different variables involved. This is a conservative approach that can maximize the retention of available information in the data set.

[0101] The processing of abnormal values ​​and outliers can usually be achieved through the 3Sigma principle (3σ principle, also known as Laida's rule). This discriminant processing principle and method is limited to the processing of sample data with normal or approximately normal distribution. Principle: Calculate the original data to obtain the standard deviation, and then determine an interval according to a certain probability. The probability of the value being distributed in (μ-σ, μ+σ) is 0.6827; the probability of the value being distributed in (μ-2σ, μ+2σ) is 0.9545; the probability of the value being distributed in (μ-3σ, μ+3σ) is 0.9973; therefore, the probability of the data being in (μ-3σ, μ+3σ) is less than 0.01, and the data can be confirmed as an abnormal value.

[0102] Step S202, clustering the cleaned transaction volume data and business feature data as samples to obtain a plurality of cluster data;

[0103] After data cleaning, clustering can be used to extract the different common features of transaction volume data and business feature data, and the samples containing common features can be classified into one category, so that cluster data with a larger amount of data can be obtained.

[0104] Among them, the clustering processing schemes that can be adopted are mainly the following.

[0105] K-means clustering. K-means clustering uses k as a parameter to divide n objects into k clusters, so that the similarity within a cluster is high, while the similarity between clusters is low. The processing of the k-means algorithm mainly includes randomly selecting k objects, each of which initially represents the average value or center of a cluster; for each remaining object, it is assigned to the nearest cluster according to its distance from the center of each cluster; and then the average value of each cluster is recalculated. This process is repeated until the criterion function converges and clustering is completed.

[0106] Agglomerative hierarchical clustering: The strategy of agglomerative hierarchical clustering is to first treat each object as a cluster, and then gradually merge the closest clusters to obtain larger and larger clusters until all objects are in one cluster or a certain termination condition is met, completing the clustering.

[0107] Self-Organizing Maps (SOM) clustering. SOM clustering assumes that there are some topological structures or orders in the input objects, and can achieve dimensionality reduction mapping from the input space (n-dimensional) to the output plane (2-dimensional). Its mapping has the property of preserving topological features and has a strong theoretical connection with actual brain processing. The processing of SOM clustering mainly includes SOM network initialization, assigning initial values ​​to the weights of each node in the output layer; randomly selecting input vectors from the input samples, and finding the weight vector with the smallest distance to the input vector; defining the winning unit, adjusting the weights in the vicinity of the winning unit to make it closer to the input vector; providing new samples and training; shrinking the neighborhood radius, reducing the learning rate, and repeating until it is less than the allowed value, completing clustering, and outputting the clustering results.

[0108] Fuzzy C Means (FCM) clustering. The FCM algorithm is an algorithm that uses membership to determine the degree of clustering to which each data point belongs. The main processing steps of FCM clustering include standardizing the data matrix; establishing a fuzzy similarity matrix and initializing the membership matrix; the algorithm starts iterating until the objective function converges to a minimum value and clustering is completed; according to the iteration results, the class to which the data belongs is determined by the final membership matrix, and the final clustering results are displayed.

[0109] After obtaining the preliminary clustering results, unimodal verification and single verification can be performed to further improve the validity of the data.

[0110] Single-peak verification. It can be understood that for the cluster data classified under each common feature, in this application, its distribution is considered to be roughly a trend of rising and then decreasing, which is similar to the normal distribution, that is, a "single-peak trend". If there is a "peakless" or "double-peak" situation, it can be considered that the current common feature is wrong or deviates from the actual common feature. The common feature can be ignored or deleted, or go back to the front and re-perform data cleaning and / or clustering processing.

[0111] by Figure 3 Taking the schematic diagram of a scenario of unimodal verification of the present application as an example, when distinguishing business scenarios and clustering business scenarios according to the X variable, combined with the transaction volume of each category of business scenarios, it can be seen that different unimodal trends are presented.

[0112] Taking a set of instance data as an example, in the unimodal test, the mixed Gaussian model can be used to fit the distribution of the business scenario, with parameters n_component = 2, weights_init = [0.1, 0.9], means_init = [[business scenario], [business scenario]];

[0113] The corresponding verification rules are as follows:

[0114] No double peaks: the lower peak mean is within the range of [0.714-1.286] the higher peak mean;

[0115] The double peaks are not obvious: 1. The mean of the low peak is outside the mean of the high peak [0.714-1.286]; 2. The weight of the low peak is less than or equal to 0.607 of the high peak;

[0116] The double peaks are obvious: 1. The mean of the low peak is outside the mean of the high peak [0.714-1.286]; 2. The weight of the low peak is greater than 0.607 of the high peak;

[0117] Normal distribution can be used to fit the transaction probability function when there is no double peak or the double peak is not obvious. If the double peak is obvious, data analysis and cleaning can be performed.

[0118] If all tests are passed, it can be considered that the division of business scenarios is relatively reasonable.

[0119] Single verification. It is understandable that in the transaction volume data and business feature data of the logistics route, due to the manual behavior of relevant staff, the literal meaning of some fields may not be consistent with their actual meaning or actual situation. In this case, the present application can also verify whether the transaction volume data and business feature data conform to the dynamic change trend of the actual logistics business through a single verification, and make corrections.

[0120] Taking the user as a logistics node to the node itself as an example, the logistics node can sign multiple logistics routes and contract the logistics transportation operations involved in these logistics routes. Among them, there is a "renewal" behavior in the application scenario. The logistics node can continue to contract the previous logistics route through the "renewal" behavior in the new business cycle or after the business time agreed upon in the historical logistics route expires. Part of the business content can also be adjusted. If the renewal situation is true, there will be several transaction tasks that are completed at the same price; if it is a non-renewal situation, the prices are more dispersed. Therefore, this application can use a single test to clean up the transaction data at the logistics node that is essentially a "renewal situation" but marked as a "non-renewal situation".

[0121] In another exemplary implementation, in the clustering process, taking the clustering process of data cleaning, single peak verification, and single verification mentioned above as an example, the specific clustering process is Figure 4 Based on a schematic diagram of a scenario of clustering processing in this application, as shown Figure 5 Another schematic diagram of a clustering process of the present application is shown, which may specifically include:

[0122] Step S501, clustering the cleaned transaction volume data, cleaned business scenario feature data, and cleaned date feature data as samples to obtain first cluster data;

[0123] In the present application, the clustering process can be mainly divided into two links. The first link targets the business scenario feature data and performs clustering based on the business scenario features. The second link targets the path feature data and quotation feature data and performs clustering based on the path features and quotation features. Through these two links, two-layer clustering processing is implemented to complete the clustering process in the present application.

[0124] Step S502, when the first cluster data passes the unimodal check and the single check, the first cluster data, the cleaned trading volume data, the cleaned path feature data, the cleaned quotation feature data and the cleaned date feature data are used as samples for clustering processing to obtain the second cluster data;

[0125] In the present application, based on the transaction volume data, business scenario feature data and date feature data, after the first stage of preliminary clustering of the business scenario features, unimodal verification and single verification can also be performed on them. If the unimodal verification or single verification fails, you can go back to the previous step and re-perform data cleaning and / or clustering to further improve the validity of the data.

[0126] Among them, the unimodal verification and single verification mentioned here can refer to the description of unimodal verification and single verification in the above content, and the details will not be repeated here.

[0127] When the first link outputs the first cluster data that has passed the unimodal check and the single check, the second link can continue to cluster the path characteristics and quotation characteristics based on the first cluster data, combined with the transaction volume data, path feature data, quotation feature data and date feature data.

[0128] Step S503: when the second cluster data passes the unimodal check and the single check, the second cluster data is confirmed as a plurality of cluster data.

[0129] Similar to the first stage, based on the first clustering cluster data, transaction volume data, path feature number, quotation feature data and date feature data, after the preliminary clustering of the path features and quotation features in the second stage, they can also be subjected to unimodal verification and single verification. If they fail the unimodal verification or single verification, they can go back to the previous stage and re-perform data cleaning and / or clustering to further improve the validity of the data.

[0130] Among them, the unimodal verification and single verification mentioned here can refer to the description of unimodal verification and single verification in the above content, and the details will not be repeated here.

[0131] In this implementation, due to the use of two-layer clustering links, the first layer of clustering processing is performed on the business scenario features, and the second layer of clustering processing is performed on the path features and quotation features, thereby realizing a two-layer clustering fusion framework. Compared with directly clustering these three features, clustering is divided into two links, which not only facilitates the refined processing of clustering and improves the refinement and hierarchical effect of clustering, but also in practical applications, especially when the user is the node itself of the logistics node, because different business scenarios are classified first, and then the quotation features and path features are further classified based on different business scenarios, for the logistics node, more powerful and detailed data support can be obtained, which is convenient for continuing to match the corresponding paths and quotations under the condition of matching the business scenarios, and then determining the target logistics route corresponding to the logistics node, which is very beneficial for the automatic matching processing of logistics routes in the process of logistics nodes contracting logistics routes.

[0132] In yet another exemplary implementation, see Figure 6 Another flow chart of clustering processing of the present application is shown. The second step of clustering processing, that is, based on the first clustering cluster data, combined with the volume data, the path feature data, the quotation feature data and the date feature data, the process of clustering the path features and the quotation features may specifically include:

[0133] Step S601, when the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line characteristic data after data cleaning, and the date characteristic data after data cleaning are used as samples for clustering processing to obtain the first sub-cluster data, wherein the line characteristic data identifies the line direction;

[0134] In the present application, the second step of the clustering process mentioned above, that is, based on the first cluster cluster data, combined with the transaction volume data, path feature data, quotation feature data and date feature data, clustering is performed on the path features and quotation features, and the following can also be performed: Figure 7 Another scenario schematic diagram of clustering processing of the present application is shown, corresponding to path features and quotation features, which are divided into two aspects, one for path features and the other for quotation features.

[0135] In this application, for the clustering of path features, the feature of "flow direction" can also be introduced. The path feature can have the characteristics of direction on the basis of the basic numerical value. That is to say, some data in the line feature data can exist in the form of scalar, such as the starting logistics node and its region, the target logistics node and its region, the logistics node of the route and its region, the longitude and latitude change range, the altitude change range, and the line length;

[0136] Part of the route characteristic data can also exist in the form of vectors to reflect, indicate or identify the route direction of the corresponding logistics route. For example, a logistics node H that a logistics route passes through can also be marked with the previous logistics node G and the next logistics node H of the logistics node H in the logistics route to indicate the route direction involved in the logistics node H in the logistics route. Of course, in addition to adjacent logistics nodes, other logistics nodes passed through in the logistics route and even other logistics routes passing through the logistics node H can also be marked;

[0137] For another example, the route length is marked with the starting logistics node and the target logistics node, and even some logistics nodes or all logistics nodes along the way can be marked. Based on the marked logistics nodes, combined with the location, coordinates, identification (ID) and other identification information of these logistics nodes, the location information of these logistics nodes can be directly or indirectly extracted, so as to determine the overall or partial route direction of the corresponding logistics route.

[0138] Of course, in addition to determining the route direction of the logistics route by marking the relevant logistics nodes, the route direction of the logistics route can also be determined by directly marking the route trajectory, or other methods can be used to mark the route direction of the logistics route, which is not limited here.

[0139] After determining the route direction of the logistics route, in the process of clustering, new clustering features can be extracted in combination with the route direction, namely the short-distance flow direction feature in the present application. The short-distance flow direction feature can indicate the directional distance between logistics nodes. In combination with the short-distance flow direction feature, path division and path classification are performed on the basis of the first cluster cluster data, transaction volume data, route characteristic data and date characteristic data, so that the first sub-cluster cluster data obtained by clustering, each cluster cluster has a finer degree of precision. Due to the directionality between logistics nodes, it is more conducive to matching two logistics nodes with different distances and different directions in different business scenarios, so that the target logistics route obtained by stringing together multiple groups of two logistics nodes can have higher accuracy.

[0140] For example, the longitude and latitude information of the corresponding nodes of the logistics route and their corresponding flow directions can be obtained; the mileage, fuel cost, and total cost of vehicles of different tonnages fully loaded and empty between any two nodes are calculated respectively; the short-distance node set deptSet of each node is calculated using the total empty-driving cost; the starting node of the route is recorded as i, and the arrival node is recorded as j, then its short-distance flow direction set is:

[0141] L ij = {l mn |0.9*d(l ij}≤d(l mn)≤1.1*d(l ij ),m∈deptSet(i),n∈deptSet(j),m,n∈full network point set}, where l mn is a logistics route with starting point m and end point n, d(l mn ) is the mileage of the logistics route.

[0142] Step S602, clustering the first cluster data, the cleaned transaction volume data, the cleaned quotation feature data, and the cleaned date feature data as samples to obtain second sub-cluster data, wherein the quotation feature data indicates the route direction;

[0143] Similar to the route characteristic data marked with route direction in the above content, the quotation characteristic data can also be marked with route direction. Therefore, after determining the route direction of the logistics route, in the process of clustering, new clustering features can be extracted in combination with the route direction, that is, the near-price flow direction feature in the present application. The near-price flow direction feature can indicate the directional spacing between logistics nodes. In combination with the near-price flow direction feature, quotation division and quotation classification are performed on the basis of the first cluster cluster data, transaction volume data, quotation characteristic data and date characteristic data, so that the second sub-cluster cluster data obtained by clustering, each cluster cluster has a finer degree of precision. Due to the directionality between logistics nodes, it is more conducive to matching different quotations in different business scenarios and pairwise logistics nodes in different directions, so that the target quotation of the target logistics route obtained by stringing together multiple groups of two logistics nodes can have higher accuracy.

[0144] For the specific marking method and specific description of the line direction, please refer to the description of the line characteristic data marked with the line direction in the above content, and the details will not be repeated here.

[0145] For example, the quotation clustering can be performed on the month (peak month & off-peak month), with the median quotation of the month as the clustering object: the median price of the earliest month is selected as the initial cluster center in chronological order; the distance between each object and each seed cluster center is calculated in chronological order, if the distance is less than 1.05*the mean of all objects and greater than 0.95*the mean of all objects, the object is assigned to the cluster center closest to it, and each time a sample is assigned, the cluster center of the cluster will be recalculated according to the existing objects in the cluster; otherwise, a new cluster center is created. A cluster center is created, and the two parts of the process are repeated until all objects are assigned and clustering is completed; in all month classes, the transaction prices of any two route tasks are subjected to levene variance test and double sample mean t test respectively. If both tests pass, they are marked as close price routes in the month class; if any test fails, they are marked as non-close price routes in the month class; for any two routes, if they have been close price routes to each other once in all month classes and have never been non-close price routes to each other, they are marked as close price routes to each other.

[0146] Step S603: confirming the first sub-cluster cluster data and the second sub-cluster cluster data as the second cluster cluster data.

[0147] After clustering the path features and the quotation features respectively, the first sub-clustering cluster data and the second sub-clustering cluster data obtained can be used as the second clustering cluster data outputted in the second step of the clustering process of the present application.

[0148] When the second cluster cluster data passes the unimodal check and the single check, it can be used as the final multiple cluster cluster data. If it fails, the second cluster cluster data that fails the unimodal check or the single check can be returned to the front and re-perform data cleaning and / or clustering processing.

[0149] After obtaining a plurality of cluster data, for example, the second cluster data mentioned above, a logistics route generation strategy can be constructed based on these cluster data. The logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

[0150] Exemplarily, the logistics route generation strategy can be implemented through a neural network model. Specifically, the second cluster data can be used as a training sample and input into an initialized neural network model for forward propagation. The loss function is then calculated based on the target logistics route matched by the neural network model. Backward propagation is performed based on the loss function to adjust the parameters of the neural network model. The training can be completed through multiple forward and backward propagations to meet the training requirements in terms of number of training times, training accuracy, etc. The trained neural network model can be used as a logistics route generation model, that is, it can be used as a logistics route generation strategy.

[0151] As another exemplary implementation, in the process of constructing a logistics route generation strategy, refer to Figure 7 A flow chart of a logistics route generation strategy constructed by the present application may include:

[0152] Step S701, extracting distribution values ​​of logistics routes with different route quotations from logistics routes with the same business scenario, date and route length from multiple cluster data as corresponding user acceptance values;

[0153] In the previous clustering process, corresponding to the logistics nodes in the line feature data, the clustering cluster data obtained by the clustering process can be presented based on the logistics routes strung together by two or more logistics nodes. For example, the logistics route in the clustering cluster data of X1 business scenario, B1 date, and C line length is the shortest logistics route formed by only 2 logistics nodes. Of course, the logistics route in the clustering cluster data of X1 business scenario, B1 date, and C line length can also be a logistics route formed by 3, 4, or other numbers of logistics nodes, which can be adjusted according to actual needs.

[0154] In this case, when constructing the logistics route generation strategy, after clustering the transaction volume data and business feature data of the logistics route, the logistics route generation strategy is constructed based on the obtained cluster data and the initial logistics route as a unit. In the logistics route generation strategy, the target logistics route corresponding to the business scenario, date, route feature, and route quotation can be matched according to the input logistics itinerary.

[0155] Furthermore, in the clustering process, each route in the initial logistics route can be decomposed into different sub-logistics routes according to the specific route characteristics such as logistics nodes and route directions in the route characteristic data. For example, if the X logistics route contains N logistics nodes, it can be decomposed into (N-1) sub-logistics routes, each of which is the shortest logistics route formed by 2 logistics nodes. Then, the logistics route generation strategy is constructed based on these sub-logistics routes. In the logistics route generation strategy, multiple sub-logistics routes corresponding to business scenarios, dates, route characteristics, and route quotations can be matched according to the input logistics itinerary, and these sub-logistics routes are strung together to form a complete logistics route as the target logistics route corresponding to the input logistics itinerary.

[0156] In the process of constructing the logistics route generation strategy, the user acceptance value can also be introduced. The user acceptance value is used to indicate the user's acceptance of the logistics route. The higher the user acceptance value, the higher the user's acceptance of the logistics route. Combined with the detection of the user acceptance value, the target logistics route generated by the logistics route generation strategy according to the input logistics itinerary can have a higher user acceptance value. The higher the user acceptance value of the target logistics route, the more likely the user is to choose the target logistics route, and the lower the possibility that the user continues to adjust the target logistics route. Therefore, it can effectively reduce the situation where the user adjusts the target logistics route to a certain extent, and improve the accuracy of the route.

[0157] Specifically, in the cluster data, the distribution values ​​of logistics routes with different route quotations in logistics routes with the same business scenarios, the same dates and the same route lengths can be extracted as the corresponding user acceptance values. The distribution values ​​can also be understood as probability density.

[0158] Step S702, building a logistics route set based on business scenarios, dates, route lengths, route quotations, and user acceptance values ​​as the characteristics, as a logistics route generation strategy. The logistics route set includes candidate logistics routes with different business scenarios, different dates, different route lengths, different route quotations, and different user acceptance values.

[0159] After obtaining the data of logistics routes for different business scenarios, dates, route lengths and user reception values, the data of these logistics routes can be used as a basis and these logistics routes can be used as candidate logistics routes to form a logistics route set, that is, the logistics route generation strategy in this application.

[0160] After obtaining the logistics route generation strategy, it can be put into practical application. The corresponding target logistics route can be generated according to the target logistics itinerary input by the user. For example, the client can generate a target logistics itinerary based on the user operation triggered by the logistics itinerary operation options displayed by the user on the client. The client then reports the target logistics itinerary to the server, and the server generates and feeds back the corresponding target logistics route based on the logistics route generation strategy generated by this application.

[0161] In particular, after building a logistics route set, multiple candidate logistics routes can be sorted in the logistics route set according to the size of the corresponding expected benefits.

[0162] In this way, in subsequent practical applications, when matching the target logistics routes corresponding to the user's target logistics itinerary through the logistics route set, if the user acceptance value of the matched candidate logistics routes reaches the threshold, the server can send these target logistics routes whose user acceptance values ​​reach the threshold to the client, and for the candidate logistics routes whose user acceptance values ​​do not reach the threshold, the server can actively make adjustments to improve the user acceptance values ​​of these candidate logistics routes. For example, on the basis of meeting the logistics route costs, adjust their quotations, including downward or upward adjustments, which can improve the user acceptance value to a certain extent, among which the downward adjustment is generally used.

[0163] Therefore, if multiple candidate logistics routes are sorted according to the expected returns in the logistics route set in the early stage, the server can give priority to logistics routes with high expected returns when making price adjustments later, and adjust the quotations to the quotations of these logistics routes.

[0164] Taking the user as the logistics node itself as an example, if the target logistics route obtained based on the reported target logistics itinerary has a high expected return, or even the highest expected return, it is obviously helpful in this case for the logistics node to contract the target logistics route as soon as possible, thereby achieving the purpose of improving the accuracy of the target logistics route.

[0165] In order to facilitate better implementation of the method for determining the logistics route generation strategy provided by the present application, the present application also provides a device for determining the logistics route generation strategy.

[0166] See also Figure 8 , Figure 8 This is a schematic diagram of a structure of a device for determining a logistics route generation strategy of the present application. In the present application, the device for determining a logistics route generation strategy 800 may specifically include the following structure:

[0167] The transceiver unit 801 is used to obtain the transaction volume data of the logistics route and the business characteristic data of the logistics route;

[0168] The processing unit 802 is used to cluster the transaction volume data and the business characteristic data as samples to obtain multiple cluster data; based on the multiple cluster data, a logistics route generation strategy is constructed, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to the input target logistics itinerary.

[0169] In an exemplary implementation, the processing unit 802 is specifically configured to:

[0170] Perform data cleaning on transaction volume data and business feature data respectively;

[0171] The cleaned transaction volume data and business feature data are used as samples for clustering processing to obtain multiple cluster data.

[0172] In yet another exemplary implementation, the service characteristic data includes at least one of line characteristic data, service scenario characteristic data, date characteristic data, and quotation characteristic data.

[0173] In yet another exemplary implementation, the processing unit 802 is specifically configured to:

[0174] The transaction volume data after data cleaning, the business scenario feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain first cluster cluster data;

[0175] When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the path feature data after data cleaning, the quotation feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain the second cluster data;

[0176] When the second cluster data passes the unimodal check and the single check, the second cluster data is confirmed as a plurality of cluster data.

[0177] In yet another exemplary implementation, the processing unit 802 is specifically configured to:

[0178] When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain the first sub-cluster data, wherein the line feature data identifies the line direction;

[0179] The first cluster data, the cleaned trading volume data, the cleaned quotation feature data and the cleaned date feature data are clustered as samples to obtain the second sub-cluster data, wherein the quotation feature data indicates an upper line trend;

[0180] The first sub-cluster cluster data and the second sub-cluster cluster data are confirmed as the second cluster cluster data.

[0181] In yet another exemplary implementation, the processing unit 802 is specifically configured to:

[0182] From multiple cluster data, the distribution values ​​of logistics routes with different route quotations in logistics routes with the same business scenarios, the same dates and the same route lengths are extracted as the corresponding user acceptance values;

[0183] Based on the business scenario, date, route length, route quotation and user acceptance value, a logistics route set is constructed as a logistics route generation strategy, wherein the logistics route set includes candidate logistics routes with different business scenarios, different dates, different route lengths, different route quotations and different user acceptance values.

[0184] In yet another exemplary implementation, the processing unit 802 is further configured to:

[0185] In the logistics route set, multiple candidate logistics routes are sorted according to the size of the corresponding expected benefits.

[0186] This application also provides a device for determining a logistics route generation strategy, see Fig. 9 , Fig. 9 A schematic diagram of a structure of a device for determining a logistics route generation strategy of the present application is shown. Specifically, the device for determining a logistics route generation strategy of the present application includes a processor 901, and the processor 901 is used to execute a computer program stored in a memory 902 to implement the following Figures 1 to 7 Corresponding to the steps of the method for determining the logistics route generation strategy in any embodiment; or, the processor 901 is used to execute the computer program stored in the memory 902 to implement the following Figure 8 The functions of each unit in the corresponding embodiment, such as Figure 8 The hardware structure corresponding to the transceiver unit 801 is the input and output device 903, the hardware structure corresponding to the processing unit 802 is the processor 901, and the memory 902 is used to store the processor 901 to execute the above Figures 1 to 7 A computer program required for the method for determining the logistics route generation strategy in any embodiment.

[0187] Exemplarily, the computer program may be divided into one or more modules / units, one or more modules / units are stored in the memory 902, and executed by the processor 901 to complete the present application. One or more modules / units may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the computer device.

[0188] The device for determining the logistics route generation strategy may include, but is not limited to, the processor 901 and the memory 902. Those skilled in the art will appreciate that the illustration is merely an example of the device for determining the logistics route generation strategy, and does not constitute a limitation on the device for determining the logistics route generation strategy, and may include more or fewer components than shown in the illustration, or a combination of certain components, or different components, for example, the device may also include input and output devices, network access devices, buses, etc., and the processor 901, the memory 902, the input and output devices, and the network access devices, etc. are connected via a bus.

[0189] The processor 901 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the device for determining the logistics route generation strategy, and uses various interfaces and lines to connect various parts of the entire device.

[0190] The memory 902 can be used to store computer programs and / or modules. The processor 901 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 902 and calling the data stored in the memory 902. The memory 902 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data (such as audio data, video data, etc.) created by the use of the device determined according to the logistics route generation strategy. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0191] When the processor 901 is used to execute the computer program stored in the memory 902, the following functions can be implemented:

[0192] Obtaining transaction volume data and business characteristic data of logistics routes;

[0193] The transaction volume data and the business characteristic data are clustered as samples to obtain multiple cluster data;

[0194] Based on multiple cluster data, a logistics route generation strategy is constructed, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

[0195] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described logistics route generation strategy determination device, equipment and corresponding units can refer to the following. Figures 1 to 7 The description of the method for determining the logistics route generation strategy in any embodiment will not be repeated here.

[0196] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0197] To this end, the present application provides a computer-readable storage medium, in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the present application as follows: Figures 1 to 7 For the steps in the method for determining the logistics route generation strategy in any embodiment, the specific operations can be referred to as follows: Figures 1 to 7 The description of the method for determining the logistics route generation strategy in any embodiment will not be repeated here.

[0198] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0199] Due to the instructions stored in the computer-readable storage medium, the present application can be executed. Figures 1 to 7 Corresponding to the steps in the method for determining the logistics route generation strategy in any embodiment, therefore, the present application can be implemented as follows Figures 1 to 7 The beneficial effects that can be achieved by the method for determining the logistics route generation strategy in any embodiment are detailed in the previous description and will not be repeated here.

[0200] The above is a detailed introduction to the method, device, equipment and computer-readable storage medium for determining the logistics route generation strategy provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for determining a logistics route generation strategy, characterized in that: The method comprises: Acquire transaction volume data of a logistics route and business characteristic data of the logistics route; the business characteristic data includes at least one of route characteristic data, business scenario characteristic data, date characteristic data, and quotation characteristic data; Performing data cleaning on the transaction volume data and the business characteristic data respectively; The cleaned transaction volume data, the cleaned business scenario feature data, and the cleaned date feature data are used as samples for clustering processing to obtain first cluster data; When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line characteristic data after data cleaning, the quotation characteristic data after data cleaning, and the date characteristic data after data cleaning are used as samples for clustering processing to obtain second cluster data; When the second cluster data passes the unimodal check and the single check, confirming the second cluster data as a plurality of cluster data; Based on the multiple cluster data, a logistics route generation strategy is constructed, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

2. The method according to claim 1, characterized in that: When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line characteristic data after data cleaning, the quotation characteristic data after data cleaning, and the date characteristic data after data cleaning are used as samples for clustering processing to obtain the second cluster data, including: When the first cluster data passes the unimodal check and the single check, the first cluster data, the transaction volume data after data cleaning, the line feature data after data cleaning, and the date feature data after data cleaning are used as samples for clustering processing to obtain first sub-cluster data, wherein the line feature data identifies the line direction; The first cluster data, the cleaned trading volume data, the cleaned quotation feature data and the cleaned date feature data are clustered as samples to obtain second sub-cluster data, wherein the quotation feature data is marked with the above line direction; The first sub-cluster cluster data and the second sub-cluster cluster data are confirmed as the second cluster cluster data.

3. The method according to claim 1 or 2, characterized in that: The logistics route generation strategy constructed based on the plurality of cluster data includes: Extracting, from the plurality of cluster data, distribution values ​​of logistics routes with different route quotations in logistics routes with the same business scenario, the same date and the same route length as corresponding user acceptance values; Based on the business scenario, date, route length, route quotation and user acceptance value, a logistics route set is constructed as a logistics route generation strategy, wherein the logistics route set includes candidate logistics routes with different business scenarios, different dates, different route lengths, different route quotations and different user acceptance values.

4. The method according to claim 3, characterized in that: After building a logistics route set based on the business scenario, date, route length, route quotation and user acceptance value as a logistics route generation strategy, the method further includes: In the logistics route set, the plurality of candidate logistics routes are sorted according to the size of the corresponding expected benefits.

5. A device for determining a logistics route generation strategy, characterized in that: The device comprises: A transceiver unit, used to obtain transaction volume data of a logistics route and business characteristic data of the logistics route; the business characteristic data includes at least one of route characteristic data, business scenario characteristic data, date characteristic data and quotation characteristic data; A processing unit is used to perform data cleaning on the transaction volume data and the business characteristic data respectively; cluster the cleaned transaction volume data, the cleaned business scenario characteristic data and the cleaned date characteristic data as samples to obtain first cluster cluster data; when the first cluster cluster data passes the unimodal check and the single check, cluster the first cluster cluster data, the cleaned transaction volume data, the cleaned route characteristic data, the cleaned quotation characteristic data and the cleaned date characteristic data as samples to obtain second cluster cluster data; when the second cluster cluster data passes the unimodal check and the single check, confirm the second cluster cluster data as multiple cluster cluster data; based on the multiple cluster cluster data, construct a logistics route generation strategy, wherein the logistics route generation strategy is used to generate a corresponding target logistics route according to an input target logistics itinerary.

6. A device for determining a logistics route generation strategy, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program, and the processor executes the method according to any one of claims 1 to 4 when calling the computer program in the memory.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the method according to any one of claims 1 to 4.

Images