Disclosure of Invention
In order to overcome the defects in the prior art, the invention provides a method, a device, equipment and a readable storage medium for predicting the transportation volume of a highway and a railway.
In order to achieve the purpose of the invention, the invention adopts the technical scheme that:
in a first aspect, the invention provides a highway-railway combined transportation volume prediction method, which comprises the following steps:
s1, obtaining road goods source data and preprocessing the data;
s2, clustering the preprocessed road freight source data by adopting a DBSCAN clustering algorithm based on an optimal contour coefficient to obtain a single-day road freight demand point clustering result;
s3, reverse geocode matching is carried out on the single-day highway freight demand point clustering result to obtain a geographic position matching result;
s4, mining the geographic position matching result by adopting a frequent item set mining algorithm to obtain a hot spot area required by road freight;
s5, selecting alternate points of the highway-railway combined transportation hub nodes, constructing a hub-spoke type highway-railway combined transportation hub node site selection model by combining with a highway freight demand hot spot area, and solving to obtain the highway-railway combined transportation hub points and service areas thereof;
s6, excavating freight characteristics in the service area of each link point of the highway-railway combined transportation pivot to obtain the service cargo type of each link point of the highway-railway combined transportation pivot;
and S7, extracting the cargo quantity of each type of service cargo of each highway-railway combined transportation pivot point from the highway cargo source data, and predicting the cargo quantity at the next moment by adopting an integrated sliding average autoregressive model.
Optionally, step S2 specifically includes the following sub-steps:
s21, setting a value range and a transformation step length of an initial neighborhood radius and a minimum point number;
s22, randomly selecting an initial demand point from demand set points;
s23, judging whether the demand point with the minimum point number exists in the neighborhood radius range of the demand point;
if yes, clustering all the existing demand points into a cluster, and skipping to the step S24;
otherwise, judging the demand point as a noise point, and jumping to the step S22;
s24, traversing each demand point in the cluster, and judging whether the demand point with the minimum point number exists in the neighborhood radius range of the selected demand point;
if yes, merging all the existing demand points into the cluster, and jumping to the step S25;
otherwise, directly jumping to the step S25;
s25, judging whether the clusters have unrepeated demand points or not;
if yes, jumping to step S24;
otherwise, jumping to step S26;
s26, judging whether unprocessed demand points exist in the demand set points;
if yes, jumping to the step S22;
otherwise, jumping to step S27;
s27, judging whether the neighborhood radius and the minimum point number reach the maximum value or not;
if yes, jumping to step S28;
otherwise, updating the neighborhood radius and the minimum point number according to the conversion step length, and jumping to the step S22;
s28, calculating the contour coefficient of each cluster and marking cluster marks to obtain the longitude and latitude coordinates of each demand point and the cluster label of each demand point.
Optionally, the calculation method of the profile coefficient of each cluster is:
calculating the spherical distance from each demand point to other demand points in the cluster to which the demand point belongs according to the longitude and latitude coordinates of the demand points, and calculating the average distance;
calculating the spherical distance from each demand point to all demand points in other clusters according to the longitude and latitude coordinates of the demand points, and calculating the average distance;
calculating the contour coefficient of each demand point according to the average distance from each demand point to other demand points in the cluster to which the demand point belongs and the average distance from each demand point to all demand points in other clusters;
and averaging the contour coefficients of all the demand points in each cluster to obtain the contour coefficient of each cluster.
Optionally, the formula for calculating the spherical distance is as follows:
wherein,
dis the spherical distance of the two required points,
rwhich is the radius of the earth, is,
as is the latitude of the two demand points,
the longitude of two demand points.
Optionally, the calculation formula of the profile coefficient of each demand point is as follows:
wherein,
is a first
iThe profile factor of each of the demand points,
is as follows
iThe average distance of a demand point to other demand points within its cluster,
is a first
iThe minimum average distance from each demand point to all demand points in other clusters;
is a first
iEach demand point going to others in other clusters
jAverage distance of individual demand points.
Optionally, the method for constructing the hub and spoke type road-rail transport hub node site selection model includes:
the method comprises the steps of constructing a hub and rail transport hub node addressing model by taking the minimum of highway and railway transport construction operation cost and transport cost as an objective function, taking the maximum number of construction pivot point points, the requirement of each highway transport requirement point is met by one transport pivot point, the transport volume of the requirement points does not exceed the total requirement quantity of the requirement points, all transport requirements of the requirement points are met, the transport volume of the transport pivot point is equal to the sum of the transport volume of each requirement point to the pivot point, the total collection transport volume of selected transport pivot alternative points does not exceed the maximum transport capacity of the transport pivot point, whether the pivot points provide transport services for the requirement points or not, whether alternative pivot point is selected, the quantity of goods transported from the requirement points to another requirement point and the quantity of goods transported to the pivot points as constraint conditions.
Optionally, the hub node site selection model of the spoke type highway-railway combined transport hub specifically includes:
s.t.
wherein,
is a set of alternative points of the highway-railway combined transportation hub,
the cost is fixed for the construction of the hub,
ffor the transfer of the handling charges per unit of cargo volume at the alternative pivot point,
for selecting points for highway-railway combined transportation hub
jThe maximum transport capacity of the transport means (c),
a 0-1 decision variable for whether the alternative pivot point is selected,
in order to be a set of demand points for road transportation,
for highway transportation demand points
iAlternative points to highway-railway combined transport hub
jThe amount of cargo to be transported is,
for selecting points for highway-railway combined transportation hub
jWhether to the highway transportation demand point
iA 0-1 decision variable for providing intermodal services,
for highway transportation demand points
iAlternative point of road-rail transport hub
jThe spherical distance of (a) is greater than (b),
Cunit transportation cost;
Sthe maximum number of the pivot points is constructed,
for highway transportation demand points
iThe total volume of traffic of (a) is,
for transportation to alternate points of road-rail transport hub
jThe amount of cargo.
In a second aspect, the present invention provides a device for predicting the transportation capacity of a highway/railway combined transport, including:
the data preprocessing module is used for acquiring the highway goods source data and preprocessing the data;
the data clustering module is used for clustering the preprocessed road cargo source data by adopting a DBSCAN clustering algorithm based on an optimal contour coefficient to obtain a clustering result of the single-day road cargo demand points;
the data matching module is used for carrying out reverse geocode matching on the clustering result of the single-day highway freight demand points to obtain a geographic position matching result;
the data mining module is used for mining the geographic position matching result by adopting a frequent item set mining algorithm to obtain a hot spot area required by road freight;
the model building module is used for selecting alternative points of the highway-railway combined transportation hub node, building a hub and spoke type highway-railway combined transportation hub node site selection model by combining a highway freight demand hot spot area, and solving to obtain a highway-railway combined transportation hub point and a service area thereof;
the freight characteristic mining module is used for mining freight characteristics in the service area of each highway-railway combined transportation pivot point to obtain the service cargo type of each highway-railway combined transportation pivot point;
and the cargo quantity prediction module is used for extracting the cargo quantities of various types of service cargos at the highway and railway transportation pivot points from the highway cargo source data and predicting the cargo quantity at the next moment by adopting an integrated moving average autoregressive model.
In a third aspect, the present invention provides a device for predicting the transportation capacity of a highway/railway combined transport, including: a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program to realize the steps of the method for predicting the combined transportation capacity of the highway and railway.
In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, the computer program, when being executed by a processor, implementing the steps of the method for predicting the transportation volume of a highway and a railway as described above.
The invention has the following beneficial effects:
according to the method, firstly, the regional freight characteristics are analyzed by using the freight source data, then, the highway freight demand gathering region is identified by a clustering and frequent item set mining method, on the basis, a hub-and-spoke type highway-railway transport hub node site selection model is constructed and solved to obtain the distribution position and the service region of the highway-railway transport hub node, the type of the suitable railway freight in the region is mined, and finally, the freight quantity of the next moment is predicted according to the freight quantity of each type of service freight at each highway-railway transport hub node, so that the prediction cost can be effectively reduced, and the prediction precision of the highway-railway transport quantity is improved.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
Example 1
As shown in fig. 1, an embodiment of the present invention provides a method for predicting a transportation volume of a highway and a railway, including the following steps S1 to S7:
s1, obtaining road goods source data and preprocessing the data;
in an optional embodiment of the invention, the highway source data is firstly obtained, by taking the example of collecting original data of a highway trunk transport online transaction platform in China as an example, when the original data is delivered by a cargo owner, a json character string format with a timestamp directly returned by the platform is low in readability and difficult to directly process, so that the json character string is analyzed and processed by using an SQL language to respectively obtain 380 pieces of arrival source data with 58 fields and 330 more than ten thousand pieces of sending source data, and the arrival source data and the sending source data are respectively stored in new tables Arr and Dep in a MySQL database through a database management software Navicat for MySQL, so that the operations of directly calling data and increasing, deleting and revising are facilitated.
According to the method, after the road source data are acquired, data preprocessing is needed, fields with unknown meanings, excessive null values or useless values are deleted, and related fields such as source creation time, creation timestamp, delivery place, destination, cargo type, cargo name, required vehicle length, required transportation volume or required capacity, OD longitude and latitude coordinates and the like are reserved, wherein the fields are shown in table 1.
TABLE 1 original data field name, meaning and data type,
Then, in addition to the longitude and latitude, for the situations that repeated delivery may exist, an abnormal value exists in a source of goods, and the like, deletion operation needs to be performed on such data lines in advance, and determination conditions are respectively set as follows:
(1) The repeated goods source refers to that a certain shipper delivers the same goods for multiple times, the frequency of the situation is many, the reason is that a shipper is urgent to deliver the goods, or misoperation is caused, or the quantity of the goods in the lot is too much, and multiple trucks are required to transport the goods, so that the repeated goods source with the same five fields of the departure district and county, the destination district and county and the quantity of the goods is deleted according to the id of the shipper and the type of the secondary goods, and a row of data with the minimum timestamp (timestamp) is reserved;
(2) The method comprises the steps that a shipper fills in the weight or volume of goods during shipment, most shippers fill in the weight of the goods, the shipper fills in a volume field only when the goods belong to light-weight goods, two fields of part of shippers are filled together, the number of goods sources which only fill in the volume and have the weight of 0 is counted to be smaller, and the data lines are deleted for the convenience of subsequent calculation;
(3) And (3) deleting data lines with overlarge weight (weight), wherein part of owners of the goods are enterprises or logistics companies, the delivery volume of the owners of the goods is far larger than that of the owners of the individual goods, the owners of the individual goods can release all the transportation volume to a goods source at one time, and the data belongs to normal data. Meanwhile, the situation that part of the shippers are randomly filled exists, data with the shipping weight being larger than or equal to 999 is defined as abnormal, and the row is deleted.
Finally, the data is preprocessed by the following data conversion and field processing:
(1) Converting the delivery time stamp of each line of the source data into a time format like 2021/6/25;
(2) The above data were summed for cargo weight in hours according to the "day" and "hour" fields and sorted in ascending order by time stamp.
(3) In order to quantify the increasing and decreasing trend of the original data, a field trend is added, which means the increasing and decreasing trend of the demand at the moment and is denoted as tr, and as shown in the following formula, the trend is increasing when tr =1, otherwise, the trend is decreasing. Where represents the amount of cargo at that moment, defining that the value of tr is also 0 when t = 0.
Finally, a data table with the size of (514, 7) is obtained, and the fields are "datatime", "weight", "county", "month", "day", "hour", "tend", and "tend", respectively.
S2, clustering the preprocessed road cargo source data by adopting a DBSCAN clustering algorithm based on an optimal contour coefficient to obtain a clustering result of the single-day road cargo demand points;
in an alternative embodiment of the present invention, the present invention first explains the relevant principle of the DBSCAN clustering algorithm.
DBSCAN (sensitivity-Based Spatial Clustering of Applications with Noise) is a Spatial Clustering algorithm Based on Density. By defining a maximum set of density-connected points, the algorithm can divide an area with a certain density into one cluster, and can find clusters of arbitrary shapes in spatial data with noise. The DBSCAN algorithm has the advantages of no need of setting the number of clusters a priori, capability of dividing a data set with a complex shape, finding noise and abnormal points in data and the like, but good and bad clustering effect and scanning neighborhood radius: (eps) And the minimum number of contained points (min) in the neighborhoodPts) The two parameters are closely related and need to be adjusted and optimized according to actual problems and clustering results.
The definition for density in DBSCAN is a certain radiusepsThe number of points included in the range, i.e. a certain spatial distance range, in the present inventionepsNumber of inner highway freight demand points minPtsBased on this set concept the following:
data points are first classified into three categories:
(1) Core point: if there is a certain demand point n, its radiusepsWithin a circular range of (1), including at least minPtsN points are core objects if other demand points are needed;
(2) Boundary points are as follows: if the demand point n is a core object, and the m point is in the neighborhood range and is not the core object, the demand point m is called as the boundary point of the demand point n;
(3) Noise point: and the points in all the demand point sets which do not belong to the core point or the boundary point do not belong to any cluster in the clustering result.
Definition of the relationship between data points based on density:
(1) The density is up to: for a set of demand points N for shipment, if m is at NepsIn the neighborhood of the radius, and n points are core points, the direct density from n points to m points is calledSo as to obtain;
(2) The density can reach: for a freight demand point N, there is a series of sample points
Wherein the direct density of two adjacent demand points can be reached, it can be called
From
The density can be reached;
(3) Density connection: for a freight demand point set, if an s point is respectively reachable with the n point and the m point, the n is connected with the m density;
(4) Clustering: the clustering result comprises a set of all density-connected demand points.
The clustering effect of the DBSCAN algorithm depends on the radius of two parameter fieldsepsAnd minimum number of points minPtsWhen the parameter is the minimum point minPtsRadius of area at the time of fixationepsThe unreasonable setting of the cluster can cause excessive or too few core points, and directly influence the number of clusters to be too small or too large; radius of field of parameterepsFixed, minimum number of points minPtsThe unreasonable arrangement of the cluster also directly influences the judgment of the core points in the cluster and the cluster quantity. In the prior art, the parameter is determined by acquiring the number of stable clustering clusters according to the experience of practical problems or continuously adjusting the parameter, but the time cost of manual trial and error is higher, so that the invention inputs the parameter within a certain range by introducing an index 'profile coefficient' for evaluating a clustering result, returns the profile coefficient until finding the input parameter corresponding to the maximum profile coefficient, and further obtains a better clustering result.
As shown in fig. 2, step S2 of the present invention specifically includes the following sub-steps:
s21, setting initial neighborhood radiusepsAnd minimum number of points minPtsAnd the corresponding transformation step sizes L1 and L2;
s22, randomly selecting an initial demand point n from demand set points;
s23, judging the neighborhood radius of the demand point nepsWhether there is a minimum number of points min within the rangePtsA number of other demand points;
if yes, clustering all the existing demand points into a cluster N, and jumping to the step S24;
otherwise, judging the demand point as a noise point, and skipping to the step S22;
s24, traversing other demand points in the cluster N
Judging the selected demand points
Neighborhood radius of
epsWhether there is a minimum number of points min within the range
PtsA number of other demand points;
if yes, merging all the existing demand points into the cluster N, and jumping to the step S25;
otherwise, directly jumping to the step S25;
s25, judging whether the demand points which are not traversed exist in the cluster N;
if yes, jumping to step S24;
otherwise, jumping to step S26;
s26, judging whether unprocessed demand points exist in the demand set points;
if yes, jumping to the step S22;
otherwise, jumping to step S27;
s27, judging the radius of the neighborhoodepsAnd minimum points minPtsWhether a maximum value is reached;
if yes, jumping to step S28;
otherwise, the neighborhood radius and the minimum point number are updated according to the conversion step length, namelyeps=eps+L1,minPts=minPts+ L2, and go to step S22;
s28, calculating the contour coefficient of each cluster and marking cluster marks to obtain the longitude and latitude coordinates of each demand point and the cluster label of each demand point.
Specifically, the contour Coefficient (Silhouette Coefficient) provided by the invention is an evaluation mode for evaluating the advantages and disadvantages of the clustering effect. By comparing the similarity of samples in clusters
And inter-cluster sample similarity
And evaluating the reasonable degree of each sample in the clustering result belonging to the current cluster.
The calculation method for calculating the contour coefficient of each cluster comprises the following steps:
calculating the spherical distance from each demand point to other demand points in the cluster to which the demand point belongs according to the longitude and latitude coordinates of the demand points, and calculating the average distance;
calculating the spherical distance from each demand point to all demand points in other clusters according to the longitude and latitude coordinates of the demand points, and calculating the average distance;
calculating the contour coefficient of each demand point according to the average distance from each demand point to other demand points in the cluster to which the demand point belongs and the average distance from each demand point to all demand points in other clusters;
and averaging the contour coefficients of all the demand points in each cluster to obtain the contour coefficient of each cluster.
The calculation formula of the spherical distance is as follows:
wherein,
dis the spherical distance of the two required points,
rwhich is the radius of the earth, is,
as is the latitude of the two demand points,
the longitude of two demand points.
The calculation formula of the profile coefficient of each demand point is as follows:
wherein,
is as follows
iThe outline coefficient of each demand point has the value range of (-1,1);
is as follows
iThe average distance from each demand point to other demand points in the cluster to which the demand point belongs, and the smaller the value is, the greater the similarity degree in the cluster is, the demand point is
iThe greater the degree of belonging to the cluster;
is a first
iThe average distance from each demand point to all demand points in other clusters is smaller, the smaller the value is, the greater the similarity degree between clusters is, the demand point is
iMay belong to other clusters;
is as follows
iEach demand point going to others in other clusters
jAverage distance of individual demand points.
S3, reverse geocode matching is carried out on the single-day highway freight demand point clustering result to obtain a geographic position matching result;
in an optional embodiment of the invention, the method is based on the clustered highway freight demand hot spots of a single day, and the hot spot areas which frequently appear in a period and have stable and large number of goods sources are mined from the time dimension analysis. Therefore, a batch of highway freight demand centers are determined, goods with stable and large-quantity transportation demand characteristics in the area range of the demand centers are mined, and a reference is provided for a railway transportation department to find a highway stable goods source.
Firstly, each cluster in the clustering result needs to be endowed with geographic significance, and because the clustering result of the DBSCAN has no definite clustering center, the invention calculates the average longitude and latitude of each cluster to obtain longitude and latitude coordinate pairs with the number equal to that of the clusters, utilizes a map api interface to carry out reverse geocoding query to obtain the administrative division name corresponding to each longitude and latitude coordinate pair, and approximately represents the geographic position and the coverage area of the cluster. In order to reduce the matching error between the clustering result and the geographic administrative division, the range of the address obtained by reverse geocoding query is expanded to the level of a local city, a county and a district. And repeating the calculation and matching operation on all the data to obtain the highway freight demand center set taking days as units.
S4, mining the geographic position matching result by adopting a frequent item set mining algorithm to obtain a hot spot area required by road freight;
in an alternative embodiment of the invention, considering that the distribution of highway freight demand centers in days is not fixed, it is not reasonable to take the aggregate of all demand centers in all data as the road-rail combined transportation service object of the railway station. In order to ensure that the freight demand point can stably generate the freight demand for a long time as much as possible, the address set which frequently appears in the whole time span covered by the data needs to be found out. The invention solves the problem based on the frequent pattern mining technology, and relevant concept definitions are given based on the problem background as follows:
(1) A transaction database: transactions are a subset of global items, and collections of transactions are organized into a transaction database. The set of demand center addresses for a day may be referred to herein as a transaction;
(2) Frequent item set: a collection of items in the data set that frequently occur simultaneously, namely a collection of freight demand points that frequently occur within a day within a data coverage time span;
(3) The support degree is as follows: the number of occurrences of a certain freight demand point set in the data set is proportional. When the support degree of a certain freight demand point set D is greater than the preset minimum support degree, the set D is called a frequent item set and comprisesxThe collection of items is called frequentxAn item.
The invention adopts FP-Growth algorithm to find out all Frequent item sets from the FPTree by storing data in a Frequent Pattern Tree (FPTree) and then utilizing a recursion method, and the basic flow is as follows:
(1) Scanning a data complete set, finding out a frequent 1 item set (only comprising one geographic position), and sequencing the frequent 1 item set in a descending manner according to the support degree until the support degree is equal to the minimum support degree;
(2) Scanning a data complete set, sequencing a geographical position set taking days as a unit according to the support degree calculated in the step (1), and inserting the geographical position set into a tree taking null as a root node in sequence to construct a frequent pattern tree;
(3) In the frequent pattern tree, searching the geographical position item on the prefix path of the frequent pattern tree from the residual frequent item set 1 in the step (1), constructing a conditional frequent pattern tree, recursively constructing the conditional frequent pattern tree until only one item remains in the tree structure, and performing permutation and combination according to the conditional frequent pattern tree to obtain all frequent item sets.
A transaction database is constructed according to the freight demand address, a day is taken as a transaction, and a frequent item set is mined by setting larger support degrees of 0.8,0.9 and 1 in consideration of the stability of the freight demand
S5, selecting alternate points of the highway-railway combined transportation hub nodes, constructing a hub-spoke type highway-railway combined transportation hub node site selection model by combining with a highway freight demand hot spot area, and solving to obtain the highway-railway combined transportation hub points and service areas thereof;
in an optional embodiment of the invention, on the basis of the identification result of the demand hot spot region, the alternative selection points of the highway-railway transport hub are selected, a highway-railway transport site selection distribution model is constructed and solved, and finally the service region, the main goods and the development emphasis point of each highway-railway transport hub are determined.
In order to screen out stations with construction conditions and operation capacity of a highway-railway combined transportation hub, the invention sets the following screening rules:
(1) And (4) station grade. The railway station grade approval method provides three indexes of getting on and off the railway station on a daily basis, changing the number of passengers into the number of passengers, the number of packages in transit, the number of vehicles in loading and unloading on a daily basis and the number of vehicles with dispatching work on a daily basis to evaluate the station grade, so that the station grade directly shows the transport capacity of the station. Other stations except for the special station and the first-class station are not considered herein;
(2) The station geographical location. And partial special stations and first-class stations are built in the main urban area due to passenger transport requirements, do not have the condition of building hubs, and simultaneously consider the requirement of relieving the non-capital function of Beijing. Deleting stations such as Beijing, beijing West station, tianjin West station and the like;
(3) And (5) station operation range. According to the public data of the Chinese railway 95306 website, the handling of the transportation business of the railway freight yard is limited by partial stations, for example, the stone mountain station only handles the steel transportation business, and stations which are built based on industrial and mining enterprises and have narrow transportation business surfaces are excluded.
The multi-type combined transportation network is constructed according to the radial principle, so that the logistics cost can be effectively reduced, and the transportation efficiency is improved. The scattered small-batch cargo flows of road freight are concentrated on a large highway-railway combined transportation hub through short-distance highway transportation, and then medium-and long-distance trunk transportation is carried out by utilizing railways, so that the advantages of various transportation modes are fully utilized, and the large-scale economic benefit of transportation is generated. The mode of the combined transportation of the highway and the railway designed by the invention is shown in figure 3.
Based on the highway-railway combined transportation mode, the longitude and latitude coordinates of the highway transportation demand point and the alternative highway-railway combined transportation pivot point, the demand of the highway demand point, and the fixed cost and the maximum capacity of the alternative highway-railway combined transportation pivot point are known. The method aims to construct a site selection distribution model, select sites of highway and railway combined transportation pivot points, and make a decision on highway transportation demand points served by each pivot point so as to find a site selection distribution scheme with the minimum total cost (construction cost and transportation cost) and meet regional freight requirements.
In order to improve the modeling and solving efficiency, the invention simplifies the process of actual highway-railway combined transport site selection allocation, and makes the following assumptions:
(1) The short-distance highway transportation distance is measured and calculated by utilizing road network data, but redundant transportation distance caused by road conditions and driver behaviors is not considered;
(2) The requirement of each road transportation demand point can be met by only one pivot point, and the behavior of transportation to a plurality of pivot points for scattered intermodal transportation does not exist;
(3) Only the fixed cost of building the link points of the intermodal transportation and the transit loading and unloading cost are considered, and other operation costs are not considered.
The invention discloses a method for constructing a hub node site selection model of a hub-spoke type highway-railway combined transportation hub, which comprises the following steps:
the method comprises the steps of constructing a hub and rail transport hub node addressing model by taking the minimum of highway and railway transport construction operation cost and transport cost as an objective function, taking the maximum number of construction pivot point points, the requirement of each highway transport requirement point is met by one transport pivot point, the transport volume of the requirement points does not exceed the total requirement quantity of the requirement points, all transport requirements of the requirement points are met, the transport volume of the transport pivot point is equal to the sum of the transport volume of each requirement point to the pivot point, the total collection transport volume of selected transport pivot alternative points does not exceed the maximum transport capacity of the transport pivot point, whether the pivot points provide transport services for the requirement points or not, whether alternative pivot point is selected, the quantity of goods transported from the requirement points to another requirement point and the quantity of goods transported to the pivot points as constraint conditions.
The hub and spoke type road and rail transport hub node site selection model constructed by the invention specifically comprises the following steps:
s.t.
wherein,
is a set of alternative points of the highway-railway combined transportation hub,
the cost is fixed for the construction of the hub,
fthe loading and unloading cost is transferred in unit cargo quantity of the alternative pivot point,
for selection point of road-rail transport hub
jThe maximum transport capacity of the transport means,
a 0-1 decision variable for whether the alternative pivot point is selected,
is a set of demand points for road transport,
for highway transportation demand points
iAlternative points to highway-railway combined transport hub
jThe amount of cargo to be transported is,
for selection point of road-rail transport hub
jWhether to the highway transportation demand point
iA 0-1 decision variable for providing intermodal services,
for highway transportation demand points
iAlternative point of road-rail transport hub
jThe spherical distance of (a) is greater than (b),
Cis the unit transportation cost;
Sthe maximum number of the pivot points is constructed,
for highway transportation demand points
iThe total volume of traffic of (2) is,
for transporting to alternate points of highway-railway combined transportation hub
jThe amount of cargo.
S6, excavating freight characteristics in the service area of each highway-railway combined transportation pivot point to obtain the service cargo type of each highway-railway combined transportation pivot point;
in an optional embodiment of the invention, after the key hub and the service range thereof are determined, the freight characteristics in the area range are mined, so that the service cargo type can be comprehensively considered, the highway-railway combined transportation volume can be more comprehensively predicted, and the prediction accuracy is improved.
As the railway department mainly pays attention to the characteristics of stable and large-batch cargo types and cargo volumes of scattered cargo transportation, the invention carries out the cargo-classified transportation volume statistics on the road transportation cargos in the service range of each hub point; in addition, as the economic benefit of railway transportation on scale cannot be effectively reflected in short-distance transportation, the average distance characteristic of the goods in the region range needs to be calculated. And screening road cargo sources suitable for the intermodal operation at the pivot point by combining the classified freight volume and the average freight distance.
The rules for screening the proper sources are set as follows:
(1) Detailed recording is carried out on the weight of each item by the original data, abnormal data are deleted, and the items with the single daily freight volume ratio larger than 10% are preliminarily screened out;
(2) And calculating the average road transportation distance according to the goods, and calculating the median of the goods-classified transportation in order to prevent the interference of the extreme value of the transportation distance data. And setting the goods as the alternative goods type when the average distance is more than 500 kilometers and the average distance of the goods is more than the median of the distance. It means that the majority of the sources of the cargo over a distance of 500 km, such cargo is considered suitable for rail transport.
And establishing a query statement in the MySQL database according to the rules, and screening the regional cargo types. The result of the main cargo type can be obtained, wherein the primary cargo types are arranged according to the proportion order of the cargo quantity.
S7, the cargo quantity of each type of service cargo at each highway-railway combined transportation pivot point is extracted from the highway cargo source data, and the cargo quantity at the next moment is predicted by adopting an integrated sliding average autoregressive model.
In an alternative embodiment of the present invention, the integrated moving average autoregressive model prediction employed by the present invention integrates the autoregressive term AR and the moving average term MA to predict the current, namely:
wherein
Is a constant number of times, and is,
is a white noise sequence and is a white noise sequence,
is one
An autoregressive model of order, considering the current date as equal to the sum of the weighted average of the historical data and an error disturbance term,
is one
And (3) a moving average model of an order, wherein current data is considered to be equal to the sum of the weighted average of the historical error disturbance and the historical average.
Form a
Model, wherein
Is the difference order. The model considers current date to be related to both historical data and historical errors.
The method is based on an integrated moving average autoregressive model, and the cargo quantity of each type of service cargo at each highway-railway combined transportation pivot point at the next moment is obtained according to the cargo quantity prediction of each type of service cargo at each highway-railway combined transportation pivot point.
Example 2
As shown in fig. 4, an embodiment of the present invention provides a device for predicting transportation volumes of a highway and a railway based on the method for predicting transportation volumes of a highway and a railway described in embodiment 1, including:
the data preprocessing module is used for acquiring the goods source data of the highway and preprocessing the data;
the data clustering module is used for clustering the preprocessed road cargo source data by adopting a DBSCAN clustering algorithm based on an optimal contour coefficient to obtain a clustering result of the single-day road cargo demand points;
the data matching module is used for carrying out reverse geocode matching on the clustering result of the single-day highway freight demand points to obtain a geographic position matching result;
the data mining module is used for mining the geographic position matching result by adopting a frequent item set mining algorithm to obtain a highway freight requirement hot spot area;
the model building module is used for selecting alternative points of the highway-railway combined transportation hub nodes, building a hub-spoke type highway-railway combined transportation hub node site selection model by combining with a highway freight demand hot spot area, and solving to obtain the highway-railway combined transportation hub points and service areas thereof;
the freight characteristic mining module is used for mining freight characteristics in the service area of each highway-railway combined transportation pivot point to obtain the service cargo type of each highway-railway combined transportation pivot point;
and the cargo quantity prediction module is used for extracting the cargo quantities of various types of service cargos at the highway and railway transportation pivot points from the highway cargo source data and predicting the cargo quantity at the next moment by adopting an integrated moving average autoregressive model.
The device for predicting the combined transportation capacity of the highway and railway provided by the embodiment 2 of the invention has the beneficial effect of the method for predicting the combined transportation capacity of the highway and railway in the embodiment 1.
Example 3
The embodiment of the invention provides a prediction device for the combined transportation volume of a highway and a railway based on the prediction method for the combined transportation volume of the highway and the railway described in the embodiment 1, which comprises the following steps: a memory and a processor;
the memory is used for storing programs;
the processor is used for executing the program to realize the steps of the method for predicting the combined transportation capacity of the highway and railway.
The device for predicting the combined transportation capacity of the highway and railway provided by the embodiment 3 of the invention has the beneficial effect of the method for predicting the combined transportation capacity of the highway and railway in the embodiment 1.
Example 4
An embodiment of the present invention provides a computer-readable storage medium, which stores a computer program that, when executed by a processor, implements the steps of the method for predicting the amount of transportation of a utility grid as described in embodiment 1.
Embodiment 4 of the present invention provides a computer-readable storage medium, which has the beneficial effect of the method for predicting the combined transportation volume of the highway and railway in embodiment 1.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The principle and the implementation mode of the invention are explained by applying specific embodiments in the invention, and the description of the embodiments is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
It will be appreciated by those of ordinary skill in the art that the embodiments described herein are intended to assist the reader in understanding the principles of the invention and are to be construed as being without limitation to such specifically recited embodiments and examples. Those skilled in the art can make various other specific changes and combinations based on the teachings of the present invention without departing from the spirit of the invention, and these changes and combinations are within the scope of the invention.