Method and device for determining demand data of air route, storage medium and program product

Abstract

The invention discloses a method and a device for determining demand data of an air route, a storage medium and a program product. The method comprises the steps that a to-be-predicted route pair of the flight equipment is acquired, and the to-be-predicted route is used for representing a moving area of the flight equipment; determining the use frequency of the to-be-predicted route pair; at least one sub-demand prediction model matched with the use frequency is called from the demand prediction model, the to-be-predicted airline pair is predicted, demand data of the to-be-predicted airline pair is obtained, the demand data is used for representing the freight rate demand of the to-be-predicted airline pair, and the sub-demand prediction models are multiple airline pair samples based on different use frequency samples; and training a demand data sample corresponding to the sample by the route. The technical problem that the accuracy of determining the freight rate demand is low is solved.

Description

Technical Field

[0001] The present application relates to the aviation field, and in particular, to a method, device, storage medium and program product for determining demand data of an air route. Background Art

[0002] Currently, the rapid expansion of the aviation industry is accompanied by a surge in the number of routes and increased complexity in freight demand forecasting.

[0003] In related technologies, the prediction of route demand usually uses machine learning models to process sales data to complete the prediction of demand. However, this method needs to process a relatively large amount of data during the prediction process, and there is a technical problem of low accuracy in determining freight rate demand.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present application provide a method, device, storage medium and program product for determining demand data of a route, so as to at least solve the technical problem of low accuracy in determining freight rate demand.

[0006] According to one aspect of an embodiment of the present application, a method for determining demand data of a route is provided, and the method may include: obtaining a pair of routes to be predicted for a flight device, wherein the route to be predicted is used to characterize an area where the flight device moves; determining a usage frequency of the route pair to be predicted; and calling at least one sub-demand prediction model that matches the usage frequency from a demand prediction model to predict the route pair to be predicted, thereby obtaining demand data for the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand for the route pair to be predicted, and the sub-demand prediction model is trained based on a plurality of route pair samples with different usage frequency samples, and demand data samples corresponding to the route pair samples.

[0007] Optionally, from the demand forecasting model, at least one sub-demand forecasting model that matches the usage frequency is called to forecast the route pair to be predicted, and demand data for the route pair to be predicted is obtained, including: determining the origin and destination of the route pair to be predicted; calling at least one sub-demand forecasting model that matches the usage frequency, origin and destination, and forecasting the route pair to be predicted to obtain demand data.

[0008] Optionally, at least one sub-demand forecasting model matching the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being greater than the frequency threshold, and the origin and the destination belonging to different areas, the first sub-demand forecasting model is called to predict the route pair to be predicted and obtain demand data, wherein the first sub-demand forecasting model is trained based on route pair samples in the usage frequency samples whose usage frequency is greater than the frequency threshold, and the demand data corresponding to the route pair samples.

[0009] Optionally, at least one sub-demand forecasting model that matches the usage frequency, the origin and the destination is called to predict the route pair to be predicted to obtain demand data, including: in response to the usage frequency being greater than the frequency threshold, and the origin and the destination belong to the same area, the first sub-demand forecasting model is called to predict the route pair to be predicted to obtain a first prediction result, and the second sub-demand forecasting model that matches the same area is called to predict the route pair to be predicted to obtain a second prediction result, wherein the second sub-demand forecasting model is trained based on route pair samples in the same area and is different from the first sub-demand forecasting model; the first prediction result and the second prediction result are weighted to obtain demand data.

[0010] Optionally, at least one sub-demand forecasting model matching the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being less than the frequency threshold and the origin and the destination belonging to different areas, a third sub-demand forecasting model matching the origin is called to predict the route pair to be predicted and obtain demand data, or a fourth sub-demand forecasting model matching the destination is called to predict the route pair to be predicted and obtain demand data.

[0011] Optionally, at least one sub-demand forecasting model matching the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being less than the frequency threshold and the origin and the destination belonging to the same area, a second sub-demand forecasting model matching the same area is called to predict the route pair to be predicted and obtain demand data.

[0012] According to another aspect of an embodiment of the present application, a device for determining demand data of a route is also provided, and the device may include: an acquisition unit, used to acquire a pair of routes to be predicted of a flight device, wherein the route to be predicted is used to characterize an area where the flight device moves; a determination unit, used to determine a usage frequency of the route pair to be predicted; and a processing unit, used to call at least one sub-demand forecasting model that matches the usage frequency from a demand forecasting model, to forecast the route pair to be predicted, and obtain demand data of the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand of the route pair to be predicted, and the sub-demand forecasting model is trained based on multiple route pair samples with different usage frequency samples, and demand data samples corresponding to the route pair samples.

[0013] According to another aspect of an embodiment of the present application, a computer-readable storage medium is also provided, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the method for determining the demand data of the route of the embodiment of the present application.

[0014] According to another aspect of an embodiment of the present application, a processor is further provided, the processor being used to run a program, wherein when the program is run by the processor, the method for determining demand data of the route of an embodiment of the present application is executed.

[0015] According to another aspect of an embodiment of the present application, a computer program product is further provided. The computer program product includes computer instructions, wherein the computer instructions, when executed by a processor, implement the method for determining demand data of the route of the embodiment of the present application.

[0016] In an embodiment of the present application, a pair of routes to be predicted of a flight device is obtained, wherein the route to be predicted is used to characterize the area where the flight device moves; the frequency of use of the route to be predicted is determined; from the demand prediction model, at least one sub-demand prediction model that matches the frequency of use is called to predict the route pair to be predicted, and demand data of the route pair to be predicted is obtained, wherein the demand data is used to characterize the freight demand of the route pair to be predicted, and the sub-demand prediction model is trained based on multiple route pair samples with different frequency of use samples, and demand data samples corresponding to the route pair samples. That is to say, in an embodiment of the present application, different sub-demand prediction models are trained based on route pair samples with different frequencies. After obtaining the route pair to be predicted, the sub-demand prediction model that matches the route pair to be predicted can be called to predict the route pair to be predicted, so as to obtain demand data (that is, freight demand) corresponding to the route pair to be predicted. The method constructs different sub-demand prediction models for different types of route pairs to be predicted, and uses the corresponding sub-demand prediction models to predict the route pair to be predicted to obtain a determined prediction result, thereby solving the technical problem of low accuracy in determining freight demand and achieving the technical effect of improving the accuracy of determining freight demand. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a flow chart of a method for determining demand data of a route according to an embodiment of the present application;

[0019] Figure 2 is a schematic diagram of a model operation structure diagram according to an embodiment of the present application;

[0020] Figure 3 It is a schematic diagram of domestic route demand drawn according to the route demand distribution of a certain month according to an embodiment of the present application;

[0021] Figure 4 is a schematic diagram of a similarity analysis according to an embodiment of the present application;

[0022] Figure 5 is a schematic diagram of a device for determining demand data of a route according to an embodiment of the present application;

[0023] Figure 6 is a structural block diagram of a computer terminal according to an embodiment of the present application;

[0024] Figure 7 It is a block diagram of an electronic device for a method for determining demand data of a route according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application 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 data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising 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.

[0027] According to an embodiment of the present application, an embodiment of a method for determining demand data of an air route is provided. The steps shown in the flowchart of the accompanying drawing can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0028] In this embodiment, a method for determining demand data of a route is proposed. The method can train different sub-demand prediction models based on route pair samples of different frequencies. After obtaining the route pair to be predicted, the sub-demand prediction model matching the route pair to be predicted can be called to predict the route pair to be predicted, thereby obtaining demand data (that is, freight rate demand) corresponding to the route pair to be predicted. The method constructs different sub-demand prediction models for different types of route pairs to be predicted, and uses the corresponding sub-demand prediction models to predict the route pair to be predicted to obtain a definite prediction result, thereby solving the technical problem of low accuracy in determining freight rate demand and achieving the technical effect of improving the accuracy of determining freight rate demand.

[0029] Figure 1 is a flow chart of a method for determining demand data of a route according to an embodiment of the present application, such as Figure 1 As shown, the method may include the following steps:

[0030] Step S102: obtaining a pair of to-be-predicted flight paths of the flying device, wherein the to-be-predicted flight paths are used to characterize a region where the flying device moves.

[0031] In the technical solution provided in the above step S102 of the present application, the above-mentioned flight equipment can be an airplane, a flying vehicle, etc. The above-mentioned route pair to be predicted can be used to determine the area where the flight equipment moves, and can be a route pair of flight equipment, or a route origin-destination (Origin-Destination, referred to as OD), which can refer to a specific combination of origin and destination, and can be a city, a specific airport, or a special place identifier, which can be used to construct a route. The above-mentioned route may include one or more transit stops, with the final origin destination as a route. For example, if the route pair to be predicted is from Hangzhou to Beijing via Jilin, then Hangzhou to Jilin is a route. It should be noted that this is only an example, and there is no specific restriction on the type of flight equipment and the type of route pair to be predicted.

[0032] Optionally, the route pair to be predicted is determined by user input, selection, etc. It should be noted that this is only an example, and there is no specific limitation on the method of obtaining the route pair to be predicted.

[0033] Step S104, determining the usage frequency of the route pair to be predicted.

[0034] In the technical solution provided in the above step S104 of the present application, the above usage frequency can be used to determine the popularity of the route pair to be predicted, and can be used to characterize the number of routes of the route pair to be predicted. For example, the more routes the route pair to be predicted corresponds to, the higher the usage frequency of the route pair to be predicted, that is, the more popular the route pair to be predicted is.

[0035] Optionally, based on the number of route pairs to be predicted, the usage frequency of the route pairs to be predicted is determined, or the usage frequency of the route pairs to be predicted in the historical period can be determined. For example, in the past month, there are 10 routes flying between Beijing and Guangzhou, and the usage frequency of the route pair to be predicted (Beijing-Guangzhou) can be determined to be 10. It should be noted that this is only an example, and there is no specific limitation on the size and determination method of the usage frequency.

[0036] Step S106, from the demand forecasting model, call at least one sub-demand forecasting model that matches the usage frequency, predict the route pair to be predicted, and obtain demand data for the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand for the route pair to be predicted, and the sub-demand forecasting model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples.

[0037] In the technical solution provided in the above step S106 of the present application, the above demand forecasting model may include multiple sub-demand forecasting models. The above sub-demand forecasting model can be used to forecast the route pairs to be predicted with different usage frequencies, and can be trained based on multiple route pair samples in different usage frequency samples, and demand data samples corresponding to the route pair samples, and can be the current demand forecasting model. It should be noted that this is only an example, and there is no specific limitation on the method of constructing the sub-forecasting model.

[0038] Optionally, the above-mentioned demand data may be a fare demand, a demand for a type of fare product, and may include the characteristics of the route with a stopover, the sales amount, etc., and also include its binding rules, such as refund and change restrictions, usage restrictions, baggage restrictions, etc. In this embodiment, the demand data may refer to the sales amount, transfer station and other information in the fare demand. It should be noted that this is only an example, and there is no specific restriction on the type of demand data.

[0039] Optionally, the demand data samples and route pair samples can be determined based on historical sales data. Different usage frequency samples can be determined based on historical sales data.

[0040] Through the above-mentioned steps S102 and S106 of the present application, the route pairs to be predicted of the flight equipment are obtained, wherein the route pairs to be predicted are used to characterize the area where the flight equipment moves; the usage frequency of the route pairs to be predicted is determined; from the demand prediction model, at least one sub-demand prediction model that matches the usage frequency is called to predict the route pairs to be predicted, and the demand data of the route pairs to be predicted is obtained, wherein the demand data is used to characterize the freight rate demand of the route pairs to be predicted, and the sub-demand prediction model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples. That is to say, in the embodiment of the present application, different sub-demand forecasting models are trained based on route pair samples of different frequencies. After obtaining the route pair to be predicted, the sub-demand forecasting model matching the route pair to be predicted can be called to predict the route pair to be predicted, thereby obtaining the demand data (that is, freight rate demand) corresponding to the route pair to be predicted. The method constructs different sub-demand forecasting models for different types of route pairs to be predicted, and uses the corresponding sub-demand forecasting models to predict the route pairs to be predicted to obtain a definite forecasting result, thereby solving the technical problem of low accuracy in determining freight rate demand and achieving the technical effect of improving the accuracy in determining freight rate demand.

[0041] The above method of this embodiment is further introduced below.

[0042] As an optional implementation, step S106, from the demand forecasting model, calls at least one sub-demand forecasting model that matches the usage frequency, predicts the route pair to be predicted, and obtains demand data for the route pair to be predicted, including: determining the origin and destination of the route pair to be predicted; calling at least one sub-demand forecasting model that matches the usage frequency, origin and destination, predicts the route pair to be predicted, and obtains demand data.

[0043] In this embodiment, the origin and destination of the route to be predicted are determined. Based on the origin, destination and usage frequency, at least one sub-demand prediction model corresponding to the route to be predicted is determined, and the sub-demand prediction model is used to predict the route to be predicted to obtain demand data.

[0044] Optionally, since different prediction models are used for popular routes and non-popular routes, the sub-demand prediction model to be used for prediction may be determined based on the usage frequency of the route to be predicted.

[0045] Optionally, before training at least one sub-demand forecasting model, different regions need to be divided to determine different types of origins and destinations, so that sub-demand forecasting models corresponding to the same types of origins and destinations can be trained. After obtaining the route pair to be predicted, based on the usage frequency, origin and destination of the route pair to be predicted, the corresponding sub-demand forecasting model can be called to obtain the demand data corresponding to the route pair to be predicted.

[0046] In this embodiment, before training the sub-demand forecasting model, the route OD regional network can be split according to the characteristics of domestic routes. Taking some areas in China as an example, the division method of the origin and destination is illustrated. It should be noted that this is only an example, and there is no specific restriction on the division method based on the division method. In addition to dividing some areas in China, other areas such as the United States and the United Kingdom can also be divided. As long as the origin and destination in the route pair are divided, the method of training the sub-demand forecasting model based on the historical sales data corresponding to the divided origin and destination should be within the scope of protection of this application.

[0047] Optionally, during the course of historical development, routes have formed a network structure centered on some hub cities, and there are also many point-to-point flights, forming multiple administrations. Therefore, the route pairs can be split according to the above classification method. The process of classifying route pairs may include: (i) For routes within the area covered by each administration / region, the actual origin and destination location identifiers are used as route pairs; (ii) If a route crosses regions, OD conversion is performed, and the part within the current research area uses the original location identifier, and the part within the non-location slice area uses the regional code as the identifier. Table 1 is a table of regions and corresponding cities. As shown in Table 1, Beijing, Tianjin, Hebei, Shanxi, and Inner Mongolia can be divided into the North China region; Heilongjiang, Jilin, and Liaoning can be divided into the Northeast region; Shandong, Jiangsu, Anhui, Shanghai, Zhejiang, Jiangxi, and Fujian can be divided into the East China region.

[0048] It should be noted that this is only an example, and there are no specific restrictions on the method of regional division, the content of the division, the code of the area, and the type of cities included, and it is not limited to the following classification method.

[0049] Table 1 Regions and corresponding cities

[0050] Area and code Included cities North China QY1 Beijing, Tianjin, Hebei, Shanxi, Inner Mongolia Northeast Region QY2 Heilongjiang, Jilin, Liaoning East China QY3 Shandong, Jiangsu, Anhui, Shanghai, Zhejiang, Jiangxi, Fujian Central and Southern Region QY4 Henan, Hubei, Hunan, Guangdong, Guangxi Zhuang Autonomous Region, Hainan Southwest Region QY5 Sichuan, Chongqing, Guizhou, Yunnan, Tibet Northwest Region QY6 Qinghai, Gansu, Ningxia Hui Autonomous Region, Shaanxi Xinjiang Region QY7 Xinjiang

[0051] Optionally, Table 2 is an example table of regional splitting. As shown in Table 2, if the origin and destination of the route pair sample are Beijing and Hohhot respectively, it can be determined that the route pair sample can be used to train the demand forecast model corresponding to the North China region. If the origin and destination of the route pair sample are Beijing and Shanghai respectively, it can be determined that the route pair sample can be used to train the demand forecast model for popular routes. If the origin and destination of the route pair sample are Hohhot and Baotou respectively, the route pair sample can be used to train the sub-demand forecast model corresponding to the North China region. If the origin and destination of the route pair sample are Beijing and Shenzhen respectively, the route pair sample can be used to train the demand forecast model for popular routes.

[0052] Table 2 Area splitting example table

[0053]

[0054] Optionally, through the above regional division, the number of routes that need to be studied within each regional scope can be significantly reduced, and the route splitting and route volume reduction can be achieved while retaining business relationships.

[0055] Optionally, the files after the above route data processing can be trained separately using 7 graphics processing units (GPU) to train the sub-demand forecast model corresponding to the North China region, the sub-demand forecast model corresponding to the Northeast region, the sub-demand forecast model corresponding to the East China region, the sub-demand forecast model corresponding to the Central and South regions, the sub-demand forecast model corresponding to the Southwest region, the sub-demand forecast model corresponding to the Northwest region, and the sub-demand forecast model corresponding to the Xinjiang region. It should be noted that the number of GPU units can be adjusted according to the division of regions, and there is no specific restriction on the number of GPU units here.

[0056] Figure 2 is a schematic diagram of a model operation structure diagram according to an embodiment of the present application, such as Figure 2 As shown, a demand forecasting model can be trained based on historical sales data. The demand forecasting model may include at least one sub-demand forecasting model, such as a demand forecasting model 201 for region 1, a demand forecasting model 202 for region 2, a demand forecasting model 203 for region 3, ..., a demand forecasting model 205 for region 7, and a demand forecasting model 204 for popular routes. It should be noted that the number and type of sub-demand forecasting models are only for illustrative purposes and are not specifically limited here.

[0057] For example, if the graphics processor is 8 graphics cards, the 8 graphics cards can be split into demand forecasting models corresponding to "7 major regions + popular routes". This splitting method can be run on different graphics cards for model training and optimization. The basic model reasoning process without route similarity analysis is relatively fast, and its accuracy is acceptable, so it can be used as a base model. On this basis, when the GPU is idle, the reasoning analysis of route similarity within the corresponding model range can be performed. This part of the work is time-consuming, but it can improve the accuracy of model operation. On the basis of the existing model (which can be a model containing only the base model, or a "base model + route similarity model" that has been iterated using some data), historical sales data can be continuously introduced to train and update model parameters, and regularly synchronized to the currently running model to achieve continuous optimization of the model. In this way, sub-demand forecasting models deployed on different graphics cards can be trained. When the route pair to be predicted is obtained, based on the frequency of use of the route pair to be predicted, the corresponding sub-demand forecasting model can be called to predict the route pair to be predicted to obtain the demand data corresponding to the route pair to be predicted.

[0058] Optionally, after obtaining the route pair to be predicted, a corresponding sub-demand prediction model can be determined based on the usage frequency, origin and destination of the route pair to be predicted, and the route pair to be predicted is predicted using the sub-demand prediction model to obtain corresponding demand data.

[0059] The following is a further example of the process of determining at least one sub-demand forecasting model that matches the route pair to be predicted based on the usage frequency, the origin and the destination, and forecasting the route pair to be predicted based on the sub-demand forecasting model to obtain demand data.

[0060] As an optional implementation, at least one sub-demand forecasting model that matches the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being greater than the frequency threshold, and the origin and the destination belonging to different areas, the first sub-demand forecasting model is called to predict the route pair to be predicted and obtain demand data, wherein the first sub-demand forecasting model is trained based on route pair samples in the usage frequency sample whose usage frequency is greater than the frequency threshold, and the demand data template corresponding to the route pair samples.

[0061] In this embodiment, if the usage frequency of the route to be predicted is greater than the frequency threshold, and the origin and the destination belong to different regions, that is, the origin and the destination belong to different regions, then the first sub-demand prediction model can be called up, and the route pair to be predicted can be predicted using the first sub-demand prediction model to obtain demand data. The first sub-demand prediction model can be trained based on the route pair samples whose usage frequency is greater than the frequency threshold in the usage frequency samples, and the demand data samples corresponding to the route pair samples, and can be a popular route demand prediction model. The usage frequency samples can be historical sales data, and can include route pair samples at different usage frequencies.

[0062] Optionally, different regions can be divided in advance. When a route pair to be predicted is obtained, it can be determined whether the origin and destination in the route pair to be predicted belong to the same region. If they belong to different regions and the route pair to be predicted is a popular route pair, a pre-trained popular route demand prediction model (i.e., the first sub-demand prediction model) can be used to predict the route pair to be predicted to obtain demand data. The first sub-demand prediction model can be trained based on a route pair sample constructed based on the origin and destination, or can be trained based on a route pair sample similar to the origin and destination.

[0063] For example, if the route pair to be predicted involves two regions, such as Beijing to Shanghai, where Beijing belongs to the North China region and Shanghai belongs to the sliding region. It can be determined that part of the data of the route pair to be predicted is classified into a route in the East China region (QY3) where Beijing and Shanghai are located in the North China regional model (QY1). Therefore, the prediction results of the sub-demand prediction model corresponding to the East China region or the sub-demand prediction model corresponding to the North China region cannot be used as a reflection of the demand from Beijing to Shanghai. Therefore, the results of the hot route demand prediction model can be used as the final prediction result (that is, demand data).

[0064] Figure 3 1 is a schematic diagram of the demand for domestic routes drawn according to the distribution of the demand for routes in a certain month according to an embodiment of the present application. The horizontal axis can be the route code, the vertical axis can be the demand for the route, and the dotted line represents the dividing line of 2,000 air tickets sold per month. Figure 3 As shown, route demand has an obvious long-tail effect, that is, a large part of concentrated demand is concentrated at the head, while there is also a large part of scattered, customized and differentiated demand distributed at the tail.

[0065] Optionally, by Figure 3 It can be seen that the changes in demand for popular routes are relatively stable and have a higher value for concentrated research. Therefore, in synchronization with the network splitting combined with business logic, hot routes can be studied separately in parallel. Hot route pairs can be trained separately to obtain the sub-demand prediction model corresponding to the hot route pairs. Further, the sub-demand prediction model can be used to predict the hot route pairs to obtain the demand data of the hot route pairs.

[0066] like Figure 3 As shown in the figure, the monthly sales volume is 2,000 tickets involving 1,544 routes. This is within the critical value range where a single card can operate with a time slice of more than 4 hours. This part of demand and products is generally stable, and can be analyzed with monthly data, and the demand and its changes can be predicted with a time slice of 6 hours or 8 hours. Therefore, this part of route data, that is, the popular route data, can be extracted, and a small graph can be drawn separately and the demand forecast analysis can be performed to effectively learn the features of a large amount of data and retain the data information to the maximum extent. The routes involved in this part of the data can be trained using a separate GPU unit.

[0067] As an optional implementation, at least one sub-demand forecasting model that matches the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being greater than the frequency threshold, and the origin and the destination belonging to the same area, the first sub-demand forecasting model is called to predict the route pair to be predicted and obtain a first prediction result, and the second sub-demand forecasting model that matches the same area is called to predict the route pair to be predicted and obtain a second prediction result, wherein the second sub-demand forecasting model is trained based on route pair samples in the same area and is different from the first sub-demand forecasting model; the first prediction result and the second prediction result are weighted to obtain demand data.

[0068] In this embodiment, if the usage frequency is greater than the frequency threshold, and the origin and the destination belong to the same area, the first sub-demand forecasting model can be called to predict the route pair to be predicted to obtain a first prediction result. At the same time, the second sub-demand forecasting model matching the same area can be called to predict the route pair to be predicted to obtain a second prediction result. The first prediction result and the second prediction result can be weighted to obtain demand data. Among them, the above-mentioned second sub-demand forecasting model can be obtained by training based on route pair samples in the same area, and can be a regional demand forecasting model.

[0069] Optionally, if the usage frequency is greater than the frequency threshold, it can be determined that the route pair to be predicted is a popular route pair, and it can be further determined whether the origin and destination in the route pair to be predicted belong to the same area. If they belong to the same area, the second sub-demand forecasting model corresponding to the area can be called to predict the route pair to be predicted to obtain a second prediction result, and the first sub-demand forecasting model can be called to predict the route pair to be predicted to obtain a first prediction result. Based on the first prediction result and the second prediction result, demand data can be obtained.

[0070] For example, if the route pair to be predicted is a popular route pair and only involves a single region, such as the route between Shenzhen and Wuhan, both belong to the Central and Southern regions (QY4). At this time, the regional prediction model and the popular route prediction model of the Central and Southern regions both have prediction results for this route. Since the demand changes for popular routes are relatively stable, the results obtained by the popular route demand prediction model are more accurate. The current final prediction result can be weighted by the demand prediction model for popular routes and the regional demand prediction model, and the weights can be 0.7 and 0.3 respectively, so as to obtain the final demand data. It should be noted that the size of the numbers here is only for example, and can be adjusted according to the actual operating conditions, or dynamic adjustment of weights can be achieved through other methods such as Kalman filters.

[0071] Optionally, a route pair to be predicted is obtained, and it is determined whether the route pair to be predicted is a popular route. This can be determined by the usage frequency corresponding to the route pair to be predicted. That is, if the usage frequency is greater than the frequency threshold, the route pair to be predicted can be determined to be a popular route. If the usage frequency is less than or equal to the frequency threshold, the route pair to be predicted can be determined to be an unpopular route. If the route pair to be predicted is a popular route, it can be further determined whether the origin and the destination belong to the same area. If the origin and the destination belong to the same area, the final demand data can be determined based on the results of the first sub-demand forecasting model and the second sub-demand forecasting model. If the origin and the destination belong to different areas, the demand data can be determined based on the first sub-demand forecasting model.

[0072] As an optional implementation, at least one sub-demand forecasting model that matches the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being less than the frequency threshold and the origin and the destination belonging to different areas, the third sub-demand forecasting model that matches the origin is called to predict the route pair to be predicted and obtain demand data, or the fourth sub-demand forecasting model that matches the destination is called to predict the route pair to be predicted and obtain demand data.

[0073] In this embodiment, if the usage frequency is less than the frequency threshold and the origin and destination belong to different areas, the third sub-demand forecasting model matching the origin can be called to predict the route pair to be predicted to obtain demand data, or the fourth sub-demand forecasting model matching the destination can be called to obtain demand data.

[0074] Optionally, the above-mentioned third sub-demand forecasting model can be trained based on route pair samples of the departure place and route pair samples corresponding to the region belonging to the same area as the departure place. For example, if the departure place is Beijing, the third sub-demand forecasting model corresponding to the departure place can be trained based on route pair samples corresponding to the North China region.

[0075] Optionally, if the route pair to be predicted involves two regions, the current model can only count the statistics from the origin or destination to other regions, and there is no separate data for this part. However, generally speaking, the number of non-popular routes involving cross-regions is small, so this method does not lose too much data information.

[0076] As an optional implementation, at least one sub-demand forecasting model matching the usage frequency, the origin and the destination is called to predict the route pair to be predicted and obtain demand data, including: in response to the usage frequency being less than the frequency threshold and the origin and the destination belonging to the same area, calling a second sub-demand forecasting model matching the same area to predict the route pair to be predicted and obtain demand data.

[0077] In this embodiment, if the route pair to be predicted is not a popular route pair, that is, the usage frequency of the route pair to be predicted is less than the frequency threshold, it can be determined whether the origin and the destination belong to the same area. If the origin and the destination belong to the same area, the second sub-demand forecasting model corresponding to the same area can be called to predict the route to be predicted to obtain demand data.

[0078] Optionally, if the route pair to be predicted only involves a single region, a regional model (ie, the second sub-demand forecasting model) may be used to obtain a forecasting result.

[0079] The following is a further introduction to a method for determining demand data for a route proposed in an embodiment of the present application.

[0080] At present, the forecast of route demand is the basis of airline revenue management. The aviation industry is developing rapidly, with more than 15,000 routes. Even if we only consider the sales data of completed transactions, the amount of data in the entire market is huge.

[0081] In related technologies, a graph model is usually introduced to study the connectivity of routes, i.e., graph networks, for learning geographical characteristics; or a Markov chain model is introduced to predict the continuity and timeliness of demand, i.e., the impact of the demand in the previous time slice on the demand in the next time slice; or a similarity comparison model is introduced to classify similar routes, and enhance the model's ability to understand changes in similar routes. However, building a model for analysis of the entire market data places high demands on hardware equipment, and the entire process from data extraction to preprocessing to machine learning takes a long time. The overall time span of the amount of data that a general GPU server can support for model training is short, and the above method makes it difficult to learn the annual demand change characteristics of routes in the entire market. Even if only data with a short time span is used, data statistics and feature compression must be performed before the model can be used for learning.

[0082] From the above, it can be seen that under the current architecture, the bottlenecks affecting performance may include the following aspects: (i) The biggest factor is the number of routes. The increase in the number of routes will lead to an exponential increase in the model parameters, which is the decisive factor that makes the model difficult to run. The current practical results show that a single graphics card with 32G video memory can only support the construction and training process of a network diagram containing about 1,500 routes. This part of the limitation is an absolute limitation of hardware resources. More routes cannot be loaded onto the same GPU card under the current model state. (ii) The calculation process of route similarity. In order to better learn routes with similar change rules and understand the market status, it is necessary to analyze the similarity of routes. The impact of this process on model performance is mainly concentrated on the running time of the model. Comparing each route one by one according to the time slice will take up a lot of computing time. This process is also the biggest time bottleneck process of this model. At the same time, its results are used as the input of subsequent model training, and occupy the video memory at the same time as the training data, which also limits the amount of data that the model can handle at the storage level. (III) Statistical time slice range. Currently, even without introducing route similarity results, a general 32 GB GPU memory can only support the analysis of demand statistics for time slices of 4 hours or more for the critical 1,500 routes. It is difficult to further characterize more route detail factors.

[0083] It can be seen that the above method cannot combine business characteristics to effectively mine freight demand information, cannot retain and mine information to the maximum extent and support distributed learning processes, and therefore has a technical problem of low accuracy in determining freight demand.

[0084] To solve the above problems, in this embodiment, the current demand forecasting model is optimized from the following dimensions combined with business characteristics: (a) domestic routes are divided into multiple regions according to the management bureau to which their administrative regions belong, and for each region, the routes involved in the region are split and preprocessed, and the generated data files are imported into the corresponding models to train the sub-demand forecasting models corresponding to different regions; (b) from all the data, the popular route data (which can be routes with monthly sales of more than 2,000 tickets) are extracted and deployed separately on a card for learning to generate a sub-demand forecasting model for predicting popular routes; (c) for routes and popular routes in the region, on the basis of the current existing prediction model, when the corresponding GPU is idle, the route similarity and historical data are introduced for learning to regularly update and optimize the model parameters of the sub-demand forecasting model. Through the above content, the technical problem of low accuracy in determining freight demand is solved, and the technical effect of improving the accuracy of determining freight demand is achieved.

[0085] In this embodiment, the route similarity analysis process is a further optimization of the first two analyses, and may include similarity analysis of route area networks and similarity analysis of popular routes. A dynamic time warping (DTW) algorithm may be used to complete the route similarity analysis process.

[0086] Alternatively, affected by climate change and hot events, the changes in demand in a city and its neighboring cities during certain time periods are related. For example, during the Chinese New Year, the routes of popular cities and their neighboring cities will show a clear growth trend in demand, and the time of change is also similar. For another example, after Zibo barbecue becomes popular, it will also drive the growth of tourism demand in surrounding cities. Similarity analysis of routes in the region can capture these trends well. Then we can better learn and understand the data, and understand the changes in related route demand caused by time and event factors.

[0087] Alternatively, popular routes usually include hub cities, and their changes themselves have strong correlations, so it is more necessary to consider the similarity of their routes. The above popular routes can be determined based on route pairs.

[0088] Optionally, the DTW algorithm may be an algorithm for comparing the frame shifts of two time series. Through the DTW algorithm, a comparative analysis of the similarities of two routes may be performed, thereby learning the correlation between the routes, and synchronous or asynchronous change rules, thereby achieving more accurate predictions. Figure 4 is a schematic diagram of a similarity analysis according to an embodiment of the present application, such as Figure 4 As shown in FIG. 1 , DTW similarity analysis can be performed on two routes using a 6-hour time slice.

[0089] Optionally, due to the large amount of sales data for the route, the pair-by-pair comparison using the DTW algorithm is relatively slow. In order to better support the timeliness of demand forecasting, this part of the calculation and parameter learning can be performed asynchronously and updated in stages to the model actually used.

[0090] In this embodiment, combined with the business characteristics of domestic air transport, a method for route demand forecasting business splitting and distributed learning is proposed, which supports the mining of route geometric characteristics and time-varying similarity characteristics, can significantly improve the timeliness of the forecast, fully retain data business information, learn the knowledge contained in the data, and effectively solve the complex geographic association map caused by the large number of routes. Difficult to construct, split and related problems.

[0091] According to an embodiment of the present application, a device for determining demand data of a route is also provided. It should be noted that the device for determining demand data of a route of this embodiment can be used to execute the method for determining demand data of a route of embodiment 1 of the present application.

[0092] Figure 5 is a schematic diagram of a device for determining demand data of a route according to an embodiment of the present application. Figure 5 As shown, the device 50 for determining the demand data of the route may include: an acquisition unit 502 , a determination unit 504 and a processing unit 506 .

[0093] The acquisition unit 502 is used to acquire a pair of to-be-predicted flight paths of the flying device, wherein the to-be-predicted flight paths are used to characterize an area where the flying device moves.

[0094] The determination unit 504 is used to determine the usage frequency of the route pair to be predicted.

[0095] Processing unit 506 is used to call at least one sub-demand forecasting model that matches the usage frequency from the demand forecasting model, predict the route pair to be predicted, and obtain demand data for the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand for the route pair to be predicted, and the sub-demand forecasting model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples.

[0096] The device for determining the demand data of the route of this embodiment obtains the route pair to be predicted of the flight equipment through the acquisition unit 502, wherein the route to be predicted is used to characterize the area where the flight equipment moves; determines the usage frequency of the route pair to be predicted through the determination unit 504; and calls at least one sub-demand prediction model that matches the usage frequency from the demand prediction model through the processing unit 506 to predict the route pair to be predicted, and obtains the demand data of the route pair to be predicted, wherein the demand data is used to characterize the fare demand of the route pair to be predicted, and the sub-demand prediction model is obtained by training multiple route pair samples based on samples of different usage frequencies and demand data samples corresponding to the route pair samples, thereby solving the technical problem of low accuracy in determining the fare demand and achieving the technical effect of improving the accuracy in determining the fare demand.

[0097] The embodiment of the present application may provide a computer terminal, which may be any computer terminal device in a computer terminal group. Optionally, in this embodiment, the computer terminal may also be replaced by a terminal device such as a mobile terminal.

[0098] Optionally, in this embodiment, the computer terminal may be located in at least one network device among a plurality of network devices of the computer network.

[0099] In this embodiment, the above-mentioned computer terminal can execute the program code of the following steps in the method for determining the demand data of the route: obtaining the route pair to be predicted of the flight equipment, wherein the route to be predicted is used to characterize the area where the flight equipment moves; determining the usage frequency of the route pair to be predicted; calling at least one sub-demand prediction model that matches the usage frequency from the demand prediction model, predicting the route pair to be predicted, and obtaining the demand data of the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand of the route pair to be predicted, and the sub-demand prediction model is trained based on multiple route pair samples with different usage frequency samples, and demand data samples corresponding to the route pair samples.

[0100] Optionally, Figure 6 is a structural block diagram of a computer terminal according to an embodiment of the present application, such as Figure 6 As shown, the computer terminal 608 may include: one or more (only one is shown in the figure) processors 602 , a memory 604 , and a transmission device 606 .

[0101] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the method and device for determining the demand data of the route in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned method for determining the demand data of the route. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer terminal 608 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0102] The processor can call the information and application programs stored in the memory through the transmission device to execute the following steps: obtain the route pair to be predicted of the flight equipment, wherein the route to be predicted is used to characterize the area where the flight equipment moves; determine the usage frequency of the route pair to be predicted; call at least one sub-demand prediction model that matches the usage frequency from the demand prediction model to predict the route pair to be predicted, and obtain the demand data of the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand of the route pair to be predicted, and the sub-demand prediction model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples.

[0103] It can be understood by those skilled in the art that Figure 6The structure shown is for illustration only, and the computer terminal 608 may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (MID for short), a PAD, or other terminal devices. Figure 6 The structure of the computer terminal 608 is not limited. For example, the computer terminal 608 may also include Figure 6 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Figure 6 Different configurations shown.

[0104] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0105] According to an embodiment of the present application, a computer-readable storage medium is further provided, the computer-readable storage medium comprising a stored program, wherein the program executes the method for determining demand data of the route in the above embodiment.

[0106] Optionally, in this embodiment, the computer-readable storage medium may be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the mobile terminals in a mobile terminal group.

[0107] Optionally, in this embodiment, the computer-readable storage medium is configured to store program code for executing the following steps: obtaining a pair of routes to be predicted for the flight equipment, wherein the route to be predicted is used to characterize the area where the flight equipment moves; determining the usage frequency of the route pair to be predicted; and calling at least one sub-demand forecasting model that matches the usage frequency from the demand forecasting model to forecast the route pair to be predicted, and obtaining demand data for the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand for the route pair to be predicted, and the sub-demand forecasting model is trained based on multiple route pair samples with different usage frequency samples, and demand data samples corresponding to the route pair samples.

[0108] Optionally, the computer-readable storage medium may also execute program code for the following steps: determining the origin and destination of the route pair to be predicted; calling at least one sub-demand prediction model that matches the usage frequency, origin and destination, predicting the route pair to be predicted, and obtaining demand data.

[0109] Optionally, the computer-readable storage medium can also execute the program code of the following steps: in response to the usage frequency being greater than the frequency threshold, and the departure place and the destination belonging to different areas, calling the first sub-demand forecasting model, forecasting the route pair to be predicted, and obtaining demand data, wherein the first sub-demand forecasting model is trained based on route pair samples in the usage frequency samples whose usage frequency is greater than the frequency threshold, and the demand data corresponding to the route pair samples.

[0110] Optionally, the computer-readable storage medium may also execute the following steps of program code: in response to the usage frequency being greater than a frequency threshold, and the departure place and the destination belonging to the same area, calling the first sub-demand forecasting model to predict the route pair to be predicted, and obtaining a first prediction result, and calling the second sub-demand forecasting model matching the same area to predict the route pair to be predicted, and obtaining a second prediction result, wherein the second sub-demand forecasting model is trained based on route pair samples in the same area, and is different from the first sub-demand forecasting model; weighting the first prediction result and the second prediction result to obtain demand data.

[0111] Optionally, the computer-readable storage medium may also execute program code for the following steps: in response to the usage frequency being less than a frequency threshold, and the origin and destination belonging to the same area, calling a second sub-demand forecasting model matching the same area, forecasting the route pair to be predicted, and obtaining demand data.

[0112] In this embodiment, different sub-demand forecasting models are trained based on route pair samples of different frequencies. After obtaining the route pair to be predicted, the sub-demand forecasting model matching the route pair to be predicted can be called to predict the route pair to be predicted, thereby obtaining the demand data (i.e., freight rate demand) corresponding to the route pair to be predicted. This method constructs different sub-demand forecasting models for different types of route pairs to be predicted, and uses the corresponding sub-demand forecasting models to predict the route pairs to be predicted to obtain a definite forecasting result, thereby solving the technical problem of low accuracy in determining freight rate demand and achieving the technical effect of improving the accuracy in determining freight rate demand.

[0113] According to an embodiment of the present application, a processor is also provided, which is used to run a program, wherein when the program is run by the processor, the method for determining the demand data of the route in the above embodiment is executed.

[0114] Optionally, in this embodiment, the computer terminal may be located in at least one network device among a plurality of network devices of the computer network.

[0115] In this embodiment, the above-mentioned computer terminal can execute the program code of the following steps in the method for determining the demand data of the route: obtaining the route pair to be predicted of the flight equipment, wherein the route to be predicted is used to characterize the area where the flight equipment moves; determining the usage frequency of the route pair to be predicted; calling at least one sub-demand prediction model that matches the usage frequency from the demand prediction model, predicting the route pair to be predicted, and obtaining the demand data of the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand of the route pair to be predicted, and the sub-demand prediction model is trained based on multiple route pair samples with different usage frequency samples, and demand data samples corresponding to the route pair samples.

[0116] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the method and device for determining the demand data of the route in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned method for determining the demand data of the route. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0117] The processor can call the information and application programs stored in the memory through the transmission device to execute the following steps: obtain the route pair to be predicted of the flight equipment, wherein the route to be predicted is used to characterize the area where the flight equipment moves; determine the usage frequency of the route pair to be predicted; call at least one sub-demand prediction model that matches the usage frequency from the demand prediction model to predict the route pair to be predicted, and obtain the demand data of the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand of the route pair to be predicted, and the sub-demand prediction model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples.

[0118] Optionally, the processor may also execute the following program code: determining the origin and destination of the route pair to be predicted; calling at least one sub-demand prediction model that matches the usage frequency, origin and destination, predicting the route pair to be predicted, and obtaining demand data.

[0119] Optionally, the processor may also execute the program code of the following steps: in response to a usage frequency greater than a frequency threshold, and the departure place and the destination belong to different areas, calling the first sub-demand forecasting model, forecasting the route pair to be predicted, and obtaining demand data, wherein the first sub-demand forecasting model is trained based on route pair samples in the usage frequency samples whose usage frequency is greater than the frequency threshold, and the demand data corresponding to the route pair samples.

[0120] Optionally, the processor may also execute the following steps of program code: in response to the usage frequency being greater than a frequency threshold, and the departure place and the destination belonging to the same area, calling the first sub-demand forecasting model to predict the route pair to be predicted, and obtaining a first prediction result, and calling the second sub-demand forecasting model matching the same area, and predicting the route pair to be predicted, and obtaining a second prediction result, wherein the second sub-demand forecasting model is trained based on route pair samples in the same area, and is different from the first sub-demand forecasting model; weighting the first prediction result and the second prediction result to obtain demand data.

[0121] Optionally, the processor may also execute the program code of the following steps: in response to the usage frequency being less than a frequency threshold, and the origin and destination belonging to different regions, calling the third sub-demand forecasting model that matches the origin, predicting the route pair to be predicted, and obtaining demand data, or calling the fourth sub-demand forecasting model that matches the destination, predicting the route pair to be predicted, and obtaining demand data.

[0122] Optionally, the processor may also execute the following program code: in response to the usage frequency being less than a frequency threshold, and the origin and destination belonging to the same area, calling a second sub-demand forecasting model matching the same area, forecasting the route pair to be predicted, and obtaining demand data.

[0123] By adopting the embodiment of the present application, different sub-demand forecasting models are trained based on route pair samples of different frequencies. After obtaining the route pair to be predicted, the sub-demand forecasting model matching the route pair to be predicted can be called to predict the route pair to be predicted, thereby obtaining the demand data (that is, freight rate demand) corresponding to the route pair to be predicted. The method constructs different sub-demand forecasting models for different types of route pairs to be predicted, and uses the corresponding sub-demand forecasting models to predict the route pairs to be predicted to obtain a definite forecasting result, thereby solving the technical problem of low accuracy in determining freight rate demand and achieving the technical effect of improving the accuracy of determining freight rate demand.

[0124] According to an embodiment of the present application, a computer program product is also provided. The computer program product includes computer instructions, wherein when the computer instructions are executed by a processor, the method for determining the demand data of the route in the above embodiment is implemented.

[0125] An embodiment of the present application may provide an electronic device, which may include a memory and a processor.

[0126] Figure 7 1 is a block diagram of an electronic device for determining a method for demand data of a route according to an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0127] like Figure 7 As shown, the device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 to a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0128] A number of components in the device 700 are connected to the I / O interface 705, including: an input unit 706, such as a keyboard, a mouse, etc.; an output unit 707, such as various types of displays, speakers, etc.; a storage unit 708, such as a disk, an optical disk, etc.; and a communication unit 709, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0129] The computing unit 701 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 701 performs the various methods and processes described above, such as a verification method for data. For example, in some embodiments, the verification method for data may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 708. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 700 via the ROM 702 and / or the communication unit 709. When the computer program is loaded into the RAM 703 and executed by the computing unit 701, one or more steps of the verification method for data described above may be executed. Alternatively, in other embodiments, the computing unit 701 may be configured to execute the data verification method in any other appropriate manner (for example, by means of firmware).

[0130] According to an embodiment of the present application, a method for determining demand data for a route is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0131] Various embodiments of the systems and techniques described above herein may be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs that are executable and / or interpreted on a programmable system including at least one programmable processor that may be a dedicated or general purpose programmable processor that may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0132] The program code for implementing the method of the present application can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0133] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer 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 foregoing.

[0134] To provide interaction with a user, the systems and techniques described herein may be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or an LCD (liquid crystal display, monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a path ball) through which the user can provide input to the computer. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0135] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a Local Area Network (LAN), a Wide Area Network (WAN), and the Internet.

[0136] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0137] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0138] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0140] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0141] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0142] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0143] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for determining demand data of a route, characterized in that: include: Acquire a pair of to-be-predicted flight paths of an aircraft, wherein the to-be-predicted flight paths are used to characterize an area where the aircraft moves; Determining the usage frequency of the to-be-predicted route pair; From the demand forecasting model, call at least one sub-demand forecasting model that matches the usage frequency, predict the route pair to be predicted, and obtain demand data for the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand for the route pair to be predicted, and the sub-demand forecasting model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples.

2. The method according to claim 1, characterized in that The step of calling at least one sub-demand forecasting model that matches the usage frequency from the demand forecasting model to forecast the route pair to be forecasted, and obtaining demand data for the route pair to be forecasted, includes: Determining the origin and destination of the route pair to be predicted; At least one sub-demand forecasting model matching the usage frequency, the origin and the destination is called to forecast the route pair to be forecasted to obtain the demand data.

3. The method according to claim 2, characterized in that The calling of at least one sub-demand forecasting model matching the usage frequency, the origin, and the destination to forecast the to-be-forecasted route pair to obtain the demand data includes: In response to the usage frequency being greater than the frequency threshold, and the departure place and the destination belonging to different areas, the first sub-demand forecasting model is called up to predict the route pair to be predicted to obtain the demand data, wherein the first sub-demand forecasting model is trained based on the route pair samples in the usage frequency samples whose usage frequency is greater than the frequency threshold, and the demand data samples corresponding to the route pair samples.

4. The method according to claim 3, characterized in that The calling of at least one sub-demand forecasting model matching the usage frequency, the origin, and the destination, forecasting the to-be-forecasted route pair, and obtaining the demand data includes: In response to the usage frequency being greater than the frequency threshold, and the origin and the destination belonging to the same area, calling the first sub-demand forecasting model to forecast the route pair to be forecasted to obtain a first forecast result, and calling the second sub-demand forecasting model matching the same area to forecast the route pair to be forecasted to obtain a second forecast result, wherein the second sub-demand forecasting model is trained based on the route pair samples in the same area and is different from the first sub-demand forecasting model; The first prediction result and the second prediction result are weighted to obtain the demand data.

5. The method according to claim 2, characterized in that: Invoking at least one sub-demand forecasting model that matches the usage frequency, the origin, and the destination to forecast the to-be-forecasted route pair to obtain the demand data includes: In response to the usage frequency being less than the frequency threshold, and the origin and the destination belonging to different areas, the third sub-demand forecasting model matching the origin is called to predict the route pair to be predicted to obtain the demand data, or the fourth sub-demand forecasting model matching the destination is called to predict the route pair to be predicted to obtain the demand data.

6. The method according to claim 2, characterized in that The calling of at least one sub-demand forecasting model matching the usage frequency, the origin, and the destination, forecasting the to-be-forecasted route pair, and obtaining the demand data includes: In response to the usage frequency being less than a frequency threshold, and the origin and the destination belonging to the same area, a second sub-demand forecasting model matching the same area is retrieved to forecast the route pair to be forecasted, thereby obtaining the demand data.

7. A device for determining demand data of a route, characterized in that: include: An acquisition unit, configured to acquire a pair of to-be-predicted flight paths of an aircraft, wherein the to-be-predicted flight paths are used to characterize an area where the aircraft moves; A determination unit, used to determine the usage frequency of the to-be-predicted route pair; A processing unit is used to call at least one sub-demand forecasting model that matches the usage frequency from the demand forecasting model, forecast the route pair to be predicted, and obtain demand data for the route pair to be predicted, wherein the demand data is used to characterize the freight rate demand for the route pair to be predicted, and the sub-demand forecasting model is trained based on multiple route pair samples with different usage frequency samples and demand data samples corresponding to the route pair samples.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 6.

9. A computer program product, characterized in that The method comprises computer instructions, which implement the method according to any one of claims 1 to 6 when executed by a processor.

10. A processor, characterized in that: The processor is used to run a program, wherein the program executes the method according to any one of claims 1 to 6 when run by the processor.

Images