A bus user trip purpose prediction method and device based on multi-source data

Abstract

The application relates to a bus user travel destination prediction method and device based on multi-source data, which comprises the following steps: establishing feature engineering based on multi-source data joint analysis of urban residents, wherein the multi-source data comprises travel survey data, bus card data and interest point data; dividing records with regular bus travel mode in the travel survey data into first-class travel data and second-class travel data according to whether the home address and the work address in the records are both filled in, and respectively constructing a first model and a second model by using a machine learning model; identifying the home station address and the work station address of each user in the bus card data, and dividing the bus card data into first-class and second-class card data according to whether the home station address and the work station address of the user can both be identified; and predicting the travel destination by using the first model and the second model for the classified card data respectively. Compared with the prior art, the application has the advantages of high prediction accuracy and high efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of bus user travel purpose prediction, and particularly relates to a bus user travel purpose prediction method based on multi-source data. BACKGROUND

[0002] In the field of transportation, data-driven research methods are rapidly emerging, and optimizing public transportation services through big data has become a new trend. However, public transportation card data often lacks information about individual travel purposes, making it difficult to understand the motivations and reasons behind user travel behavior and limiting the in-depth understanding of user behavior by public transportation operators.

[0003] At the same time, resident travel survey data contains rich individual activity spatiotemporal characteristics and travel purpose information, which is of great significance for analyzing the relationship between public transportation user travel behavior and travel purpose. These survey data often require a lot of manpower and resources, resulting in a small sample size. Although card data can be automatically collected, it cannot directly obtain the travel purpose of the user. If the advantages of these two types of data can be combined, a large-scale card data set containing user travel purposes can be constructed, thereby laying the foundation for studying the relationship between user travel demand and travel motivation. Current research has not effectively integrated these two types of data, which has led to challenges in inferring the travel purpose of regular public transportation users.

[0004] After searching, Chinese invention patent publication No. CN111968018A discloses a subway travel purpose identification method and device, which can improve the identification accuracy of travel purpose. The method includes: obtaining subway travel samples, and obtaining land use feature information around subway stations; dividing the subway travel samples into a training set and a test set; taking the travel feature information in the training set and the land use feature information around the corresponding travel start and end subway stations as input, and taking the travel purpose as a prediction variable, training a classifier, and verifying the trained classifier based on the test set to obtain the verified classifier; inputting the target travel feature information corresponding to each target travel record in the public transportation smart card data and the land use feature information around the corresponding target travel start and end subway stations into the verified classifier to obtain the travel purpose identification result of each target travel record. The existing patent has the problem of not classifying travel data, resulting in low identification accuracy.

[0005] How to accurately predict the travel purpose of public transportation users has become a technical problem to be solved. SUMMARY

[0006] The purpose of the present application is to overcome the defects of the prior art and provide a bus user travel purpose prediction method based on multi-source data.

[0007] The object of the present application can be achieved by the following technical solutions:

[0008] According to one aspect of the present application, a bus user travel purpose prediction method based on multi-source data is provided, which comprises the following steps:

[0009] Step S1, establishing feature engineering based on multi-source data joint analysis of urban residents, wherein the multi-source data comprises travel survey data, bus card data and point of interest data;

[0010] Step S2, dividing the records of travel survey data with regular bus travel mode into first-class travel data and second-class travel data according to whether the home address and work address in the records are filled in, and respectively constructing a first model and a second model by using a machine learning model;

[0011] Step S3, identifying the home station address and work station address of each user in the bus card data, and dividing the first-class card data and the second-class card data according to whether the home station address and the work station address of the user can be identified;

[0012] Step S4, predicting the travel purpose by using the first model and the second model for the card data classified in step S3 respectively.

[0013] Preferably, the process of establishing feature engineering comprises: calculating the proportion of different types of points of interest within the threshold range of the surrounding distance, the departure time cosine value, the arrival time cosine value and the travel time as the characteristics of the user's travel according to the travel origin and the travel destination of the travel survey data and the boarding station and the alighting station of the bus card data.

[0014] Preferably, if the home address and the work address of a piece of the travel survey data are both not missing, it is divided into first-class travel data, otherwise it is divided into second-class travel data.

[0015] Preferably, the process of constructing the first model comprises: excluding the records of the first-class travel data with the travel purpose of commuting or returning home, dividing the remaining data into a training set and a test set, and training and generating the first model by using a machine learning model.

[0016] The process of constructing the second model comprises: selecting all the second-class travel data, dividing them into a training set and a test set, and training and generating the second model by using a machine learning model.

[0017] More preferably, the machine learning model is an XGBoost model.

[0018] Preferably, the step S3 of identifying the home site address and the work site address of each user in the bus card data specifically comprises: in the bus card data of a period of time, the departure site with the most first-time travel times and a proportion of the times to all days higher than a first threshold is identified as the home site address of the user, and the departure site with the most last-time travel times and a proportion of the times to all days higher than a second threshold is identified as the work site address of the user.

[0019] Preferably, the process of dividing into the first type of card data and the second type of card data comprises:

[0020] If the home site address and the work site address can be identified in a bus travel record, the bus travel record is divided into the first type of card data, otherwise, the bus travel record is divided into the second type of card data.

[0021] Preferably, the step S4 comprises:

[0022] For the first type of card data, if the distance between the alighting site and the home site address is less than a third threshold, the travel purpose is identified as "going home"; if the distance between the alighting site and the work site address is less than a fourth threshold, the travel purpose is identified as "commuting"; and for the first type of card data with no identified travel purpose, a first model is used to predict the travel purpose.

[0023] For the second type of card data, a second model is used to predict the travel purpose.

[0024] Preferably, the method further comprises evaluating the prediction result based on multi-dimensional evaluation indexes, comprising:

[0025] For the travel survey data, the prediction result is evaluated from the aspects of accuracy, precision and recall rate;

[0026] For the bus card data, the prediction result is evaluated by calculating the Jensen-Shannon divergence.

[0027] According to another aspect of the present application, an electronic device is provided, comprising a memory and a processor, the memory has a computer program stored thereon, and the processor implements the method when executing the program.

[0028] Compared with the prior art, the present application has the following beneficial effects:

[0029] 1) The application uses multi-source data joint analysis to establish bus user travel feature engineering, and analyzes the relationship between travel space-time characteristics and travel purposes; the travel survey data and bus card data are classified according to whether the user's home address and work address are filled in or recognized, and different types of travel data are respectively used to build and train their own prediction models. The different types of card data are applied to the trained different types of prediction model to predict the travel purpose of the bus user, and the prediction accuracy is more accurate than the previous non-zoned model.

[0030] 2) The application classifies and processes the travel survey data and bus card data, and then uses different models to predict different types of bus card data. Fine classification makes the output more accurate travel purpose prediction, which helps to dig the deep mechanism and motivation behind the bus user travel behavior, and obtains more scientific and accurate user travel behavior mechanism analysis. It is of great significance for accurately grasping the user travel demand and promoting the efficient and reasonable allocation of resources for public transport enterprises.

[0031] 3) The application uses rules to identify the travel purpose of commuting and returning home for the first type of card data, and then uses the first model to output the prediction result of other travel purposes. The hierarchical prediction enhances the accuracy and efficiency of the prediction result. DETAILED DESCRIPTION

[0032] Figure 1 The flowchart of the bus user travel purpose prediction method in the application is shown.

[0033] Figure 2 The flowchart of the bus card data travel purpose identification in the application is shown.

[0034] Figure 3 The confusion matrix of the first model and the travel purpose prediction result in the application is shown.

[0035] Figure 4 The confusion matrix of the second model travel purpose prediction result in the application is shown.

[0036] Figure 5 The proportion distribution diagram of the travel purpose in the travel survey data and the bus card data in the application is shown. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, not all embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the application.

[0038] The bus card data does not contain user travel purpose information, and existing methods cannot effectively fuse multi-source data, and the application can accurately grasp user demand by establishing appropriate travel purpose inference feature engineering, and can be used to improve the service level of public transport enterprises.

[0039] Embodiment 1

[0040] The present embodiment relates to a bus user travel purpose prediction method based on multi-source data, which comprises: Figure 1 The method comprises:

[0041] Step S1, obtaining travel survey data D survey , bus card data D card and interest point data, including: for the travel survey data, the travel starting point and the travel ending point, and the boarding station and the alighting station of the bus card data, according to the proportion of different types of interest points within the distance threshold range around the starting point and the ending point (or the boarding station and the alighting station) and the departure time and the arrival time (or the boarding time and the alighting time), a feature engineering is established.

[0042] The interest point data is obtained by using the API of an electronic map (such as Baidu Map, Gaode Map, etc.), including six categories of interest points: catering, company, leisure and entertainment, science and education, business and residence, and medical care;

[0043] For the travel survey data D survey , the travel starting point coordinates and the travel ending point coordinates of each travel survey data are obtained. All interest points in the interest point data with a distance less than 600 meters from the corresponding station are selected as set A, the proportions of all different types of interest points in set A are calculated, and the cosine values of the departure time and the arrival time and the travel time are calculated as the features of the travel;

[0044] For the bus card data D card , the boarding station coordinates and the alighting station coordinates of each card data are obtained. All interest points in the interest point data with a distance less than 600 meters from the corresponding station are selected as set B, the proportions of all different types of interest points in set B are calculated, and the cosine values of the departure time and the arrival time and the travel time are calculated as the features of the travel record.

[0045] Step S2, from the obtained travel survey data, the records with the travel mode being regular public transportation, including buses, subways and other urban public transportation, are selected. According to whether the home address and the work address of the user belong to the travel survey data, all travel records are divided into first type and second type travel data, and first type and second type travel purpose division machine learning models are trained respectively, corresponding to first model and second model, including:

[0046] Step S2-1, for the travel survey data Dsurvey Classify the data, if the home address and work address of a trip data are both not missing, it is classified as the first type of trip data, otherwise it is classified as the second type of trip data;

[0047] Step S2-2, remove the records of the first type of trip data whose travel purpose is commuting or going home, take the remaining data (the travel purpose of the first type of trip data is neither commuting nor going home), and divide the training set and the test set, and use machine learning training to generate the first model; select all the second type of trip data, divide the training set and the test set, and use machine learning training to generate the second model.

[0048] Step S3: For the obtained bus card data, identify the home station address and work station address of each user, and according to whether the home station address and work station address of the user to which the card data belongs can be identified, divide all card records into first type and second type of card data, including:

[0049] Step S31, identify the home station address and work station address of each user: filter the bus card data D card , identify the home station and work station of each user for a longer period of time (such as a month). The identification rule is: the departure station with the most number of first trips in multiple days and the proportion of the number of times to all days is higher than the threshold value is identified as the home station address s home of the user, and the departure station with the most number of last trips in multiple days and the proportion of the number of times to all days is higher than the threshold value is identified as the work station address s work of the user.

[0050] Step S32, classify the results of step S31: if the home station address and work station address of a resident can be identified, it is classified as the first type of card data, otherwise it is classified as the second type of card data.

[0051] Step S33, for the first type of card data, if the alighting station s al and the home station address s home satisfy dis(s al , s home )≤600 meters, it is identified as the purpose of "going home", and if the alighting station s al and the work station address s work satisfy dis(s al , s work )≤600 meters, it is identified as the purpose of "commuting". For the remaining two types of card data that are not identified, use the two models trained in step S2 to identify the travel purpose, such as Figure 2 .

[0052] Step S4: applying the first model and the second model to the first type and the second type of card swiping data respectively, and evaluating the effectiveness of the inference model by using multi-dimensional evaluation indexes such as accuracy, precision, recall, and Jensen-Shannon divergence.

[0053] On the test set of the travel survey data D survey The accuracy (Acc), precision (Pre), and recall (Rec) of the final model are calculated, and the calculation formula is as follows:

[0054]

[0055] Wherein, TP, FP, FN, and TN are the number of true positives, true negatives, false positives, and false negatives, respectively, and K is the total number of categories.

[0056] On the bus card swiping data D card The Jensen-Shannon divergence of the final inference model result is calculated, and the calculation formula is as follows:

[0057]

[0058] Wherein, p is the true distribution of travel purpose on the resident travel survey data, q is the estimated distribution of travel purpose on the card swiping data, D KL (a||b) is the Kullback-Leibler divergence between distribution a and distribution b.

[0059] The present application establishes a bus user single trip feature engineering by using multi-source data, analyzes the relationship between travel space-time characteristics and travel purpose, and constructs two models of bus user travel purpose, which is helpful to mine the deep mechanism and motivation behind the bus user travel behavior, and obtain more scientific and accurate user travel behavior mechanism analysis, which has important significance for accurately grasping user travel demand and promoting efficient and reasonable resource allocation of bus enterprises.

[0060] Embodiment 2

[0061] The present embodiment also relates to a bus user travel purpose prediction method based on multi-source data, and a verification test of bus user travel purpose prediction based on semi-supervised learning.

[0062] In the present embodiment, the selected research data includes the travel survey data of a certain city in 2020, the bus card swiping data from October to December in 2018, and the interest point data of the city in 2018.

[0063] The original travel survey data contains 11 categories of travel purposes, which are merged into four categories: commuting, returning home, entertainment and medical treatment. The proportions of these four purposes are 41.4%, 48.5%, 9.6% and 0.5% respectively.

[0064] For the travel survey data and the bus card data, feature engineering is established according to the proportion of different types of interest points within a range of 600 meters around the origin and destination (or the boarding and alighting stations), the departure time, the arrival time (or the alighting time) and the travel time. The selection of detailed feature variables is shown in Table 1.

[0065] Table 1

[0066] Variable Description Pickup time The position of the pickup time in the 24 hours of a day (translated into a cosine value) Dropoff time The position of the dropoff time in the 24 hours of a day (translated into a cosine value) Travel time The time spent on this segment of travel Origin_restaurant The proportion of POIs of the type restaurant in the vicinity of the origin (pickup station) Origin_leisure The proportion of POIs of the type leisure in the vicinity of the origin (pickup station) Origin_education The proportion of POIs of the type education in the vicinity of the origin (pickup station) Origin_healthcare The proportion of POIs of the type healthcare in the vicinity of the origin (pickup station) Origin_business The proportion of POIs of the type business in the vicinity of the origin (pickup station) Origin_residential The proportion of POIs of the type residential in the vicinity of the origin (pickup station) Destination_restaurant The proportion of POIs of the type restaurant in the vicinity of the destination (dropoff station) Destination_leisure The proportion of POIs of the type leisure in the vicinity of the destination (dropoff station) Destination_education The proportion of POIs of the type education in the vicinity of the destination (dropoff station) Destination_healthcare The proportion of POIs of the type healthcare in the vicinity of the destination (dropoff station) Destination_business The proportion of POIs of the type business in the vicinity of the destination (dropoff station) Destination_residential The proportion of POIs of the type residential in the vicinity of the destination (dropoff station)

[0067] For the travel survey data, all travel records are divided into two categories according to whether the home address and the work address of the user belong to the first category or the second category. The records of the first category are used to train the first travel purpose inference model, i.e. the first model. The second travel purpose inference model is trained using all the second category data, i.e. the second model. The division ratio of the training set and the test set in the above two training processes is 80% and 20%. XGBoost models are trained based on the two training sets, and the hyperparameters are determined by grid search. The trained models are the first model and the second model, and the specific hyperparameter values are shown in Table 2.

[0068] Table 2

[0069] Hyperparameters First model Second model Number of estimators 140 140 Maximum depth 7 7 Minimum child weight 2 8 Subsample rate 0.6 0.8 Learning rate 0.1 0.1

[0070] For the bus card data, the home station address and the work station address of each user are identified based on three months of card data. The identification rule is as follows: the departure station with the most number of first trips and the proportion of which is higher than 0.6 is identified as the home station address s home , and the departure station with the most number of last trips and the proportion of which is higher than 0.6 is identified as the work station address s work .

[0071] According to whether the home station address and the work station address of the user belonging to the bus card data can be identified, all card records are divided into two categories: the first category and the second category. If the home station address and the work station address of a resident can be identified, it is classified as the first category of card data, otherwise it is classified as the second category of card data.

[0072] For the first category of card data, if the alighting station s al and s homes al s home )≤ 600 meters, it is identified as a "home" purpose, if its drop-off site s al and s work satisfy dis(s al , s work )≤ 600 meters, it is identified as a "commute" purpose.

[0073] Thereafter, the first model and the second model are applied to the remaining un-identified first-class and second-class card swiping data, respectively. The identification process, as shown in Figure 2 , includes the following steps:

[0074] Step S3-1, determine whether the home site address and the work site address can be identified, if yes, it is determined as first-class travel data, and step S3-2 is executed; otherwise, it is determined as second-class travel data, and step S3-4 is executed.

[0075] Step S3-2, identify the home site address and the work site address of each user based on the identification rule, and identify the travel purpose, wherein the travel purpose includes commute and home, and step S3-3 is executed.

[0076] Step S3-3, use the trained first model to predict the travel purpose of the remaining first-class travel data, and the predicted travel purpose includes leisure and medical treatment, and the process ends.

[0077] Step S3-4, use the trained second model to predict the travel purpose of the second-class travel data, and the predicted travel purpose includes commute, home, leisure and medical treatment, and the process ends.

[0078] The hierarchical identification and prediction of steps S3-2 and S3-3 reduce the prediction time of the first-class travel data and improve the efficiency.

[0079] For the constructed travel purpose identification model, the effectiveness of the travel model is evaluated by using multi-dimensional evaluation indexes such as accuracy, precision, recall, and Jensen-Shannon divergence. The prediction results of the first model and the second model of the present patent are compared with the true values on the test set, and the obtained confusion matrix is shown in Figure 3 and Figure 4 . The accuracy of the first model and the second model on their respective data sets is 86.7% and 78.3%, respectively. Considering the overall prediction process of the model (including the result based on the identification rule), the accuracy can reach 92.7%. Training the corresponding model for different classified travel data can achieve better accuracy in predicting the travel purpose than the model without distinguishing the data.

[0080] The trip survey data and the prediction method of the present application are in the proportion distribution of trip purposes, wherein the prediction method of the present application comprises prediction based on bus card data of weekdays and non-weekdays, as shown in Figure 5 It can be seen that the distributions are basically consistent, the proportion of weekday commuting is higher than that of non-weekday commuting, and the proportion of weekday leisure and entertainment is lower than that of non-weekday leisure and entertainment, thus verifying the accuracy and rationality of the prediction method of the present application.

[0081] Embodiment 3

[0082] The electronic device of the present application includes a central processing unit (CPU) that can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or computer program instructions loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM, and RAM are connected to each other through a bus. An input / output (I / O) interface is also connected to the bus.

[0083] A plurality of components in the device are connected to the I / O interface, including: an input unit such as a keyboard, a mouse, etc.; an output unit such as various types of displays, a speaker, etc.; a storage unit such as a magnetic disk, an optical disk, etc.; and a communication unit such as a network card, a modem, a wireless communication transceiver, etc. The communication unit allows the device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0084] The processing unit performs the various methods and processes described above, such as methods S1-S4. For example, in some embodiments, methods S1-S4 can be implemented as a computer software program that is tangibly embodied in a machine-readable medium, such as the storage unit. In some embodiments, part or all of the computer program can be loaded and / or installed on the device via the ROM and / or the communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1-S4 described above can be performed. Alternatively, in other embodiments, the CPU can be configured to perform methods S1-S4 by any other appropriate means (e.g., by means of firmware).

[0085] The functions described above herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that can be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system on a chip (SOCs), complex programmable logic devices (CPLDs), etc.

[0086] Program code for carrying out methods of the present application can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces a means for implementing the functions / operations specified in the flowchart diagrams and / or block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, or entirely on a remote machine or server.

[0087] In the context of the present application, a machine-readable medium can be a tangible medium that can contain or store program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable storage medium can include, without limitation, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination of the foregoing.

[0088] The above description is only specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for predicting a travel purpose of a bus user based on multi-source data, characterized in that, The method comprises the following steps: Step S1, establishing feature engineering based on joint analysis of multi-source data of urban residents, wherein the multi-source data comprises travel survey data, bus card data and point of interest data; Step S2, selecting records with regular bus travel mode in the travel survey data, dividing them into first-type travel data and second-type travel data according to whether the home address and work address in the records are filled in, and respectively constructing a first model and a second model by using a machine learning model; Step S3, identifying the home station address and work station address of each user in the bus card data, and dividing them into first-type card data and second-type card data according to whether the home station address and work station address of the user can be identified; Step S4, predicting the travel purpose of the card data classified in step S3 by using the first model and the second model respectively; If the home address and work address of the travel survey data are not missing, the travel survey data are divided into first-type travel data, otherwise, the travel survey data are divided into second-type travel data; The construction process of the first model comprises: excluding records with travel purpose of commuting or returning home in the first-type travel data, dividing the remaining data into a training set and a test set, and training and generating the first model by using a machine learning model; The construction process of the second model comprises: selecting all second-type travel data, dividing them into a training set and a test set, and training and generating the second model by using a machine learning model; the machine learning model is an XGBoost model; The process of dividing into first-type card data and second-type card data comprises: If the home station address and work station address in a bus travel record can be identified, the bus travel record is divided into first-type card data, otherwise, the bus travel record is divided into second-type card data; The step S4 comprises: For the first-type card data, if the distance between the alighting station and the home station address is less than or equal to a third threshold value, the first-type card data is identified as "returning home"; if the distance between the alighting station and the work station address is less than or equal to a fourth threshold value, the first-type card data is identified as "commuting"; for the first-type card data with unidentified travel purpose, the first model is used to predict the travel purpose; For the second-type card data, the second model is used to predict the travel purpose. 2.The bus user trip destination prediction method based on multi-source data according to claim 1, characterized in that, The process of establishing feature engineering comprises: calculating the proportion of different types of points of interest within a threshold range of distance, the departure time cosine value, the arrival time cosine value and the travel time of the user as the features of the user's travel according to the travel origin and the travel destination of the travel survey data and the boarding station and the alighting station of the bus card data. 3.The bus user trip destination prediction method based on multi-source data according to claim 1, characterized in that, In the step S3, the home station address and the work station address of each user in the bus card data are identified as follows: in a period of bus card data, the departure station with the most first-time travel times and the proportion of the most first-time travel times being higher than a first threshold value in all days is identified as the home station address of the user, and the departure station with the most last-time travel times and the proportion of the most last-time travel times being higher than a second threshold value in all days is identified as the work station address of the user. 4.The bus user trip destination prediction method based on multi-source data according to claim 1, characterized in that, The method further comprises evaluating the prediction results based on multi-dimensional evaluation indexes, comprising: For the trip survey data, the prediction results are evaluated from the aspects of accuracy, precision and recall rate; For the bus card swiping data, the prediction results are evaluated by calculating the Jensen-Shannon divergence.

5. An electronic device comprising a memory and a processor, said memory having stored thereon a computer program, characterized in that, The processor implements the method of any one of claims 1-4 when executing the program.

Images