Variable Step-Size Inference Method and Device for Task Resource Requirements Based on Network Topology Structure

Abstract

The present application relates to a variable step-size inference method and device for task resource requirements based on a network topology structure. The method includes: establishing a resource attribute processing model based on the requirement attributes of tasks, then establishing a directed acyclic network that describes the resource requirements in a task plan according to the resource requirement attributes and the resource processing mode, and finally designing a variable step-size resource requirement inference algorithm according to the resource consumption mode to achieve the statistical calculation of the resource requirements of the task plan. Using this method can achieve variable step-size inference of resource requirements, reduce the number of resource calculation steps, and improve the calculation efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of space station on-orbit operation mission planning, and in particular to a variable-step reasoning method and device for mission resource requirements based on a network topology structure. Background Art

[0002] In the short-term operation planning of the Chinese space station, the planning objects cover various tasks within a mission cycle (about 6 months), such as spacecraft launch, major on-orbit application / experiment, space station platform maintenance and repair, astronaut health and security, extravehicular missions, and robotic arm operations. Due to the wide variety of short-term operation missions and complex requirements of the space station, the resource supply and demand relationship is much more complicated than that of general space missions. As the operation time increases, the types of on-orbit missions will gradually increase, and the corresponding resource types will also increase. Among the existing space mission planning methods, most only consider limited types of resources and cannot cope with the needs of future resource expansion. Therefore, in view of the complex characteristics of mission resources, designing a resource description model with expansion capabilities and corresponding resource reasoning methods is of great significance to solving the problem of short-term operation mission planning of the space station.

[0003] Mission implementation requires resource support. Whether the short-term operation mission plan of the space station can meet the resource constraints directly determines its rationality and feasibility. For resource constraint judgment, the current method is to first obtain a set of temporary plans that meet the task timing and logical relationships, and then make resource constraint judgments, and then iterate and correct the plans until the final plan is generated. There are two problems with this method. First, each type of resource has an independent processing method, resulting in poor resource scalability; second, there are a large number of iterative resource constraint judgments and conflict correction steps in the planning, which affect the planning efficiency, so new exploration is needed. Summary of the invention

[0004] Based on this, it is necessary to provide a variable-step reasoning method and device for task resource requirements based on network topology structure, which can realize variable-step reasoning of resource requirements, reduce the number of resource calculation steps, and improve computing efficiency, in response to the above technical problems.

[0005] A variable-step reasoning method for task resource requirements based on a network topology structure, the method comprising:

[0006] Obtaining a temporary planning task plan, wherein the temporary planning task plan satisfies time constraints and logical constraints;

[0007] Using the resource processing mode model to classify the attribute types of resources required to execute each task in the temporary planning task plan, and matching the corresponding processing plan according to the classification results;

[0008] Select all tasks related to a certain type of resource from the temporary planning task plan, extract the information of each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing plan, and the information of each processing stage in the tasks;

[0009] Sort each node according to the processing time sequence of each node in the resource requirement network, and generate a node sequence set;

[0010] Recommend and process the resource flow according to the nodes in the node sequence set. Each time it advances to a node, it checks according to the resource constraints. If the check result of the current node is satisfied, it continues to advance the resource flow according to the node sequence and checks the resource constraints for the next node;

[0011] If the resource constraint check result of the current node is not satisfied, adjust the task corresponding to the node, and after the adjustment, return to the previous node adjacent to the node to re-check the resource constraints, and continue to advance the resource flow according to the node sequence;

[0012] Until all nodes in the node sequence set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step inference of task resource requirements.

[0013] In one embodiment, the resources required to execute a task include platform resources. Then, use the platform resource processing mode model to divide the resources required to execute the task into 8 attribute types, including:

[0014] Shared continuous recoverable resources, shared discrete recoverable resources, exclusive discrete recoverable resources, exclusive continuous recoverable resources, shared continuous non-recoverable resources, shared discrete non-recoverable resources, exclusive discrete non-recoverable resources, and exclusive continuous non-recoverable resources;

[0015] Among them, shared resources can support multiple tasks simultaneously, while exclusive resources can only support one task at the same time. The processing volume value of discrete resources remains a fixed value that does not change during the entire task processing stage, and the processing volume of continuous resources is proportional to the task processing time. When recoverable resources are consumed, the consumed resources will be released at the end of the processing stage, and the total resource amount remains unchanged. When non-recoverable resources are consumed, the total resource amount gradually decreases as the resources are used, and need to be replenished through productive tasks.

[0016] In one embodiment, the directed and acyclic resource requirement network includes multiple nodes and directed edges connecting each node, and is expressed as:

[0017] Net R (V Resource ,ER , E P )

[0018] In the above formula, V Resource represents the set of nodes in the resource processing stage, and E R represents the set of directed edges in the internal resource processing stage of the task, and E P represents the set of directed edges in the resource processing stage between tasks.

[0019] In one embodiment, the nominal processing time, the earliest available processing time, and the latest available processing time representing the node are marked under each node in the resource demand network, and the task ID to which the node belongs is also marked;

[0020] When the resource is continuous, each segment of the resource processing stage can be disassembled into a stage start node, a transition node, and a stage end node, and the resource processing amount at the stage start point is marked at the stage start node, the resource processing amount at the stage transition point is marked at the transition node, and the resource processing amount at the stage end point is marked at the stage end node;

[0021] When the resource is discrete, each segment of the resource processing stage only retains the stage start node and the stage end node, and the resource processing amount at the stage start point is marked at the stage start node, and the resource processing amount at the stage end point is marked at the stage end node.

[0022] In one embodiment, the resource constraint includes a platform resource constraint, and the platform resource constraint includes a usage attribute constraint and a recovery attribute constraint;

[0023] The usage attribute constraint is expressed as:

[0024]

[0025] In the above formula, represents the number of tasks with a demand for resource k at time t, and M k is the maximum number of tasks that resource k can support at the same time, where the exclusive resource is 1 and the shared resource is ∞, K represents the set of all resources, and T represents the set of all times on the time line;

[0026] The recovery attribute constraint includes a recoverable resource constraint and a non-recoverable resource constraint. Among them, the recoverable resource constraint is expressed as:

[0027]

[0028] In the above formula, represents the consumption of the recoverable resource k by task i at time t, Denote the available resource quantity of recoverable resource k at time t, and A represents the set of all tasks;

[0029] The non-recoverable resource constraint is expressed as:

[0030]

[0031] In the above formula, Denote the consumption of non-recoverable resource k by task i at time t, Denote the quantity of resource k replenished by task i in j time periods, Denote the available resource quantity of resource k in the p-th time period.

[0032] In one embodiment, if the resource constraint check result of the current node is not satisfied, the adjustment of the task corresponding to the node includes:

[0033] Obtain the current time. If the current time is greater than or equal to the nominal processing time of the stage start node of the task to which the current node belongs and less than or equal to the nominal processing time of the stage end node, then this resource processing stage is a conflict stage, and the task to which it belongs is a resource conflict task;

[0034] According to the conflict resolution strategy, determine the task to be adjusted and its adjustment information;

[0035] According to the time constraint and logical constraint of the task, re-determine the execution time of the task to be adjusted and its subsequent related tasks, and generate an updated temporary task plan;

[0036] According to the updated temporary task plan, update the resource processing stage time information of the adjusted task and its subsequent tasks in the resource demand network, and update the node sequence set.

[0037] In one embodiment, the resources required to execute a task further include man-hour resources;

[0038] Using the man-hour resource processing mode model, divide the man-hours required to execute a task into exclusive discrete resources, and adopt a man-hour balanced scheduling strategy to arrange the man-hours for each task, and then process the man-hour resources according to the exclusive discrete non-recoverable resources.

[0039] In one embodiment, the resource constraint includes a man-hour resource constraint, and the man-hour resource constraint restricts man-hours from three aspects: the working time range of personnel, the working duration, and the number of working days.

[0040] A task resource demand variable step size inference device based on a network topology structure, the device includes:

[0041] A task plan acquisition module, which is used to acquire a temporary planning task plan that meets time constraints and logical constraints;

[0042] A task attribute type classification module, which is used to classify the attribute types of the resources required for each task in the temporary planning task plan by using a resource processing mode model, and match corresponding processing solutions according to the classification results;

[0043] A resource requirement network construction module, which is used to select all tasks related to a certain type of resource from the temporary planning task plan, extract information on each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing solution, and the information on each processing stage in the tasks;

[0044] A node order set generation module, which is used to sort each node according to the processing time order of each node in the resource requirement network and generate a node order set;

[0045] A node resource constraint check module, which is used to recommend and process the resource flow according to the nodes in the node order set, and check according to the resource constraints every time it advances to a node. If the check result of the current node is satisfied, the resource flow is continued according to the node order, and the resource constraints of the next node are checked;

[0046] A task adjustment module, which is used to adjust the task corresponding to the node if the resource constraint check result of the current node is not satisfied, and after the adjustment, return to the previous node adjacent to the node to re-check the resource constraints, and continue to advance the resource flow according to the node order;

[0047] A resource requirement variable step-size inference completion module, which is used to output the resource constraint consistency check results of each node until all nodes in the node order set satisfy the resource constraint checks, so as to complete the task resource requirement variable step-size inference.

[0048] A computer device, including a memory and a processor, where the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0049] Acquire a temporary planning task plan that meets time constraints and logical constraints;

[0050] Classify the attribute types of the resources required for each task in the temporary planning task plan by using a resource processing mode model, and match corresponding processing solutions according to the classification results;

[0051] Select all tasks related to a certain type of resource from the temporary planning task plan, extract the information of each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing plan, and the information of each processing stage in the tasks;

[0052] Sort each node according to the processing time sequence of each node in the resource requirement network, and generate a node sequence set;

[0053] Recommend and process the resource flow according to the nodes in the node sequence set. Each time a node is advanced, check according to the resource constraints. If the check result of the current node is satisfied, continue to advance the resource flow according to the node sequence and check the resource constraints for the next node;

[0054] If the resource constraint check result of the current node is not satisfied, adjust the task corresponding to the node, and after adjustment, return to the previous node adjacent to the node to re-check the resource constraints, and continue to advance the resource flow according to the node sequence;

[0055] Until all nodes in the node sequence set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step-size inference of task resource requirements.

[0056] A computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0057] Obtain a temporary planning task plan, and the temporary planning task plan satisfies time constraints and logical constraints;

[0058] Use the resource processing mode model to classify the attribute types of resources required for each task in the temporary planning task plan, and match the corresponding processing plan according to the classification results;

[0059] Select all tasks related to a certain type of resource from the temporary planning task plan, extract the information of each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing plan, and the information of each processing stage in the tasks;

[0060] Sort each node according to the processing time sequence of each node in the resource requirement network, and generate a node sequence set;

[0061] Recommend and process the resource flow according to the nodes in the node sequence set. Each time a node is advanced, check according to the resource constraints. If the check result of the current node is satisfied, continue to advance the resource flow according to the node sequence and check the resource constraints for the next node;

[0062] If the resource constraint check result of the current node is not satisfied, adjust the task corresponding to the node, and after the adjustment, go back to the previous node adjacent to the node to re-perform the resource constraint check, and continue to advance the resource flow according to the node order;

[0063] Until all nodes in the node order set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step-size inference of the task resource requirements.

[0064] The above-mentioned variable step-size inference method and device for task resource requirements based on the network topology structure establish a resource attribute processing model based on the demand attributes of task resources, and then establish a directed and acyclic network describing the resource requirements in the task plan according to the resource demand attributes and resource processing modes. Finally, according to the resource consumption mode, a variable step-size resource demand inference algorithm is designed to realize the statistical calculation of the resource requirements of the task plan. Using this method can realize the variable step-size inference of resource requirements, reduce the number of resource calculation steps, and improve the calculation efficiency. Description of the Drawings

[0065] Figure 1 It is a schematic flow chart of the variable step-size inference method for task resource requirements based on the network topology structure in an embodiment;

[0066] Figure 2 It is a schematic diagram of multi-consumption of platform resources in an embodiment;

[0067] Figure 3 It is a schematic diagram of the man-hour balanced scheduling strategy in an embodiment;

[0068] Figure 4 It is a schematic diagram of the resource usage attribute constraints in an embodiment;

[0069] Figure 5 It is a schematic diagram of the propellant resource consumption in an embodiment;

[0070] Figure 6 It is a schematic diagram of the resource requirement network in an embodiment;

[0071] Figure 7 It is a Gantt chart of task execution with resource constraints satisfied in an experimental simulation;

[0072] Figure 8 It is a schematic diagram of the consumption curves of various resources in an experimental simulation;

[0073] Figure 9 It is a structural block diagram of the variable step-size inference device for task resource requirements based on the network topology structure in an embodiment;

[0074] Figure 10It is the internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0075] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0076] As Figure 1 shown, a variable step-size inference method for task resource requirements based on a network topology structure is provided, including the following steps:

[0077] Step S100, obtain a temporary planning task plan, and the temporary planning task plan meets time constraints and logical constraints;

[0078] Step S110, classify the attribute types of the resources required for each task in the temporary planning task plan by using a resource processing mode model, and match the corresponding processing plan according to the classification result;

[0079] Step S120, select all tasks related to a certain type of resource from the temporary planning task plan, extract the information of each processing stage in the task, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing plan and the information of each processing stage in the task;

[0080] Step S130, sort each node according to the processing time sequence of the nodes in the resource requirement network, and generate a node sequence set;

[0081] Step S140, recommend and process the resource flow according to the nodes in the node sequence set, and check according to the resource constraints every time a node is advanced. If the check result of the current node is satisfied, continue to advance the resource flow according to the node sequence and check the resource constraints for the next node;

[0082] Step S150, if the resource constraint check result of the current node is not satisfied, adjust the task corresponding to the node, and after the adjustment, return to the previous node adjacent to the node to re-check the resource constraints and continue to advance the resource flow according to the node sequence;

[0083] Step S160, until all nodes in the node sequence set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step-size inference of task resource requirements.

[0084] In this embodiment, a resource processing mode model is established by analyzing the attributes of task resource requirements. The task is classified according to the attributes by using the resource processing mode model, and the corresponding processing mode of the attributes is matched. In fact, the resource information of each task is converted into a processable mode.

[0085] In this embodiment, the temporary planning task plan that requires resource requirement step inference has satisfied the time constraint and the logical constraint. That is to say, the planning task plan can be further adjusted by the method proposed in this article to satisfy the resource constraint on the basis of satisfying the time constraint and the logical constraint, making the plan more reasonable in the process.

[0086] In this embodiment, if the resources required to execute the task include platform resources, the resources required to execute the task are divided into 8 attribute types by using the platform resource processing mode model, including: shared continuous recoverable resources, shared discrete recoverable resources, exclusive discrete recoverable resources, exclusive continuous recoverable resources, shared continuous non-recoverable resources, shared discrete non-recoverable resources, exclusive discrete non-recoverable resources, and exclusive continuous non-recoverable resources.

[0087] Among them, shared resources can support multiple tasks at the same time, while exclusive resources can only support one task at the same moment. The processing amount of discrete resources remains a fixed value that does not change during the entire task processing stage, and the processing amount of continuous resources is proportional to the task processing time. When recoverable resources are consumed, the consumed resources will be released at the end of the processing stage, and the total amount of resources remains unchanged. When non-recoverable resources are consumed, the total amount of resources gradually decreases with the use of resources and needs to be replenished through productive tasks.

[0088] Specifically, it can be known from the resource requirement attributes that the processing of platform resources for short-term operation tasks of the space station adopts a multi-category and multi-stage method, that is, the same task can process multiple types of platform resources during execution, and the processing of each type of resource can be divided into multiple stages. As Figure 2 shown, the task consumes electric power in 3 stages, and the power of each stage is different.

[0089] The processing mode of each processing stage of platform resources during task execution depends on the resource attributes. Each type of platform resource can be divided from 3 attribute dimensions: usage attribute, numerical attribute, and recovery attribute. Table 1 shows the attribute classification of some platform resources in the short-term operation task planning of the space station.

[0090] Table 1 Platform Resource Attribute Classification in Short-Term Operation of the Space Station

[0091]

[0092] As can be seen from Table 1, although the resources belong to different categories, there are resources with the same attributes, such as load suspension points, scientific experiment cabinets, and extravehicular experimental devices, etc. They all belong to exclusive, discrete, and recoverable resources. For resources with the same attributes, the same resource processing method can be adopted. Therefore, in actual planning, the resource processing mode can be determined and processed only according to the resource attributes, without specifically identifying the resource categories.

[0093] There are a total of 8 attribute combination methods in the 3 attribute dimensions of usage attribute, numerical attribute, and recovery attribute, which are shared continuous recoverable resources, shared discrete recoverable resources, exclusive discrete recoverable resources, exclusive continuous recoverable resources, shared continuous non-recoverable resources, shared discrete non-recoverable resources, exclusive discrete non-recoverable resources, and exclusive continuous non-recoverable resources. The characteristics of the processing mode of each combined attribute are shown in Table 2, where: shared resources can support multiple tasks simultaneously; exclusive resources can only support one task at the same time; discrete resources have a fixed value of the processing quantity during the entire processing stage and remain unchanged; the processing quantity value of continuous resources is proportional to the processing time; when recoverable resources are consumed, the consumed resources will be released at the end of the processing stage, and the total amount of resources remains unchanged; when non-recoverable resources are consumed, the total amount of resources gradually decreases with the use of resources and needs to be supplemented through productive tasks. For example, the on-orbit data download task will release the on-orbit data storage space.

[0094] Table 2 Processing Mode of Resources with Combined Attributes

[0095]

[0096] Each type of platform resource will correspond to a group of the above 8 combined attributes according to its own physical properties as its processing mode, and is equipped with a unique ID as the only identifier of the resource. During the planning process, the algorithm will automatically identify the resource processing mode and consume or replenish according to the resource requirement attributes of the task.

[0097] In addition, when processing nested mode resources similar to electrical energy resources (the usage power needs to be set), it is generally necessary to convert them into the form of "single resource + time".

[0098] In this embodiment, the resources required to execute the task also include man-hour resources. The man-hour resources required to execute the task are divided into exclusive discrete resources by using the man-hour resource processing mode model. After arranging the man-hours of each task by using the man-hour equalization scheduling strategy, the man-hour resources are processed according to the exclusive discrete non-recoverable resources.

[0099] Specifically, man-hour resources can be classified as exclusive and discrete resources from the perspective of attribute dimension. Except for the astronaut rotation mission, tasks generally consume man-hour resources in multiple stages. Therefore, only the consumption mode of man-hour resources will be discussed below. There are three differences between man-hour resources and platform resources: (1) The demand for man-hour resources in tasks adopts the "skill + number of people" mode; (2) Astronaut work and rest time needs to be considered in the constraints of man-hour resources; (3) Man-hour resources recover on a daily basis and do not recover within a day. Therefore, man-hour resources have their own independent consumption mode.

[0100] The consumption of man-hour resources can be divided into two steps. One is to determine the staff, and the other is to allocate the working hours. Since in the short-term operation task planning of the space station:

[0101] When a task has no demand for astronaut skills, astronauts are not distinguished. Therefore, when allocating astronaut tasks, a man-hour balanced scheduling strategy is adopted, that is, when a task has a man-hour demand, the idle astronauts are sorted in ascending order of the total working hours, and according to the required number of people, the astronauts with less working hours are preferentially selected to complete the corresponding tasks. The strategy process is as Figure 3 shown.

[0102] After determining the working astronauts, the man-hour can be consumed as an exclusive, discrete, and non-recoverable resource, that is, the occupied astronauts cannot complete other tasks at the same time. The working hours are the single-person working hours required for the task. After the task is completed, the working hours of the astronauts on the same day are reduced until the working hours are restored on the second day.

[0103] In this embodiment, after re-describing the platform resources and man-hour resources using the resource processing mode model for the tasks, when subsequently performing variable-step deduction of the resource requirements of each task using the resource demand network, it is also necessary to construct the resource constraints of the platform resources and man-hour resources respectively.

[0104] Specifically, the platform resource constraints stem from the limited energy and equipment on the space station. A reasonable short-term operation task plan for the space station needs to perform resource constraint judgment to meet the resource requirements of the tasks. The following will analyze the resource constraints that need to be considered in the short-term operation task planning of the space station.

[0105] In this embodiment, the platform resource constraints include usage attribute constraints and recovery attribute constraints.

[0106] Among them, the usage attributes are divided into exclusive and shared, which mainly describe the number of tasks that the resource can support at the same time. Therefore, the usage attribute constraint can be expressed as:

[0107]

[0108] In formula (1), represents the number of tasks with a demand for resource k at time t, Mk is the maximum number of tasks that resource k can support at the same time, where the exclusive resource is 1 and the shared resource is ∞. K represents the set of all resources, and T represents the set of all moments on the time line. As Figure 4 shown, the consumption process of the robotic arm resources is given. Since the robotic arm is an exclusive resource, it can support at most one task execution at any moment.

[0109] Among them, the recovery property constraints include the constraints for recoverable resources and non-recoverable resources. The main difference lies in whether the resources will be released after being consumed, resulting in consumption accumulation.

[0110] Among them, for recoverable resources, since the resources will be released when they are used up, it is only necessary to meet the constraint of the instant available amount of resources, that is:

[0111]

[0112] In formula (2), represents the consumption amount of recoverable resource k by task i at time t, represents the available resource amount of recoverable resource k at time t, and A represents the set of all tasks.

[0113] For non-recoverable resources, since the consumption amount will gradually accumulate, it is necessary to meet the cumulative usage constraint. In the short-term operation task planning of the space station, the resources adopt a multi-stage supply method, that is, the supply amount of each type of resource may be slightly different in different time periods, rather than remaining unchanged. Therefore, it is necessary to ensure that the cumulative value of the consumption amount of non-recoverable resources in each time period does not exceed the total amount limit in the current time period:

[0114]

[0115] In formula (3), represents the consumption amount of non-recoverable resource k by task i at time t, represents the quantity of resource k replenished by task i in j time periods, represents the available resource amount of resource k in the p-th time period.

[0116] As Figure 5 shown, the consumption process of the propellant is given. The resource upper limit is 1000 kg. When task 5 is executed, due to the over-limit of resource consumption, a constraint conflict occurs.

[0117] Specifically, regarding the human-hour resources constraint, as the main body of mission execution, the working human hours of astronauts are a relatively special and extremely important type of resource. When planning and scheduling the human hours of astronauts, multiple factors need to be comprehensively considered. Currently, during the short-term operation of the space station, the human-hour resources of astronauts are mainly constrained from three aspects: the working time range, working duration, and working days of astronauts.

[0118] Among them, to ensure the physical and mental health of astronauts, the work and rest time of astronauts during their stay in space need to be consistent with that on the ground. Therefore, two time points are defined: the starting time point of the day TP Day and the starting time point of the night TP Night , [TP Day , TP Night is the daily working time interval of astronauts. Except in special cases where overtime is required, tasks should be arranged to be carried out during working hours as much as possible. Then the working time range constraint is:

[0119] T AstWork ∈[TP Day , TP Night (4)

[0120] Among them, the daily working duration of astronauts should not exceed the upper limit of their working duration, unless they need to carry out emergency tasks or other tasks with higher priorities. Taking the International Space Station as an example, the daily working duration of astronauts is 6 hours, and the rest of the time is used for eating, exercising, and dealing with personal affairs, etc. Then the working duration constraint of astronauts is expressed as:

[0121]

[0122] In formula (5), H AstWork represents the daily working duration of astronauts, represents the maximum daily working duration of astronauts.

[0123] Among them, the working days of astronauts per week are generally 6 days. Except in special cases, rest time is generally not occupied. Then the working days constraint of astronauts is expressed as:

[0124]

[0125] In formula (6), D AstWork represents the working days of astronauts per week, represents the maximum working days of astronauts per week.

[0126] After converting the resource information of each task in the temporary planning task plan using the resource demand pattern model, in step S120, all tasks related to a certain type of resource are selected from the temporary planning task plan, and the information of each processing stage in the tasks is extracted. A corresponding directed and acyclic resource demand network is constructed according to the attribute type of the resource, the corresponding processing scheme, and the information of each processing stage in the tasks.

[0127] In this embodiment, a corresponding resource demand network can be constructed for each type of resource according to the temporary planning task plan.

[0128] Specifically, the resource demand network is a directed and acyclic network that describes the resource requirements in the short-term operation task plan of the space station. It is mainly used for checking the consistency of resource constraints in the short-term operation task plan of the space station, assisting in the continuous iterative correction of the plan to obtain the final task plan, as Figure 6 shown in the network schematic diagram. During the planning process, each type of resource has a resource demand network to describe its evolving set over time during the execution of the short-term operation task plan of the space station, such as the power consumption resource demand network, the heat dissipation resource demand network, etc.

[0129] Since the short-term operation tasks of the space station can perform multi-stage processing on the same resource, in order to directly display the resource processing process, the tasks are disassembled into independent resource processing stages in the resource demand network. The time information and resource processing information of each stage are intuitively displayed in the network, and the resource attributes can also be reflected in the network through the configuration of nodes and related information.

[0130] In this embodiment, the nodes and directed edges in the resource demand network are represented by a triple as:

[0131] Net R (V Resource ,E R ,E P )(7)

[0132] In formula (7), V Resource represents the node set of the resource processing stage, E R represents the set of directed edges of the resource processing stage within the task, and E P represents the set of directed edges of the resource processing stage between tasks.

[0133] Specifically, each node in the resource demand network is marked with the nominal processing time, the earliest available processing time, and the latest available processing time of the node, and is also marked with the task ID to which the node belongs;

[0134] When the resource is continuous, each resource processing stage can be disassembled into a stage start node, a transition node, and a stage end node. InFigure 6 Among them, they are respectively represented by and . Under the three types of nodes, a parameter set is correspondingly marked as: and respectively represent the nominal processing time, the earliest available processing time, and the latest available processing time of each type of node.

[0135] In the resource network, the processing of the resource flow only occurs at the nodes. Therefore, at each node, the starting node of the task phase to which this node belongs, the starting point of the marking phase, and the resource processing volume are marked At the transition node, the transition point of the phase and the resource processing volume are marked And at the terminating node of the phase, the terminating point of the phase and the resource processing volume are marked

[0136] Furthermore, the resource processing volume of continuous resources changes with time during the entire resource processing phase. Therefore, to accurately describe the resource processing process, between and a phase transition point is added and can be obtained through, while needs to be calculated through the planning step size t step as follows:

[0137]

[0138] In formula (8), m represents the serial number of the phase transition. Through the above processing, a continuous resource processing process is approximated as multiple discrete processing nodes, and the number of phase transition points is:

[0139]

[0140] Since for continuous resources, is the processing rate per unit time of this phase. Therefore, when the phase processing mode is the resource consumption mode:

[0141]

[0142]

[0143] When the resource recovery attribute is recoverability:

[0144]

[0145] If it is a non-recoverable resource, then is zero.

[0146] When the resources are discrete, only the starting node and the ending node of each resource processing stage are retained, and the resource processing volume at the starting point of the stage is marked at the starting node of the stage, and the resource processing volume at the ending point of the stage is marked at the ending node of the stage.

[0147] Furthermore, since the resource processing volume of discrete resources is fixed throughout the resource processing stage, only the starting point of the stage is retained in the network and the ending point of the stage and can be obtained respectively through the following formulas:

[0148]

[0149]

[0150] In formulas (13) and (14), represents the starting point of the task to which this processing stage belongs, represents the time interval between the starting point of the task and the start time of the resource processing period, represents the time interval between the starting point of the task and the end time of the resource processing period, represents the earliest starting point of the task, represents the latest starting point of the task. It should be noted here that resource processing is only carried out during the execution of the task, so t ns and t ne need to ensure that they are within the task execution interval, that is:

[0151]

[0152] In formula (15), represents the end point of the task.

[0153] The resource processing volume at the starting point of the stage needs to be determined according to the processing mode of this stage. If this stage is a consumption mode, then:

[0154]

[0155] In formula (16), represents the resource processing volume of this stage. For discrete resources, it is the total processing volume. If it is a production mode, then:

[0156]

[0157] And the resource processing volume at the ending point of the stage depends on the recovery attribute of the resource. If it is a recoverable resource, the resource is released at the end of the resource processing:

[0158]

[0159] If it is a non-recoverable resource, then it is 0.

[0160] In the triple, E R is the set of directed edges inside the resource processing stage. In Figure 6 it is represented by and is directed from the node with an earlier processing time to the node with a later processing time. The capacity between nodes of E R is +∞, and the flow is not restricted during reasoning. E P represents the set of directed edges between resource processing stages, which is represented by in the figure and is used to map the timing relationship between tasks.

[0161] The resource demand network is an ordered mapping of resource demands in the short-term operation task plan of the space station. The premise for constructing the resource demand network is that the time constraints and logical constraints between tasks are satisfied, that is, the time and logical constraint reasoning processing of the short-term operation task plan of the space station has been completed. The horizontal axis in the resource demand network is the time axis, and the nodes are strictly arranged in the order of processing time horizontally.

[0162] Next, based on the resource demand network constructed corresponding to each resource, reasoning for demand recommendation is performed, and its reasoning process is divided into platform resource demand reasoning and man-hour resource demand reasoning.

[0163] Among them, for the demand reasoning of platform resources, only the resource flow recommendation needs to be performed in the order of node time, and consistency checking is performed according to the platform resource constraints. This process is described in steps S130 to S160.

[0164] In this embodiment, when performing resource demand reasoning based on the resource demand network, if the resource constraint check result of the current node is not satisfied, the adjustment of the task corresponding to this node includes: obtaining the current time. If the current time is greater than or equal to the nominal processing time of the stage start node of the task to which the current node belongs and less than or equal to the nominal processing time of the stage end node, then this resource processing stage is a conflict stage, and the task to which it belongs is a resource conflict task. According to the conflict resolution strategy, determine the task to be adjusted and its adjustment information. According to the time constraints and logical constraints of the task, re-determine the execution times of the task to be adjusted and its subsequent related tasks, generate an updated temporary task plan, and based on the updated temporary task plan, update the time information of the resource processing stages of the adjusted task and its subsequent tasks in the resource demand network, and update the node sequence set.

[0165] In this embodiment, the specific steps of the platform resource reasoning algorithm are as shown in Algorithm 1:

[0166]

[0167]

[0168] For the above algorithm, the resource flow is only processed at nodes and remains unchanged between nodes, enabling the resource inference to advance with variable step sizes, reducing the number of resource calculation steps and improving the calculation efficiency. There will be multiple resource demand networks in the planning. Since each network can be inferred independently and without mutual influence, parallel inference can be performed.

[0169] The processing time of each node in Step 2 is not necessarily t ns 、t nm or t ne , and any time point in or can also be selected as the processing time.

[0170] The conflict resolution strategy in Step 5-(2) is formulated by comprehensively considering factors such as the priority, importance, and time redundancy of conflicting tasks. In this embodiment, the existing strategy can be adopted for the conflict resolution strategy.

[0171] In the resource demand network, according to the relationship between the task start point end point and the current time t, the resource processing nodes can be divided into three categories: nodes belonging to tasks that have been completed, that is, nodes belonging to tasks in the execution process, that is, nodes belonging to tasks that have not been executed, that is,

[0172] In the resource demand network, task adjustment affects the nodes of tasks in the execution process and tasks that have not been executed, but has no impact on the nodes of tasks that have been completed. Therefore, in each calculation process of resource flow advancement after task adjustment, only the time needs to be rolled back to the processing time of the nearest non-affected node in Step 5-(5), that is, the last stage node of the previous completed task.

[0173] In this embodiment, similarly, the man-hour requirements of tasks are also divided by stage. Therefore, the man-hour resources can also construct a resource demand network, and the specific steps of the man-hour resource demand inference algorithm are as shown in Algorithm 2:

[0174]

[0175]

[0176] For the above algorithm, in Step 3, since the man-hours are recovered periodically on a daily basis, it is necessary to divide the resource processing nodes on a daily basis and perform man-hour reasoning on a daily basis. In Step 4, since there is no distinction between astronauts during the short-term operation of the space station, when performing man-hour resource reasoning, compared with only one resource flow for the first-class platform resources, the man-hour resource flow will be differentiated according to the number of astronauts on the station. Each differentiated resource flow represents the working man-hours of an astronaut. When performing specific resource node processing, astronauts are selected according to the man-hour balanced scheduling strategy and the working man-hours are arranged.

[0177] When the execution time of a task is adjusted due to constraint conflicts, the resource processing nodes will be adjusted accordingly, and the days they are in may change. Therefore, when the task is adjusted, it is necessary to repeat Step 3 to update the days where the nodes are located.

[0178] To verify the effectiveness of the variable step-size reasoning method for task resource requirements in this paper, experimental simulations were also carried out according to this method. Four types of constraints were set in the planning scenario. The attributes of each type of resource and the upper limit of resource consumption in each stage of the planning are shown in Table 3, where the start time and end time of each stage are the time intervals relative to the start time of the scenario.

[0179] Among the 177 tasks in the planning scenario, 28 tasks with close execution intervals were selected and relevant resource processing information was configured, as shown in Table 4, where the start time and end time of the resource processing stage are the time intervals relative to the start time of the task. Some tasks are set for multi-stage and multi-type resource processing.

[0180] Among the 28 selected short-term operation tasks of the space station, Tasks 3, 9, 10, 13, 78, and 122 are repetitive tasks. After being disassembled, they become 35 tasks. The resource processing information of the disassembled tasks is the same as that of the original tasks. The specific information of each disassembled task is shown in Table 5.

[0181] Table 3 Resource Information in the Short-Term Operation Task Planning of the Space Station

[0182]

[0183]

[0184] Table 4 Task Resource Processing Information in the Short-Term Operation Task Planning of the Space Station (Time Unit: Hours)

[0185]

[0186] Table 5 Disassembly Information of Repetitive Tasks

[0187]

[0188] According to the method proposed in this paper, the material demand network corresponding to each material is constructed. After reasoning about the nodes in the material demand network according to resource constraints, the following results are obtained:

[0189] The designed resource demand reasoning algorithm is used to process 35 tasks, of which 33 tasks meet the resource constraint consistency after time adjustment. Table 6 shows the comparison of task execution time before and after reasoning. The execution time in the table is the interval time with the start time of the planning scenario. Before reasoning, each task uses the nominal time as the execution time. Figure 7 The Gantt chart of the solution execution is given, and the blue part is the task adjustment time after resource constraint reasoning. By comparing the data in the chart, it can be seen that for most tasks without resource constraint conflicts, the execution time is the same as the execution time obtained after time logic constraint reasoning.

[0190] Table 6 Task execution time information after resource requirement reasoning (time unit: hour)

[0191]

[0192]

[0193] Through comprehensive analysis of Table 4 and Figure 8 Based on the information in , we can get the adjustment reasons of 5 adjusted tasks and the cutting reasons of 2 cut tasks, as shown in Table 7 and Table 8. By comparing the above two tables, we can see that the adjusted tasks basically have a wider adjustable range, and the resources that cause conflicts are generally recoverable resources. It is precisely because there is no consumption accumulation in the consumption process of recoverable resources that when tasks have resource constraint conflicts, they can meet resource constraint consistency by adjusting the timeline to stagger the immediate resource consumption peak. For non-recoverable resources, unless they are cut or moved to productive tasks for execution, the consumption peak cannot be eliminated no matter how they are adjusted.

[0194] Table 7 Reasons for adjusting the adjusted tasks

[0195]

[0196] Table 8 Task pruning reasons

[0197]

[0198] like Figure 8 As shown in the figure, the consumption curves of various resources are given, from which we can clearly see the changes in resource consumption before and after resource demand reasoning. It can be seen that before reasoning, there were time periods when power, robotic arms, and propellants violated resource constraints, which were completely eliminated after reasoning. Figure 8 (c) andFigure 8 (d) It can be seen that there are 1 - 2 times of resource replenishment for the propellant and on - orbit storage space, so that the resource constraints of subsequent tasks can be met.

[0199] It can be seen from the experimental simulation that the resource requirement reasoning method based on the resource requirement network can effectively solve various resource conflict situations in the solution and meet the constraint consistency requirements of the solution.

[0200] The above - mentioned variable - step - size reasoning method for task resource requirements based on the network topology structure establishes a resource attribute processing model through the demand attributes of task resources, then builds a directed acyclic network describing the resource requirements in the task solution according to the resource demand attributes and resource processing modes, and finally designs a variable - step - size resource requirement reasoning algorithm according to the resource consumption mode to realize the resource requirement statistics of the task solution. Using this method can realize variable - step - size reasoning of resource requirements, reduce the number of resource calculation steps, and improve the calculation efficiency.

[0201] It should be understood that although Figure 1 each step in the flowchart of Figure 1 is shown in sequence according to the indication of the arrow, these steps do not necessarily execute in the order indicated by the arrow. Unless there is a clear description in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover,

[0202] In one embodiment, as Figure 9 shown, a variable - step - size reasoning device for task resource requirements based on the network topology structure is provided, including: a task solution acquisition module 200, a task attribute type classification module 210, a resource requirement network construction module 220, a node order set generation module 230, a node resource constraint check module 240, a task adjustment module 250, and a resource requirement variable - step - size reasoning completion module 260, where:

[0203] The task solution acquisition module 200 is used to acquire a temporarily planned task solution, and the temporarily planned task solution meets the time constraint and logical constraint;

[0204] The task attribute type classification module 210 is used to classify the attribute types of resources required for each task in the temporarily planned task solution by using a resource processing mode model, and match the corresponding processing solutions according to the classification results;

[0205] A resource requirement network construction module 220, configured to select all tasks related to a certain type of resource from the temporary planning task solution, extract information on each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing solution, and the information on each processing stage in the tasks;

[0206] A node sequence set generation module 230, configured to sort each node according to the processing time sequence of each node in the resource requirement network, and generate a node sequence set;

[0207] A node resource constraint check module 240, configured to recommend and process a resource flow according to the nodes in the node sequence set, and perform a check according to resource constraints every time a node is advanced. If the check result of the current node is satisfied, the resource flow is continued according to the node sequence, and a resource constraint check is performed on the next node;

[0208] A task adjustment module 250, configured to, if the resource constraint check result of the current node is not satisfied, adjust the task corresponding to the node, and after the adjustment, return to the previous node adjacent to the node to re-perform the resource constraint check, and continue to advance the resource flow according to the node sequence;

[0209] A resource requirement variable step-size inference completion module 260, configured to output the resource constraint consistency check results of each node until all nodes in the node sequence set satisfy the resource constraint check, so as to complete the task resource requirement variable step-size inference.

[0210] For the specific limitations of the task resource requirement variable step-size inference device based on the network topology structure, reference can be made to the limitations of the task resource requirement variable step-size inference method based on the network topology structure in the above text, which will not be elaborated here. Each module in the above task resource requirement variable step-size inference device based on the network topology structure can be implemented in whole or in part through software, hardware, and their combination. The above modules can be embedded in the processor in the computer device in hardware form or independent of it, or stored in the memory in the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.

[0211] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as Figure 10As shown in the figure. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a variable step-size inference method for task resource requirements based on a network topology. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0212] Those skilled in the art can understand that Figure 10 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0213] In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, the following steps are implemented:

[0214] Obtain a temporary planning task solution, and the temporary planning task solution satisfies time constraints and logical constraints;

[0215] Use a resource processing mode model to classify the attribute types of resources required for each task in the temporary planning task solution, and match the corresponding processing solutions according to the classification results;

[0216] Select all tasks related to a certain type of resource from the temporary planning task solution, extract the information of each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing solution, and the information of each processing stage in the tasks;

[0217] Sort each node according to the processing time sequence of each node in the resource requirement network, and generate a node sequence set;

[0218] Recommend and process the resource flow according to the nodes in the node sequence set. Each time a node is advanced, check according to the resource constraints. If the check result of the current node is satisfied, continue to advance the resource flow according to the node sequence and check the resource constraints for the next node;

[0219] If the resource constraint check result of the current node is not satisfied, adjust the task corresponding to this node, and after the adjustment, return to the previous node adjacent to this node to re-perform the resource constraint check, and continue to advance the resource flow in the order of the nodes;

[0220] Until all the nodes in the node sequence set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step-size reasoning of the task resource requirements.

[0221] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

[0222] Obtain a temporary planning task solution, where the temporary planning task solution satisfies time constraints and logical constraints;

[0223] Use a resource processing mode model to classify the attribute types of the resources required for each task in the temporary planning task solution, and match the corresponding processing solutions according to the classification results;

[0224] Select all the tasks related to a certain type of resource from the temporary planning task solution, extract the information of each processing stage in the tasks, and construct a corresponding directed and acyclic resource requirement network according to the attribute type of the resource, the corresponding processing solution, and the information of each processing stage in the tasks;

[0225] Sort each node according to the processing time order of the nodes in the resource requirement network, and generate a node sequence set;

[0226] Recommend and process the resource flow according to the nodes in the node sequence set. Each time a node is advanced, check according to the resource constraints. If the check result of the current node is satisfied, continue to advance the resource flow in the order of the nodes and check the resource constraints of the next node;

[0227] If the resource constraint check result of the current node is not satisfied, adjust the task corresponding to this node, and after the adjustment, return to the previous node adjacent to this node to re-perform the resource constraint check, and continue to advance the resource flow in the order of the nodes;

[0228] Until all the nodes in the node sequence set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step-size reasoning of the task resource requirements.

[0229] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above various methods. Among them, any reference to a memory, storage, database, or other medium used in the various embodiments provided in this application can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0230] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0231] The above-described embodiments merely represent several implementation manners of this application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of this application, several modifications and improvements can still be made, and these all belong to the protection scope of this application. Therefore, the protection scope of the patent of this application should be subject to the appended claims.

Claims

1. A variable step-size inference method for task resource requirements based on network topology, characterized in that The method includes: Obtaining a temporary planning task plan that meets time constraints and logical constraints; Using a resource processing mode model to classify the attribute types of resources required for each task in the temporary planning task plan, and matching corresponding processing solutions according to the classification results; Selecting all tasks related to a certain type of resource from the temporary planning task plan, extracting information on each processing stage in the tasks, and constructing a corresponding directed and acyclic resource requirement network based on the attribute type of the resource, the corresponding processing solution, and the information on each processing stage in the tasks; Sorting each node according to the processing time sequence of the nodes in the resource requirement network, and generating a node sequence set; Recommending and processing the resource flow according to the nodes in the node sequence set, and checking according to resource constraints every time a node is advanced. If the check result of the current node is satisfied, the resource flow is continued according to the node sequence, and the next node is checked for resource constraints; If the resource constraint check result of the current node is not satisfied, adjust the task corresponding to the node, and after adjustment, return to the previous node adjacent to the node to re-check the resource constraints, and continue to advance the resource flow according to the node sequence; Until all nodes in the node sequence set satisfy the resource constraint check, output the resource constraint consistency check results of each node to complete the variable step-size inference of task resource requirements.

2. The variable step-size inference method for task resource requirements based on network topology structure according to claim 1, wherein If the resources required for task execution include platform resources, use the platform resource processing mode model to classify the resources required for task execution into 8 attribute types, including: Shared continuous recoverable resources, shared discrete recoverable resources, exclusive discrete recoverable resources, exclusive continuous recoverable resources, shared continuous non-recoverable resources, shared discrete non-recoverable resources, exclusive discrete non-recoverable resources, and exclusive continuous non-recoverable resources; Among them, shared resources can support multiple tasks simultaneously, while exclusive resources can only support one task at the same time. The processing volume of discrete resources remains a fixed value that does not change during the entire task processing stage, and the processing volume of continuous resources is proportional to the task processing time. When recoverable resources are consumed, the consumed resources will be released at the end of the processing stage, and the total amount of resources remains unchanged. When non-recoverable resources are consumed, the total amount of resources gradually decreases with the use of resources and needs to be replenished through productive tasks.

3. The variable step-size inference method for task resource requirements based on network topology structure according to claim 2, wherein The directed and acyclic resource requirement network includes multiple nodes and directed edges connecting each node, and is represented as: Net R (V Resource ,E R ,E P ) In the above formula, V Resource represents the set of nodes in the resource processing stage, and E R represents the set of directed edges in the internal resource processing stage of the task, and E P represents the set of directed edges in the resource processing stage between tasks.

4. The step-size variable inference method for task resource requirements based on network topology structure according to claim 3, wherein Each node in the resource requirement network is marked with the nominal processing time, the earliest available processing time, and the latest available processing time of the node, and is also marked with the task ID to which the node belongs; When the resource is continuous, each resource processing stage can be disassembled into a stage start node, a transition node, and a stage end node, and the resource processing volume at the stage start point is marked at the stage start node, the resource processing volume at the stage transition point is marked at the transition node, and the resource processing volume at the stage end point is marked at the stage end node; When the resources are discrete, only the starting node and the ending node of each resource processing stage are retained, and the resource processing amount at the starting point of the stage is marked at the starting node of the stage, and the resource processing amount at the ending point of the stage is marked at the ending node of the stage.

5. The step-size variable inference method for task resource requirements based on network topology structure according to claim 4, wherein The resource constraints include platform resource constraints, and the platform resource constraints include usage attribute constraints and recovery attribute constraints; The usage attribute constraint is expressed as: In the above formula, represents the number of tasks that have a demand for resource k at time t, and M k is the maximum number of tasks that resource k can support at the same time, where the exclusive resource is 1 and the shared resource is ∞. K represents the set of all resources, and T represents the set of all time points on the time line; The recovery attribute constraint includes the recoverable resource constraint and the non-recoverable resource constraint. Among them, the recoverable resource constraint is expressed as: In the above formula, represents the consumption of recoverable resource k by task i at time t, represents the available resource amount of recoverable resource k at time t, and A represents the set of all tasks; The non-recoverable resource constraint is expressed as: In the above formula, represents the consumption of non-recoverable resource k by task i at time t, represents the quantity of resource k replenished by task i in j time periods, represents the available quantity of resource k in the p-th time period.

6. The variable step-size inference method for task resource requirements based on network topology structure according to claim 5, wherein If the resource constraint check result of the current node is not satisfied, the adjustment of the task corresponding to this node includes: Obtain the current time. If the current time is greater than or equal to the nominal processing time of the starting node of the stage to which the current node belongs and less than or equal to the nominal processing time of the ending node of the stage, then this resource processing stage is a conflict stage, and the task to which it belongs is a resource conflict task; According to the conflict resolution strategy, determine the task to be adjusted and its adjustment information; According to the time constraints and logical constraints of the task, re-determine the execution times of the task to be adjusted and its subsequent related tasks, and generate an updated temporary task plan; According to the updated temporary task plan, update the time information of the resource processing stages of the adjusted task and its subsequent tasks in the resource demand network, and update the node sequence set.

7. The variable step-size inference method for task resource requirements based on network topology structure according to any one of claims 1-6, characterized in that The resources required to execute a task also include man-hour resources; Using the man-hour resource processing mode model, the man-hours required to execute a task are divided into exclusive discrete resources, and after the man-hours of each task are arranged using the man-hour balanced scheduling strategy, the man-hour resources are processed according to the exclusive discrete non-recoverable resources.

8. The variable step-size inference method for task resource requirements based on network topology structure according to claim 7, wherein The resource constraints include man-hour resource constraints, and the man-hour resource constraints constrain man-hours from three aspects: the working time range of personnel, the working duration, and the number of working days.

9. A task resource requirement variable step-size inference device based on a network topology structure, characterized in that, The device includes: A task plan acquisition module, configured to acquire a temporary planned task plan that satisfies time constraints and logical constraints; A task attribute type classification module, configured to classify the attribute types of the resources required to execute each task in the temporary planned task plan using a resource processing mode model, and match corresponding processing solutions according to the classification results; A resource demand network construction module, configured to select all tasks related to a certain type of resource from the temporary planned task plan, extract the information of each processing stage in the tasks, and construct a corresponding directed and acyclic resource demand network according to the attribute type of the resource, the corresponding processing solution, and the information of each processing stage in the tasks; A node sequence set generation module, configured to sort each node according to the processing time sequence of each node in the resource demand network, and generate a node sequence set; A node resource constraint check module, configured to recommend and process the resource flow according to the nodes in the node sequence set, and perform a check according to the resource constraints every time a node is advanced. If the check result of the current node is satisfied, the resource flow is continued according to the node sequence, and the resource constraints of the next node are checked; The task adjustment module is used to adjust the task corresponding to the node if the resource constraint check result of the current node is not satisfied, and after the adjustment, return to the previous node adjacent to the node to re-perform the resource constraint check, and continue to advance the resource flow in the order of the nodes; The resource requirement variable step inference completion module is used to output the resource constraint consistency check results of each node until all the nodes in the node sequence set satisfy the resource constraint check, so as to complete the task resource requirement variable step inference.

Images