Satellite autonomous task planning method

Abstract

The invention relates to the technical field of satellite autonomous task planning, in particular to a satellite autonomous task planning method, which comprises the following steps of: S1, adopting a dual-core processor on a satellite, the processor 1 is used for executing satellite service and attitude and orbit control tasks, satellite service autonomous planning tasks and instruction sequences expanded by the autonomous planning tasks, and also can undertake external temporary task planning, and the processor 2 is mainly used for performing regional target planning and anchoring target library task planning; s2, periodically scanning a point target in the local anchoring task, and preliminarily calculating an executable time window and resource dependence by using a rapid planning algorithm; the satellite autonomous task planning method provided by the invention has the characteristics of low CPU (Central Processing Unit) resource consumption, meticulous task priority granularity, short large-scale task processing time, high task forecasting precision, high emergent task responsivity and the like, and is suitable for the development trend that the number of satellites is greatly increased and the task complexity and uncertainty are increased at present.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite autonomous mission planning, and in particular to a satellite autonomous mission planning method. Background Art

[0002] With the booming aerospace market and technology, the number of in-orbit remote sensing constellation satellites continues to increase. More and more imaging tasks, such as power facility inspections, post-disaster assessments, change detection, situational awareness, and other scenarios require the use of constellation-level satellite resources for collaborative implementation, which will increase the complexity of constellation-level mission planning exponentially. The relevant technical difficulties are reflected in the following aspects:

[0003] 1. The contradiction between the huge amount of computation required for constellation mission planning and the timeliness of mission planning

[0004] Assuming that the current constellation has M satellites and N areas to be imaged, in order to fully utilize satellite resources, it is necessary to perform computational evaluation on the accessibility of each satellite for each mission, and the total number of computational evaluations is M×N times. As the size of the constellation increases and the number of target areas increases, the amount of computation will increase exponentially. For example, if a constellation of 100 satellites wants to capture 3,000 targets, 300,000 evaluation operations are required. Considering that each operation must traverse the orbit of a single satellite for several days and the geometric coverage relationship, the time-consuming cost of the computation will be unbearable, which will have a great negative impact on the timeliness of mission planning.

[0005] 2. Traditional constellation mission planning methods are difficult to adapt to satellite elastic expansion and node damage

[0006] As the constellation is being built, the number of satellites in the constellation may be temporarily expanded or reduced due to expiration of their service life. Traditional constellation mission planning generally relies on central node planning and deployment to complete the entire constellation mission under the premise that satellite resources are relatively fixed. The applicability of this architecture under a flexible networking architecture will be greatly reduced.

[0007] 3. Priority-based target selection rules are difficult to adapt to the ever-changing tasks

[0008] There are many types and sources of imaging tasks. When faced with sudden tasks, how to automatically use limited satellite resources to achieve optimal coverage cost-effectiveness in an unmanned environment is an urgent problem that needs to be solved. However, traditional priority-based task scheduling rules are difficult to achieve optimal results in this scenario. Summary of the invention

[0009] The purpose of the present invention is to solve the problems existing in the prior art and to propose a satellite autonomous mission planning method.

[0010] In order to achieve the above object, the present invention adopts the following technical solutions:

[0011] A satellite autonomous mission planning method comprises the following steps:

[0012] S1, dual-core processors are used on the satellite, where the tasks running on processor 1 are the main satellite tasks, which are used for satellite and attitude and orbit control tasks, satellite autonomous planning tasks, and the execution of instruction sequences expanded from each main planning task. It can also undertake external temporary task planning. Processor 2 is mainly used for regional target planning and anchor target library task planning;

[0013] S2. Point targets in the local anchoring task will be scanned regularly, and the executable time window and resource dependency will be preliminarily calculated using the fast planning algorithm;

[0014] S3. The regional planning tasks in the local target database will be planned once every control cycle. If the planning of this cycle is not completed, the intermediate state results will be retained and the planning will continue in the next cycle. During planning, the target area will be divided into grids of specified resolution, and each grid point will be marked to record its current coverage status. The possible coverage time period will be quickly found through the simplified orbit model and vertical projection collision detection algorithm, and then the orbit position will be quickly traversed to determine the epoch time corresponding to the start and end of the imaging task while leaving a certain margin in the time period. At the same time, the newly added coverage grid points will be counted, thereby generating multiple single-point imaging tasks. Finally, the single-point task that can be executed within 48 hours and contributes the most to the coverage increment is selected and inserted into the queue of points to be planned, and the next planning will be carried out after the task is executed or abandoned.

[0015] S4. When an external temporary task arrives, it is first verified and then different planning strategies are adopted according to its urgency: for non-urgent tasks, the processing is similar to that of anchor tasks, and the initial executable time window and resource dependency are calculated using a fast planning algorithm. If it can be carried out within the specified time, it will enter the queue of points to be planned, and then execute the next step. Otherwise, the task cannot be executed through inter-satellite feedback and the operation is terminated;

[0016] S5. The task management process regularly and quickly traverses the queue of points to be planned, dynamically adjusts the task score according to different target types and waiting times, and calculates the resource dependencies required to execute the task;

[0017] S6. Traverse the planned task queue to check whether the current point task to be planned overlaps with the time window and dependent resources of the planned task;

[0018] S7. If there is no task conflict, the current task to be planned is directly inserted into the planned task queue. If the task is an external temporary task, it will be confirmed and fed back through intersatellite communication. Otherwise, the next step will be executed.

[0019] S8. If a task conflict is detected, try to see if the tasks can be merged. If the tasks can be merged, merge the two conflicting tasks and insert the merged task into the planned task queue. If the task is an external temporary task, confirm and feedback will be given through intersatellite communication. Otherwise, proceed to the next step.

[0020] S9. If the tasks conflict and cannot be merged, the high-scoring tasks will be inserted into the planned task queue according to the task scores, and the low-scoring tasks will be replaced; the first-come-first-served principle will be adopted for tasks with the same task scores; if the replaced tasks are external temporary tasks, they will be abandoned through the intersatellite feedback task so that they can be reassigned by the constellation;

[0021] S10. To avoid imaging or imaging deviation caused by excessive orbit extrapolation error due to planning too long in advance, a new forecast is conducted 5 minutes before the planned mission execution to reduce the orbit extrapolation error as much as possible, and key execution parameters such as the imaging center time are regenerated and calculated based on the accurate forecast results;

[0022] S11, after reaching the task execution window, it will automatically generate a command sequence according to the task execution parameters and autonomously complete all operations required for the task;

[0023] S12, after the mission is executed, feedback the mission execution status through the satellite;

[0024] S13. After the imaging task is completed, an instruction will be automatically generated according to the task information to create a target recognition or data transmission task to complete the subsequent work.

[0025] Preferably, in step S2, the algorithm only considers the simplified circular orbit model of the J2 perturbation and uses spherical geometric relationships to calculate possible imaging positions, and uses the dichotomy method to quickly find the orbital position and epoch time that are closest to the center of the satellite imaging range of each circle within the planning time range, and combines the side viewing angle range and other information to determine whether the imaging conditions are met. If it can be carried out within 48 hours, it will enter the queue of points to be planned.

[0026] Preferably, in step S4, for urgent tasks, the precise orbit prediction model is directly used with the aid of binary search to traverse the orbit. Since the time span is very small, the accuracy and calculation time can be guaranteed at the same time. If it can be executed within 48 hours, jump to step S6. Otherwise, the task cannot be executed through intersatellite feedback and the operation is terminated.

[0027] Preferably, in step S5, the task management process time is once every 250 milliseconds.

[0028] Compared with the prior art, the present invention has the following advantages:

[0029] The satellite autonomous mission planning method proposed in the present invention has the characteristics of low CPU resource consumption, fine task priority granularity, fast large-scale task processing time, high task prediction accuracy, high responsiveness to sudden tasks, etc. It is suitable for the current development trend of a substantial increase in the number of satellites and increasing task complexity and uncertainty. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A schematic diagram of task allocation between processor 1 and processor 2 in a satellite autonomous task planning method proposed by the present invention;

[0031] Figure 2 The present invention provides a schematic diagram of the execution flow of an autonomous mission planning task in a satellite autonomous mission planning method proposed by the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Reference Figure 1-Figure 2 , a satellite autonomous mission planning method, comprising the following steps:

[0034] S1, dual-core processors are used on the satellite, where the tasks running on processor 1 are the main satellite tasks, which are used for satellite and attitude and orbit control tasks, satellite autonomous planning tasks, and the execution of instruction sequences expanded from each main planning task. It can also undertake external temporary task planning. Processor 2 is mainly used for regional target planning and anchor target library task planning;

[0035] S2. Point targets in the local anchoring task will be scanned regularly, and the executable time window and resource dependency will be preliminarily calculated using the fast planning algorithm;

[0036] S3. The regional planning tasks in the local target database will be planned once every control cycle. If the planning of this cycle is not completed, the intermediate state results will be retained and the planning will continue in the next cycle. During planning, the target area will be divided into grids of specified resolution, and each grid point will be marked to record its current coverage status. The possible coverage time period will be quickly found through the simplified orbit model and vertical projection collision detection algorithm, and then the orbit position will be quickly traversed to determine the epoch time corresponding to the start and end of the imaging task while leaving a certain margin in the time period. At the same time, the newly added coverage grid points will be counted, thereby generating multiple single-point imaging tasks. Finally, the single-point task that can be executed within 48 hours and contributes the most to the coverage increment is selected and inserted into the queue of points to be planned, and the next planning will be carried out after the task is executed or abandoned.

[0037] S4. When an external temporary task arrives, it is first verified and then different planning strategies are adopted according to its urgency: for non-urgent tasks, the processing is similar to that of anchor tasks, and the initial executable time window and resource dependency are calculated using a fast planning algorithm. If it can be carried out within the specified time, it will enter the queue of points to be planned, and then execute the next step. Otherwise, the task cannot be executed through inter-satellite feedback and the operation is terminated;

[0038] S5. The task management process regularly and quickly traverses the queue of points to be planned, dynamically adjusts the task score according to different target types and waiting times, and calculates the resource dependencies required to execute the task;

[0039] S6. Traverse the planned task queue to check whether the current point task to be planned overlaps with the time window and dependent resources of the planned task;

[0040] S7. If there is no task conflict, the current task to be planned is directly inserted into the planned task queue. If the task is an external temporary task, it will be confirmed and fed back through intersatellite communication. Otherwise, the next step will be executed.

[0041] S8. If a task conflict is detected, try to see if the tasks can be merged. If the tasks can be merged, merge the two conflicting tasks and insert the merged task into the planned task queue. If the task is an external temporary task, confirm and feedback will be given through intersatellite communication. Otherwise, proceed to the next step.

[0042] S9. If the tasks conflict and cannot be merged, the high-scoring tasks will be inserted into the planned task queue according to the task scores, and the low-scoring tasks will be replaced; the first-come-first-served principle will be adopted for tasks with the same task scores; if the replaced tasks are external temporary tasks, they will be abandoned through the intersatellite feedback task so that they can be reassigned by the constellation;

[0043] S10. To avoid imaging or imaging deviation caused by excessive orbit extrapolation error due to planning too long in advance, a new forecast is conducted 5 minutes before the planned mission execution to reduce the orbit extrapolation error as much as possible, and key execution parameters such as the imaging center time are regenerated and calculated based on the accurate forecast results;

[0044] S11, after reaching the task execution window, it will automatically generate a command sequence according to the task execution parameters and autonomously complete all operations required for the task;

[0045] S12, after the mission is executed, feedback the mission execution status through the satellite;

[0046] S13. After the imaging task is completed, an instruction will be automatically generated according to the task information to create a target recognition or data transmission task to complete the subsequent work.

[0047] In step S2, the algorithm only considers the simplified circular orbit model of the J2 perturbation and uses spherical geometric relationships to calculate possible imaging positions. It uses the dichotomy method to quickly find the orbital position and epoch time that are closest to the center of the satellite imaging range of each circle within the planning time range. It combines information such as the side viewing angle range to determine whether the imaging conditions are met. If it can be carried out within 48 hours, it will enter the queue of points to be planned.

[0048] In step S4, for urgent tasks, the precise orbit prediction model is directly used with the aid of binary search to traverse the orbit. Since the time span is very small, the accuracy and calculation time can be guaranteed at the same time. If it can be executed within 48 hours, it jumps to step S6. Otherwise, the task cannot be executed through intersatellite feedback and the operation is terminated.

[0049] In step S5, the task management process is performed once every 250 milliseconds.

[0050] Further, such as Figure 2 As shown in the figure, tasks in the waiting task queue and the planned task queue can be deleted or terminated through instructions, and the flexible insertion, replacement, deletion and termination of tasks can be realized by combining the previous management mechanism. In addition, the execution process of each task is recorded in the full process task log, and the current status of the task can be switched and displayed by specifying the telemetry package. The automatically generated instruction sequence can automatically generate a binary file or text file for download and verification, providing a handle for task tracking management and even ground debugging.

[0051] The satellite autonomous mission planning method proposed in the present invention has the characteristics of low CPU resource consumption, fine task priority granularity, fast large-scale task processing time, high task prediction accuracy, high responsiveness to sudden tasks, etc. It is suitable for the current development trend of a substantial increase in the number of satellites and increasing task complexity and uncertainty.

[0052] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A satellite autonomous mission planning method, characterized in that: The following steps are involved: S1, dual-core processors are used on the satellite, where the tasks running on processor 1 are the main satellite tasks, which are used for satellite and attitude and orbit control tasks, satellite autonomous planning tasks, and the execution of instruction sequences expanded from each main planning task. It can also undertake external temporary task planning. Processor 2 is mainly used for regional target planning and anchor target library task planning; S2. Point targets in the local anchoring task will be scanned regularly, and the executable time window and resource dependency will be preliminarily calculated using the fast planning algorithm; S3. The regional planning tasks in the local target database will be planned once in each control cycle. If the planning in this cycle is not completed, the intermediate state results will be retained and the planning will continue in the next cycle. During planning, the target area will be divided into grids of specified resolution, and each grid point will have a mark to record its current coverage status. By simplifying the orbit model and the vertical projection collision detection algorithm, the possible coverage time period is quickly found, and then the orbit position is quickly traversed to determine the epoch time corresponding to the start and end of the imaging task while leaving a certain margin in the time period. At the same time, the newly added coverage grid points are counted to generate multiple single-point imaging tasks. Finally, the single-point task that can be executed within 48 hours and contributes the most to the coverage increment is selected and inserted into the queue of points to be planned, and the next planning is carried out after the task is executed or abandoned. S4. When an external temporary task arrives, it is first verified and then different planning strategies are adopted according to its urgency: for non-urgent tasks, the processing is similar to that of anchor tasks, and the initial executable time window and resource dependency are calculated using a fast planning algorithm. If it can be carried out within the specified time, it will enter the queue of points to be planned, and then execute the next step. Otherwise, the task cannot be executed through inter-satellite feedback and the operation is terminated; S5. The task management process regularly and quickly traverses the queue of points to be planned, dynamically adjusts the task score according to different target types and waiting times, and calculates the resource dependencies required to execute the task; S6. Traverse the planned task queue to check whether the current point task to be planned overlaps with the time window and dependent resources of the planned task; S7. If there is no task conflict, the current task to be planned is directly inserted into the planned task queue. If the task is an external temporary task, it will be confirmed and fed back through intersatellite communication. Otherwise, the next step will be executed. S8. If a task conflict is detected, try to see if the tasks can be merged. If the tasks can be merged, merge the two conflicting tasks and insert the merged task into the planned task queue; If the task is an external temporary task, it will be confirmed and fed back through intersatellite communication, otherwise it will proceed to the next step; S9. If the tasks conflict and cannot be merged, the high-scoring tasks will be inserted into the planned task queue according to the task scores, and the low-scoring tasks will be replaced; the first-come-first-served principle will be adopted for tasks with the same task scores; if the replaced tasks are external temporary tasks, they will be abandoned through the intersatellite feedback task so that they can be reassigned by the constellation; S10. To avoid imaging or imaging deviation caused by excessive orbit extrapolation error due to planning too long in advance, a new forecast is conducted 5 minutes before the planned mission execution to reduce the orbit extrapolation error as much as possible, and key execution parameters such as the imaging center time are regenerated and calculated based on the accurate forecast results; S11, after reaching the task execution window, it will automatically generate a command sequence according to the task execution parameters and autonomously complete all operations required for the task; S12, after the mission is executed, feedback the mission execution status through the satellite; S13. After the imaging task is completed, an instruction will be automatically generated according to the task information to create a target recognition or data transmission task to complete the subsequent work.

2. A satellite autonomous mission planning method according to claim 1, characterized in that: In step S2, the algorithm only considers the simplified circular orbit model of the J2 perturbation and uses spherical geometric relationships to calculate possible imaging positions, and uses the dichotomy method to quickly find the orbital position and epoch time closest to the target at the center of the satellite imaging range of each circle within the planning time range, and combines information such as the side viewing angle range to determine whether the imaging conditions are met. If it can be carried out within 48 hours, it will enter the queue of points to be planned.

3. A satellite autonomous mission planning method according to claim 1, characterized in that: In step S4, for urgent tasks, the precise orbit prediction model is directly used with the aid of binary search to traverse the orbit. Since the time span is very small, the accuracy and calculation time can be guaranteed at the same time. If it can be executed within 48 hours, it jumps to step S6. Otherwise, the task cannot be executed through inter-satellite feedback and the operation is terminated.

4. A satellite autonomous mission planning method according to claim 1, characterized in that: In step S5, the task management process is performed once every 250 milliseconds.

Images