A method and device for updating task flow work template

Abstract

The present application provides a method and device for updating a task flow work template, the method comprising: for each task flow in the work template, obtaining and storing the operation information of the task flow when it passes through the work node in the work template; wherein the operation information includes: the operation time, whether to enter the next work node; inputting the operation information obtained within the stored preset time into the prediction model to obtain the operation time and pass rate of each work node; and updating the work template according to the obtained operation time and pass rate of each work node in the work template. The method can save computing resources while speeding up the processing speed of the task flow.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method and device for updating a task flow work template. Background Art

[0002] With the development of informatization, some traditional businesses have been moved to the cloud. With the development of cloud computing, more and more complex business links have been informatized, digitized and intelligentized.

[0003] Faced with so many complex business scenarios, in order to meet the needs of rapid business access, some tools for developers have appeared on the market. For some businesses that need to use workflow engines, the current tools on the market all work according to the processes defined by developers or business personnel. This approach makes the business less flexible, and the data will pass through some unnecessary process nodes, causing pressure on the service cluster and a waste of computing resources.

[0004] Since the working mode of the workflow engine on the market is relatively rigid, it is mainly implemented through the following technologies:

[0005] Shark: The system and functions are relatively complex;

[0006] Osworkflow: A relatively flexible and lightweight framework, but it is not very friendly in terms of process modeling. You need to manually write XML files to define process files.

[0007] SWF: Also, it cannot support complex processes.

[0008] However, all of the above technologies have a common technical defect: the work template cannot be adjusted dynamically, that is, the work order cannot be automatically adjusted according to its own business status. Summary of the invention

[0009] In view of this, the present application provides a method and device for updating a task flow work template, which can save computing resources while speeding up the processing of the task flow.

[0010] To solve the above technical problems, the technical solution of this application is implemented as follows:

[0011] A method for updating a task flow work template, the method comprising:

[0012] For each task flow in the work template, the running information of the task flow when it passes through the work node in the work template is obtained and stored; wherein the running information includes: running time, whether to enter the next work node;

[0013] Input the operation information acquired within the preset time into the prediction model to obtain the operation time and pass rate of each working node;

[0014] The work template is updated according to the obtained running time and pass rate of each work node in the work template.

[0015] A task flow work template updating device, the device comprising: an acquisition unit, a storage unit, a prediction unit and an update unit;

[0016] The acquisition unit is used to acquire, for each task flow in the work template, the running information of the task flow when it passes through the work node in the work template;

[0017] The storage unit is used to store the operation information of each working node acquired by the acquisition unit; wherein the operation information includes: the operation time and whether to enter the next working node;

[0018] The prediction unit is used to input the operation information acquired within the preset time stored in the storage unit into the prediction model to obtain the operation time and pass rate of each working node;

[0019] The updating unit is used to update the working template according to the running time and the passing rate of each working node in the working template obtained by the predicting unit.

[0020] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the task flow work template updating method as described above when executing the program.

[0021] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the task flow work template updating method as described above.

[0022] It can be seen from the above technical solution that in this application, by obtaining the running information of each working node when the task flow passes through, the running time and pass rate of each working node can be predicted through the prediction model, and then the work template can be dynamically adjusted according to different business needs, thereby saving computing resources while speeding up the processing speed of the task flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a schematic diagram of the task flow work template update process in an embodiment of the present application;

[0024] Figure 2 It is a schematic diagram of the structure of GRU;

[0025] Figure 3 This is a schematic diagram of signal transduction in the embodiments of this application;

[0026] Figure 4 This is the original workflow template diagram;

[0027] Figure 5 For Update Figure 4 The working template after the working template shown;

[0028] Figure 6 For Update Figure 5 The working template after the working template shown;

[0029] Figure 7 For Update Figure 4 The working template after the working template shown;

[0030] Figure 8 Schematic diagram of the device structure applied to the above technology in the embodiment of the present application. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the technical solution of the present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0032] In the embodiment of the present application, before implementing the work template update process, it is necessary to first obtain the operation information generated when the task flow in the work template passes through the work node within the preset time T.

[0033] As for how long it takes to obtain the information, that is, the preset time setting can be set according to actual needs. Usually, the operation information is obtained for a certain period of time. By default, it can be predicted how long the change trend of the corresponding operation information of each work node in the work template will be in the future.

[0034] For each task flow in the work template, the running information of the task flow when it passes through the work node in the work template is obtained and stored; wherein the running information includes: running time, whether to enter the next work node.

[0035] The running time here refers to the running time of a task flow on a working node;

[0036] Here, whether to enter the next working node can be defined as whether to pass at the working node. If entering the next working node, it is confirmed as passing at the node; if not entering the next working node, it is confirmed as failing at the working node.

[0037] When obtaining the running information of the task flow passing through each working node, a globally unique identifier is assigned to each running information. That is to say, for each working node, when different task flows pass through, the running information of the task flow on the working node needs to be recorded, and a globally unique identifier is assigned to the running information.

[0038] The specific implementation process of allocating a globally unique identifier to each operation information in the embodiment of the present application is given below:

[0039] Assign a unique identifier, i.e., a unique UUID, to each working node in the working template. When the task flow passes through each node, assign a unique identifier to the task flow. Combine the identifier of the task flow and the identifier of the working node into a globally unique identifier corresponding to the running information generated by the task flow on a node.

[0040] Here, the unique identifier of the working node only needs to be able to uniquely identify each working node, and the unique identifier of the task flow only needs to be able to uniquely identify each task flow; the combination of the two can globally uniquely identify a running information, and can identify which task flow is running on which node.

[0041] When storing each operation information, the operation information is decomposed according to different dimensions, that is, divided into two dimensions of information, one dimension is the operation time; the other dimension is whether to enter the next work node.

[0042] The following is a detailed description of the task flow work template update process in the embodiment of the present application in conjunction with the accompanying drawings.

[0043] It should be noted that the devices involved in the embodiments of the present application that execute the task flow work template update entity may include but are not limited to personal computers (PC), personal digital assistants (PDA), wireless handheld devices, tablet computers, mobile phones, etc.

[0044] The execution subject of the following embodiments may be a task flow work template updating device, which may be an application located in a local device, or may be a functional unit such as a plug-in or software development kit (SDK) located in an application located in a local device. The embodiments of the present invention do not specifically limit this.

[0045] See also Figure 1 , Figure 1 This is a schematic diagram of the task flow work template update process in the embodiment of this application. The specific steps are:

[0046] Step 101: Input the operation information acquired within the preset time into the prediction model to obtain the operation time and pass rate of each working node.

[0047] The running time and pass rate of each working node obtained by the prediction model, that is, the running time and pass rate of each node predicted by the prediction model. If a node has no running time and / or no pass rate, they can be recorded as 0.

[0048] The preset time here can be set according to actual needs. If the workflow tasks are frequent, the preset time can be set smaller; otherwise, it can be set larger.

[0049] In the specific implementation of this application, the prediction model is a trend prediction system based on a recurrent neural network algorithm; wherein, a reset gate is used to replace the input gate and forget gate in the long short-term memory system (Long Short-Term Memory, LSTM), which can avoid the problem of too many parameters caused by the complexity of the traditional LSTM.

[0050] Recurrent networks are implemented using Gated Recurrent Unit (GRU). Figure 2 , Figure 2 The schematic diagram of the GRU structure is shown in Figure 2. h is a memory unit, which corresponds to the Constant Error Carousel (CEC) unit of LSTM. The forward process of GRU is the signal transmission after the time sequence is expanded. For details, see Figure 3 , Figure 3 This is a schematic diagram of signal transduction in an embodiment of the present application.

[0051] Figure 3 In the example, GRU contains only two gate structures: input gate z and reset gate r. Input gate z and reset gate r receive the same input: the input signal at the current moment and the context at the previous moment. The inputs of these two gate structures are defined by the following formulas (1) and (2), respectively. t is the input signal at the current moment, and is the context at the previous moment.

[0052] z t =σ(W z x t +U z h t-1 ) (1)

[0053] r t =σ(W r x t +U r h t-1 ) (2)

[0054] The hidden layer activation state of the GRU unit is defined by the following formula (3). ⊙ is a bitwise multiplication operation. U and W are fixed parameters.

[0055]

[0056] The final output of GRU is determined by the previous moment context, hidden layer activation, W, and reset port activation. The following formula (4) is used as the current moment context and is saved in the memory unit h.

[0057]

[0058] In the embodiment of the present application, two GRUs with output vectors of length 512 are used as the internal structure of the recurrent neural network; the initial parameters of the GRU memory unit are obtained by back-propagation learning.

[0059] Through the GRU unit-based recurrent neural network, the acquired and stored data is analyzed and predicted to obtain the trend of the running time of each working node and the change of the pass rate. Based on this data, the template is dynamically adjusted.

[0060] When the operation information acquired within the preset storage time is input into the prediction model, the operation time in the operation information and whether the node has been stored are input into the prediction model as two-dimensional sample data.

[0061] In the embodiment of the present application, the device for acquiring operation information can be the same device as the working template updating device, or can be a different device. At different device times, the operation information is acquired by the device for storing and acquiring data.

[0062] Step 102: Update the work template according to the obtained running time and pass rate of each work node in the work template.

[0063] The pass rate in this step is the proportion of task flows that pass through the work node and enter the next work node. For example, among 100 task flows, 60 task flows pass through work node 2 and enter other work nodes, such as work node 3. Then, the pass rate for work node 1 is 60%.

[0064] If it is determined that there is a dependency relationship between the working nodes in the working template, and the number N of the nodes with the dependency relationship is determined, then the working template is updated according to the obtained operation information of each working node in the working template, including:

[0065] Determine the number N of nodes with dependencies, and divide the nodes in the working template into N echelons from the beginning to the end of the working template; and the nodes with dependencies are placed in corresponding echelons according to their dependencies; wherein N is an integer greater than 1;

[0066] For each work node that has no dependency in the work template, if it is determined that the difference between the running time and pass rate of the work node and the running time and pass rate of any work node that has a dependency is within a preset range, the work node is placed in the echelon corresponding to the work node that has a dependency;

[0067] Among them, the working nodes in each echelon are in parallel relationship, and the first echelon, the second echelon and the third echelon are in series relationship.

[0068] If it is determined that there is no dependency relationship between the working nodes in the working template, updating the working template according to the obtained running information of each working node in the working template includes:

[0069] From the beginning to the end of the work template, it is divided into three echelons, namely the first echelon, the second echelon and the third echelon;

[0070] For each work node in the work template, if it is determined that the pass rate of the work node is less than the first preset value and the running time is less than the second preset value, the work node is placed in the first echelon;

[0071] If it is determined that the pass rate of the working node is less than the first preset time and the running time is greater than the second preset value, the working node is placed in the second echelon;

[0072] If it is determined that the pass rate of the working node is greater than the first preset time, the working node is placed in the third echelon;

[0073] Among them, the working nodes in each echelon are in parallel relationship, and the first echelon, the second echelon and the third echelon are in series relationship.

[0074] In the embodiment of the present application, work nodes with dependent relationships can only be placed in different echelons according to their dependencies and the order of the work template from the beginning to the end, and work nodes with similar running time and pass rate are placed in the same echelon; for each work node in the work template that has no dependent relationship, the node with the lowest pass rate and relatively short running time is moved forward, which saves computing resources and speeds up the processing speed of the process.

[0075] The following describes in detail the updating process of the work template with reference to a specific example.

[0076] See also Figure 4 , Figure 4 This is a schematic diagram of the original workflow template. Figure 4 The structure of a serial working node is given. Take 8 working nodes as an example, namely: working node 1, working node 2...working node 8. Assume that there is a dependency relationship between working nodes 1, 5 and 8 in the working template. Each working node runs a piece of code or calls some interfaces. The task flow is obtained by running the task flow on these 8 working nodes, and the obtained operation information is input into the prediction model for prediction.

[0077] Since the nodes with a dependency relationship are determined to be work node 1, work node 5, and work node 8, there are three nodes in total; they are divided into three echelons according to the order from the beginning to the end of the work template, with work node 1 placed in the first echelon, work node 5 placed in the second echelon, and work node 8 placed in the third echelon;

[0078] Assuming that the running time and pass rate of each node obtained through the prediction model, when the difference between the running time and pass rate of work node 2, work node 3 and work node 4 and the running time and pass rate of work node 1 is the smallest, the work node is placed in the echelon (first echelon) corresponding to the work node (work node 1) with which there is a dependency relationship.

[0079] When the difference between the running time and the pass rate of the working node 6 and the working node 5 is the smallest, the working node is placed in the echelon (the second echelon) corresponding to the working node (working node 5) with which there is a dependency relationship.

[0080] When the difference between the running time and the pass rate of working node 7 and working node 8 is the smallest, the working node is placed in the echelon (the third echelon) corresponding to the working node (working node 8) with which there is a dependency relationship.

[0081] Adjust according to the above results Figure 4 The updated work template is as follows Figure 5 As shown, Figure 5 For Update Figure 4 The job template following the job template shown.

[0082] right Figure 5 The work template in obtains the running information of each work node at the preset time, and inputs the running information of each node into the preset model to obtain the predicted running time and pass rate of each work node.

[0083] Assuming that the difference between the running time and pass rate of each node obtained by the prediction model is the smallest, the working node is placed in the echelon (first echelon) corresponding to the working node (working node 1) with which there is a dependency relationship;

[0084] Among other working nodes, the working node with the smallest difference in running time and passing rate between working node 1, working node 5 and working node 8 remains unchanged and is not adjusted.

[0085] Adjust according to the above results Figure 5 The updated work template is as follows Figure 6 As shown, Figure 6 For Update Figure 5 The job template following the job template shown.

[0086] Assuming that Figure 4Taking the initial working template shown as an example, and assuming that there is no dependency between all working nodes, the specific implementation method given in this application is to divide it into three gradient levels, and determine the gradient to which each working node belongs according to the above rules. Assuming that the pass rate of each working node in the working template is less than the first preset value, and the running time is less than the second preset value, all working nodes are placed in the first echelon.

[0087] Adjust according to the above results Figure 5 The updated work template is as follows Figure 7 As shown, Figure 7 For Update Figure 4 The job template following the job template shown.

[0088] Based on the same inventive concept, the present application also proposes a task flow work template updating device. Figure 8 , Figure 8 This is a schematic diagram of the structure of a device applied to the above technology in an embodiment of the present application. The device includes: an acquisition unit 801, a storage unit 802, a prediction unit 803 and an update unit 804;

[0089] The acquisition unit 801 is used to acquire, for each task flow in the work template, the running information of the task flow when it passes through the work node in the work template;

[0090] The storage unit 802 is used to store the operation information of each working node obtained by the obtaining unit 801; wherein the operation information includes: the running time and whether to enter the next working node;

[0091] The prediction unit 803 is used to input the operation information acquired within the preset time stored in the storage unit 802 into the prediction model to obtain the operation time and pass rate of each working node;

[0092] The updating unit 804 is used to update the work template according to the running time and the pass rate of each work node in the work template obtained by the predicting unit 803.

[0093] Preferably,

[0094] The acquisition unit 801 is specifically configured to allocate a globally unique identifier to each piece of running information when acquiring the running information of the task flow passing through each working node.

[0095] Preferably,

[0096] The prediction model is a trend prediction system based on a recurrent neural network algorithm, wherein a reset gate is used to replace the input gate and the forget gate in the LSTM.

[0097] Preferably,

[0098] Two GRUs with output vectors of length 512 are used as the internal structure of the recurrent neural network; the initial parameters of the GRU memory unit are learned by back propagation.

[0099] Preferably,

[0100] The prediction unit 803 is specifically configured to input the operation information acquired within the preset time into the prediction model, and input the operation time and whether the node has been stored in the operation information as two-dimensional sample parameters into the prediction model.

[0101] Preferably,

[0102] The updating unit 804 is specifically used to update the working template according to the obtained running time and pass rate of each working node in the working template. If it is determined that there is a dependency relationship between the working nodes in the working template, the number N of nodes with the dependency relationship is determined, and each node in the working template is divided into N echelons from the beginning to the end of the working template; and the nodes with the dependency relationship are placed in the corresponding echelons according to their dependency relationship; wherein N is an integer greater than 1; for each working node in the working template that does not have a dependency relationship, if it is determined that the running time and pass rate of the working node are within a preset range of the difference between the running time and pass rate of any working node with a dependency relationship, the working node is placed in the echelon corresponding to the working node with the dependency relationship; wherein the working nodes in each echelon are in a parallel relationship, and the first echelon, the second echelon and the third echelon are in a series relationship.

[0103] Preferably,

[0104] The updating unit 804 is specifically used to update the working template according to the obtained running time and pass rate of each working node in the working template. If it is determined that there is no dependency between the working nodes in the working template, the working template is divided into three echelons from the beginning to the end, namely the first echelon, the second echelon and the third echelon; for each working node in the working template, if it is determined that the pass rate of the working node is less than the first preset value and the running time is less than the second preset value, the working node is placed in the first echelon; if it is determined that the pass rate of the working node is less than the first preset time and the running time is greater than the second preset value, the working node is placed in the second echelon; if it is determined that the pass rate of the working node is greater than the first preset time, the working node is placed in the third echelon; wherein, the working nodes in each echelon are in a parallel relationship, and the first echelon, the second echelon and the third echelon are in a series relationship.

[0105] The units in the above-mentioned embodiments may be integrated into one body or deployed separately; they may be combined into one unit or further divided into multiple sub-units.

[0106] In addition, an embodiment of the present application also proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the task flow work template updating method as described above are implemented.

[0107] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the aforementioned task flow work template updating method.

[0108] To sum up, this application obtains the running information of each work node when the task flow passes through, and can predict the running time and pass rate of each work node through the prediction model, and then can dynamically adjust the work template according to different business needs, thereby saving computing resources while speeding up the processing speed of the task flow.

[0109] In the embodiment of the present application, work nodes with dependent relationships can only be placed in different echelons according to their dependencies and the order of the work template from the beginning to the end, and work nodes with similar running time and pass rate are placed in the same echelon; for each work node in the work template that has no dependent relationship, the node with the lowest pass rate and relatively short running time is moved forward, which saves computing resources and speeds up the processing speed of the process.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A task flow work template updating method, characterized in that: The method includes: For each task flow in the work template, the running information of the task flow when passing through the work node in the work template is obtained and stored; wherein the running information includes: running time, whether to enter the next work node; when obtaining the running information of the task flow passing through each work node, a globally unique identifier is assigned to each running information; the globally unique identifier is an identifier obtained by combining the identifier of the task flow and the identifier of the work node; Input the operation information acquired within the preset time into the prediction model to obtain the operation time and pass rate of each working node; And update the working template according to the obtained running time and pass rate of each working node in the working template; wherein the pass rate is the proportion of the task flow passing through the working node to enter the next working node; the updating of the working template according to the obtained running time and pass rate of each working node in the working template includes: if it is determined that there is a dependency relationship between the working nodes in the working template, then determine the number N of nodes with dependencies, and divide each node in the working template into N echelons from the beginning to the end of the working template; and the nodes with dependencies are placed in corresponding echelons according to their dependencies; wherein N is an integer greater than 1; for each working node in the working template that does not have a dependency relationship, if it is determined that the difference between the running time and pass rate of the working node and the running time and pass rate of any working node with dependencies is within a preset range, place the working node in the echelon corresponding to the working node with dependencies; wherein the working nodes in each echelon are in a parallel relationship, and the first echelon, the second echelon and the third echelon are in a series relationship.

2. The method according to claim 1, characterized in that The prediction model is a trend prediction system based on a recurrent neural network algorithm, wherein a reset gate is used to replace an input gate and a forget gate in a long short-term memory system LSTM.

3. The method according to claim 2, characterized in that Two gated recurrent units (GRUs) with output vectors of length 512 are used as the internal structure of the recurrent neural network; the initial parameters of the GRU memory unit are learned by back-propagation.

4. The method according to claim 1, characterized in that: When the operation information acquired within the preset time is input into the prediction model, the operation time in the operation information and whether the node has been stored are input into the prediction model as two-dimensional sample parameters.

5. The method according to any one of claims 1 to 4, characterized in that: The updating of the work template according to the obtained running time and pass rate of each work node in the work template comprises: If it is determined that there is no dependency between the work nodes in the work template, the work template is divided into three echelons from the beginning to the end, namely the first echelon, the second echelon and the third echelon; For each work node in the work template, if it is determined that the pass rate of the work node is less than the first preset value and the running time is less than the second preset value, the work node is placed in the first echelon; If it is determined that the pass rate of the working node is less than the first preset time and the running time is greater than the second preset value, the working node is placed in the second echelon; If it is determined that the pass rate of the working node is greater than the first preset time, the working node is placed in the third echelon; Among them, the working nodes in each echelon are in parallel relationship, and the first echelon, the second echelon and the third echelon are in series relationship.

6. A task flow work template updating device, characterized in that: The device comprises: an acquisition unit, a storage unit, a prediction unit and an update unit; The acquisition unit is used to acquire, for each task flow in the work template, the operation information of the task flow when it passes through the work node in the work template; when acquiring the operation information of the task flow passing through each work node, assign a globally unique identifier to each operation information; the globally unique identifier is an identifier obtained by combining the identifier of the task flow and the identifier of the work node; The storage unit is used to store the operation information of each working node acquired by the acquisition unit; wherein the operation information includes: the operation time and whether to enter the next working node; The prediction unit is used to input the operation information acquired within the preset time stored in the storage unit into the prediction model to obtain the operation time and pass rate of each working node; The updating unit is used to update the work template according to the running time and the pass rate of each work node in the work template obtained by the prediction unit; wherein the pass rate is the proportion of the task flow passing through the work node to the next work node; The updating unit is specifically used to update the working template according to the obtained running time and pass rate of each working node in the working template. If it is determined that there is a dependency relationship between the working nodes in the working template, the number N of nodes with dependency relationships is determined, and each node in the working template is divided into N echelons from the beginning to the end of the working template; and the nodes with dependency relationships are placed in corresponding echelons according to their dependency relationships; wherein N is an integer greater than 1; for each working node in the working template that does not have a dependency relationship, if it is determined that the difference between the running time and pass rate of the working node and the running time and pass rate of any working node with dependency relationship is within a preset range, the working node is placed in the echelon corresponding to the working node with dependency relationship; wherein the working nodes in each echelon are in a parallel relationship, and the first echelon, the second echelon and the third echelon are in a series relationship.

7. The device according to claim 6, characterized in that The prediction model is a trend prediction system based on a recurrent neural network algorithm, wherein a reset gate is used to replace an input gate and a forget gate in a long short-term memory system LSTM.

8. The device according to claim 7, characterized in that Two gated recurrent units (GRUs) with output vectors of length 512 are used as the internal structure of the recurrent neural network; the initial parameters of the GRU memory unit are learned by back-propagation.

9. The device according to claim 6, characterized in that The prediction unit is specifically used to input the operation information acquired within the preset time into the prediction model, and input the operation time and whether the node has been stored in the operation information as two-dimensional sample parameters into the prediction model.

10. The device according to any one of claims 6 to 9, characterized in that: The updating unit is specifically configured to update the work template according to the obtained running time and pass rate of each work node in the work template, and if it is determined that there is no dependency relationship between the work nodes in the work template, divide the work template into three echelons from the beginning to the end, namely the first echelon, the second echelon and the third echelon; For each work node in the work template, if it is determined that the pass rate of the work node is less than the first preset value and the running time is less than the second preset value, the work node is placed in the first echelon; If it is determined that the pass rate of the working node is less than the first preset time and the running time is greater than the second preset value, the working node is placed in the second echelon; if it is determined that the pass rate of the working node is greater than the first preset time, the working node is placed in the third echelon; wherein the working nodes in each echelon are in a parallel relationship, and the first echelon, the second echelon and the third echelon are in a series relationship.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 5 is implemented.

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

Images