Multi-device cooperative task execution management method and system for power distribution scenarios

Abstract

The application provides a multi-device cooperative task execution management method and system for power distribution scenarios, and relates to the technical field of distributed control. The method comprises the following steps: loading power distribution scenario demand targets to a task rule library, and decomposing to obtain a plurality of atomic sub-tasks; performing task parallel dependence analysis, and constructing a task sequential execution chain; performing device reuse analysis, and obtaining P reuse time chains of P reuse devices and W task execution time windows of W non-reuse devices; performing task execution control fitting, and outputting a plurality of task execution control parameters; obtaining a task cooperative control chain; and driving the P reuse devices and the W non-reuse devices to perform device space-time cooperative control management. The application solves the technical problem that the multi-device task scheduling method of the prior art usually ignores the space-time coordination problem between devices, resource contention and time conflicts occur when devices are scheduled, the scheduling efficiency of the devices is reduced, and the task execution efficiency in the power distribution scenario is low.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of distributed control, in particular to a multi-device cooperative task execution management method and system for power distribution scenarios. BACKGROUND

[0002] Tasks in power distribution scenarios usually involve the cooperation of multiple devices, such as device monitoring, state updating, data collection, etc. These tasks are sometimes executed in parallel and sometimes sequentially dependent. Device scheduling needs to consider not only the capabilities and locations of devices, but also the dependency between tasks, device availability, and other factors. Therefore, task scheduling and device cooperative execution require a highly optimized management method.

[0003] In the prior art, the spatio-temporal coordination problem between devices is usually ignored when executing tasks. Conflicts, path overlaps, and execution timing conflicts between devices are not effectively solved, resulting in resource contention and time conflicts during device scheduling. This causes delays in task execution and waste of device resources, which cannot meet the real-time and efficient task execution requirements. Especially when multiple tasks are executed in parallel, conflicts between devices cannot be avoided, resulting in low efficiency of task execution in power distribution scenarios. SUMMARY

[0004] The present application provides a multi-device cooperative task execution management method and system for power distribution scenarios, aiming to solve the technical problem that the multi-device task scheduling method of the prior art usually ignores the spatio-temporal coordination problem between devices, resulting in resource contention and time conflicts during device scheduling, which reduces the efficiency of device scheduling and further reduces the efficiency of task execution in power distribution scenarios.

[0005] The first aspect of the present application provides a multi-device cooperative task execution management method for power distribution scenarios, the method comprising: loading real-time uploaded power distribution scenario demand targets to a task rule library to obtain a plurality of atomic-level subtasks; performing task parallel dependency analysis on the plurality of atomic-level subtasks to construct a task sequential execution chain, wherein the task sequential execution chain comprises a plurality of parallel subtask groups in series; performing device multiplexing analysis based on the task sequential execution chain to obtain P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices; fitting task execution control parameters based on building information of the power distribution scenario; associating the plurality of task execution control parameters based on the task sequential execution chain to obtain a task cooperative control chain; and driving the P multiplexing devices and W non-multiplexing devices to perform spatio-temporal coordination control management based on the task cooperative control chain, taking the P multiplexing time chains and W task execution time windows as device parallel scheduling constraints.

[0006] In a second aspect, a multi-device cooperative task execution management system for a power distribution scenario is provided. The system is used for the multi-device cooperative task execution management method for a power distribution scenario. The system includes a subtask acquisition module configured to obtain a plurality of atomic subtasks by loading a real-time uploaded power distribution scenario demand target into a task rule library; a task parallel dependency analysis module configured to perform task parallel dependency analysis on the plurality of atomic subtasks and construct a task sequential execution chain, wherein the task sequential execution chain includes a plurality of parallel subtask groups in series; a device multiplexing analysis module configured to perform device multiplexing analysis based on the task sequential execution chain to obtain P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices; a task execution control fitting module configured to perform task execution control fitting on the plurality of atomic subtasks according to power distribution scenario building information to output a plurality of task execution control parameters; a cooperative control chain acquisition module configured to associate the plurality of task execution control parameters based on the task sequential execution chain to obtain a task cooperative control chain; and a control management module configured to drive the P multiplexing devices and the W non-multiplexing devices to perform device space-time cooperative control management based on the task cooperative control chain, with the P multiplexing time chains and the W task execution time windows as device parallel scheduling constraints.

[0007] One or more technical solutions provided in the present application have at least the following beneficial effects:

[0008] By loading the real-time uploaded power distribution scene demand target to the task rule library, and decomposing the task into multiple atomic sub-tasks, it is ensured that the task can be refined and executed step by step, thereby enhancing the flexibility and response speed of the system; the task parallel dependence analysis is performed on multiple atomic sub-tasks, and the task sequential execution chain is constructed, which improves the efficiency of parallel execution and avoids the execution delay between tasks; through device reuse analysis, P reuse time chains of P reuse devices and P task execution time windows of W non-reuse devices are generated, so that the reuse devices and non-reuse devices can work effectively without interfering with each other during task execution, and the balance between device use and task execution is optimized; according to the power distribution scene building information, the task execution is controlled and fitted to generate task execution control parameters, which ensures that the device executes the task according to the optimal path, and this process reduces the device conflict caused by space layout and physical obstacles, and ensures efficient task completion; by associating the task execution control parameters, the task coordination control chain is generated, which clearly defines the order, path and coordination mode of device execution, which provides a systematic control scheme for device coordination and improves the coordination of devices in task execution; based on the task coordination control chain and device scheduling constraints, the device can realize space-time coordination control management during task execution, avoid execution bottlenecks caused by time conflicts or path overlaps, ensure efficient cooperation between devices, and reduce conflicts and delays that may occur during task execution.

[0009] The above description is only a summary of the technical scheme of the present application. In order to more clearly understand the technical means of the present application, the following specific embodiments of the present application can be implemented according to the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0010] Figure 1 The power distribution scene-oriented multi-device cooperative task execution management method flowchart provided by the embodiments of the present application.

[0011] Figure 2 The power distribution scene-oriented multi-device cooperative task execution management system structure diagram provided by the embodiments of the present application.

[0012] Mark explanation: sub-task acquisition module 10, task parallel dependence analysis module 20, device reuse analysis module 30, task execution control fitting module 40, cooperative control chain acquisition module 50, control management module 60. DETAILED DESCRIPTION

[0013] This application provides a method and system for managing the collaborative task execution of multiple devices in a power distribution scenario. It solves the technical problem that existing multi-device task scheduling methods often neglect the spatiotemporal coordination between devices, leading to resource contention and time conflicts during device scheduling, resulting in reduced scheduling efficiency and consequently low task execution efficiency in power distribution scenarios.

[0014] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0015] Example 1, as Figure 1 As shown in the embodiments of this application, a multi-device collaborative task execution management method for power distribution scenarios is provided, the method including:

[0016] By loading the real-time uploaded power distribution scenario requirements into the task rule library, multiple atomic-level subtasks are obtained.

[0017] The power distribution scenario requirement objective refers to the task or operation objective that needs to be completed in the system. For example, a task may require data collection at a specific location, starting equipment, or monitoring the status of equipment. The power distribution scenario requirement objective uploaded in real time is analyzed in detail to extract the key elements of the task and match them with the rules in the task rule base to determine how each task should be decomposed into smaller atomic-level subtasks. For example, if the task objective involves the operation of device A, this operation needs to be broken down into multiple subtasks, such as device startup, data collection, and operation feedback. Based on the decomposition method defined in the task rule base, the power distribution scenario requirement objective is decomposed into multiple atomic-level subtasks. The atomic-level subtask is the most basic and indivisible unit, which includes specific execution actions, time, equipment requirements, etc.

[0018] Perform parallel dependency analysis on the multiple atomic-level subtasks to construct a sequential execution chain of tasks, wherein the sequential execution chain of tasks includes multiple parallel subtask groups in sequence.

[0019] Analyze the parallel dependencies of multiple atomic-level subtasks to determine which atomic-level subtasks can be executed in parallel and which must be executed sequentially. For example, device A and device B can execute a subtask group simultaneously, while device C must be executed after devices A and B have completed their tasks. Find the dependencies by traversing the task connection matrix and mark the order of tasks.

[0020] The task dependency relationship is converted into a task topology graph, each node representing a subtask, and each edge representing a dependency relationship. According to the task topology graph, it is analyzed which subtasks can be executed in parallel and which subtasks need to be executed in series. By analyzing these task dependency relationships, a task sequential execution chain is constructed. Among them, the parallel subtask group refers to those tasks that can be executed at the same time and are not dependent on each other, for example, "parallel group 1{device A, device B}→serial task 1{device C}", which means that device A and device B can perform tasks at the same time, and device C must wait for device A and B to complete before starting.

[0021] Based on the task sequential execution chain, device multiplexing analysis is performed to obtain P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices.

[0022] According to the execution requirements of each atomic-level subtask, the corresponding device attributes are matched, including device ID, task execution reference time consumption, etc. Combined with the task dependency relationship in the task sequential execution chain, it is analyzed whether the device can be multiplexed. For example, if some atomic-level subtasks in the task sequence need to operate on the same device at different time periods, these atomic-level subtasks will share the device. Through device multiplexing analysis, P multiplexing time chains of P multiplexing devices are generated, i.e. the usage time period of each device, which is used to ensure that the device usage is reasonably arranged without conflict. P is the number of multiplexing devices and is a positive integer. For those devices that cannot be multiplexed, i.e. W non-multiplexing devices, W corresponding task execution time windows are allocated, which means that these non-multiplexing devices can only work in a specific time period in the task execution chain. W is the number of non-multiplexing devices and is a positive integer.

[0023] According to the power distribution scene building information, the task execution control fitting of the plurality of atomic-level subtasks is performed, and a plurality of task execution control parameters are output.

[0024] The power distribution scene building information is analyzed, including building structure, such as geographic and physical layout of the power distribution scene, and coordinate distribution of existing physical obstacles. The operation target of the power distribution scene is projected into the spatial structure to determine the data collection position during task execution, for example, some tasks need to collect data or perform actions at specific positions, and the collection sites are determined through the power distribution scene building information. With physical obstacles and building structure as constraints, the task execution control parameters of each task are determined, for example, the device needs to avoid certain physical obstacles during execution, so the path or action sequence of the device needs to be adjusted. Based on the position of the device, the task requirements and the physical obstacles, a plurality of task execution control parameters are fitted and output.

[0025] According to the task sequential execution chain, the plurality of task execution control parameters are associated to obtain a task cooperative control chain.

[0026] The task execution control parameters of each parallel task group are associated with the related tasks in the task sequential execution chain, which has determined which atomic sub-tasks are parallel and which are serial, so that the execution order of the devices in each task, resource allocation and device configuration can be adjusted based on this information. Specifically, the multiple task execution control parameters are merged according to the structure of the task sequential execution chain, and based on the parallel task groups and serial task groups in the task sequential execution chain, it is clear which devices need to be executed simultaneously and which devices need to be executed sequentially, so as to aggregate the task execution control parameters of each task group to ensure that the devices within the task group can execute their respective tasks in coordination. Finally, a task coordination control chain is obtained, which means that the execution order, execution condition and coordination mode of all devices are precisely defined, ensuring that the execution of the entire task is efficient and coordinated.

[0027] The P multiplexing devices and W non-multiplexing devices are driven to perform device space-time coordination control management based on the task coordination control chain under the constraints of the P multiplexing time chains and W task execution time windows.

[0028] For the P multiplexing time chains, the corresponding P multiplexing devices will be called multiple times in the task, so their usage time needs to be planned and coordinated to avoid conflicts; for the W task execution time windows, the task execution time windows of the corresponding W non-multiplexing devices are already fixed, so it is necessary to ensure that these non-multiplexing devices execute tasks within a specific time period.

[0029] Based on the task coordination control chain, the devices are scheduled in parallel, specifically, the parallel execution of the devices is arranged according to the constraints of the P multiplexing time chains and W task execution time windows, for example, the task coordination control chain specifies that device A and device B execute a certain task at the same time, while device C must execute after A and B have finished.

[0030] Through the task coordination control chain, the P multiplexing devices and W non-multiplexing devices are driven to execute tasks according to the specified time and space conditions, where the scheduling of multiplexing devices needs to consider their switching and coordination between different tasks, and non-multiplexing devices strictly execute according to the plan, for example, devices A and B need to execute tasks in the same spatial area, their scheduling needs to ensure that there is no conflict at the same time, while also avoiding physical obstacles.

[0031] Further, the task parallel dependency analysis is performed on the multiple atomic sub-tasks to construct a task sequential execution chain, and the method comprises:

[0032] The plurality of atomic-level sub-tasks are traversed through the task connection association matrix to obtain a plurality of dependent sub-tasks; the plurality of atomic-level sub-tasks are connected according to the plurality of dependent sub-tasks to output a task topology graph; a plurality of parallel sub-task groups without dependent relationships are obtained by decomposing the task topology graph; and the plurality of parallel sub-task groups are serially connected according to the inter-group dependent relationship of the plurality of parallel sub-task groups to obtain the task sequential execution chain.

[0033] The task connection association matrix is a two-dimensional array, wherein each element represents the connection relationship between one atomic-level sub-task and another atomic-level sub-task, such as dependency, sequence, etc. For example, if atomic-level sub-task 1 must be executed after atomic-level sub-task 2, the corresponding position of the matrix is marked as “dependent” in the task connection association matrix. By traversing the task connection association matrix, a plurality of dependent sub-tasks are identified, which are composed of a plurality of atomic-level sub-tasks. The plurality of atomic-level sub-tasks in each dependent sub-task group have a dependent relationship with each other and must be executed in a specific order. For example, if atomic-level sub-task 1 depends on atomic-level sub-task 2, and atomic-level sub-task 2 depends on atomic-level sub-task 3, sub-tasks 1, 2 and 3 will be divided into a group, and the tasks in the group will be executed one by one according to the dependent relationship.

[0034] Out-degree refers to the number of atomic-level sub-tasks pointing to other atomic-level sub-tasks in the dependent chain. By counting the out-degree of each atomic-level sub-task, the dependent relationship of the task can be determined. The task topology graph is a directed acyclic graph that shows the dependent relationship of all atomic-level sub-tasks, ensuring that each atomic-level sub-task is executed in the correct order. In the task topology graph, each atomic-level sub-task is a node, and the dependent relationship between atomic-level sub-tasks is represented by a directed edge. The direction of the arrow indicates the direction of the dependent relationship. For example, if task 1 depends on tasks 2 and 3, in the task topology graph, tasks 2 and 3 will point to task 1, indicating that task 1 must wait for tasks 2 and 3 to complete before execution.

[0035] In the task topology graph, any task without a dependent relationship can be considered parallel. By analyzing the task topology graph, a set of sub-tasks without a dependent relationship is identified. For example, if tasks A, B and C have no dependent relationship, they can be executed in parallel in the same time period. A plurality of parallel sub-task groups are obtained by decomposing the task topology graph, and the tasks in each parallel sub-task group are independent and can be executed simultaneously.

[0036] There are inter-group dependencies between multiple parallel sub-task groups, for example, the execution of a certain parallel sub-task group depends on the completion of another parallel sub-task group. By analyzing the dependencies between parallel sub-task groups, the order of task execution is determined. By serially arranging multiple parallel sub-task groups according to their dependencies, a task sequential execution chain is finally obtained, which ensures that each task is executed in the correct order and optimizes the parallelism and dependencies of the tasks.

[0037] Further, the multiple task execution control parameters are associated with the task sequential execution chain to obtain a task cooperative control chain, and the method comprises:

[0038] According to the multiple parallel sub-task groups, the multiple task execution control parameters are aggregated to obtain multiple multi-device cooperative control parameters. According to the connection relationship of the multiple parallel sub-task groups in the task sequential execution chain, the multiple multi-device cooperative control parameters are mapped and spliced to obtain the task cooperative control chain. With the P multiplexing time chains and W task execution time windows as device parallel scheduling constraints, the P multiplexing devices and W non-multiplexing devices are driven to perform device space-time cooperative control management in the multiple parallel sub-task groups based on the task cooperative control chain.

[0039] Each parallel sub-task group has corresponding task execution control parameters describing the operation requirements of the devices in the parallel sub-task group. In any parallel sub-task group, multiple devices need to cooperatively execute tasks. The aggregated multi-device cooperative control parameters ensure the cooperative work between multiple devices, ensuring that they can effectively cooperate within the same time period. The resource sharing, scheduling conflicts and cooperative execution efficiency between devices are considered during aggregation. After merging the multiple task execution control parameters of multiple parallel sub-task groups, multiple multi-device cooperative control parameters are generated. These multi-device cooperative control parameters define how devices coordinate work during parallel execution, ensuring that tasks can be completed simultaneously and efficiently.

[0040] In the task sequential execution chain, the execution order of multiple parallel sub-task groups is dependent, for example, a parallel sub-task group must be executed after another parallel sub-task group. By analyzing the connection relationship in the task sequential execution chain, the multiple multi-device cooperative control parameters of multiple parallel sub-task groups are mapped and spliced. The mapping process refers to integrating the multiple multi-device cooperative control parameters of multiple parallel sub-task groups into a unified execution framework. Splicing refers to connecting multiple multi-device cooperative control parameters in order to form a complete task cooperative control chain. The task cooperative control chain defines the scheduling, execution time, coordination method, etc. of all devices in the entire task process, ensuring that multiple devices can work cooperatively and successfully complete the task.

[0041] Through the task cooperative control chain, the execution order of the P multiplexing devices and the W non-multiplexing devices is scheduled according to the constraint conditions of the P multiplexing time chains and the W task execution time windows, and a plurality of multi-device cooperative control parameters in the task cooperative control chain guide the scheduling, execution time, coordination mode and the like of the devices. During the execution of a plurality of parallel subtask groups, the devices are ensured to be cooperative in time and space, for example, the scheduling of the multiplexing devices needs to ensure that they can be smoothly switched between a plurality of parallel subtask groups, and the non-multiplexing devices need to execute tasks according to fixed time windows, and the time and space cooperative control management ensures that there is no conflict between devices, and each device can complete the task according to the specified time and space conditions.

[0042] Further, based on the task sequential execution chain, device multiplexing analysis is performed to obtain P multiplexing time chains of the P multiplexing devices and W task execution time windows of the W non-multiplexing devices, and the method comprises:

[0043] In the task attribute library matching, a plurality of task execution attributes of the plurality of atomic-level subtasks are obtained, wherein the task execution attributes comprise a device ID and a task execution reference time consumption; and according to the plurality of task execution attributes, serial task device multiplexing analysis is performed on the task sequential execution chain to obtain P multiplexing time chains of the P multiplexing devices and W task execution time windows of the W non-multiplexing devices.

[0044] The task attribute library contains various information required for task execution, in particular, the task execution attributes of each atomic-level subtask, wherein the task execution attributes comprise a device ID and a task execution reference time consumption, the device ID indicates the device required to execute the atomic-level subtask, and each atomic-level subtask depends on certain specific devices; the task execution reference time consumption indicates the basic time required for the execution of each atomic-level subtask, that is, the time required to complete the atomic-level subtask under ideal conditions without considering any external factors, and this attribute is used to evaluate the task execution time and device scheduling.

[0045] In the task sequential execution chain, different device requirements may exist between any tasks, and for the devices that can be multiplexed, they can be used in rotation between multiple tasks to improve device utilization, and serial task device multiplexing analysis is to determine which devices can be shared between multiple tasks and which devices must be used alone.

[0046] The P multiplexing devices are devices that are used multiple times during task execution, for each multiplexing device, the multiplexing time period thereof between different tasks in the task sequential execution chain is analyzed and determined, and based on the device ID and the task execution reference time consumption in the task execution attributes, the multiplexing time chain of each multiplexing device is calculated, the P multiplexing time chains refer to the use time periods of the P multiplexing devices between different tasks, and it is necessary to ensure that the multiplexing devices between different tasks do not have time conflicts.

[0047] W non-multiplexed devices are devices that cannot be multiplexed in the task sequential execution chain, each non-multiplexed device can only execute one task, so each non-multiplexed device has a fixed task execution time window, i.e. it must complete the task within a specific time period, based on the device ID and task execution reference duration in the task execution attribute, the task execution time window of each non-multiplexed device is calculated, each task execution time window is fixed and cannot conflict with other tasks.

[0048] Further, the method further comprises:

[0049] extracting a plurality of task execution reference durations and a plurality of device IDs from the plurality of task execution attributes; converting the plurality of task execution reference durations into a plurality of task execution time windows according to the execution order characteristics of the plurality of atomic sub-tasks in the task sequential execution chain; separating Q multiplexed devices and an initial non-multiplexed device set according to the recurrence frequency of the plurality of device IDs; extracting Q sets of task execution time windows associated with the Q multiplexed devices from the plurality of task execution time windows, and then performing time overlap conflict resolution on the Q sets of task execution time windows to generate the P multiplexed time chains of the P multiplexed devices; incorporating Q-P multiplexed devices excluded in the conflict resolution process into the initial non-multiplexed device set to obtain the W non-multiplexed devices; and calling the W task execution time windows corresponding to the W non-multiplexed devices from the plurality of task execution time windows with the P multiplexed time chains as time occupation constraints.

[0050] For each atomic sub-task, the task execution reference duration and the device ID are extracted from the corresponding task execution attribute.

[0051] The task sequential execution chain defines the execution order of a plurality of atomic sub-tasks, by analyzing the execution order of these tasks, the start time of each atomic sub-task is determined, the task execution time window represents the time range occupied by the atomic sub-task during execution, by combining the task execution reference duration of each task with the start time of the task, the task execution time window of each atomic sub-task is determined.

[0052] Iterate through the device IDs in the task execution attribute, and count the recurrence frequency of each device ID, the recurrence frequency represents the number of times the device is repeatedly used in multiple tasks, devices with high device recurrence frequency have high multiplexability. According to the recurrence frequency, the device IDs are divided into two categories, wherein the recurrence frequency of the Q multiplexed devices is high and shared among multiple tasks, Q is the number of multiplexed devices, which is a positive integer; the recurrence frequency of the initial non-multiplexed device set is low and cannot be shared among multiple tasks.

[0053] According to the plurality of task execution time windows, Q sets of task execution time windows corresponding to Q multiplexing devices are extracted, and any set of task execution time windows represents a plurality of time ranges in which the corresponding multiplexing device executes a plurality of tasks. When the plurality of task execution time windows of any multiplexing device overlap, it may cause conflict use of the multiplexing device. In order to solve this conflict, the overlapping conditions need to be eliminated to ensure that the multiplexing device will not be used by multiple tasks in the same time period.

[0054] For example, assume that multiplexing device A is applied to task 1 and task 2, and the execution time windows of task 1 and task 2 overlap, that is, multiplexing device A needs to execute the two tasks in the same time period. Then device A cannot complete the two tasks at the same time, resulting in a time conflict. In this case, multiplexing device A, which cannot execute the tasks, is excluded from the multiplexing device set, that is, the original multiplexing device A is applied to one of task 1 or task 2 as a non-multiplexing device, and the remaining tasks are executed by other devices.

[0055] After conflict resolution, P multiplexing time chains of P multiplexing devices are generated, P being a positive integer less than Q. These multiplexing time chains represent the time allocation of each multiplexing device during task execution, ensuring that the use time period of each device does not conflict.

[0056] After the conflict resolution process, Q-P multiplexing devices that cannot be multiplexed have been excluded from the plurality of multiplexing devices. The Q-P multiplexing devices are reclassified as non-multiplexing devices. These devices cannot be shared multiple times during task execution, so their use is fixed to a certain task and they are no longer multiplexed among multiple tasks. These devices are added to the initial non-multiplexing device set, and finally W non-multiplexing devices are obtained.

[0057] In the scheduling process, the P multiplexing time chains of the P multiplexing devices are used as time occupancy constraints, which means that other tasks cannot use the device during the use time period of each multiplexing device. These multiplexing time chains provide a time window for device scheduling, ensuring that tasks will not be delayed due to device conflicts. Based on the time windows of the tasks in the task order execution chain, W task execution time windows of W non-multiplexing devices are called according to the P multiplexing time chains. The time windows of non-multiplexing devices have been fixed in the previous steps, so they must execute tasks within the specified time. At this time, the multiplexing time chain of the multiplexing device and the task execution time window of the non-multiplexing device are used as the constraint condition for scheduling, ensuring that the devices will not conflict during task execution.

[0058] Further, the plurality of atomic sub-tasks are fitted for task execution control according to the building information of the power distribution scene, and a plurality of task execution control parameters are output. The method comprises:

[0059] The power distribution scene building information is analyzed to obtain a power distribution scene spatial structure and a physical obstacle coordinate distribution; an operation target of the power distribution scene demand target is projected to the power distribution scene spatial structure to locate a task data collection site; W task execution control parameters are obtained by performing device control parameter fitting according to the task data collection site and W device starting points of the W non-multiplexing devices, with the physical obstacle coordinate distribution as a spatial constraint condition; P starting position sequences are obtained by performing starting point updating of the P multiplexing devices according to the P multiplexing time chains; and P sets of task execution control parameters are obtained by performing multi-level progressive control parameter fitting according to the task data collection site and the P starting position sequences, with the physical obstacle coordinate distribution as a spatial constraint condition, wherein the W task execution control parameters and the P sets of task execution control parameters constitute the plurality of task execution control parameters.

[0060] The power distribution scene spatial structure refers to the layout of a building, walls, the size of a room, a passageway, a door and a window and other spatial elements, which all affect the layout of devices and task execution. For example, some devices need to be operated in a specific area or must avoid some areas. The physical obstacles include walls, columns, device spacing, doors and the like. The physical obstacle coordinate distribution directly affects the movement and task execution of devices and provides necessary constraints for subsequent spatial planning of tasks.

[0061] The power distribution scene demand target refers to a task or an operation target that needs to be completed in the system. For example, a task requires collecting data at a specific location, starting a device or monitoring the status of a device and the like. These operation targets are projected to the power distribution scene spatial structure. For example, if a task requires collecting device data in a specific area, a suitable collection location needs to be selected in the area. The power distribution scene spatial structure provides a physical environment for task execution, and the demand target provides the operation requirements of the task. According to the operation target and the power distribution scene spatial structure, a task data collection site is determined. The task data collection site is a specific location where a device needs to collect information when executing a task. The device must be positioned and dispatched according to these locations when executing a task.

[0062] The physical obstacle coordinate distribution is used to ensure that the device does not collide with obstacles when executing a task. For example, the path of the device needs to avoid obstacles during task execution to ensure normal operation of the device. Each task requires the device to start from different device starting points. For non-multiplexing devices, the device starting point is fixed and is usually the starting position required by the task. The task data collection site affects the path selection, speed control and operation mode of the device. The device control parameters need to be adapted to the execution requirements of the task according to the task data collection site.

[0063] According to the physical obstacle coordinate distribution, the task data collection site, and the device starting point, device control parameters are fitted, including the device starting time, path planning, execution speed, moving direction, etc. For example, if the task requires the device to start from position A, collect data through certain specific areas, and avoid obstacles, the device control parameters will be fitted according to these requirements. The fitting of device control parameters ensures that the device can travel according to the predetermined path and avoid physical obstacles when performing the task, and finally complete the task on time. Through the fitting process, W task execution control parameters of W non-multiplexing devices are finally obtained, and each task execution control parameter ensures that the device can complete the task in the predetermined space while meeting all physical constraints and task requirements.

[0064] According to the use time of the device in the P multiplexing time chain, the starting position of the device is calculated, and the starting position of the device is adjusted according to the change of the multiplexing time chain, especially when the device switches between multiple tasks. For example, if the device needs to perform task B immediately after completing task A, the starting position of task B should be updated at the end position of task A to ensure seamless connection. Through the above analysis, P starting position sequences of P multiplexing devices are generated, which provide accurate time and space information for the starting position of each multiplexing device for subsequent task scheduling and control.

[0065] Similarly, taking the physical obstacle coordinate distribution as the spatial constraint condition, according to the task data collection site, combined with the P starting position sequences, multi-level progressive control parameter fitting is performed, that is, a multi-level progressive control method is used to gradually fit the device control parameters. The progressive fitting process is carried out step by step, starting from the initial position of the device, calculating the path, speed, and execution time of the device, etc., until the task is completed. Through progressive control parameter fitting, P sets of task execution control parameters are finally obtained, which include device path, execution time, speed, etc. information, ensuring that the device can successfully complete the task.

[0066] Combining the W task execution control parameters of the W non-multiplexing devices with the P task execution control parameters of the P multiplexing devices, a plurality of task execution control parameters are finally obtained, which provide complete information for task scheduling and device control, ensuring that the device can be scheduled and executed as expected.

[0067] Further, the plurality of atomic-level sub-tasks are used to traverse the task connection association matrix to obtain a plurality of dependent sub-tasks. Previously, the method comprises:

[0068] The plurality of atomic-level sub-tasks are combined to obtain a plurality of groups of atomic-level sub-tasks; a plurality of first connection direction recurrence frequencies of the plurality of groups of atomic-level sub-tasks are extracted from the historical task log; a plurality of first total execution frequencies of a plurality of first sub-tasks in the plurality of groups of atomic-level sub-tasks are extracted from the historical task log; the plurality of first connection direction recurrence frequencies are divided by the plurality of first total execution frequencies to obtain a plurality of first connection direction correlation strength values of the plurality of groups of atomic-level sub-tasks; a plurality of second connection direction correlation strength values of the plurality of groups of atomic-level sub-tasks are calculated by analogy; and the task connection correlation matrix is constructed based on a filtering result of the plurality of first connection direction correlation strength values and the plurality of second connection direction correlation strength values based on a preset correlation strength threshold.

[0069] Through the decomposition process of the task, each task is decomposed into a plurality of atomic-level sub-tasks, and the plurality of atomic-level sub-tasks are combined and enumerated, that is, any two atomic-level sub-tasks are combined two by two to obtain a plurality of groups of atomic-level sub-tasks for subsequent dependency analysis.

[0070] The historical task log records the information of the past executed tasks, including the start time, the end time, the used device, and the dependency relationship between the atomic-level sub-tasks. The connection direction recurrence frequency represents the frequency of the execution of two atomic-level sub-tasks in a specific order in the historical task log. For example, the atomic-level sub-task A1 of the task A is usually executed before the atomic-level sub-task B1 of the task B, and the recurrence frequency of this order can be represented as the first connection direction recurrence frequency of the atomic-level sub-task A1 to the atomic-level sub-task B1.

[0071] For the plurality of groups of atomic-level sub-tasks, a plurality of first sub-tasks are determined, that is, in each group of atomic-level sub-tasks, the atomic-level sub-task in the front is taken as the first sub-task. For example, in the previous example, the atomic-level sub-task A1 is taken as the first sub-task, and the execution of the first sub-task in each task execution is found in the historical task log. The first total execution frequency represents the frequency of the execution of the corresponding first sub-task in the past task execution process.

[0072] The first connection direction recurrence frequency is divided by the corresponding first total execution frequency to obtain the first connection direction correlation strength value of each group of atomic-level sub-tasks, which reflects the strength of the dependency relationship between the two atomic-level sub-tasks.

[0073] The first connection direction correlation strength value is the dependency relationship strength from the front atomic-level sub-task to the rear atomic-level sub-task in any group of atomic-level sub-tasks, and the second connection direction correlation strength value is the dependency relationship strength from the rear atomic-level sub-task to the front atomic-level sub-task. The second connection direction correlation strength value is calculated using a similar calculation process as the first connection direction correlation strength value.

[0074] In order to screen out effective connection relationships, a correlation strength threshold is preset, and only when the correlation strength of a certain connection direction is greater than the correlation strength threshold, it is considered that it is an effective dependent relationship that needs to be considered in task scheduling. Based on the correlation strength threshold, the plurality of first connection direction correlation strength values and the plurality of second connection direction correlation strength values are screened, and those connection relationships with correlation strength exceeding the correlation strength threshold are retained, and in this way, weak dependent relationships can be filtered out, and only key dependencies in task execution are concerned. According to the screened connection direction, a task connection correlation matrix is constructed, and each matrix element of the matrix represents whether there is a dependent relationship between two atomic-level sub-tasks, and the strength of the dependent relationship.

[0075] Further, the plurality of atomic-level sub-tasks are used to traverse the task connection correlation matrix to obtain a plurality of dependent sub-tasks, and the method comprises:

[0076] A task correlation threshold is predefined, wherein the task correlation threshold is greater than A times the correlation strength threshold; during the process of traversing the task connection correlation matrix using the plurality of atomic-level sub-tasks, the task correlation threshold is used for dependent sub-task correlation screening, and a direct dependent pair set is output; and the plurality of atomic-level sub-tasks are aggregated according to the direct dependent pair set, and the plurality of dependent sub-tasks are output.

[0077] A task correlation threshold is predefined, which is greater than A times the correlation strength threshold, and A is a preset amplification coefficient for limiting the relationship between the task correlation threshold and the correlation strength threshold. For example, if A times the correlation strength threshold is 0.6, the task correlation threshold should be greater than 0.6, which can be set to 0.7 or other suitable values.

[0078] In the task connection correlation matrix, each row and each column represents an atomic-level sub-task, and the elements in the matrix represent the correlation strength between two atomic-level sub-tasks. By traversing the task connection correlation matrix, the correlation between each two atomic-level sub-tasks is viewed. For each element in the matrix, the correlation strength is compared with the predefined task correlation threshold. If the correlation strength of a pair of atomic-level sub-tasks is greater than the set task correlation threshold, it is considered that there is an effective dependent relationship between the two atomic-level sub-tasks, which belongs to a direct dependent pair. Through the screening process, a direct dependent pair set, i.e. those task pairs with strong dependent relationships, is output.

[0079] From the direct dependency pairs in the set, it can be found that there are higher-level dependency relationships between some atomic-level sub-tasks, for example, if A depends on B and B depends on C, then there is an indirect dependency relationship between A and C. By aggregating these direct dependency pairs, larger dependency groups can be formed, each consisting of multiple atomic-level sub-tasks, and the execution order and dependency relationship between them have been determined through direct dependency pairs. After aggregation, multiple sets of dependent sub-tasks are output, each set of dependent sub-tasks reflecting the overall dependency relationship between multiple atomic-level sub-tasks.

[0080] Further, in the process of driving the P multiplexing devices and the W non-multiplexing devices to perform space-time collaborative control management based on the task collaborative control chain, the P multiplexing devices are subjected to continuous smooth transition processing in the cross-parallel sub-task group switching process.

[0081] Multiplexing devices usually switch between multiple tasks, for example, device A needs to execute task group 1 first, and then switch to task group 2 for execution. Since task group 1 and task group 2 may be executed in parallel, it means that device A needs to transition smoothly when switching task groups to avoid delays or interruptions in the device during the task switching process. Illustratively, during the switching process, the scheduling strategy of the multiplexing device is dynamically adjusted, and if there is a time overlap or path conflict in the task group switching, the execution order or start time of the multiplexing device is adjusted in real time to avoid conflicts in task execution; to avoid the multiplexing device immediately changing the execution state during the task switching process, a gradual switching strategy can be used, for example, when task group 1 is approaching completion, the multiplexing device can prepare resources or configurations for task group 2 in advance to reduce the delay of switching. Through the driving of the task collaborative control chain, the multiplexing device can switch according to the time constraints, path planning, execution order, etc. between task groups, optimize execution efficiency, and ensure smooth transition between parallel task groups.

[0082] Further, the method further comprises:

[0083] According to the W task execution control parameters, W first task energy consumptions are calculated; based on the W non-multiplexing IDs, W historical energy consumption record sets and W theoretical energy consumption test sets are retrieved; based on the W historical energy consumption record sets and the W theoretical energy consumption test sets, device energy consumption fluctuations are solved to obtain a first reference fluctuation scale; the W first task energy consumptions are compensated using the first reference fluctuation scale to obtain W first steady-state energy consumptions; and the W non-multiplexing devices are retrieved and called using the W non-multiplexing IDs and the W first steady-state energy consumptions.

[0084] According to W task execution control parameters, the energy consumed by each non-multiplexing device when executing a task is calculated through a predefined energy consumption model or an empirical formula. Common calculation methods include calculating the task energy consumption according to the product of the device power and the execution time. Finally, W first task energy consumptions are obtained, which provide input for subsequent steps.

[0085] The non-multiplexing ID is a unique device ID for each non-multiplexing device. Through these IDs, the energy consumption data of the corresponding device can be extracted from historical energy consumption data and theoretical energy consumption test data sets. The W historical energy consumption record sets contain energy consumption data of the device when executing tasks in the past, which is derived from previous task execution, device performance monitoring, operation logs, etc. The W theoretical energy consumption test sets contain energy consumption data estimated according to the specifications and task requirements of the device, which is usually obtained through theoretical calculation or simulation in a standard experimental environment.

[0086] By comparing the W historical energy consumption record sets and the W theoretical energy consumption test sets, the fluctuation of the energy consumption of the device during task execution is analyzed. For example, if the energy consumption of a device fluctuates greatly in different tasks, it indicates that the load or task execution environment of the device is unstable. The energy consumption fluctuation range of the device is calculated through mathematical methods such as standard deviation and coefficient of variation, which represents the range of energy consumption changes of the device during task execution as the first reference fluctuation scale.

[0087] The purpose of compensation is to determine the necessary energy consumption required for task execution. According to the first reference fluctuation scale, the W first task energy consumptions are adjusted. For example, if the energy consumption of a device fluctuates greatly, a compensation factor based on the fluctuation scale is added or subtracted to compensate for it, so that the energy consumption is more in line with the actual energy consumption demand during operation. After compensation, W first steady-state energy consumptions are obtained, which reflect the actual energy consumption of the device during task execution.

[0088] According to the W non-multiplexing IDs and the first steady-state energy consumption of each device, the devices that can provide corresponding energy consumption are searched from the device pool to ensure that the steady-state energy consumption of each device is within the specified range. Through retrieval and calling, W non-multiplexing devices are obtained to ensure that the task can be executed smoothly.

[0089] Further, the method further comprises:

[0090] Smooth transition fitting is performed on the P sets of task execution control parameters to obtain P task execution control chains. P second task energy consumptions are calculated based on the P task execution control chains. P second reference fluctuation scales are obtained by solving the energy consumption fluctuation of the device based on historical energy consumption data. The compensation results of the P second task energy consumptions are matched by using the P second reference fluctuation scales to obtain the P multiplexing devices.

[0091] Since the multiplexing device needs to switch between different tasks when performing multiple tasks, the transition process may not be smooth. By using a smoothing transition fitting method such as linear interpolation, smooth curve fitting, etc., sudden changes of the device during task switching are reduced, making the execution behavior of the device more stable. By using the fitting method, the P sets of task execution control parameters are smoothly transitioned during task switching, ensuring seamless connection of the device to the next task. After smoothing transition fitting, P task execution control chains are generated, which describe the smooth transition and execution control mode of each multiplexing device during multiple task execution.

[0092] The P task execution control chains describe various control parameters of the device during task execution, including task execution duration, load, resource consumption, etc. These parameters are the basis for calculating energy consumption. According to the P task execution control chains, the energy consumption of the multiplexing device during task execution is calculated. The calculation formula is power multiplied by execution time multiplied by load coefficient, where power is the power consumption of the device, execution time is the execution duration of the task, and load coefficient reflects the load condition of the device during task execution. After calculation, P second task energy consumptions, i.e., energy consumption values of the P multiplexing devices during task execution, are obtained.

[0093] Based on historical task execution records, historical energy consumption data of the device during past task execution are obtained, which include energy consumption fluctuation information of the device under different tasks. By comparing the energy consumption fluctuation range of the device in historical tasks, the variability of the device energy consumption is analyzed, for example, the standard deviation or coefficient of variation of the energy consumption of each device is calculated to measure the fluctuation degree of the energy consumption. Through calculation, the second reference fluctuation scale of each multiplexing device is obtained, which represents the reference value of the energy consumption fluctuation of the device during task execution.

[0094] The P second task energy consumptions are compensated using the P second reference fluctuation scales, specifically, the second task energy consumptions are added or subtracted by a fluctuation compensation factor. The fluctuation compensation factor is calculated based on the second reference fluctuation scale and aims to eliminate the energy consumption fluctuation caused by changes in device load or task switching.

[0095] According to the compensated energy consumption data, device retrieval and matching are performed. The device retrieval and matching are performed by comparing the steady-state energy consumption of the device with the requirements of the task energy consumption, to ensure that the selected device can execute the task within the expected energy consumption range. Finally, P multiplexing devices are obtained, thereby ensuring the energy efficiency of the system and the smooth completion of the task.

[0096] In summary, the multi-device collaborative task execution management method for power distribution scenarios provided by the embodiments of the present application has the following technical effects:

[0097] By loading the real-time uploaded power distribution scene demand target into the task rule library and decomposing the task into multiple atomic sub-tasks, the flexibility and response speed of the system are enhanced; the multiple atomic sub-tasks are analyzed for task parallel dependence, and a task sequential execution chain is constructed, which improves the efficiency of parallel execution and avoids execution delay between tasks; through device reuse analysis, P reuse time chains of P reuse devices and P task execution time windows of W non-reuse devices are generated, so that reuse devices and non-reuse devices can work effectively without interfering with each other during task execution, and the balance between device use and task execution is optimized; according to the power distribution scene building information, task execution control fitting is performed to generate task execution control parameters, ensuring that devices execute tasks according to the optimal path, which reduces device conflicts caused by spatial layout and physical obstacles and ensures efficient task completion; by associating task execution control parameters, a task coordination control chain is generated to determine the order, path and coordination method of device execution, which provides a systematic control scheme for device coordination and improves the coordination of devices in task execution; based on the task coordination control chain and device scheduling constraints, devices can achieve spatiotemporal coordination control management during task execution, avoiding execution bottlenecks caused by time conflicts or path overlaps, ensuring efficient collaboration between devices and reducing conflicts and delays that may occur during task execution.

[0098] In the second embodiment, based on the same inventive concept as the power distribution scene-oriented multi-device collaborative task execution management method in the foregoing embodiments, as shown in the following Figure 2 The embodiments of the present application provide a power distribution scene-oriented multi-device collaborative task execution management system, which comprises:

[0099] The subtask acquisition module 10 is configured to obtain a plurality of atomic-level subtasks by loading a real-time uploaded power distribution scene demand target into a task rule library; the task parallel dependency analysis module 20 is configured to perform task parallel dependency analysis on the plurality of atomic-level subtasks and construct a task sequential execution chain, wherein the task sequential execution chain comprises a plurality of parallel subtask groups in series; the device multiplexing analysis module 30 is configured to perform device multiplexing analysis based on the task sequential execution chain to obtain P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices; the task execution control fitting module 40 is configured to perform task execution control fitting on the plurality of atomic-level subtasks according to power distribution scene building information and output a plurality of task execution control parameters; the collaborative control chain acquisition module 50 is configured to associate the plurality of task execution control parameters according to the task sequential execution chain to obtain a task collaborative control chain; and the control management module 60 is configured to take the P multiplexing time chains and the W task execution time windows as device parallel scheduling constraints and drive the P multiplexing devices and the W non-multiplexing devices to perform device space-time collaborative control management based on the task collaborative control chain.

[0100] Further, the task parallel dependency analysis module 20 is configured to perform the following operation steps:

[0101] The plurality of atomic-level subtasks are traversed through a task connection association matrix to obtain a plurality of groups of dependent subtasks; the plurality of atomic-level subtasks are connected according to out-degree of the plurality of groups of dependent subtasks to output a task topology graph; a plurality of parallel subtask groups without dependency are obtained by decomposing the task topology graph; and the plurality of parallel subtask groups are serially connected according to inter-group dependency of the plurality of parallel subtask groups to obtain the task sequential execution chain.

[0102] Further, the collaborative control chain acquisition module 50 is configured to perform the following operation steps:

[0103] The plurality of task execution control parameters are aggregated according to the plurality of parallel subtask groups to obtain a plurality of multi-device collaborative control parameters; the plurality of multi-device collaborative control parameters are mapped and spliced according to a connection relationship of the plurality of parallel subtask groups in the task sequential execution chain to obtain the task collaborative control chain; and the P multiplexing time chains and the W task execution time windows are taken as device parallel scheduling constraints, and the P multiplexing devices and the W non-multiplexing devices are driven to perform device space-time collaborative control management based on the task collaborative control chain.

[0104] Further, the device multiplexing analysis module 30 is configured to perform the following operation steps:

[0105] a plurality of task execution attributes of the plurality of atomic sub-tasks are matched from the task attribute library, wherein the task execution attributes include device IDs and task execution benchmark time consumptions; serial task device multiplexing analysis is performed on the task sequential execution chain according to the plurality of task execution attributes, and P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices are obtained.

[0106] Further, the device multiplexing analysis module 30 is configured to perform the following operation steps:

[0107] a plurality of task execution benchmark time consumptions and a plurality of device IDs are extracted from the plurality of task execution attributes; the plurality of task execution benchmark time consumptions are converted into a plurality of task execution time windows according to execution order characteristics of the plurality of atomic sub-tasks in the task sequential execution chain; Q multiplexing devices and an initial non-multiplexing device set are separated according to recurrence frequencies of the plurality of device IDs; Q groups of task execution time windows associated with the Q multiplexing devices are extracted from the plurality of task execution time windows, and the Q groups of task execution time windows are subjected to time overlap conflict resolution to generate the P multiplexing time chains of the P multiplexing devices; Q-P multiplexing devices excluded in the conflict resolution process are incorporated into the initial non-multiplexing device set to obtain the W non-multiplexing devices; the P multiplexing time chains are taken as time occupation constraints, and the W task execution time windows corresponding to the W non-multiplexing devices are called from the plurality of task execution time windows.

[0108] Further, the task execution control fitting module 40 is configured to perform the following operation steps:

[0109] the power distribution scene building information is parsed to obtain a power distribution scene spatial structure and a physical obstacle coordinate distribution; the operation target of the power distribution scene demand target is projected to the power distribution scene spatial structure to locate a task data collection site; device control parameter fitting is performed according to the task data collection site and W device starting points of the W non-multiplexing devices with the physical obstacle coordinate distribution as a spatial constraint condition to obtain W task execution control parameters; starting point updating of the P multiplexing devices is performed according to the P multiplexing time chains to obtain P starting position sequences; multi-level progressive control parameter fitting is performed according to the task data collection site and the P starting position sequences with the physical obstacle coordinate distribution as a spatial constraint condition to obtain P groups of task execution control parameters, wherein the W task execution control parameters and the P groups of task execution control parameters constitute the plurality of task execution control parameters.

[0110] Further, the task parallel dependence analysis module 20 is configured to perform the following operation steps:

[0111] The multiple atomic-level sub-tasks are enumerated to obtain multiple groups of atomic-level sub-tasks; multiple first connection direction recurrence frequencies of the multiple groups of atomic-level sub-tasks are extracted from a historical task log; multiple first total execution frequencies of multiple first sub-tasks in the multiple groups of atomic-level sub-tasks are extracted from the historical task log; the multiple first connection direction recurrence frequencies are mapped divided by the multiple first total execution frequencies to obtain multiple first connection direction correlation strength values of the multiple groups of atomic-level sub-tasks; multiple second connection direction correlation strength values of the multiple groups of atomic-level sub-tasks are calculated by analogy; after the multiple first connection direction correlation strength values and the multiple second connection direction correlation strength values are filtered based on a preset correlation strength threshold, the task connection correlation matrix is constructed based on a filtering result.

[0112] Further, the task parallel dependency analysis module 20 is configured to perform the following operation steps:

[0113] A task correlation threshold is predefined, wherein the task correlation threshold is greater than A times the correlation strength threshold; during the traversal of the task connection correlation matrix by the multiple atomic-level sub-tasks, the task correlation threshold is used for correlation dependency sub-task filtering, and a direct dependency pair set is output; and the multiple atomic-level sub-tasks are aggregated according to the direct dependency pair set, and the multiple groups of dependency sub-tasks are output.

[0114] Further, during the driving of the P multiplexing devices and the W non-multiplexing devices to perform the device space-time collaborative control management process based on the task collaborative control chain, a continuous smooth transition process of cross-parallel sub-task group switching is performed on the P multiplexing devices.

[0115] Further, the device multiplexing analysis module 30 is configured to perform the following operation steps:

[0116] W first task energy consumptions are calculated according to the W task execution control parameters; W historical energy consumption record sets and W theoretical energy consumption test sets are obtained by searching based on W non-multiplexing IDs; device energy consumption fluctuations are solved based on the W historical energy consumption record sets and the W theoretical energy consumption test sets, to obtain a first reference fluctuation scale; W first stable energy consumptions are obtained by compensating the W first task energy consumptions by using the first reference fluctuation scale; and the W non-multiplexing devices are searched and called by using the W non-multiplexing IDs and the W first stable energy consumptions.

[0117] Further, the device multiplexing analysis module 30 is configured to perform the following operation steps:

[0118] The P group of task execution control parameters are smoothly transitioned and fitted to obtain P task execution control chains; P second task energy consumptions are calculated according to the P task execution control chains; P second reference fluctuation scales are obtained by solving device energy consumption fluctuation based on historical energy consumption data; and the P multiplexing devices are obtained by performing device retrieval matching on compensation results of the P second task energy consumptions using the P second reference fluctuation scales.

[0119] Through the foregoing detailed description of the multi-device cooperative task execution management method for the power distribution scenario, those skilled in the art can clearly understand the multi-device cooperative task execution management system for the power distribution scenario in the embodiments. Since the system corresponds to the method disclosed in the embodiments, the system is described relatively simply, and the relevant parts can be referred to the method part.

[0120] The above description of disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the application. Therefore, the present application will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for managing multi-device cooperative task execution for power distribution scenarios, characterized in that, The method comprises: by loading the real-time uploaded power distribution scene demand target to the task rule library, a plurality of atomic level sub-tasks are decomposed; task parallel dependence analysis is performed on the plurality of atomic level sub-tasks, and a task sequential execution chain is constructed, wherein the task sequential execution chain comprises a plurality of parallel sub-task groups in series; based on the task sequential execution chain, device multiplexing analysis is performed, P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices are obtained; task execution control fitting is performed on the plurality of atomic level sub-tasks according to the power distribution scene building information, and a plurality of task execution control parameters are output; the plurality of task execution control parameters are associated with the task sequential execution chain, and a task collaborative control chain is obtained; the P multiplexing time chains and the W task execution time windows are taken as device parallel scheduling constraints, and the P multiplexing devices and the W non-multiplexing devices are driven to perform device space-time collaborative control management based on the task collaborative control chain; task parallel dependence analysis is performed on the plurality of atomic level sub-tasks, and a task sequential execution chain is constructed, the method comprising: a plurality of dependent sub-tasks are obtained by traversing the task connection association matrix with the plurality of atomic level sub-tasks; the plurality of atomic level sub-tasks are connected according to the out-degree of the plurality of dependent sub-tasks, and a task topology graph is output; a plurality of parallel sub-task groups without dependence relationship are obtained by decomposing the task topology graph; the plurality of parallel sub-task groups are serialized according to the inter-group dependence relationship of the plurality of parallel sub-task groups, and the task sequential execution chain is obtained; the plurality of task execution control parameters are associated with the task sequential execution chain, and a task collaborative control chain is obtained, the method comprising: the plurality of task execution control parameters are aggregated according to the plurality of parallel sub-task groups, and a plurality of multi-device collaborative control parameters are obtained; the plurality of multi-device collaborative control parameters are mapped and spliced according to the connection relationship of the plurality of parallel sub-task groups in the task sequential execution chain, and the task collaborative control chain is obtained; the P multiplexing time chains and the W task execution time windows are taken as device parallel scheduling constraints, and the P multiplexing devices and the W non-multiplexing devices are driven to perform device space-time collaborative control management based on the task collaborative control chain in the plurality of parallel sub-task groups; based on the task sequential execution chain, device multiplexing analysis is performed, P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices are obtained, the method comprising: a plurality of task execution attributes of the plurality of atomic level sub-tasks are matched in the task attribute library, wherein the task execution attributes comprise device ID and task execution reference time consumption; according to the plurality of task execution attributes, serial task device multiplexing analysis is performed on the task sequential execution chain, P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices are obtained.

2. The multi-device coordinated task execution management method for power distribution scenarios of claim 1, wherein, The method further comprises: a plurality of task execution reference time consumptions and a plurality of device IDs are extracted from the plurality of task execution attributes; according to the execution order characteristics of the plurality of atomic level sub-tasks in the task sequential execution chain, the plurality of task execution reference time consumptions are converted into a plurality of task execution time windows; According to the recurrence frequency of the plurality of device IDs, separate Q multiplexing devices and an initial non-multiplexing device set; After extracting Q sets of task execution time windows associated with the Q multiplexing devices from the plurality of task execution time windows, perform time overlap conflict resolution on the Q sets of task execution time windows to generate the P multiplexing time chains of the P multiplexing devices; Integrate Q-P multiplexing devices excluded in the conflict resolution process into the initial non-multiplexing device set to obtain the W non-multiplexing devices; Call the W task execution time windows corresponding to the W non-multiplexing devices from the plurality of task execution time windows as time occupation constraints.

3. The power distribution scenario oriented multi-device cooperative task execution management method of claim 1, wherein, According to the power distribution scene building information, perform task execution control fitting on the plurality of atomic-level subtasks, and output a plurality of task execution control parameters, the method comprising: Parse the power distribution scene building information to obtain the power distribution scene spatial structure and the physical obstacle coordinate distribution; Project the operation target of the power distribution scene demand target to the power distribution scene spatial structure to locate the task data collection site; Under the spatial constraint condition of the physical obstacle coordinate distribution, perform device control parameter fitting according to the task data collection site and the W device starting points of the W non-multiplexing devices to obtain W task execution control parameters; Update the starting points of the P multiplexing devices according to the P multiplexing time chains to obtain P starting position sequences; Under the spatial constraint condition of the physical obstacle coordinate distribution, perform multi-level progressive control parameter fitting according to the task data collection site and the P starting position sequences to obtain P sets of task execution control parameters, wherein the W task execution control parameters and the P sets of task execution control parameters constitute the plurality of task execution control parameters.

4. The power distribution scenario oriented multi-device cooperative task execution management method of claim 1, wherein, Traverse the task connection association matrix using the plurality of atomic-level subtasks to obtain a plurality of dependent subtasks, and before that, the method comprises: Combine and enumerate the plurality of atomic-level subtasks to obtain a plurality of groups of atomic-level subtasks; Extract a plurality of first connection direction recurrence frequencies of the plurality of groups of atomic-level subtasks from historical task logs; Extract a plurality of first total execution frequencies of a plurality of first subtasks in the plurality of groups of atomic-level subtasks from the historical task logs; Map the plurality of first connection direction recurrence frequencies by the plurality of first total execution frequencies to obtain a plurality of first connection direction association strength values of the plurality of groups of atomic-level subtasks; Analogically calculate a plurality of second connection direction association strength values of the plurality of groups of atomic-level subtasks; After screening the plurality of first connection direction association strength values and the plurality of second connection direction association strength values based on a preset association strength threshold, construct the task connection association matrix based on the screening result.

5. The multi-device coordinated task execution management method for power distribution scenarios of claim 4, wherein, Traverse the task connection association matrix using the plurality of atomic-level subtasks to obtain a plurality of dependent subtasks, and the method comprises: Predefine a task association threshold, wherein the task association threshold is greater than A times the association strength threshold; During the process of traversing the task connection association matrix using the plurality of atomic-level subtasks, use the task association threshold to screen associated dependent subtasks, and output a direct dependent pair set; The direct dependency pair set is aggregated according to the plurality of atomic subtasks, and the plurality of sets of dependent subtasks are output.

6. The power distribution scenario oriented multi-device cooperative task execution management method of claim 1, wherein, In the device space collaborative control management process performed by the P multiplexing devices and the W non-multiplexing devices based on the task collaborative control chain, the P multiplexing devices are subjected to continuous smooth transition processing of the cross-parallel subtask group switching process.

7. The multi-device coordinated task execution management method for power distribution scenarios of claim 3, wherein, The method further comprises: W first task energy consumptions are calculated according to the W task execution control parameters; W historical energy consumption record sets and W theoretical energy consumption test sets are obtained based on W non-multiplexing IDs; First reference fluctuation scales are obtained by solving device energy consumption fluctuations based on the W historical energy consumption record sets and the W theoretical energy consumption test sets; W first stable energy consumptions are obtained by compensating the W first task energy consumptions using the first reference fluctuation scales; The W non-multiplexing devices are called using W non-multiplexing IDs and the W first stable energy consumptions.

8. The multi-device coordinated task execution management method for power distribution scenarios of claim 3, wherein, The method further comprises: P task execution control chains are obtained by smooth transition fitting of the P sets of task execution control parameters; P second task energy consumptions are calculated according to the P task execution control chains; P second reference fluctuation scales are obtained by solving device energy consumption fluctuations based on historical energy consumption data; The P multiplexing devices are obtained by device retrieval matching of compensation results of the P second task energy consumptions using the P second reference fluctuation scales.

9. A multi-device coordinated task execution management system for power distribution scenarios, characterized in that, A system for implementing the power distribution scene-oriented multi-device collaborative task execution management method of any one of claims 1-8, the system comprising: A subtask acquisition module configured to obtain a plurality of atomic subtasks by loading real-time uploaded power distribution scene demand targets to a task rule library; A task parallel dependency analysis module configured to perform task parallel dependency analysis on the plurality of atomic subtasks, and construct a task sequential execution chain, wherein the task sequential execution chain comprises a plurality of parallel subtask groups in series; A device multiplexing analysis module configured to perform device multiplexing analysis based on the task sequential execution chain, and obtain P multiplexing time chains of P multiplexing devices and W task execution time windows of W non-multiplexing devices; A task execution control fitting module configured to perform task execution control fitting on the plurality of atomic subtasks according to power distribution scene building information, and output a plurality of task execution control parameters; A collaborative control chain acquisition module configured to obtain a task collaborative control chain by associating the plurality of task execution control parameters with the task sequential execution chain; A control management module configured to drive the P multiplexing devices and the W non-multiplexing devices to perform device space collaborative control management based on the task collaborative control chain, with the P multiplexing time chains and the W task execution time windows as device parallel scheduling constraints.

Images