Hierarchical systems, methods for processing hierarchical systems

Abstract

The device nodes receive requests from the higher-level planning computer, assign tasks quickly, and execute them autonomously. [Solution] In a hierarchical system, the lower execution system creates a logical group formation recipe and sends it to the equipment node, receives an execution plan including evaluation indicators from the upper planning system server, creates an execution plan recipe including evaluation indicators converted into work for the lower workspace, and sends it to the equipment node. The lower execution system server and logical group formation recipe receive multiple equipment nodes that select candidates to be the overall leader and logical group leader for each logical group. When an equipment node that has become a logical group leader receives a request from another equipment node to join the logical group, it updates the current allocation status of the equipment node in the work environment management table. When other equipment nodes receive an execution plan recipe, they volunteer to be the equipment node that executes the work, receive a work selection notification from the leader, and execute the work when the received selection notification includes a work ID.

Description

Technical Field

[0001] The present invention relates to a hierarchical system including an upper planning system and a lower execution system, and to a processing method for a hierarchical system.

Background Art

[0002] Patent Document 1 of prior art discloses a method for providing multi-site orchestration in a public network for factory automation.

[0003] The method provides communication and computation functions to a plurality of sites configured to communicate with each other using slices of a public network, constructs a multi-site orchestration model based on initial performance of communication between different sites, determines chances of choreography between different sites by using the multi-site orchestration model, triggers choreography between different sites, evaluates the performance of choreography between different sites, and updates the multi-site orchestration model.

Prior Art Literature

Patent Literature

[0004]

Patent Document 1

Summary of the Invention

Problem to be Solved by the Invention

[0005] With the scaling-up of systems, it has become difficult to absorb system configuration changes of lower execution systems and dynamic changes in environments, and to suppress influences on upper planning system processing.

[0006] In addition, after obtaining an agreement on autonomous decision-making by lower execution systems, inter-group cooperation among a plurality of node configurations in different work areas is required for improving work efficiency.

[0007] For example, in conventional choreography control at lower execution levels, when multiple nodes are organized into work groups to process tasks, distributed algorithm processing is applied to the integration / modification / dissolution of work groups. However, distributed algorithm processing is computationally intensive and requires a very long time to compute.

[0008] The objective of this invention is to enable multiple device nodes to quickly assign tasks and autonomously execute them in response to requests received from a higher-level planning computer. [Means for solving the problem]

[0009] The above problem is solved by a hierarchical system comprising a higher-level planning server equipped with a processing unit and memory device, lower-level planning servers equipped with a processing unit and memory device, and multiple device nodes equipped with a processing unit and memory device, wherein the lower-level execution server processing unit creates a logical group formation recipe and sends it to the multiple device nodes, receives an execution plan including evaluation indicators from the processing unit of the higher-level planning server, creates an execution plan recipe including evaluation indicators converted into work for lower-level work areas, and sends it to the multiple device nodes, and the processing units of the multiple device nodes receive the logical group formation recipe and select candidates for overall leader and logical group leader for each logical group, the processing unit of the device node that has become a logical group leader updates the current allocation status of the device node in the work environment management table when it receives a request from a device node to join a logical group, the processing unit of the device node applies to be the device node that will execute the work when it receives an execution plan recipe and the work included in the execution plan recipe is executable, the logical group leader sends a work selection notification to the processing unit of the applicant device node, and the processing unit of the device node executes the work if the received selection notification includes a work ID. [Effects of the Invention]

[0010] Multiple device nodes can quickly assign tasks and autonomously execute them in response to requests received from higher-level planning computers. [Brief explanation of the drawing]

[0011] [Figure 1] This is an example of a hierarchical system configuration diagram in an embodiment. [Figure 2] This is an example of a configuration diagram for a lower-level execution server in an embodiment. [Figure 3] This is an example of a configuration diagram of the equipment nodes in the embodiment. [Figure 4] This is an example of an execution plan recipe in the embodiment. [Figure 5] This is an example of a logic group formation recipe in the embodiment. [Figure 6] This is an example of a work environment management table in the embodiment. [Figure 7] This figure illustrates examples of message types in the embodiment. [Figure 8] This is an example of a group member management table in the implementation example. [Figure 9] This is an example of a capability table in the embodiment. [Figure 10] This is an example of an environment-dependent management table in the implementation example. [Figure 11] This is an example of a timing chart showing the generation and processing of a logic group formation recipe in an embodiment. [Figure 12] This is an example of a timing chart showing the generation and processing of execution plan recipes in an embodiment. [Figure 13] This is an example of a timing chart showing the process of moving equipment nodes between logical groups in an embodiment. [Figure 14] This is an example flowchart showing the processing of the logic group formation recipe creation unit in the embodiment. [Figure 15] This is an example flowchart showing the processing of the logical group reshaping management unit in the embodiment. [Figure 16] This is an example flowchart showing the processing of the execution plan recipe creation unit in the embodiment. [Modes for carrying out the invention]

[0012] The present invention is not limited to the embodiments described below, and includes various modified examples and equivalent configurations within the spirit of the appended claims. For example, the above-described embodiments have been described in detail for the purpose of clearly explaining the present invention, and the present invention is not necessarily limited to those having all the configurations described.

[0013] In addition, part or all of the processing units and processing modules described in the embodiments may be implemented by hardware, for example, by designing them in an integrated circuit, or may be implemented by software through a processor interpreting and executing a program that realizes each respective function.

[0014] The information described in the embodiments may be a table or a database (DB), or may be data stored in a main memory.

[0015] Figure 1 is an example of a configuration diagram of a hierarchical system according to the embodiment. A lower execution system server 3 of a lower execution system 2 receives an execution plan from an upper planning system server 1 connected via a network.

[0016] The lower execution system server 3 that has received the execution plan sends work obtained by dividing the execution plan to a plurality of device nodes 5 connected via a network such as a wireless LAN 4. The device nodes execute the work divided according to their capabilities.

[0017] For example, in the management of a distribution warehouse, the present invention can be applied to a logistics system in which the upper planning system server 1 creates an execution plan based on rough information such as the cargo carry-in volume and date / time, and cargo carry-out volume and date / time when performing cargo carry-in work and carry-out work, a lower execution system server 3 such as VW (Virtual Warehouse) divides the execution plan, and causes device nodes such as vehicles, forklifts, and AGVs (Automatic Guided Vehicles) to execute the work.

[0018] In addition, the present invention can be applied to energy management systems such as VPPs (Virtual Power Plants) for solar power generation, to divide tasks among equipment nodes such as PVs (Photovoltaics), storage batteries, and PCS (Power Conditioning Systems) and allow them to execute autonomously.

[0019] In such a large-scale system, specifying detailed tasks for equipment node 5 from the higher-level planning server 1 places a heavy load on the higher-level planning server 1, resulting in a delay before specific work instructions are issued to equipment node 5.

[0020] Furthermore, it is difficult to keep equipment nodes operational at all times due to the need for failures, maintenance, and other factors. Therefore, it is necessary to use other equipment nodes to perform tasks when an equipment node is malfunctioning or undergoing maintenance.

[0021] This would require modifying the execution plan, which could place an additional burden on the higher-level planning server 1 in order to create the modified execution plan.

[0022] This embodiment describes a hierarchical system that solves the problems of choreographic architectures in which each device node connected to a network provides a service corresponding to each device node in response to service requests sent via their respective networks.

[0023] The first problem is that even in systems using choreographic architecture, the higher-level planning servers are becoming large in scale due to the integration of multiple systems.

[0024] To ensure system availability and scalability, the processing on higher-level planning servers should be simple and should be avoided in relation to changes in lower-level systems.

[0025] Furthermore, a centralized system that handles all processes, such as calculating target values ​​for achieving KPIs (Key Performance Indicators), creating execution plans, and determining specific execution details, on a single computer is not suitable for controlling hierarchical systems.

[0026] Furthermore, implementing processing that takes into account the dynamic changes in the lower-level execution environment leads to the lower-level execution environment becoming larger and the amount of computation required becoming enormous. This creates a problem where system downtime occurs due to processing changes on the higher-level planning server in response to configuration changes in the lower-level execution environment.

[0027] To address this problem, this embodiment adopts a hierarchical system configuration with loosely coupled layers, separating environment-dependent elements from the planning system.

[0028] Furthermore, in the planning system, server 1 calculates target values ​​for KPI achievement and creates an execution plan, while server 3, a lower-level execution system, concretizes the execution plan in accordance with the dynamic changes in the work environment.

[0029] The second problem is that in typical systems, equipment nodes are operated by work groups consisting of multiple equipment nodes within each work area. Therefore, coordination between different areas and groups is necessary to improve work efficiency.

[0030] However, when coordinating between groups, changes such as group mergers and reorganizations increase the processing load on the system, requiring time before actual work can begin and reducing efficiency.

[0031] To address this problem, in order to ensure consensus in autonomous decision-making among the various device nodes, a single overall leader for consensus assurance will be elected by all cooperating device nodes. The overall leader will streamline the decision-making process by forming logical groups for each work area and electing logical group leaders for each logical group.

[0032] Furthermore, logical group reshaping uses the equipment surplus / deficit information of the execution plan recipe that performs the actual work, and the overall leader selects the logical group leader and the equipment nodes belonging to the logical group.

[0033] Figure 2 shows an example of the configuration diagram of a lower-level execution server in the embodiment. The lower-level execution server 3 is implemented as a computer equipped with a CPU (Central Processing Unit) 10, main memory 11, external storage 12, and input / output device 13.

[0034] The CPU 10 may be a processing unit such as a GPU (Graphics Processing Unit). The main memory 11 consists of semiconductor elements such as RAM (Random Access Memory) and ROM (Read Only Memory). The external memory 12 consists of memory devices such as an HDD (Hard Disk Drive) and SSD (Solid State Drive). The input / output device 13 is an input / output device such as a display, touch panel, keyboard, and mouse.

[0035] These hardware components depend on the system to which the present invention is applied and are not limited to the hardware described above. In this embodiment, the case of implementation using a standalone computer is described as an example, but it may also be implemented using a cloud system that provides computing resources.

[0036] The main memory 11 stores the execution plan recipe creation unit 15, the logical group formation recipe creation unit 16, the logical group reformation management unit 17, the work environment management table update unit 18, the communication unit 19, and the message analysis unit 20. These processing units are implemented as software modules and executed by the CPU 10 while referring to data stored in the external memory 12. Alternatively, these processing modules may be implemented in hardware.

[0037] The external storage device 12 stores the execution plan 21, the logical group formation recipe 22, the execution plan recipe 23, and the work environment management table 24. These tables and data may be stored in a database or in memory.

[0038] The communication unit 19 receives messages sent via the network, and the message analysis unit 20 analyzes the received messages based on the type of message. Based on the results, the work environment management table update unit 18 updates the work environment management table 24 with information such as the number of devices in each area.

[0039] Figure 3 shows an example of the configuration diagram of a device node in the embodiment. Device node 5 is implemented by a computer equipped with a CPU 30, main memory 31, external memory 32, and input / output device 33.

[0040] The CPU 30 may be a processing unit such as a GPU. The main memory 11 is composed of semiconductor elements such as RAM and ROM. The external memory 12 is composed of storage devices such as HDDs and SSDs. The input / output device 33 is an input / output device such as a display, touch panel, keyboard, and mouse.

[0041] These hardware components depend on the system to which the present invention is applied and are not limited to the hardware described above. In a logistics warehouse management system, since the equipment node 5 is a vehicle such as a truck or a transport device such as a forklift, it may be implemented as part of a navigation system mounted on the vehicle or forklift, or it may be implemented as an external tablet terminal.

[0042] The main memory 31 includes an arbitration unit 35, a group management unit 36, an overall leader determination unit 37, a capability management unit 38, a leader processing unit 39, a logical group leader selection unit 40, a work assignment unit 41, a member processing unit 42, a logical group formation processing unit 43, a candidate processing unit 44, a work execution unit 45, a communication unit 46, and a message analysis unit 47. The leader processing unit 39 is a processing unit that performs logical group leader processing when a device node is selected as a logical group leader.

[0043] These processing units are implemented as software modules and executed by the CPU 30 while referencing data stored in the external storage device 32. Alternatively, these processing modules may be implemented in hardware.

[0044] The external storage device 32 stores the capability table 50, the environment-dependent management table 51, and the group member management table 52. These tables and data may be stored in a database or in memory.

[0045] Figure 4 shows an example of an execution plan recipe in the embodiment. An execution plan including evaluation indicators is received from the higher-level planning server 1 and converted into an execution plan recipe including evaluation indicators that have been converted into work content in units that can be executed by multiple device nodes in the lower-level workspace.

[0046] In this example, minimizing working time is specified as the KPI (Key Performance Indicator). Other KPIs in a logistics warehouse management system could include maximizing processing volume per unit of time or minimizing energy consumption.

[0047] Furthermore, the number of equipment nodes deployed in each work area, Area A and Area B, the task IDs of the divided work, the work content, and the surplus or shortage of equipment nodes are all recorded in a corresponding manner.

[0048] In this example, since the task involves transporting goods within a logistics warehouse, the task description includes the number and weight of the goods. Based on the task description, the equipment node can determine whether the goods are transportable.

[0049] If a negative value is listed in the surplus / deficit status, it indicates that the execution plan recipe assigned to the area lacks sufficient equipment nodes. Therefore, the logical group leader of that area must wait for equipment nodes belonging to other logical groups to volunteer to perform the task, thereby securing the necessary equipment nodes.

[0050] If a positive value is listed in the surplus / deficit status, it indicates that the execution plan recipe assigned to the area has surplus equipment. These devices may be eligible for inter-group movement to perform work in logical groups in other areas.

[0051] When the execution plan recipe is issued from the lower-level execution server 3, the divided tasks described in the execution plan recipe are executed on the device node 5.

[0052] Figure 5 shows an example of a logical group formation recipe in the embodiment. The logical group formation recipe is used to execute the execution plan from the higher-level planning server 1. For each logical group corresponding to an area that instructs the configuration of multiple device nodes, a logical group ID and a configuration number indicating the number of device nodes that make up the logical group are stored in association.

[0053] A logical group is formed when a logical group formation recipe is sent from the lower-level execution server 3 to the device node.

[0054] Figure 6 shows an example of a work environment management table in the embodiment. It stores the work area details, which show the work content for each area, the number of operational nodes, which is the number of equipment nodes operating in the area, the allocation status, which is the number of equipment nodes allocated to the area, and the owner of the equipment node, all of which are associated with each area.

[0055] When a device node moves to a different logical group, it sends information indicating the new logical group to all other device nodes and subordinate execution servers. Upon receiving this information, all other device nodes and subordinate execution servers update the current allocation status in the work environment management table 24.

[0056] Based on this work environment management table 24, it becomes possible to determine the number of equipment nodes located in a logical group (area), allowing the lower execution server 3 to notify the server of any surplus or shortage of equipment nodes in the execution plan recipe.

[0057] Figure 7 illustrates an example of message types in the embodiment. By sending the messages shown in this table to other device nodes and the lower execution system server 3 via the network, the device nodes can work together to perform tasks.

[0058] The JOIN message is used when a new device node is added to a hierarchical system and requests its participation in a logical group. LEADER is a message indicating that an equipment node is applying to become the overall leader. LD-VOTE is a message that casts a vote based on the capabilities of the device node in response to a candidate received via the LEADER message. LD-DONE is a message indicating that the overall leader has been confirmed.

[0059] ROLE is a message that requests the configuration of a device node's role. ROLE-DONE is a message that notifies the result of the role configuration of a device node. ORDER is a message that sends recipes for forming logical groups and execution plans. Nomination is a message of applying to take on a task in response to the action plan recipe. The selection notification is a message that informs the equipment node that applied for the task that it has been assigned to perform the work. It is also used to notify users that a logical group needs to be moved. The selection ACK is a message indicating that the selection notification message has been received.

[0060] The communication unit 46 receives these messages, the message analysis unit 47 analyzes the messages, and each processing unit in the main memory 31 performs processing corresponding to the messages.

[0061] Figure 8 shows an example of a group member management table in the embodiment. This table stores the node IDs of device nodes that have become members of a logical group.

[0062] Figure 9 shows an example of a capability table in the embodiment. It stores a classification indicating the type of capability and a capability value corresponding to that capability. When the capability of the device associated with the equipment node is registered in this table and an execution plan recipe is received, it is possible to determine whether the task can be executed on that equipment node and to inform the logical group leader of the time required to execute the task on that equipment node.

[0063] Figure 10 shows an example of an environment-dependent management table in the embodiment. It stores the local node ID of equipment node 5, the source group ID to which it belonged before the move, the logical group ID of the source logical group to which it belonged before the move, the current logical group ID, a Role indicating whether it is the overall leader / logical group leader / member, the coordinates of the equipment node's current location within the workspace, the progress of the work, and the operating status of the equipment node.

[0064] Based on the information stored in the Role, the device node performs the roles of Global Leader (LD), Logical Group Leader (LGL), or Member (M).

[0065] There are two methods for forming groups in the lower-level execution system. The first method, in order to guarantee decision-making consensus across all device nodes, involves selecting the necessary logical group leaders and device nodes to belong to each logical group according to a logical group formation recipe.

[0066] The second method involves the logical group leader and the overall leader selecting the logical group according to a logical group formation recipe in order to efficiently execute the execution plan from the higher-level planning server.

[0067] This is done by the leader of each logical group selecting the equipment node to move, based on the surplus / deficit information that the lower-level execution server includes as inter-group movement information in the 3 execution plan recipe, and then performing an inter-group movement to which the selected equipment node moves.

[0068] Figure 11 is an example of a timing chart showing the generation and processing of a logical group formation recipe in an embodiment. When a new device node is added to a device node group, the added device node sends a JOIN message from the communication unit 46 to all device nodes.

[0069] Upon receiving a JOIN message, the device node adds the newly added device node to the group member management table 52. Next, each device node submits a LEADER message via the communication unit 46, adding its own node's capabilities information to the message. Since performing the role of overall leader requires a certain amount of computing resources, performance / functional requirements for the device nodes submitting the application may be set in advance to prevent unnecessary applications.

[0070] Each device node votes for the overall leader from among the device nodes that have nominated themselves for the position via a LEADER message, and the overall leader is then responded to each device node with the selected node via an LC-VOTE message. A consensus-distributed algorithm is used in which the device node with the most votes via the LD-VOTE message is determined to become the overall leader.

[0071] The selected overall leader's device node ID is sent to all device nodes and the subordinate execution server 3 via an LD-DONE message. The overall leader is selected based on the processing power, communication speed, etc., but in the case of identical power levels, the device node with the smaller self-node ID may be selected as the overall leader.

[0072] The device node selected as the overall leader adds an LD (Likely a character) to the Role in the environment-dependent management table 51 to indicate that it is the overall leader.

[0073] After confirming the selection of the overall leader with an LD-DONE message, the subordinate execution server 3 creates a logical group formation recipe 22 and sends it to all device nodes with an ORDER message. The receiving device nodes determine whether they are eligible to run for logical group leader based on the capability table 50, and device nodes with sufficient functionality and capabilities run for logical group leader with a candidacy message.

[0074] The overall leader selects a logical group leader from among the nominated device nodes and notifies all device nodes and the subordinate execution server 3. The device node selected as the logical group leader adds LGL, indicating its role as logical group leader, to the Role in the environment-dependent management table 51.

[0075] Figure 12 is an example of a timing chart showing the generation and processing of execution plan recipes in an embodiment. The higher-level planning server 1 determines the KPI target value for the work and sends the equipment allocation plan to the work area corresponding to the logical group and the work details to the lower-level execution server 3.

[0076] The lower execution server 3 reads the number of device nodes allocated to the logical group from the current allocation status in the work environment management table 24. Then it compares the device allocation plan received from the higher planning server 1 with the current allocation status, and if it does not need to create a new logical group, it divides the work and creates the execution plan recipe 23.

[0077] The created execution plan recipe 23 is distributed to all equipment nodes via ORDER message. Each equipment node refers to the capability table 50 to determine whether the sent execution plan recipe 23 is executable on that node. If it is executable, it calculates the required work time and notifies all equipment nodes of its candidacy via a candidacy message.

[0078] The logical group leader selects equipment nodes that align with KPI targets. Selection results, including the task ID, are sent to equipment nodes that do not require logical group relocation, while selection results, including the destination logical group ID, are sent to equipment nodes that require logical group relocation, via selection notification messages.

[0079] The selection results will be sent to all device nodes and the lower-level execution server 3. If the KPI target is execution in the shortest time, the device node with the shortest required work time will be selected.

[0080] A device node that receives the selection result, including the task ID, returns a selection ACK message and executes the task. A device node that receives the selection result, including the destination logical group, submits a candidate message to the logical group leader of the destination logical group to volunteer for the move, and executes the task after obtaining the logical group leader's approval for the move.

[0081] When a device node moves out of a logical group, and the number of device nodes belonging to that logical group becomes zero, the logical group is deleted.

[0082] Figure 13 is an example of a timing chart showing the process of moving equipment nodes between logical groups in an embodiment. The higher-level planning server 1 determines the KPI target value for the work and sends the equipment allocation plan to the work area corresponding to the logical group and the work details to the lower-level execution server 3.

[0083] The lower execution server 3 reads the number of device nodes allocated to logical groups from the current allocation status in the work environment management table 24. Then, it compares the device allocation plan received from the higher planning server 1 with the current allocation status, and if there are any logical groups with 0 device nodes, it creates a logical group formation recipe 22 and sends it to all device nodes as an ORDER message.

[0084] The overall leader determines the logical group leader for the logical group to be generated based on the logical group formation recipe 22, notifies all device nodes of the logical group leader with a ROLE-DONE message, and reshapes the logical group.

[0085] The device node refers to the logical group formation recipe 22 and submits a nomination message to nominate itself as the logical group leader and to nominate itself to move to the created logical group. The logical group leader selected by the overall leader receives the nomination message for move, adds the nominated device node to its group, and notifies all device nodes and the subordinate execution server 3 of the device node's move.

[0086] The lower-level execution server 3 receives notification of equipment node relocation from the logical group leader, and if it meets the requirements of the equipment allocation plan, it divides the work into tasks, creates an execution plan recipe 23, and sends it to all equipment nodes.

[0087] The equipment node refers to the capability table 50 to determine whether the submitted execution plan recipe 23 can be executed on that equipment node. If it can, it calculates the required work time and notifies all equipment nodes of its candidacy with a candidacy message.

[0088] The logical group leader selects equipment nodes that align with the KPI objectives and sends selection notification messages to all equipment nodes and the subordinate execution server 3. If the KPI objective is to complete the task in the shortest possible time, the equipment node with the shortest required work time will be selected. The selected equipment node then performs the task.

[0089] Figure 14 shows an example flowchart illustrating the processing of the logical group formation recipe creation unit in the embodiment. The lower execution system server 3 receives a message from the device node (S1) and determines whether the received message is an LD-DONE message confirming the overall leader (S2). If it is any other message, it waits for another message to be received.

[0090] If the message confirms the overall leader, read the number of operational nodes from the work environment management table (S3). Create a logical group formation recipe (S4), and send the created logical group formation recipe to all nodes as an ORDER message (S5).

[0091] The system receives a message from the device node (S6), determines whether the received message is a ROL-DONE message indicating the completion of logical group formation (S7), and terminates processing if it is a ROL-DONE message indicating the completion of logical group formation. If it is any other message, the system reads the message again.

[0092] Figure 15 is an example flowchart showing the processing of the logical group reshaping management unit in the embodiment. The lower execution server 3 receives an execution plan from the higher planning server 1, which includes the area name, the number of required equipment nodes, and the work details (S10).

[0093] The current allocation status of the work environment management table 24 is read (S11), and it is determined whether the number of required equipment nodes in the execution plan is a positive number and whether the current allocation status of the target logical group is 0 (S12).

[0094] If the conditions in S12 are met, it is determined that a logical group needs to be created and the logical group formation recipe creation process is executed (S13). If the conditions in S12 are not met, the processing of the logical group reformation management unit 17 is terminated.

[0095] Figure 16 is an example flowchart showing the processing of the execution plan recipe creation unit in the embodiment. The lower execution system server 3 receives the execution plan from the upper planning system server 1 (S11). The logical group reformation management unit performs processing (S22). Since logical groups with zero device nodes among the logical groups that will perform the work require the creation of new logical groups, the logical group reformation process is performed to create the logical groups and place the necessary device nodes.

[0096] From the received execution plan, read the KPIs and node allocation plan for each work area, and determine the area name / number of units / work content (S23). Divide the work content, assign task IDs, and create an execution plan recipe 23 with each item set (S24).

[0097] The current allocation status is read from the work environment management table, the difference from the allocation plan is calculated (S25), and the calculated difference is set as the "surplus / deficit status" in the execution plan recipe 23 and sent to all equipment nodes (S26). A selection ACK message, which confirms the task assignment and logical group movement, is received from the equipment node (S27).

[0098] The device node ID and task ID for each logical group are determined, the number of allocated device nodes for each work area is calculated (S28), and the current allocation status in the work environment management table is updated (S29).

[0099] The processes described herein are intended to facilitate understanding of the present invention. Depending on the type, configuration, application, and characteristics of the storage system, necessary processes may be added or removed. Therefore, the present invention does not necessarily include all the processes described herein and is not limited to the examples provided. [Explanation of symbols]

[0100] 1. Higher-level planning server 2 Lower execution systems 3. Lower-level execution server 4 Silent LAN 5. Device Nodes 10, 30 CPU 11, 31 Main memory 12, 32 External storage device 13, 33 Input / Output Devices 15. Execution Plan Recipe Creation Department 16. Logical Group Formation Recipe Creation Department 17 Logical Group Reorganization Management Department 18. Work Environment Management Table Update Section 19, 46 Communications Department 20, 47 Message Analysis Department 21. Action Plan 22 Logical Group Formation Recipe 23 Execution Plan Recipe 24 Work Environment Management Table 35 Mediation Department 36 Group Management Department 37 Overall Leader Selection Department 38 Capability Management Department 39 Reader Processing Unit 40 Logical Group Selection Department 41 Work Assignment Department 42 Member Processing Unit 43 Logical Group Formation Processing Unit 44. Candidate Application Processing Section 45. Work Execution Department 50 Ability Table 51 Environment-dependent management table 52 Group Member Management Table

Claims

1. In a hierarchical system including a higher-level planning server equipped with a processing unit and storage device, a lower-level planning server equipped with a processing unit and storage device, and multiple device nodes equipped with processing units and storage devices, A processing unit of a lower-level execution server creates a logical group formation recipe, sends it to multiple device nodes, receives an execution plan including evaluation indicators from the processing unit of the higher-level planning server, creates an execution plan recipe including evaluation indicators converted into work for lower-level work areas, and sends it to the aforementioned multiple device nodes, The system includes a processing unit for multiple device nodes that receives the aforementioned logical group formation recipe and selects candidates for the overall leader and the logical group leader for each logical group, When the processing unit of the equipment node that has become the logical group leader receives a request from the equipment node to join the logical group, it updates the current allocation status of the equipment node in the work environment management table. When the processing unit of the aforementioned device node receives the execution plan recipe, if it is capable of performing the tasks included in the execution plan recipe, it will volunteer to be the device node that performs the tasks. The aforementioned logical group leader sends a notification of work selection to the processing unit of the nominated equipment node, The processing unit of the aforementioned equipment node is a hierarchical system that executes a task if the received selection notification includes a task ID.

2. In the hierarchical system described in claim 1, If the processing unit of the aforementioned device node includes the destination logical group ID in the selection notification sent from the logical group leader, it notifies the logical group leader of the destination logical group of the move. A hierarchical system that updates the current allocation status of the work environment management table stored in the storage device of the equipment node which is the logical group leader.

3. In the hierarchical system described in claim 1, The processing unit of the equipment node which is the logical group leader is a hierarchical system that deletes the logical group when the current allocation status of the equipment nodes included in the work environment management table is zero.

4. In the hierarchical system described in claim 1, The processing unit of the device node, which is the overall leader, sends a logical group formation recipe to multiple device nodes when a new logical group is needed. A hierarchical system that selects a logical group leader from the processing units of device nodes that have volunteered to be the logical group leader, and generates a new logical group.

5. In the hierarchical system described in claim 1, In a hierarchical system, when a device node is added to the system, the processing unit of the added device node sends information indicating the addition of the new device node to the processing units of all device nodes. The processing unit of the device node that receives the information indicating the addition sends information indicating its candidacy to the processing units of all device nodes. The device node's memory unit is equipped with a global leader determination unit that selects the global leader from among the device nodes that have received the candidacy information.

6. In the hierarchical system described in claim 5, The overall leader determination unit, executed by the processing unit of the equipment node, is a hierarchical system that selects the overall leader based on the capabilities of the processing unit of the equipment node.

7. In the hierarchical system described in claim 1, The aforementioned device node has a logical group formation processing unit that determines whether or not to run for logical group leader when it receives a group formation recipe. A hierarchical system comprising a device node that is the overall leader, and whose storage device is equipped with a logical group leader selection unit that selects a logical group leader from among the candidate device nodes.

8. In the hierarchical system described in claim 1, Upon receiving the execution plan recipe, the processing unit of the equipment node refers to a capability table containing the equipment node's functions and performance, determines the time required for execution if the execution plan recipe is executable, and sends information to the logical group leader indicating its candidacy to execute the execution plan recipe. A hierarchical system in which the logical group leader selects the equipment node that will execute the execution plan recipe from among the equipment nodes that have volunteered based on the evaluation indicators.

9. In a processing method for a hierarchical system including a higher-level planning server equipped with a processing unit and a storage device, a lower-level planning server equipped with a processing unit and a storage device, and multiple device nodes equipped with processing units and storage devices, The processing unit of the lower execution server creates a logical group formation recipe and sends it to multiple device nodes, receives an execution plan including evaluation indicators from the processing unit of the higher planning server, creates an execution plan recipe including evaluation indicators converted into work for the lower workspace, and sends it to the multiple device nodes. The processing units of multiple device nodes receive the logical group formation recipe and select candidates for the overall leader and the logical group leader for each logical group. When the processing unit of the equipment node that has become the logical group leader receives a request from the equipment node to join the logical group, it updates the current allocation status of the equipment node in the work environment management table. When the processing unit of the aforementioned device node receives the execution plan recipe, if it is capable of performing the tasks included in the execution plan recipe, it will volunteer to be the device node that performs the tasks. The aforementioned logical group leader sends a notification of work selection to the processing unit of the nominated equipment node, A processing method for a hierarchical system in which the processing unit of the aforementioned device node executes a task if the received selection notification includes a task ID.

Images