A method, device and storage medium for estimating the average volume of express delivery

Abstract

This scheme discloses a method, device and storage medium for estimating the average volume of delivered express parcels, the method comprising: based on the measurement data of the express parcels to be delivered from a certain origin distribution location to a certain destination distribution location on a certain day, preliminarily derive the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location on that day; establish a norm approximation model to solve the correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location; based on the correction coefficient and the preliminarily derived average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location on that day, obtain the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location on that day. The method of the present invention can scientifically estimate the product structure data between each distribution point in the express delivery network.

Description

Technical Field

[0001] The present invention relates to the field of logistics distribution, and in particular to a method, a device and a storage medium for estimating the average volume of delivered express items. Background Art

[0002] In logistics and distribution, product structure is used to represent the number of express parcels that can be accommodated in one cubic meter of volume. As a basic parameter used in logistics and distribution, product structure represents the mapping of the number of express parcels to volume, and its numerical value is equal to the inverse of the average volume of the express parcels.

[0003] There are usually two types of product structures. One is the line segment product structure. A line segment refers to a section between two distribution locations that are directly accessible by vehicles. A line segment usually carries express parcels from multiple origin distribution locations and destination distribution locations. For example, there is a shuttle bus running on the Jinan-Hangzhou line segment, which may carry express parcels from Jinan to Hangzhou, Jining to Hangzhou, and Jinan to Jiaxing. The line segment product structure is the reciprocal of the average volume of all express parcels carried by the line segment. The other is the origin-destination product structure, which is the reciprocal of the average volume of all express parcels from the origin distribution location to the destination distribution location.

[0004] In the express logistics network, since most express shipments are dumped, the loading of most vehicles is restricted by volume. We can usually easily obtain the quantity of each origin distribution to the destination distribution. If we want to calculate the number of vehicles required, we need to convert the quantity of pieces into volume through product structure. Since it is difficult to obtain the exact volume of each express shipment, and express shipments with different origins and destinations usually have different product structures, we need a scientific method to estimate the product structure data between each distribution in the express network. Summary of the invention

[0005] One purpose of this solution is to provide a method for estimating the average volume of delivered express parcels.

[0006] Another object of the present invention is to provide a device for estimating the average volume of delivered express items.

[0007] The third objective of this solution is to provide a storage medium.

[0008] To achieve the above objectives, this plan is as follows:

[0009] A method for estimating an average volume of a delivery package, the method comprising:

[0010] Based on the obtained measurement data of express parcels to be delivered from a certain origin distribution location to a certain destination distribution location on a certain day, the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on that day is preliminarily obtained;

[0011] A norm approximation model is established to solve a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0012] Based on the correction coefficient and the preliminary obtained average volume of express parcels to be delivered from the originating distribution location to the destination distribution location on that day, the average volume of express parcels to be delivered from the originating distribution location to the destination distribution location on that day is obtained.

[0013] Preferably, the establishment of a norm approximation model to solve the correction coefficient for correcting the average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location includes:

[0014] Based on the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the first-level correction coefficient;

[0015] Based on the first-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the second-level correction coefficient;

[0016] Based on the first-level correction coefficient, the second-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the third-level correction coefficient;

[0017] The result of multiplying the first-level correction coefficient, the second-level correction coefficient and the third-level correction coefficient is used as a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0018] To reduce the number of parameters, in each level of fitting, the correction factor for deliveries with a number of parcels not exceeding 1,000 is always fixed to 1.

[0019] Preferably, the method further comprises screening out fully loaded vehicles based on the acquired express measurement data;

[0020] The screening out of fully loaded vehicles includes:

[0021] Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping;

[0022] Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained;

[0023] Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group;

[0024] All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered;

[0025] After the second grouping, sort each group of transport vehicles from most to least according to the number of express parcels loaded on each transport vehicle, and select the transport vehicles ranked in the first 30% of the group into the second screening group;

[0026] The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group;

[0027] or

[0028] After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked in the first 30% of the group of transport vehicles are selected into the third screening group;

[0029] The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group;

[0030] Combine the vehicles of the first screening group, the second screening group and the third screening group, and group all the combined vehicles for the third time according to the rated loading volume and the transport line segment of each vehicle, find out the vehicle with the least number of parcels or the smallest weight of parcels loaded in each group after the third grouping, and use the number of parcels or the weight of parcels loaded on the vehicle as the screening threshold of the group, and confirm the vehicle in the group whose number of parcels or the weight of parcels loaded exceeds the screening threshold as a fully loaded vehicle.

[0031] Preferably, the first grouping is to group the express parcels to be delivered having the same origin distribution location and destination distribution location into one group;

[0032] The second grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group;

[0033] The third grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group.

[0034] Preferably, the measurement data of the express delivery includes the scanning date of the express delivery, the origin distribution location, the destination distribution location, and the measured volume of the express delivery.

[0035] In a second aspect, a device for estimating an average volume of delivered express items is provided, comprising:

[0036] A data acquisition unit, based on the acquired measurement data of express parcels to be delivered from a certain origin distribution location to a certain destination distribution location on a certain day, preliminarily obtains an average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on that day;

[0037] A fitting unit is used to establish a norm approximation model and solve a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0038] Based on the correction coefficient and the preliminary obtained average volume of express parcels to be delivered from the originating distribution location to the destination distribution location on that day, the average volume of express parcels to be delivered from the originating distribution location to the destination distribution location is obtained.

[0039] Among them, solving the final correction coefficient through the norm approximation model includes:

[0040] Based on the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the first-level correction coefficient;

[0041] Based on the first-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the second-level correction coefficient;

[0042] Based on the first-level correction coefficient, the second-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the third-level correction coefficient;

[0043] The result of multiplying the first-level correction coefficient, the second-level correction coefficient and the third-level correction coefficient is used as a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0044] To reduce the number of parameters, in each level of fitting, the correction factor for deliveries with a number of parcels not exceeding 1,000 is always fixed to 1.

[0045] Preferably, the device further comprises a data screening unit for screening out fully loaded vehicles, comprising the following steps:

[0046] Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping;

[0047] Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained;

[0048] Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group;

[0049] All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered;

[0050] After the second grouping, sort each group of transport vehicles from most to least according to the number of express parcels loaded on each transport vehicle, and select the transport vehicles ranked in the first 30% of the group into the second screening group;

[0051] The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group;

[0052] or

[0053] After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked in the first 30% of the group of transport vehicles are selected into the third screening group;

[0054] The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group;

[0055] Combine the transport vehicles of the first screening group, the second screening group and the third screening group, and group all the combined transport vehicles for the third time according to the rated loading volume and transport line segment of each transport vehicle, find the transport vehicle with the least number of express parcels or the smallest weight of express parcels loaded in each group after the third grouping, and use the number of express parcels or the weight of express parcels loaded by the transport vehicle as the screening threshold of the group, and confirm the transport vehicle in the group whose number of express parcels or the weight of express parcels loaded exceeds the screening threshold as a fully loaded vehicle;

[0056] The first grouping is to group the express parcels to be delivered with the same origin and destination.

[0057] The second grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group;

[0058] The third grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group.

[0059] Preferably, the measurement data of the express delivery includes the scanning date of the express delivery, the origin distribution location, the destination distribution location, and the measured volume of the express delivery.

[0060] According to a third aspect, a computer-readable storage medium is provided, on which a computer program is stored, wherein the program implements the steps of the above method when executed by a processor.

[0061] The beneficial effects of this program are as follows:

[0062] This solution provides a systematic method to efficiently and accurately calculate the product structure data from each origin distribution location to each destination distribution location, so that vehicle resources can be arranged more accurately according to the number of express parcels to be processed. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the implementation of the present scheme, 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 scheme. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0064] Figure 1 A flow chart of a method for estimating the average volume of delivered express shipments;

[0065] Figure 2 Schematic diagram of a device for estimating the average volume of delivered express parcels. DETAILED DESCRIPTION

[0066] The implementation method of this solution will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this solution, rather than an exhaustive list of all the embodiments. It should be noted that the embodiments in this solution and the features in the embodiments can be combined with each other without conflict.

[0067] The terms "first", "second", etc. (if any) in the specification and claims and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way are interchangeable where appropriate so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0068] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0069] A large part of the logistics network is e-commerce parcels. Due to the industrial agglomeration effect of the origin and similar consumption habits of the destination, and distribution usually only covers a limited range (usually the city where the distribution is located and surrounding cities), express parcels sent from a certain origin distribution location to the same destination distribution location usually have similar characteristics, and the origin-destination product structure can better reflect the real situation.

[0070] The number of express shipments between distribution centers is huge and varies greatly, so the product structure is just a statistical number that represents the average size of express shipments between two locations. Express delivery networks usually have nearly 100 distribution centers, so the number of origin and destination product structures that need to be estimated is close to 10,000.

[0071] Logistics companies generally deploy a large number of DWS (Dimension Weigh Scanning) devices, which can use lasers to measure the volume of each express package. However, due to the irregular shape of express packages, the measured volume is not very accurate. In addition, the main purpose of estimating product structure is to determine the number of shuttle buses required by the number of pieces, and there will be certain extrusion deformation and gaps during the loading process. Therefore, even if the precise volume of each express package can be obtained, the product structure cannot be directly derived.

[0072] In the logistics network, the rated loading volume of each shuttle bus is known, as is the number of parcels loaded, and the origin and destination of each parcel. However, the actual loading volume of the shuttle bus is unknown, which poses a great challenge to the estimation of the product structure.

[0073] The existing method for calculating the product structure of a line segment is to divide the sum of the actual loading volumes of all buses on the line segment by the number of express parcels on the line segment. Since the actual loading volume of the buses is unknown, usually only fully loaded buses can be selected and their rated volumes can be used as the loading volumes. Based on the selected full buses and the loading details of the vehicles, the product structure of the line segment can be calculated.

[0074] The existing method for calculating the origin-destination product structure is to calculate the reciprocal of the average volume of all express shipments from the origin distribution location to the destination distribution location. Based on the selected full truck and vehicle loading details data, the least squares method is directly used to fit the average volume of each origin destination, and its reciprocal is used as the origin-destination product structure.

[0075] However, the method of calculating the segment product structure is affected by the express route. The route refers to the complete path from the origin distribution location to the destination distribution location. For example, the route from distribution location A to destination distribution location D is ABCD, which requires transit through two distribution locations B and C. Express shipments on a segment usually have multiple origin distributions and destination distributions. If the route of the express shipment changes, the product structure of the corresponding segment will also change significantly, making it difficult to apply it to future dates. The origin destination product structure is relatively stable. If the date difference is not far, the current product structure can be directly used in the future.

[0076] Moreover, the product structure of the segment cannot be calculated between two distribution locations that are not connected by shuttle buses. If two distribution locations are not directly connected by shuttle buses, it is impossible to estimate the volume of all express parcels on the segment, and there is no corresponding segment product structure.

[0077] The method for calculating the product structure of the origin and destination has too many fitting parameters. There are nearly 10,000 origins and destinations in total, but there are usually only a few thousand buses selected using one week's data, which is easy to cause overfitting. Although the error in the final line segment is not large, the difference between each flow direction and the true value may be large. Moreover, the calculation results for the same origin and destination for multiple days are unstable. The fitting is more accurate for the flow direction with a larger volume, but the deviation is larger for the flow direction with a smaller volume.

[0078] In view of the problems existing in the existing calculation methods, the inventors of this application propose a method for estimating the average volume of express delivery, such as Figure 1 As shown, the method includes:

[0079] S100 preliminarily obtains the average volume of the express parcels to be delivered from a certain originating distribution location to a certain destination distribution location on a certain day based on the obtained measurement data of the express parcels to be delivered from the originating distribution location to the destination distribution location on that day;

[0080] S200 establishes a norm approximation model to solve a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0081] S300 obtains the average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location based on the correction coefficient and the preliminarily obtained average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location on that day.

[0082] In step S100, express parcels to be delivered are grouped based on the originating distribution location and the destination distribution location, and the average volume of each group of express parcels to be delivered after grouping is obtained.

[0083] In step S200,

[0084] Solving for the correction factor includes:

[0085] Based on the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the first-level correction coefficient;

[0086] Based on the first-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the second-level correction coefficient;

[0087] Based on the first-level correction coefficient, the second-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the third-level correction coefficient;

[0088] The result of multiplying the first-level correction coefficient, the second-level correction coefficient and the third-level correction coefficient is used as a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0089] To reduce the number of parameters, in each level of fitting, the correction factor for deliveries with a number of parcels not exceeding 1,000 is always fixed to 1.

[0090] Meanwhile, step S200 also includes screening out fully loaded vehicles based on the obtained express measurement data, and the steps of screening out fully loaded vehicles are as follows:

[0091] Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping;

[0092] Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained;

[0093] Each transport vehicle may be loaded with multiple groups of express parcels to be delivered;

[0094] Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group;

[0095] All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered;

[0096] After the second grouping, each group of transport vehicles is sorted from most to least according to the number of express parcels loaded on each transport vehicle, and the transport vehicles ranked before 30% of the transport vehicles in the group are selected into the second screening group; among them, the transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group;

[0097] or

[0098] After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked before 30% of the transport vehicles in the group are selected into the third screening group; among them, the transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group;

[0099] Combine the vehicles of the first screening group, the second screening group and the third screening group, and group all the combined vehicles for the third time according to the rated loading volume and the transport line segment of each vehicle, find out the vehicle with the least number of parcels or the smallest weight of parcels loaded in each group after the third grouping, and use the number of parcels or the weight of parcels loaded on the vehicle as the screening threshold of the group, and confirm the vehicle in the group whose number of parcels or the weight of parcels loaded exceeds the screening threshold as a fully loaded vehicle.

[0100] The present invention is described below by means of specific embodiments.

[0101] Step 1: From the DWS device (automated parcel weighing, scanning and volume measurement equipment), obtain the measurement data of express parcels delivered from multiple origin distribution locations to various destination distribution locations. The following Table 1 shows the scan records of express parcels sent from a certain origin distribution location to a certain destination distribution location.

[0102] Table 1

[0103]

[0104]

[0105] After removing the outliers in the scanned data, the data is grouped according to (scanning date, origin and destination). For example, the express shipments in Table 1 are grouped into the same group, with Hangzhou Transshipment Center as the origin and Beijing Transshipment Center as the destination.

[0106] Calculate the average volume of each group of express parcels and preliminarily obtain the average volume of express parcels to be delivered each day.

[0107] Step 2: Filter fully loaded vehicles

[0108] Table 2 shows the measurement data of express parcels to be delivered loaded on a transport vehicle.

[0109] Table 2

[0110]

[0111] Since only fully loaded transport vehicles know the exact loading volume, it is necessary to filter out fully loaded vehicles from all loading data. The screening method is as follows:

[0112] Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping;

[0113] Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained;

[0114] Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group;

[0115] All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered;

[0116] After the second grouping, each group of transport vehicles is sorted from most to least according to the number of express parcels loaded on each transport vehicle, and the transport vehicles ranked before 30% of the transport vehicles in the group are selected into the second screening group; among them, the transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group;

[0117] or

[0118] After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked before 30% of the transport vehicles in the group are selected into the third screening group; among them, the transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group;

[0119] Combine the vehicles of the first screening group, the second screening group and the third screening group, and group all the combined vehicles for the third time according to the rated loading volume and the transport line segment of each vehicle, find out the vehicle with the least number of parcels or the smallest weight of parcels loaded in each group after the third grouping, and use the number of parcels or the weight of parcels loaded on the vehicle as the screening threshold of the group, and confirm the vehicle in the group whose number of parcels or the weight of parcels loaded exceeds the screening threshold as a fully loaded vehicle.

[0120] Step 3: Build a norm approximation model

[0121] Based on the various measurement data of express shipments on fully loaded vehicles, a norm approximation model is established. Since the volume change of express shipments on adjacent dates should be small, directly fitting the average volume of each origin destination may result in large differences in the results of adjacent dates. Therefore, the correction coefficient of the DWS measurement value is selected. Although the DWS measurement value is not accurate, the trend of the measurement value on different dates in the same direction can reflect the trend of the actual express shipment volume, that is, the DWS measurement value has both systematic errors and random errors. By fitting the correction coefficients of each transportation direction, the systematic error of the measurement value can be eliminated.

[0122] In order to avoid overfitting, the correction coefficients from all the originating distribution places to the destination distribution places are not directly fitted, but divided into three levels for fitting in sequence. First, the first-level correction coefficient is fitted based on the DWS measurement values ​​of each originating distribution place. At this time, there are less than 100 parameters that need to be fitted; then the correction coefficient from the originating distribution place to the province where the destination distribution place is located is fitted, that is, the second-level correction coefficient; finally, the correction coefficient from the originating distribution place to the destination distribution place is fitted, that is, the third-level correction coefficient. In order to further reduce the number of parameters, the correction coefficient for the express number not exceeding 1000 in each level is always fixed to 1. The number of variables in these three levels increases step by step, and the allowed range of variation of the correction coefficient decreases step by step. The final correction coefficient is equal to the product of the correction coefficients of the three levels, and the final product structure is equal to the reciprocal of the product of the correction coefficient and the average volume measured by DWS. The basic idea of ​​this method is that the originating distribution place plays a decisive role in the correction coefficient. For the same originating distribution place, the correction coefficients of different destination provinces and destination distribution places can only be fine-tuned.

[0123] By establishing a norm approximation model, that is, knowing the actual loading volume (volume) of each transport vehicle on each line segment and the number of express parcels loaded at each origin and destination, a set of optimal correction coefficients for the DWS measurement value is found, so that the error between the calculated volume and the actual loading volume is minimized while satisfying the upper and lower bound constraints. For a given level l, the established norm approximation model is shown in formula (1):

[0124]

[0125] The meanings of the symbols in the model represented by formula (1) are shown in Table 1. The objective function contains three terms: error term, regularization term, and slack variable penalty. || || is a certain norm, which means minimizing the sum of a certain norm of the error between the actual loading volume and the calculated volume. Generally, the 1-norm, that is, the sum of the absolute values ​​of the errors; or the 2-norm, that is, the sum of squared errors (in this case, it is a least squares problem). In order to prevent overfitting, L2 regularization is added to the objective function, that is, the difference between the correction coefficient to be fitted and the mean of the correction coefficient of the previous few days. The first constraint gives the upper and lower bounds of the correction coefficient, and the second and third constraints give the upper and lower bounds of the final product structure. In order to ensure feasibility, the slack variable s is introduced.

[0126] Table 1

[0127]

[0128]

[0129] In order to obtain better robustness and eliminate the influence of some abnormal data on the results, the 1-norm is used here. By introducing auxiliary variables, the absolute value target can be transformed into a linear constraint, as shown in formula (2), where u t This is a convex quadratic programming problem. The optimal solution can be directly obtained using the solver.

[0130]

[0131] Taking the express shipments from Zhejiang Province to Hebei Province and Henan Province as an example, according to step 1, the DWS scanning records (origin distribution, destination distribution) of all express shipments are grouped to obtain the average volume of each group. The data is shown in Table 4:

[0132] Table 4 Average volume of DWS measurements in various directions

[0133]

[0134] Table 5 Shuttle bus loading details

[0135]

[0136] Table 5 is the loading details of the shuttle buses from Zhejiang Province to Hebei Province and Henan Province. Taking Table 5 as an example, the screening process for fully loaded transport vehicles is as follows:

[0137] First, calculate the approximate loading volume of each vehicle. Taking DBGJ003603 as an example, it carries 892 pieces of express mail from Hangzhou to Luoyang and 17914 pieces of express mail from Yiwu to Luoyang. From Table 4, we know that the average volume of the express mail from Hangzhou to Luoyang is 0.00804m 3 The average volume of Yiwu-Luoyang express is 0.00405m 3, so the approximate loading volume of the transport vehicle is 892×0.00804+17914×0.00405=79.72m 3 . After calculation, the ratios of the approximate loading volumes of the three transport vehicles to the rated cubic meters are all less than 0.95, so the first screening group is empty.

[0138] Take the Yiwu-Luoyang line section in Table 5 and the rated loading volume as 135m 3 Taking the grouping as an example, because the vehicles with the same rated loading volume and the same line segment are grouped into one group, the group contains the following three transport vehicles, namely DBGJ003603, DBGJ003604, and QYGJ013652;

[0139] Sort by the number of express parcels loaded on the transport vehicles in the group, and select the transport vehicles ranked in the top 30% of the group. When selecting the transport vehicles, select according to the principle of rounding up. Since there are only 3 transport vehicles in the group, 3×30%=0.9, which is less than one vehicle, but according to rounding up, one vehicle can be selected, that is, the transport vehicle with the largest number of express parcels loaded is selected as the vehicle to enter the second screening group, that is, select DBGJ003603 to enter the second screening group;

[0140] Similarly, the third screening group selects the transport vehicle with the largest loading weight, which is still DBGJ003603.

[0141] The transport vehicles of the first screening group, the second screening group and the third screening group are combined. There is only one transport vehicle in the combined set, namely DBGJ003603, whose minimum number of loaded express parcels and weight of express parcels are 18806 and 8675.76 kg respectively. The number of express parcels and weight of express parcels loaded by this vehicle are used as the screening threshold, that is, the rated loading volume on all Yiwu-Luoyang segments is 135m 3 Among the vehicles, as long as the number of express parcels loaded or the weight of express parcels is greater than or equal to the threshold, it is confirmed as a fully loaded vehicle. In this embodiment, there is only one fully loaded vehicle, DBGJ003603. The fully loaded vehicles finally obtained are shown in Table 6.

[0142] Table 6 The fully loaded vehicles finally screened

[0143]

[0144] First, use formula (2) to calculate the correction coefficient of the first level, that is, the correction coefficient of Hangzhou and Yiwu. At this time, v lod is equal to the average volume of DWS measurements from o to d. After substituting the data, an optimization solver (such as gurobi, cplex, etc.) can be used to solve the correction coefficients of Hangzhou and Yiwu. Then continue to use formula (2) to calculate the correction coefficient of the second level. At this time, v lod= is equal to the average volume of DWS measurements from o to d multiplied by the correction coefficient of o in the first level, and the correction coefficients from Hangzhou to Hebei Province and Henan Province, and from Yiwu to Hebei Province and Henan Province can be obtained; finally, the correction coefficient of the third level is calculated using formula (2), at this time v lod It is equal to the average volume of DWS measurements from o to d multiplied by the correction factor of o in the first level, and then multiplied by the correction factor of the province from o to d in the second level. Finally, the average volume from o to d is equal to the average volume of DWS measurements from o to d multiplied by the correction factor of o obtained in the first level, and then multiplied by the correction factor of the province from o to d obtained in the second level, and finally multiplied by the correction factor of o to d obtained in the third level.

[0145] like Figure 2 As shown, this embodiment also provides a device 10 for estimating the average volume of delivered express items, the device comprising:

[0146] The data acquisition unit 11 preliminarily obtains the average volume of the express parcels to be delivered from a certain origin distribution location to a certain destination distribution location on a certain day based on the measurement data of the express parcels to be delivered from the origin distribution location to the destination distribution location on that day;

[0147] A fitting unit 13 is used to establish a norm approximation model to solve a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0148] Based on the correction coefficient and the preliminarily obtained average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location on that day, the average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location is obtained. Wherein, solving the final correction coefficient by the norm approximation model includes:

[0149] Based on the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the first-level correction coefficient;

[0150] Based on the first-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the second-level correction coefficient;

[0151] Based on the first-level correction coefficient, the second-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the third-level correction coefficient;

[0152] The result of multiplying the first-level correction coefficient, the second-level correction coefficient and the third-level correction coefficient is used as a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location;

[0153] To reduce the number of parameters, in each level of fitting, the correction factor for deliveries with a number of parcels not exceeding 1,000 is always fixed to 1.

[0154] The device also includes a data screening unit 12, which is used to screen out fully loaded vehicles, including the following steps:

[0155] Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping;

[0156] Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained;

[0157] Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group;

[0158] All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered;

[0159] After the second grouping, sort each group of transport vehicles from most to least according to the number of express parcels loaded on each transport vehicle, and select the transport vehicles ranked before 30% of the transport vehicles in the group into the second screening group; among which, the transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group; or

[0160] After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked before 30% of the transport vehicles in the group are selected into the third screening group; among them, the transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group;

[0161] Combine the vehicles of the first screening group, the second screening group and the third screening group, and group all the combined vehicles for the third time according to the rated loading volume and the transport line segment of each vehicle, find out the vehicle with the least number of parcels or the smallest weight of parcels loaded in each group after the third grouping, and use the number of parcels or the weight of parcels loaded on the vehicle as the screening threshold of the group, and confirm the vehicle in the group whose number of parcels or the weight of parcels loaded exceeds the screening threshold as a fully loaded vehicle.

[0162] The present solution further provides a computer-readable storage medium. The computer-readable storage medium is used to implement a program product of the above data acquisition method, which can be a portable compact disk read-only memory (CD-ROM) and includes program code, and can be run on a device, such as a personal computer. However, the program product of the present invention is not limited to this. In this document, a readable storage medium can be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, an apparatus or a device.

[0163] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0164] Computer readable storage media may include data signals propagated in baseband or as part of a carrier wave, which carry readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Readable signal media may also be any readable medium other than readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device.

[0165] The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0166] Program code for performing the operation of the present solution can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and also conventional procedural programming languages ​​such as "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user computing device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0167] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not limitations on the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present invention are still within the protection scope of the present invention.

Claims

1. A method for estimating the average volume of delivered express items, characterized in that: The method includes: Based on the obtained measurement data of express parcels to be delivered from a certain origin distribution location to a certain destination distribution location on a certain day, the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on that day is preliminarily obtained; A norm approximation model is established based on various measurement data of express parcels on a fully loaded vehicle, and a correction coefficient for correcting the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location is solved. The correction coefficient is divided into three levels and fitted in sequence, and the final correction coefficient is equal to the product of the correction coefficients of the three levels; Based on the correction coefficient and the preliminary average volume of express parcels to be delivered from the originating distribution location to the destination distribution location on that day, the average volume of express parcels to be delivered from the originating distribution location to the destination distribution location on that day is obtained, and the average volume of express parcels to be delivered from the originating distribution location to the destination distribution location is equal to the average volume of the DWS measurement of the destination distribution location from the originating distribution location to the destination distribution location multiplied by the correction coefficient of the originating distribution location obtained at the first level, then multiplied by the correction coefficient of the province where the originating distribution location is located obtained at the second level, and finally multiplied by the correction coefficient of the province where the originating distribution location is located obtained at the third level; The established norm approximation model is shown in formula (1): Where l is the correction coefficient fitting level, which can be 1, 2, and 3, corresponding to the first level, the second level, and the third level respectively; l is the set of all correction coefficients to be fitted under level l, the first level is all the originating distribution places, the second level is the combination of all the provinces where the originating distribution places and the destination distribution places are located, and the third level is the combination of all the originating distribution places and the destination distribution places; H is the set of all distribution places; o is the originating distribution place; d is the destination distribution place; I l (o, d) is the correction coefficient subscript corresponding to (o, d) at level l. When l = 1, I l (o,d) is o; when l = 2, I l (o,d) is the combination of the provinces where o and d are located; when l = 3, I l (o,d) is the combination of o and d; x i is the decision variable, the correction coefficient to be fitted at level l; s is the decision variable, the slack variable; T is the set of all (dates, line segments); S t V is the set of the origin and destination distribution locations of the goods loaded on date segment t; t is the actual loading volume of date segment t; Q tod v is the number of pieces sent from the origin distribution location o to the destination distribution location d in the date segment t; lod W is the average volume to be corrected from the origin distribution location o to the destination distribution location d at level l. The first level is the measured average volume from o to d. The second level is the measured average volume multiplied by the first level correction factor of o. The third level is the measured average volume multiplied by the first level correction factor of o, and then multiplied by the second level correction factor from o to the province where d is located. l is the L2 regularization weight of level l, which is 100 / 600 / 1000 for the three levels respectively; M is the large M value, 1e6; For x i The average value of the past 5 days; is the lower bound of the correction coefficient of level l, and the three levels are 0.6 / 0.7 / 0.8 respectively; is the upper limit of the correction coefficient of level l, and the three levels are 2.0 / 1.3 / 1.2 respectively; is the lower bound of the product structure at level l, and is 20 for all three levels; is the upper limit of the product structure at level l, and is 250 for all three levels.

2. The method for estimating the average volume of express delivery according to claim 1, characterized in that: The norm approximation model is established based on the various measurement data of the express parcels on the fully loaded vehicle, and the correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location is solved. The correction coefficient is divided into three levels and fitted in sequence. The final correction coefficient is equal to the product of the correction coefficients of the three levels, including: Based on the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the first-level correction coefficient; Based on the first-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the second-level correction coefficient; Based on the first-level correction coefficient, the second-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the third-level correction coefficient; The result of multiplying the first-level correction coefficient, the second-level correction coefficient and the third-level correction coefficient is used as a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location; To reduce the number of parameters, in each level of fitting, the correction factor for deliveries with a number of parcels not exceeding 1,000 is always fixed to 1.

3. The method for estimating the average volume of express delivery according to claim 1, characterized in that: The method also includes screening out fully loaded vehicles based on the obtained express measurement data; The screening out of fully loaded vehicles includes: Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping; Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained; Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group; All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered; After the second grouping, sort each group of transport vehicles from most to least according to the number of express parcels loaded on each transport vehicle, and select the transport vehicles ranked in the first 30% of the group into the second screening group; The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group; or After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked in the first 30% of the group of transport vehicles are selected into the third screening group; The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group; Combine the vehicles of the first screening group, the second screening group and the third screening group, and group all the combined vehicles for the third time according to the rated loading volume and the transport line segment of each vehicle, find out the vehicle with the least number of parcels or the smallest weight of parcels loaded in each group after the third grouping, and use the number of parcels or the weight of parcels loaded on the vehicle as the screening threshold of the group, and confirm the vehicle in the group whose number of parcels or the weight of parcels loaded exceeds the screening threshold as a fully loaded vehicle.

4. The method for estimating the average volume of delivered express items according to claim 3, characterized in that: The first grouping is to group the express parcels with the same origin and destination into one group; The second grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group; The third grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group.

5. The method for estimating the average volume of express delivery according to claim 1, characterized in that: The measurement data of the express shipment includes the scanning date of the express shipment, the originating distribution location, the destination distribution location, and the measured volume of the express shipment.

6. A device for estimating the average volume of delivered express mails applied to the method for estimating the average volume of delivered express mails as claimed in any one of claims 1 to 5, characterized in that: include A data acquisition unit, based on the acquired measurement data of express parcels to be delivered from a certain origin distribution location to a certain destination distribution location on a certain day, preliminarily obtains an average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on that day; A fitting unit is used to establish a norm approximation model and solve a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location; Based on the correction coefficient and the preliminarily obtained average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location on that day, the average volume of the express parcels to be delivered from the originating distribution location to the destination distribution location is obtained; Among them, solving the final correction coefficient through the norm approximation model includes: Based on the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the first-level correction coefficient; Based on the first-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the second-level correction coefficient; Based on the first-level correction coefficient, the second-level correction coefficient and the average volume of express parcels to be delivered from the origin distribution location to the destination distribution location on a certain day, the norm approximation model is used for fitting to obtain the third-level correction coefficient; The result of multiplying the first-level correction coefficient, the second-level correction coefficient and the third-level correction coefficient is used as a correction coefficient for correcting the average volume of the express parcels to be delivered from the origin distribution location to the destination distribution location; To reduce the number of parameters, in each level of fitting, the correction factor for deliveries with a number of parcels not exceeding 1,000 is always fixed to 1.

7. The device for estimating the average volume of delivered express items according to claim 6, characterized in that: The device also includes a data screening unit for screening out fully loaded vehicles, comprising the following steps: Perform the first grouping of the express parcels to be delivered based on the origin distribution location and the destination distribution location, and obtain the average volume of each group of express parcels after the first grouping; Based on the average volume of each group of express parcels after the first grouping and the loading data of each transport vehicle, the approximate loading volume of each transport vehicle loaded with the express parcels to be delivered is obtained; Compare the approximate loading volume of each transport vehicle with its rated loading volume. If the ratio of the approximate loading volume to the rated loading volume is greater than or equal to 0.95, select the transport vehicle into the first screening group; All transport vehicles are grouped for the second time based on the rated loading volume and transport line segment of each transport vehicle loaded with express parcels to be delivered; After the second grouping, sort each group of transport vehicles from most to least according to the number of express parcels loaded on each transport vehicle, and select the transport vehicles ranked in the first 30% of the group into the second screening group; The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the second screening group; or After the second grouping, each group of transport vehicles is sorted from most to least according to the weight of express shipments loaded on each transport vehicle, and the transport vehicles ranked in the first 30% of the group of transport vehicles are selected into the third screening group; The transport vehicles ranked before 30% of the transport vehicles in the group are rounded up to determine the transport vehicles entering the third screening group; Combine the transport vehicles of the first screening group, the second screening group and the third screening group, and group all the combined transport vehicles for the third time according to the rated loading volume and transport line segment of each transport vehicle, find the transport vehicle with the least number of express parcels or the smallest weight of express parcels loaded in each group after the third grouping, and use the number of express parcels or the weight of express parcels loaded by the transport vehicle as the screening threshold of the group, and confirm the transport vehicle in the group whose number of express parcels or the weight of express parcels loaded exceeds the screening threshold as a fully loaded vehicle; The first grouping is to group the express parcels to be delivered with the same origin and destination. The second grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group; The third grouping is to group the transport vehicles with the same rated loading volume and transport line segment into one group.

8. The device for estimating the average volume of delivered express items according to claim 6, characterized in that: The measurement data of the express shipment includes the scanning date of the express shipment, the originating distribution location, the destination distribution location, and the measured volume of the express shipment.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images