Collaborative Operation Method and Device for Space Robot Based on Information Interaction

Abstract

The present invention relates to the field of artificial intelligence technology, and particularly relates to a cooperative operation method and device for a space robot based on information interaction. The method includes: determining a plurality of space robots based on a target task, each space robot corresponding to a sub-task and an agent network respectively, and each agent network including a policy network, an evaluation network, a communication network, and an output network; constructing a joint operation model between each agent network, in the joint operation model, each agent network respectively performs information interaction with other agent networks through its communication network and output network; training the joint operation model to obtain a trained joint operation model; respectively inputting the observation data obtained by each space robot based on its sub-task into the trained joint operation model to obtain the corresponding operation strategy of each space robot. In the present invention, through the collaborative work among the robots, the target task can be better completed.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and particularly to a collaborative operation method and device for a space robot based on information interaction. Background Art

[0002] Space operation is an on-orbit activity that a space robot engages in to complete specified space actions or tasks, such as on-orbit maintenance, on-orbit assembly, etc. For the on-orbit maintenance or assembly of large space structures, multiple robots need to cooperate, such as two or more robots cooperating in handling, two or more robots cooperating in observation to obtain more observation information, or one group of robots being responsible for observation and another group of robots being responsible for operation. To achieve the collaborative work among multiple heterogeneous robots, communication connections often need to be established between the robots, and close information interaction is carried out among them to achieve information sharing and intention transmission.

[0003] In the related art, when multiple robots cooperate, the communication ability and cooperation ability between them are poor, and the operation goals cannot be well achieved. Summary of the Invention

[0004] Based on the technical problem of poor cooperation ability among multiple current space robots, the embodiments of the present invention provide a collaborative operation method and device for a space robot based on information interaction.

[0005] In a first aspect, the embodiments of the present invention provide a collaborative operation method for a space robot based on information interaction, including:

[0006] Determine multiple space robots based on a target task, each of the space robots corresponds to a subtask and an agent network, and each of the agent networks includes a policy network, an evaluation network, a communication network, and an output network;

[0007] Construct a joint operation model among each of the agent networks. In the joint operation model, each of the agent networks respectively performs information interaction with other agent networks through its communication network and output network;

[0008] Train the joint operation model to obtain a trained joint operation model;

[0009] Input the observation data obtained by each of the space robots based on its subtask into the trained joint operation model respectively to obtain the corresponding operation strategies of each of the space robots.

[0010] In a possible design, each of the agent networks respectively performs information interaction with other agent networks through its communication network and output network, including:

[0011] In a possible design, each of the agent networks interacts with other agent networks through its communication network and output network respectively, including:

[0012] Using the communication network of the current agent network to receive the status information output by each other agent network;

[0013] Using the communication network to screen and fuse each received status information, and inputting the screened and fused information into the policy network of the current agent network;

[0014] Based on the screened and fused information, the policy network outputs corresponding operation policies to the environment and outputs its status information to the output network of the current agent network;

[0015] The output network processes the received status information and sends the processed status information to the communication network of other agent networks.

[0016] In a possible design, training the joint operation model to obtain a trained joint operation model includes:

[0017] Pre-training each of the policy networks respectively to obtain trained policy networks;

[0018] Based on each of the trained policy networks, training the joint operation model to obtain a trained joint operation model.

[0019] In a possible design, pre-training each of the policy networks respectively to obtain trained policy networks includes:

[0020] For each of the policy networks, perform the following:

[0021] Based on the task type of the subtask of the agent corresponding to the current policy network, obtain a sample set corresponding to the subtask, where the sample set includes the local observation data of the agent and the operation policies corresponding to the local observation data;

[0022] Pre-training the current policy network based on the sample set and random noise to obtain a trained policy network.

[0023] In a possible design, training the joint operation model based on each of the trained policy networks to obtain a trained joint operation model includes:

[0024] Taking the network parameters of each trained policy network as the initial parameters;

[0025] Randomly select the network parameters of each of the evaluation networks, each of the communication networks, and each of the output networks, and determine the corresponding reward function according to the target task;

[0026] Use the MADDPG algorithm to train the joint operation model to obtain a trained joint operation model.

[0027] In a possible design, the random noise is Gaussian noise.

[0028] In a possible design, the network structure of the policy network is an MLP structure and / or an LSTM structure.

[0029] In a second aspect, an embodiment of the present invention further provides a cooperative operation device for a space robot based on information interaction, including:

[0030] A determination module, configured to determine a plurality of space robots based on a target task, each of the space robots corresponding to a sub-task and an intelligent agent network, and each of the intelligent agent networks including a policy network, an evaluation network, a communication network, and an output network;

[0031] A construction module, configured to construct a joint operation model between each of the intelligent agent networks. In the joint operation model, each of the intelligent agent networks exchanges information with other intelligent agent networks through its communication network and output network respectively;

[0032] A training module, configured to train the joint operation model to obtain a trained joint operation model;

[0033] An input module, configured to input the observation data obtained by each of the space robots based on its sub-task into the trained joint operation model respectively to obtain the corresponding operation strategy of each of the space robots.

[0034] In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the method described in any embodiment of this specification is implemented.

[0035] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed on a computer, the computer is made to execute the method described in any embodiment of this specification.

[0036] An embodiment of the present invention provides a cooperative operation method for space robots based on information interaction. In this method, first, multiple space robots are determined based on the target task, and each space robot is respectively used to complete a subtask. In order to enable the robots to work together, each robot corresponds to an agent network respectively, and a joint operation model between the agent networks is constructed. The joint operation model performs information interaction through the communication network and the output network in each agent network respectively. Then, the joint operation model is trained. Finally, the observation data obtained by each space robot based on its subtask is respectively input into the trained joint operation model to obtain the corresponding operation strategy for each space robot. The joint operation model can enhance the communication ability between the robots, realize information interaction, and supplement and correct the original strategies or skills of each robot based on the interaction information. In the present invention, through the cooperation of the robots, the target task can be better completed. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0038] Figure 1 is a schematic structural diagram of a cooperative operation method for space robots based on information interaction provided by an embodiment of the present invention;

[0039] Figure 2 is a hardware architecture diagram of an electronic device provided by an embodiment of the present invention;

[0040] Figure 3 is a structural diagram of a cooperative operation device for space robots based on information interaction provided by an embodiment of the present invention;

[0041] Figure 4 is a schematic structural diagram of a single agent network provided by an embodiment of the present invention;

[0042] Figure 5 is a schematic structural diagram of a joint operation model provided by an embodiment of the present invention;

[0043] Figure 6 is a schematic diagram of separately training a policy network provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] Please refer to Figure 1 , an embodiment of the present invention provides a collaborative operation method for a space robot based on information interaction. The method includes:

[0046] Step 100: Determine multiple space robots based on a target task. Each space robot corresponds to a subtask and an agent network respectively. Each agent network includes a policy network, an evaluation network, a communication network, and an output network.

[0047] Step 102: Construct a joint operation model among each agent network. In the joint operation model, each agent network performs information interaction with other agent networks through its communication network and output network respectively.

[0048] Step 104: Train the joint operation model to obtain a trained joint operation model.

[0049] Step 106: Input the observation data obtained by each space robot based on its subtask into the trained joint operation model respectively to obtain the corresponding operation strategy for each space robot.

[0050] In the embodiment of the present invention, first, multiple space robots are determined based on a target task. Each space robot is respectively used to complete a subtask. To enable the robots to work collaboratively, each robot corresponds to an agent network respectively, and a joint operation model among each agent network is constructed. The joint operation model performs information interaction through the communication network and output network in each agent network respectively. Then, the joint operation model is trained. Finally, the observation data obtained by each space robot based on its subtask is input into the trained joint operation model respectively to obtain the corresponding operation strategy for each space robot. The joint operation model can enhance the communication ability among the robots, realize information interaction, and supplement and correct the original strategies or skills of each robot based on the interaction information. In the present invention, through the collaborative work among the robots, the target task can be better completed.

[0051] The execution manners of the following described Figure 1 steps are shown.

[0052] First, for step 100, multiple space robots are determined based on the target task. Each space robot corresponds to a sub-task and an agent network respectively. Each agent network includes a policy network, an evaluation network, a communication network, and an output network.

[0053] In this step, the target task is determined based on user requirements, such as on-orbit maintenance and on-orbit assembly. After the target task is determined, it can be divided into multiple sub-tasks according to task type, task nature, and task volume. Each sub-task is assigned a robot, and each robot corresponds to an agent network. A schematic diagram of a single agent network is shown as Figure 4 shown. As can be seen from the figure, each agent network includes a policy network, an evaluation network, a communication network, and an output network. Among them, the input of the policy network is the local observation o i of the environment, and the output is a i and the intermediate state w. The input of the communication network is the state information S output by other agent networks, which is specifically determined by the communication topology structure, and the output is the hidden layer information z. The input of the output network is the intermediate state w, o i and the output is the state information s i of this agent. The evaluation network takes the global observation o of the environment as the input, and the output is the evaluation value of a i . In addition, the network structure of the policy network can adopt typical structures such as the LSTM structure or the MLP, the network structure of the communication network can adopt typical structures such as the MLP structure or the attention mechanism, and the network structure of the output network can adopt typical structures such as the MLP structure or the attention mechanism.

[0054] Then, for step 102, a joint operation model between each agent network is constructed. In the joint operation model, each agent network interacts with other agent networks through its communication network and output network respectively.

[0055] As Figure 5 shown, it is a schematic diagram of this joint operation model. As can be seen from the figure, each agent network interacts with other agent networks through its communication network and output network respectively. Through information interaction, each robot can obtain the state information of other robots, so as to provide a reference for its own operation strategy, and then make an operation strategy that is more beneficial to the overall situation.

[0056] In some embodiments, each agent network interacts with other agent networks through its communication network and output network respectively, including:

[0057] Using the communication network of the current agent network to receive the state information output by each other agent network;

[0058] The communication network filters and fuses each received status information, and inputs the filtered and fused information into the policy network of the current agent network;

[0059] Based on the filtered and fused information, the policy network outputs corresponding operation policies to the environment and outputs its status information to the output network of the current agent network;

[0060] The output network processes the received status information and sends the processed status information to the communication networks of other agent networks.

[0061] As Figure 4 shown, the status information output by the i-th output network is s i , and this status information s i is used to represent the current state of the i-th agent, and this status information can provide a reference for the operations of other agents. Since each communication network will receive status information sent by the communication networks of multiple other agent networks, it is necessary to filter and fuse the multiple received status information to obtain a comprehensive information, and input this comprehensive information into the corresponding policy network, which is beneficial for the agent to make better decisions.

[0062] In addition, the formats of the intermediate states w output by each policy network may be different. In this embodiment, the output network processes the intermediate state w to obtain the state s i , and this state s i is more conducive to the communication network to process it.

[0063] It should also be noted that the state s i output by the output network is determined based on the target task.

[0064] Next, for step 104, the joint operation model is trained to obtain a trained joint operation model.

[0065] Before using the joint operation model to guide multiple robots to work collaboratively, it needs to be trained. The specific training process is as follows:

[0066] First, each policy network is pre-trained separately to obtain a trained policy network.

[0067] Each agent network in the joint operation model can work independently. When it does not communicate with other agent networks, it can also independently guide the corresponding robot to work. When it works independently, only the policy network and the evaluation network are required. As Figure 6As shown. At this time, it is necessary to pre-train the policy network separately. The trained policy network can not only guide the robot to work alone, but also further improve the training speed and accuracy of the joint operation model when the trained policy network is integrated into the joint operation model.

[0068] In some embodiments, the process of pre-training each policy network is as follows:

[0069] For each policy network, the following is performed:

[0070] Based on the task type of the subtask of the agent corresponding to the current policy network, obtain a sample set corresponding to the subtask. The sample set includes the local observation data of the agent and the operation policy corresponding to the local observation data;

[0071] Pre-train the current policy network based on the sample set and random noise to obtain a trained policy network.

[0072] In this step, for each robot, the robot can be controlled to complete basic actions in ways such as teleoperation, so as to generate a large number of data samples (o i , a i ), that is, the sample set. Then, under the condition of taking the Gaussian random noise z ∼ N(0,1) and the sample set (o i , a i ) as inputs, the policy network is pre-trained in a supervised learning or imitation learning manner so that it learns basic actions and operation policies to obtain a trained policy network.

[0073] Then, based on each trained policy network, the joint operation model is trained to obtain a trained joint operation model. The process of joint training is as follows:

[0074] Take the network parameters of each trained policy network as the initial parameters;

[0075] Randomly select the network parameters of each evaluation network, each communication network, and each output network, and determine the corresponding reward function according to the target task;

[0076] Use the MADDPG algorithm to train the joint operation model to obtain a trained joint operation model.

[0077] Through the above training process, the network parameters of the communication network, the output network, and the evaluation network can be determined, and the network parameters of the policy network can be fine-tuned to obtain a trained joint operation model.

[0078] Through the above operation process, the joint operation model provided by the present invention can be applied to various multi-agent reinforcement learning tasks with communication. In addition, based on the idea of the conditional generation network, the pre-trained policy network can be fine-tuned with the hidden layer information z output by the communication network, enabling the rapid training of the joint operation model.

[0079] Finally, for step 106, the observation data obtained by each space robot based on its sub-task is respectively input into the trained joint operation model to obtain the corresponding operation strategy for each space robot.

[0080] Since the joint operation model can achieve information interaction between agents, the operation strategies made by each space robot are more accurate.

[0081] As Figure 2 、 Figure 3 shown, an embodiment of the present invention provides a cooperative operation device for space robots based on information interaction. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From a hardware perspective, as Figure 2 shown, it is a hardware architecture diagram of an electronic device where the cooperative operation device for space robots based on information interaction provided by an embodiment of the present invention is located. In addition to Figure 2 the processor, memory, network interface, and non-volatile memory shown, the electronic device where the device is located in the embodiment usually may also include other hardware, such as a forwarding chip responsible for processing packets, etc. Taking software implementation as an example, as Figure 3 shown, as a logically meaningful device, it is formed by the CPU of its corresponding electronic device reading the computer program in the non-volatile memory into the memory for operation.

[0082] A cooperative operation device for space robots based on information interaction provided in this embodiment includes:

[0083] A determination module 300, configured to determine multiple space robots based on a target task. Each space robot corresponds to a sub-task and an agent network respectively, and each agent network includes a policy network, an evaluation network, a communication network, and an output network;

[0084] A construction module 302, configured to construct a joint operation model between each agent network. In the joint operation model, each agent network performs information interaction with other agent networks through its communication network and output network respectively;

[0085] A training module 304, configured to train the joint operation model to obtain a trained joint operation model;

[0086] An input module 306, configured to input the observation data obtained by each space robot based on its subtask into the trained joint operation model respectively, so as to obtain the corresponding operation strategy for each space robot.

[0087] In the embodiment of the present invention, the determination module 300 can be used to execute step 102 in the above method embodiment, the construction module 302 can be used to execute step 102 in the above method embodiment, the training module 304 can be used to execute step 104 in the above method embodiment, and the input module 306 can be used to execute step 106 in the above method embodiment.

[0088] In some embodiments, each agent network respectively performs information interaction with other agent networks through its communication network and output network, including:

[0089] Receiving the state information output by each other agent network by using the communication network of the current agent network;

[0090] Screening and fusing each received state information by using the communication network, and inputting the screened and fused information into the policy network of the current agent network;

[0091] Based on the screened and fused information, the policy network outputs the corresponding operation strategy to the environment and outputs its state information to the output network of the current agent network;

[0092] The output network processes the received state information and sends the processed state information to the communication network of other agent networks.

[0093] In some embodiments, the training module 304 is configured to perform the following operations:

[0094] Pre-training each policy network respectively to obtain a trained policy network;

[0095] Based on each trained policy network, training the joint operation model to obtain a trained joint operation model.

[0096] In some embodiments, pre-training each policy network respectively to obtain a trained policy network includes:

[0097] For each policy network, the following operations are all performed:

[0098] Based on the task type of the subtask of the agent corresponding to the current policy network, obtaining a sample set corresponding to the subtask, where the sample set includes the local observation data of the agent and the operation strategy corresponding to the local observation data;

[0099] Pre-training the current policy network based on the sample set and random noise to obtain a trained policy network.

[0100] In some embodiments, based on each trained policy network, the joint operation model is trained to obtain a trained joint operation model, including:

[0101] Taking the network parameters of each trained policy network as initial parameters;

[0102] Randomly selecting the network parameters of each evaluation network, each communication network, and each output network, and determining the corresponding reward function according to the target task;

[0103] Using the MADDPG algorithm to train the joint operation model to obtain a trained joint operation model.

[0104] In some embodiments, the random noise is Gaussian noise.

[0105] In some embodiments, the network structure of the policy network is an MLP structure and / or an LSTM structure.

[0106] It can be understood that the structure illustrated in the embodiments of the present invention does not constitute a specific limitation on a cooperative operation device of a space robot based on information interaction. In other embodiments of the present invention, a cooperative operation device of a space robot based on information interaction may include more or fewer components than those illustrated, or combine certain components, or split certain components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0107] Regarding the information interaction, execution process, etc. between the various modules in the above device, since they are based on the same concept as the method embodiments of the present invention, the specific content can be referred to the description in the method embodiments of the present invention and will not be elaborated here.

[0108] The embodiments of the present invention also provide an electronic device, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, it implements a cooperative operation method of a space robot based on information interaction in any embodiment of the present invention.

[0109] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it causes the processor to execute a cooperative operation method of a space robot based on information interaction in any embodiment of the present invention.

[0110] Specifically, a system or device equipped with a storage medium can be provided, on which software program codes for implementing the functions of any one of the above embodiments are stored, and the computer (or CPU or MPU) of the system or device is caused to read and execute the program codes stored in the storage medium.

[0111] In this case, the program codes read from the storage medium itself can implement the functions of any one of the above embodiments, so the program codes and the storage medium storing the program codes constitute a part of the present invention.

[0112] Examples of the storage medium for providing the program codes include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, the program codes can be downloaded from a server computer via a communication network.

[0113] In addition, it should be clear that not only can the functions of any one of the above embodiments be implemented by executing the program codes read by the computer, but also by causing an operating system or the like operating on the computer based on the instructions of the program codes to complete part or all of the actual operations.

[0114] In addition, it can be understood that the program codes read from the storage medium are written into the memory provided in the expansion board inserted into the computer or into the memory provided in the expansion module connected to the computer, and then based on the instructions of the program codes, the CPU or the like installed on the expansion board or the expansion module is caused to execute part and all of the actual operations, so as to implement the functions of any one of the above embodiments.

[0115] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of another identical element in the process, method, article or device comprising the element.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A collaborative operation method for a space robot based on information interaction, characterized in that, Comprising: Determine multiple space robots based on the target task, each of the space robots corresponding to a subtask and an agent network respectively, and each of the agent networks including a policy network, an evaluation network, a communication network, and an output network; Construct a joint operation model among each of the agent networks. In the joint operation model, each of the agent networks performs information interaction with other agent networks through its communication network and output network respectively; Train the joint operation model to obtain a trained joint operation model; Input the observation data obtained by each of the space robots based on its subtask into the trained joint operation model respectively to obtain the corresponding operation strategy for each of the space robots; Each of the agent networks performs information interaction with other agent networks through its communication network and output network respectively, including: Use the communication network of the current agent network to receive the state information output by each of the other agent networks; Use the communication network to screen and fuse each received state information, and input the screened and fused information into the policy network of the current agent network; The policy network outputs the corresponding operation strategy to the environment based on the screened and fused information and outputs its state information to the output network of the current agent network; The output network processes the received state information and sends the processed state information to the communication network of other agent networks.

2. The method according to claim 1, wherein The training of the joint operation model to obtain a trained joint operation model includes: Pre-train each of the policy networks respectively to obtain a trained policy network; Based on each of the trained policy networks, train the joint operation model to obtain a trained joint operation model.

3. The method according to claim 2, characterized in that, The pre-training of each of the policy networks respectively to obtain a trained policy network includes: For each of the policy networks, perform: Based on the task type of the subtask of the space robot corresponding to the current policy network, obtain a sample set corresponding to the subtask, and the sample set includes the local observation data of the space robot and the operation strategy corresponding to the local observation data; Pre-train the current policy network based on the sample set and random noise to obtain a trained policy network.

4. The method according to claim 2, wherein The training of the joint operation model based on each of the trained policy networks to obtain a trained joint operation model includes: Take the network parameters of each trained policy network as the initial parameters; Randomly select the network parameters of each of the evaluation networks, each of the communication networks, and each of the output networks, and determine the corresponding reward function according to the target task; Use the MADDPG algorithm to train the joint operation model to obtain a trained joint operation model.

5. The method according to claim 3, characterized in that The random noise is Gaussian noise.

6. The method according to claim 1, wherein The network structure of the policy network is an MLP structure, and / or, an LSTM structure.

7. A collaborative operation device for a space robot based on information interaction, characterized in that, Comprising: A determination module, configured to determine multiple space robots based on a target task, each of the space robots corresponding to a sub-task and an agent network respectively, and each of the agent networks including a policy network, an evaluation network, a communication network, and an output network; A construction module, configured to construct a joint operation model among each of the agent networks. In the joint operation model, each of the agent networks respectively performs information interaction with other agent networks through its communication network and output network; A training module, configured to train the joint operation model to obtain a trained joint operation model; An input module, configured to respectively input the observation data obtained by each of the space robots based on its sub-task into the trained joint operation model to obtain the corresponding operation strategy of each of the space robots; Each of the agent networks respectively performs information interaction with other agent networks through its communication network and output network, including: Receiving the state information output by each of the other agent networks by using the communication network of the current agent network; Screening and fusing each received state information by using the communication network, and inputting the screened and fused information into the policy network of the current agent network; The policy network outputs a corresponding operation strategy to the environment based on the screened and fused information and outputs its state information to the output network of the current agent network; The output network processes the received state information and sends the processed state information to the communication network of other agent networks.

8. A computing device, including a memory and a processor, where a computer program is stored in the memory, and when the processor executes the computer program, the method according to any one of claims 1-6 is implemented.

9. A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed on a computer, the computer is made to execute the method according to any one of claims 1-6.

Images