A body-equipped intelligent agent driven satellite autonomous method, device, equipment and medium

Abstract

This application relates to the field of intelligent aerospace technology and discloses a satellite autonomy method driven by embodied intelligent agents, applicable to subsatellites belonging to an in-orbit constellation. It constructs a multi-agent framework comprising multiple autonomous intelligent agents, including at least a resource assessment agent and a resource optimization agent. In response to a resource assessment command regarding the target mission, the method acquires the subsatellite's in-orbit operating parameters. The resource assessment agent is invoked to perform resource assessment based on the in-orbit operating parameters, obtaining the mission resource assessment result, which is then fed back to the in-orbit constellation. If mission feasibility indicates that the subsatellite can execute the target mission, the method acquires a single-satellite mission plan fed back by the in-orbit constellation. The resource optimization agent is invoked to optimize resource allocation based on the single-satellite mission plan, obtaining a mission operation plan. Its beneficial effects are that it improves the intelligence level of subsatellites in remote sensing observation missions and enhances the satellite's in-orbit autonomy.

Description

Technical Field

[0001] This application relates to the field of intelligent aerospace technology, and in particular to a satellite autonomy method, apparatus, device and medium driven by an embodied intelligent agent. Background Technology

[0002] Satellite remote sensing technology is widely used in various fields such as ecological environment monitoring, climate change monitoring, and urban management. Based on its high coverage and all-weather observation capabilities, satellite remote sensing provides efficient observation methods and accurate data foundations for these fields. However, in most conventional technologies, satellites typically rely on ground stations to formulate and plan observation tasks, and then execute remote sensing observations according to the predetermined instructions after receiving the tasks from the ground stations. In this approach, if the satellite's own condition or environmental conditions within its observation range change, the satellite cannot respond quickly or adjust in real time, thus affecting the timeliness and accuracy of the observation data.

[0003] Among related technologies, the level of intelligence in satellite remote sensing technology still needs to be improved. Summary of the Invention

[0004] This application provides a satellite autonomy method, apparatus, device, and medium driven by embodied intelligent agents. By controlling sub-satellites through a multi-agent framework, the sub-satellites can independently perform actions such as resource assessment and resource optimization, thereby improving the intelligence level of sub-satellites in remote sensing observation missions and enhancing the satellite's on-orbit autonomy.

[0005] To achieve the above objectives, the main technical solutions adopted in this application include: In a first aspect, embodiments of this application provide a satellite autonomy method driven by an embodied intelligent agent, applicable to any sub-satellite belonging to an in-orbit constellation. The sub-satellite contains a multi-agent framework comprising multiple autonomous intelligent agents, wherein the autonomous intelligent agents include at least a resource assessment agent and a resource optimization agent. The method includes: In response to a resource assessment instruction regarding the target mission, the on-orbit operating parameters of the subsatellite are acquired; The resource assessment agent is invoked to perform resource assessment on the subsatellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the subsatellite, and feed back the mission resource assessment result to the on-orbit constellation; wherein, the mission resource assessment result includes the mission feasibility of the subsatellite in the target mission; If the mission feasibility indicates that the sub-satellite can perform the target mission, obtain the single-satellite mission plan for the target mission fed back to the sub-satellite by the on-orbit constellation; The resource optimization agent is invoked to optimize the on-orbit operation parameters of the sub-satellite based on the single-satellite mission plan, thereby obtaining the mission operation plan of the sub-satellite and controlling the sub-satellite to achieve on-orbit autonomy.

[0006] The satellite autonomy method driven by embodied intelligent agents proposed in this application sets up a multi-agent framework containing multiple autonomous intelligent agents in the sub-satellite. Upon receiving a resource assessment command, the resource assessment agent in the multi-agent framework performs a resource assessment on the sub-satellite, and uses the sub-satellite's mission resource assessment result as the basis for mission planning in the on-orbit constellation, thereby obtaining the single-satellite mission plan fed back by the on-orbit constellation. Finally, the resource optimization agent in the sub-satellite is invoked to optimize the resource allocation of the sub-satellite according to the single-satellite mission plan, realizing intelligent on-orbit autonomy of the sub-satellite. Compared with related technologies, this application controls the sub-satellite for parameter acquisition and resource assessment through a multi-agent framework, thereby enabling real-time acquisition of changes in the sub-satellite, which serves as the data basis for on-orbit constellation mission planning and improves the sub-satellite's responsiveness to real-time changes. On this basis, by invoking different autonomous intelligent agents in the multi-agent framework and interacting with the on-orbit constellation, real-time adjustment of remote sensing observation tasks and resource optimization for remote sensing observation tasks are realized, enabling the sub-satellite to operate independently and stably, improving the intelligence level of the sub-satellite in remote sensing observation tasks, and enhancing the satellite's on-orbit autonomy capability.

[0007] Optionally, the autonomous intelligent agent further includes an instruction parsing intelligent agent and a state-aware intelligent agent; the step of obtaining the on-orbit operating parameters of the sub-satellite in response to a resource assessment instruction regarding the target mission includes: The instruction parsing agent performs intent parsing on the resource assessment instruction to obtain the parameter acquisition object corresponding to the resource assessment instruction; The state-aware intelligent agent is invoked to collect parameters of the sub-satellite based on the parameter acquisition object, thereby obtaining the on-orbit operating parameters corresponding to the parameter acquisition object.

[0008] Optionally, the step of invoking the resource assessment agent to perform resource assessment on the subsatellite based on the on-orbit operating parameters and obtain the mission resource assessment result of the subsatellite includes: Based on the task requirements of the target task, determine the task execution standards corresponding to each of the on-orbit operation parameters; The resource assessment agent is invoked to conduct a feasibility assessment of the on-orbit operating parameters based on the task execution criteria, thereby determining the mission feasibility of the subsatellite in the target mission. The mission resource assessment results are generated based on the on-orbit operating parameters, mission feasibility, and the fault status of the sub-satellites.

[0009] Optionally, the step of invoking the resource optimization agent to optimize resource allocation for the on-orbit operating parameters of the sub-satellite based on the single-satellite mission plan, thereby obtaining the mission operation plan for the sub-satellite, includes: Constraints are extracted based on the task requirements of the target task to obtain the optimization constraints for resource allocation optimization; Based on the preset optimization objectives and the optimization constraints, the on-orbit operation parameters of the sub-satellite are planned for mission resources to obtain the mission operation scheme; wherein, the mission operation scheme includes a mission scheme obtained by rearranging multiple single-satellite missions in the single-satellite mission scheme, and the satellite resources allocated to the sub-satellite when performing each single-satellite mission.

[0010] Optionally, the sub-satellite also includes an agent invocation model; each autonomous agent is invoked in the following manner: The agent invocation model is used to perform semantic parsing on the agent invocation instructions to determine the instruction intent of the agent invocation instructions; wherein, the agent invocation instructions include the resource assessment instructions and the resource optimization instructions issued by the sub-satellite when it receives the single-satellite mission plan, and the resource optimization instructions are used to control the sub-satellite to optimize resource allocation; Based on the instruction intent, the autonomous agents in the agent framework are functionally matched to determine the agent invocation plan that conforms to the instruction intent, so as to invoke the corresponding autonomous agent.

[0011] Optionally, the method further includes: The on-orbit operating parameters are sent to the multi-agent framework, and each autonomous agent is controlled to perceive and learn the on-orbit operating parameters to obtain the dynamic operating environment status. The behavioral strategy parameters of each autonomous agent are updated and adjusted according to the dynamic operating environment state, so that each autonomous agent can adapt to the operating environment of the sub-satellite.

[0012] Optionally, the autonomous intelligent agent further includes a fault diagnosis intelligent agent and a fault repair intelligent agent; the method further includes: The fault diagnosis agent is invoked to analyze the operational status of the sub-satellite's operation log and determine the faults that occurred during the autonomous operation of the sub-satellite. The fault repair agent is invoked to repair the fault.

[0013] Secondly, embodiments of this application provide a satellite autonomous device driven by an embodied intelligent agent, applicable to any sub-satellite belonging to an on-orbit constellation. The sub-satellite contains a multi-agent framework comprising multiple autonomous intelligent agents, the autonomous intelligent agents including at least a resource assessment agent and a resource optimization agent. The device includes: The status awareness module is used to obtain the on-orbit operating parameters of the sub-satellite in response to resource assessment instructions regarding the target mission; The resource assessment module is used to invoke the resource assessment agent to perform resource assessment on the subsatellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the subsatellite, and feed back the mission resource assessment result to the on-orbit constellation; wherein, the mission resource assessment result includes the mission feasibility of the subsatellite in the target mission; The information interaction module is used to obtain the single-satellite mission plan for the target mission fed back to the sub-satellite by the on-orbit constellation when the mission feasibility indicates that the sub-satellite can perform the target mission; The resource optimization module is used to call the resource optimization agent to optimize the resource allocation of the sub-satellite's on-orbit operation parameters based on the single-satellite mission plan, thereby obtaining the sub-satellite's mission operation plan and controlling the sub-satellite to achieve on-orbit autonomy.

[0014] Thirdly, embodiments of this application provide a computer device, including: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the method described in any of the above embodiments.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer instructions, which are used to cause a computer to perform the method described in any one of the above embodiments.

[0016] Fifthly, embodiments of this application provide a computer program product, including computer instructions, which are used to cause a computer to perform the method described in any of the above embodiments. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A step diagram illustrating the satellite autonomy method driven by an embodied intelligent agent provided in the embodiments of this application; Figure 2 This is an interactive diagram of the multi-agent framework in the embodiments of this application; Figure 3 This is a flowchart illustrating the steps for obtaining on-orbit operating parameters in an embodiment of this application. Figure 4 This is a schematic diagram of the interaction of some autonomous intelligent agents in the embodiments of this application; Figure 5 This is a flowchart illustrating the steps involved in obtaining the task resource evaluation results in this application embodiment; Figure 6 This is a flowchart illustrating the steps involved in obtaining the task execution scheme in the embodiments of this application; Figure 7 This is a flowchart illustrating the steps of calling each autonomous intelligent agent in the embodiments of this application; Figure 8 This is a flowchart illustrating the steps of environmental perception and adaptation for each autonomous agent in the embodiments of this application. Figure 9 This is a flowchart illustrating the steps of fault diagnosis and repair in the embodiments of this application; Figure 10 This is a schematic diagram of the interaction of some autonomous intelligent agents in the embodiments of this application; Figure 11 This is a diagram illustrating the overall interaction framework of the multi-agent framework in this application embodiment; Figure 12 A block diagram of a satellite autonomous device driven by an embodied intelligent agent provided in the embodiments of this application; Figure 13 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Satellite remote sensing technology is widely used in various fields such as ecological environment monitoring, climate change monitoring, and urban management. Based on its high coverage and all-weather observation capabilities, satellite remote sensing provides efficient observation methods and accurate data foundations for these fields. However, in most conventional technologies, satellites typically rely on ground stations to formulate and plan observation tasks, and then execute remote sensing observations according to the predetermined instructions after receiving the tasks from the ground stations. In this approach, if the satellite's own condition or environmental conditions within its observation range change, the satellite cannot respond quickly or adjust in real time, thus affecting the timeliness and accuracy of the observation data.

[0021] In related technologies, improvements addressing the aforementioned problems focus on introducing improved algorithms to enhance the timeliness of observation mission planning, such as using online algorithms for real-time adjustments to satellite orbits. Online algorithms require frequent intervention from ground personnel in orbit control, attitude adjustment, and fault diagnosis, communicating with the satellite for real-time control. This high level of human intervention not only increases the labor costs of satellite remote sensing technology but also limits the satellite's autonomy in observation missions. Furthermore, when dealing with large-scale constellations, this method struggles to ensure that all satellites can correctly perform their observation tasks. Therefore, the level of intelligence in satellite remote sensing technology still needs improvement.

[0022] To address the aforementioned issues, this application provides a satellite autonomy method, apparatus, device, and medium driven by embodied intelligent agents, applicable to subsatellites belonging to an on-orbit constellation. It constructs a multi-agent framework comprising multiple autonomous intelligent agents, including at least a resource assessment agent and a resource optimization agent. In response to a resource assessment instruction regarding a target mission, it acquires the on-orbit operating parameters of the subsatellite; it invokes the resource assessment agent to perform resource assessment based on the on-orbit operating parameters, obtaining a mission resource assessment result, which is then fed back to the on-orbit constellation; if mission feasibility indicates that the subsatellite can execute the target mission, it acquires a single-satellite mission plan fed back by the on-orbit constellation; and it invokes the resource optimization agent to optimize resource allocation based on the single-satellite mission plan, obtaining a mission operation plan.

[0023] The satellite autonomy method driven by embodied intelligent agents provided in this application sets up a multi-agent framework containing multiple autonomous intelligent agents in the sub-satellite. When a resource assessment instruction is received, the resource assessment agent in the multi-agent framework is used to assess the resources of the sub-satellite, and the mission resource assessment result of the sub-satellite is used as the basis for mission planning of the on-orbit constellation, thereby obtaining the single-satellite mission plan fed back by the on-orbit constellation. Finally, the resource optimization agent in the sub-satellite is invoked to optimize the resource allocation of the sub-satellite according to the single-satellite mission plan, thereby realizing the intelligent on-orbit autonomy of the sub-satellite.

[0024] Compared with related technologies, this application controls subsatellites to acquire parameters and assess resources through a multi-agent framework, thereby enabling real-time acquisition of subsatellite changes. This data serves as the foundation for on-orbit constellation mission planning, enhancing the subsatellites' responsiveness to real-time changes. Furthermore, by invoking different autonomous agents within the multi-agent framework and interacting with the on-orbit constellation, real-time adjustments to remote sensing observation missions and resource optimization for these missions are achieved. This allows subsatellites to operate independently and stably, improving their intelligence level in remote sensing observation missions and enhancing their on-orbit autonomy.

[0025] The embodied intelligent agent-driven satellite autonomy method provided in this specification can be applied to spacecraft with remote sensing capabilities, such as satellites, which belong to a spacecraft system comprising multiple spacecraft for performing remote sensing observation missions. It is understood that, with adaptive modifications, the method provided in this specification can also be applied to spacecraft other than satellites to improve the on-orbit autonomy of spacecraft.

[0026] According to an embodiment of this application, an embodiment of a satellite autonomous method driven by an embodied intelligent agent is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0027] This embodiment provides a satellite autonomy method driven by an embodied intelligent agent, which can be used in the aforementioned spacecraft. (Refer to...) Figure 1 As shown, this method is applied to any subsatellite belonging to an in-orbit constellation. The subsatellite contains a multi-agent framework comprising multiple autonomous agents, which at least include a resource assessment agent and a resource optimization agent; including: S100. In response to a resource assessment instruction regarding the target mission, acquire the on-orbit operating parameters of the subsatellite.

[0028] S200. Call the resource assessment agent to conduct resource assessment on the subsatellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the subsatellite, and feed back the mission resource assessment result to the on-orbit constellation; wherein, the mission resource assessment result includes the mission feasibility of the subsatellite in the target mission.

[0029] S300. If the mission feasibility indicates that the subsatellite can perform the target mission, obtain the single-satellite mission plan regarding the target mission fed back to the subsatellite from the on-orbit constellation.

[0030] S400 invokes a resource optimization agent to optimize the on-orbit operation parameters of the sub-satellite based on the single-satellite mission plan, thereby obtaining the mission operation plan of the sub-satellite and controlling the sub-satellite to achieve on-orbit autonomy.

[0031] The on-orbit operating parameters can be status parameters related to the execution of observation tasks when the subsatellite is in orbit. These parameters represent the subsatellite's on-orbit status and determine whether it can perform the corresponding observation tasks. On-orbit operating parameters can be real-time parameters of the subsatellite in various aspects such as energy, thermal control, computing power, or communication, to comprehensively understand the subsatellite's operational status. For example, on-orbit operating parameters may include the subsatellite's power supply voltage, power supply current, CPU / GPU status, CAN bus correct frame count, CAN bus error frame count, memory usage, available disk capacity, parameter collection time, energy status, and charging / discharging time.

[0032] Specifically, after analyzing the target mission, the on-orbit constellation sends resource assessment commands to subsatellites capable of performing the target mission to evaluate the real-time status of each subsatellite. Upon receiving the resource assessment commands, the subsatellites determine the mission requirements based on the commands and use appropriate sensors to monitor their status and obtain on-orbit operational parameters. For example, these sensors may be voltage sensors, current sensors, or temperature sensors. It is understood that these on-orbit operational parameters cover multiple aspects of the subsatellite's operation, enabling comprehensive monitoring of its operational status. This provides an effective data foundation for mission planning of the on-orbit constellation and resource optimization of the subsatellites themselves, improving the stability of the subsatellite's on-orbit autonomy.

[0033] In some embodiments, the resource assessment command may also be sent by ground personnel at the ground base station to the sub-satellites to obtain the on-orbit status of each sub-satellite, monitor the health status of the sub-satellites in real time, and ensure that the sub-satellites can operate normally.

[0034] Furthermore, after obtaining the on-orbit operating parameters of the subsatellite, the resource assessment agent within the multi-agent framework is invoked to perform a resource assessment of the on-orbit operating parameters, yielding the mission resource assessment result. The resource assessment includes health status assessment and feasibility assessment. The health status assessment evaluates the health status of the subsatellite to ensure it is in normal operating condition. During the health status assessment, the resource assessment agent compares the on-orbit operating parameters with the preset parameter range for normal operation of the subsatellite, obtaining the health status assessment result based on the numerical relationship between the two. It can be understood that the health status assessment result indicates whether the subsatellite is in normal operating condition. If the health status assessment result indicates that the subsatellite is in normal operating condition, it can stably and reliably execute remote sensing observation tasks transmitted to the on-orbit constellation, improving the accuracy of remote sensing observation data. If the health status assessment result indicates that the subsatellite is not in normal operating condition, there may be potential fault risks within the subsatellite. Based on the health status assessment result, the subsatellite can be controlled to handle fault risks, or ground personnel at the ground station can be urged to handle faults in the subsatellite to eliminate fault risks, improve the operational stability of the subsatellite, and extend its on-orbit lifespan.

[0035] For example, a health status assessment may include an energy system assessment and a thermal control system assessment. The energy system assessment may include evaluation items such as charging efficiency, battery health, and load balancing. Charging efficiency can be obtained based on the average irradiance angle of the solar panel and the average charging current. The preset parameter range for the average irradiance angle of the solar panel may be within 35° ± 5°, and the preset parameter range for the average charging current may be approximately 4.5A. Battery health can be obtained based on the remaining charge during the shaded period, the slope of the charge curve, and the full charge capacity. The preset parameter range for the remaining charge during the shaded period may be approximately 215%, and the preset parameter range for the full charge capacity may be the initial design value of the energy system. Load balancing can be obtained based on the average power consumption of the load and the power consumption of the optical load. The preset parameter range for the average power consumption of the load may be within 5% of 80W, and the power consumption of the optical load should remain stable under normal conditions without intermittent peak changes.

[0036] The evaluation items for the thermal control system can include temperature stability and heater performance. Temperature stability can be obtained based on the load temperature and the rate of temperature drop during the shadow period. The preset parameter range for the load temperature can be -15℃ to +30℃, and the preset parameter range for the rate of temperature drop during the shadow period can be determined based on historical averages. Heater performance can be obtained based on the average heater PWM duty cycle and PWM waveform variation. The preset parameter range for the average heater PWM duty cycle can be between 40% and 50%, and the PWM waveform variation is continuously output under normal conditions.

[0037] A health status assessment of the subsatellite was conducted using the aforementioned evaluation items. The average illumination angle of the subsatellite's solar panels was 32°, and the average charging current was 4.2A. At this time, the subsatellite's charging efficiency decreased by 6.7%, which may be related to localized shading caused by micrometeorite impacts on the solar panels. The subsatellite's remaining battery power dropped to a minimum of 18% during the shading period. The battery curve slope was normal, but the full charge capacity decreased to 92%. The subsatellite's energy system was functioning normally, but long-term aging trends need to be monitored. The average power consumption of the subsatellite's payload was 85W. The optical payload Payload-3 exhibited intermittent power consumption peaks of 105W, lasting for 5 minutes per orbit. This indicates a load balancing fault in the subsatellite, requiring investigation into whether it is due to abnormal mission modes or hardware failure. Based on the above analysis, the overall energy system score of the subsatellite is 87, a decrease of 12% compared to the previous cycle.

[0038] The temperatures of all payloads on the subsatellite ranged from -10°C to +25°C. However, the optical payload Payload-2 experienced a temperature drop rate of 0.8°C / min during the shadow period, higher than the historical average of 0.5°C / min. This accelerated temperature drop rate during the shadow period may be due to localized failure of the thermal insulation material. The average PWM duty cycle of the subsatellite's heaters was 45%, but the PWM waveform exhibited three short pauses during the illumination period of orbit N-2. This may be due to poor contact in the drive circuit, and this should be listed as a potential fault point. Based on the above analysis, the overall score of the subsatellite's thermal control system is 91, remaining stable compared to the previous cycle.

[0039] Specifically, feasibility assessment can evaluate whether a subsatellite can perform the target mission to ensure its completion. During the feasibility assessment process, the resource assessment agent compares the on-orbit operating parameters with the mission requirements of the target mission and determines the subsatellite's mission feasibility based on whether the on-orbit operating parameters meet the requirements. It should be noted that mission feasibility indicates whether a subsatellite possesses the capability to fully execute the target mission and achieve the expected results. Mission feasibility is a binary result: if the on-orbit operating parameters meet the mission requirements, mission feasibility indicates that the subsatellite can perform the target mission, and the on-orbit constellation can include the subsatellite in its mission planning; if the on-orbit operating parameters do not meet the target mission requirements, mission feasibility indicates that the subsatellite cannot perform the target mission, and the on-orbit constellation will not consider the subsatellite during mission planning.

[0040] Furthermore, after obtaining the mission resource assessment results, which include mission feasibility, the results are fed back to the on-orbit constellation to which the sub-satellite belongs. Based on the health status and mission feasibility of each sub-satellite, the on-orbit constellation determines the feasible satellites capable of performing the target mission. Then, it performs mission planning for each feasible satellite according to the target mission, obtaining a single-satellite mission plan for each sub-satellite, which is then fed back to the sub-satellite. It should be noted that the single-satellite mission plan may include the single-satellite mission that the sub-satellite needs to perform, as well as the satellite resources allocated by the sub-satellite for each single-satellite mission.

[0041] Furthermore, after receiving the single-satellite mission plan, the subsatellite invokes a resource optimization agent to optimize resource allocation based on the single-satellite mission plan, thus obtaining the subsatellite's mission execution plan. It should be noted that the resource allocation optimization process may include adjusting the order of the single-satellite tasks or the allocated satellite resources in the single-satellite mission plan, so that the execution order of the single-satellite tasks matches the current resource status of the subsatellite. This reduces redundant actions used to ensure satellite resources during the execution of the target mission, such as charging the subsatellite's power system. Understandably, by optimizing the resource allocation of the subsatellite, the utilization rate of existing satellite resources within the subsatellite is improved, the total execution time of the mission execution plan is shortened, thereby effectively improving the execution efficiency of the target mission.

[0042] The interaction method between autonomous agents in this embodiment can be referred to Figure 2 As shown, the subsatellites belong to an on-orbit constellation, and their multi-agent framework includes at least a resource assessment agent and a resource optimization agent. After receiving on-orbit operating parameters in response to a resource assessment command, these parameters are sent to the resource assessment agent. The resource assessment agent then uses these parameters as data to perform a resource assessment on the subsatellite, obtaining the mission resource assessment result. After outputting the mission resource assessment result, the resource assessment agent sends it to the on-orbit constellation. The on-orbit constellation can then perform mission planning for its subsatellites based on the mission resource assessment result, resulting in one or more single-satellite mission plans. The on-orbit constellation sends each single-satellite mission plan to the corresponding subsatellite, which is received by the resource optimization agent within each subsatellite. The resource optimization agent optimizes resource allocation for the subsatellite based on the single-satellite mission plan, obtaining a mission operation plan for each subsatellite. The subsatellite controls each mission execution module based on the specific actions in the mission operation plan, enabling the subsatellite to effectively execute its target mission and obtain the required remote sensing observation data.

[0043] It should be noted that the multi-agent framework can also include other autonomous agents, such as an instruction parsing agent for resolving resource assessment instructions, a state-aware agent for acquiring on-orbit operating parameters, a fault diagnosis agent and a fault repair agent for handling faults within the subsatellite, and an emergency backup agent for responding to emergencies. These different autonomous agents are interconnected and exchange data and information in their respective application scenarios to control the subsatellite and achieve on-orbit autonomy.

[0044] Understandably, in this embodiment, a resource assessment agent evaluates the resources of the subsatellites and feeds back the mission resource assessment results to the on-orbit constellation. This allows the on-orbit constellation to comprehensively monitor the operational status of the subsatellites, understand their health status and execution capabilities, and use this as a data foundation for mission planning. After the on-orbit constellation feeds back the single-satellite mission plan for the target mission to the subsatellites, the subsatellites further optimize their resources through a resource optimization agent. This optimized mission operation plan controls the subsatellites to execute their missions, improving not only the efficiency of the subsatellites in executing the target mission but also extending their lifespan. Managing the subsatellite's operation through a multi-agent framework reduces the involvement of ground personnel during the subsatellite's operation, effectively improving the intelligence level within the subsatellites and enhancing their on-orbit autonomy.

[0045] The embodied agent-driven satellite autonomy method provided in this embodiment sets up a multi-agent framework containing multiple autonomous agents in the sub-satellite. When a resource assessment instruction is received, the resource assessment agent in the multi-agent framework is used to assess the resources of the sub-satellite, and the task resource assessment result of the sub-satellite is used as the basis for mission planning of the on-orbit constellation, thereby obtaining the single-satellite mission plan fed back by the on-orbit constellation. Finally, the resource optimization agent in the sub-satellite is invoked to optimize the resource allocation of the sub-satellite according to the single-satellite mission plan, thereby realizing the intelligent on-orbit autonomy of the sub-satellite.

[0046] Compared with related technologies, this application controls subsatellites to acquire parameters and assess resources through a multi-agent framework, thereby enabling real-time acquisition of subsatellite changes. This data serves as the foundation for on-orbit constellation mission planning, enhancing the subsatellites' responsiveness to real-time changes. Furthermore, by invoking different autonomous agents within the multi-agent framework and interacting with the on-orbit constellation, real-time adjustments to remote sensing observation missions and resource optimization for these missions are achieved. This allows subsatellites to operate independently and stably, improving their intelligence level in remote sensing observation missions and enhancing their on-orbit autonomy.

[0047] Reference Figure 3As shown, in one embodiment of this application, the autonomous intelligent agent further includes an instruction parsing agent and a state-aware agent; in response to a resource assessment instruction regarding the target mission, it acquires the on-orbit operating parameters of the sub-satellite, including: S110. The instruction parsing agent performs intent parsing on the resource assessment instruction to obtain the parameter acquisition object corresponding to the resource assessment instruction.

[0048] S120. Invoke the state-aware intelligent agent to collect parameters from the sub-satellite based on the parameter acquisition object, and obtain the on-orbit operating parameters corresponding to the parameter acquisition object.

[0049] The parameter acquisition object can be determined according to the requirements of the target mission, and can be one or more parameter types that determine whether the subsatellite can successfully perform the target mission, including but not limited to the subsatellite's spatiotemporal coverage capability, energy status and communication transmission capability.

[0050] Specifically, upon receiving a resource assessment instruction, the subsatellite invokes an instruction parsing agent to interpret the intent of the instruction and obtain the parameter acquisition objects corresponding to it. It can be understood that the resource assessment instruction may be given based on the target mission, containing task information related to the target mission. The instruction parsing agent parses this task information to determine the mission requirements imposed on each subsatellite by the target mission, and then determines the types of parameters to be acquired based on these requirements, thus obtaining the parameter acquisition objects.

[0051] Furthermore, by invoking the state-aware intelligent agent, one or more parameters corresponding to the parameter acquisition object are collected to obtain the on-orbit operating parameters. It should be noted that the state-aware intelligent agent can connect to various sensors within the subsatellite, calling the corresponding type of sensor based on the parameter acquisition object to obtain the collected data from each parameter acquisition object. For example, the parameter acquisition objects may include energy and thermal control status data, historical data, application scheduling data, environmental and orbital parameters, telemetry and remote control data, and payload logs, etc. Among these, energy and thermal control status data, historical data, and application scheduling data can be obtained through space-based distributed system software. This software has different data interfaces set up for different parameter acquisition objects for the state-aware intelligent agent to collect parameters. Environmental and orbital parameters, telemetry and remote control data, and payload logs can be obtained through the telemetry, tracking, and command (TT&C) platform software. This software also has different data interfaces set up for different parameter acquisition objects for the state-aware intelligent agent to collect parameters.

[0052] Furthermore, various acquired data are fed back to the state-aware agent through standardized data interfaces, where preprocessing is performed to obtain on-orbit operating parameters. It is understood that the acquired data may include parameter values ​​corresponding to the parameter acquisition objects, as well as acquisition characteristic information such as the acquisition frequency of these parameter values. The preprocessing process may include parameter refinement and format processing. Parameter refinement involves refining the parameter values ​​corresponding to each parameter acquisition object based on the acquisition characteristic information in the acquired data to obtain the temporal variation characteristics of each parameter acquisition object, thereby enabling a deeper understanding of the subsatellite's operating status. For example, for thermal control status data, in addition to obtaining the current temperature value, information such as temperature change trends and temperature fluctuations over different time periods can also be obtained.

[0053] Formatting can involve cleaning and formatting the collected data to improve its quality, thereby enhancing the accuracy and reliability of resource assessment. For example, formatting can include steps such as data cleaning, feature selection, and data standardization. Data cleaning removes erroneous and redundant data from the collected data, ensuring its accuracy. Feature selection extracts representative variables from the collected data to reduce the impact of data noise. Data standardization provides a unified standard for collected data with different dimensions, reducing errors caused by inconsistencies in measurement units.

[0054] Reference Figure 4 As shown, in some embodiments, the instruction parsing agent, the state-aware agent, and the resource assessment agent are connected sequentially and interact with each other in turn. The instruction parsing agent receives resource assessment instructions sent from outside and, after parsing, sends the parameter acquisition object to the state-aware agent. After receiving the parameter acquisition object, the state-aware agent calls the sensors to collect parameters and, after obtaining the on-orbit operating parameters, sends the on-orbit operating parameters to the resource assessment agent for resource assessment.

[0055] Reference Figure 5 As shown, in one embodiment of this application, a resource assessment agent is invoked to perform resource assessment on the subsatellite based on its on-orbit operating parameters, and the mission resource assessment result of the subsatellite is obtained, including: S210. Based on the mission requirements of the target mission, determine the mission execution standards corresponding to each on-orbit operating parameter.

[0056] S220. Call upon the resource assessment agent to conduct a feasibility assessment of the on-orbit operating parameters based on the mission execution standards, and determine the mission feasibility of the subsatellite in the target mission.

[0057] S230. Generate mission resource assessment results based on on-orbit operating parameters, mission feasibility, and subsatellite failure status.

[0058] Specifically, mission execution standards can be the parameter standards required for a subsatellite to perform the target mission. By using an instruction parsing agent to interpret the intent of resource assessment instructions, the mission requirements for each subsatellite are obtained, and the mission execution standards corresponding to each on-orbit operating parameter are determined for use in feasibility assessment of the subsatellites.

[0059] Furthermore, the resource assessment agent is invoked to compare the subsatellite's on-orbit operational parameters with the corresponding mission execution standards. If the on-orbit operational parameters meet the requirements of the mission execution standards, it means that the subsatellite meets the requirements for that parameter. If all on-orbit operational parameters meet the mission execution standards, then mission feasibility indicates that the subsatellite can successfully execute the target mission and obtain the expected mission results. If any on-orbit operational parameter of the subsatellite does not meet the mission execution standards, then mission feasibility indicates that the subsatellite cannot execute the target mission.

[0060] Furthermore, in addition to feasibility assessment, the resource assessment agent also performs a health status assessment of the subsatellites based on on-orbit operational parameters, obtaining health status assessment results. These results can include anomalies observed during subsatellite operation, operational risks, and overall health assessments. Subsatellite malfunctions can be detected by a fault diagnosis agent within the multi-agent framework to verify the health status assessment results and ensure the accuracy of the mission resource assessment. Based on on-orbit operational parameters, health status assessment, mission feasibility, and subsatellite malfunctions, mission resource assessment results are generated and fed back to the on-orbit constellation, enabling the constellation to monitor the operational status of the subsatellites.

[0061] Reference Figure 6 As shown, in one embodiment of this application, a resource optimization agent is invoked to optimize the resource allocation of the sub-satellite's on-orbit operating parameters based on the single-satellite mission plan, thereby obtaining the sub-satellite's mission operation plan, including: S410. Extract constraints based on the task requirements of the target task to obtain the optimization constraints for resource allocation optimization.

[0062] S420. Based on the preset optimization objectives and optimization constraints, perform mission resource planning on the on-orbit operation parameters of the sub-satellite to obtain a mission operation scheme; wherein, the mission operation scheme includes the mission scheme obtained by rearranging multiple single-satellite missions in the single-satellite mission scheme, as well as the satellite resources allocated to the sub-satellite when performing each single-satellite mission.

[0063] Specifically, the result of the instruction parsing agent's intent parsing of the resource assessment instruction is sent to the resource optimization agent. This allows the resource optimization agent to understand the task requirements imposed on the subsatellite by the target mission and extract constraints based on these requirements, resulting in optimized constraints for resource allocation. It is understood that the premise of resource allocation optimization is ensuring the subsatellite's successful execution of the target mission; therefore, the optimization constraints include the target mission's requirements. Furthermore, the optimization constraints can also include the subsatellite's on-orbit operating parameters to ensure that the subsatellite's operational status covers the mission operation plan, enabling the subsatellite to execute the optimized mission operation plan.

[0064] It should be noted that the preset optimization goal can be to shorten the total time for the subsatellite to complete the target mission. By optimizing the resource allocation of the subsatellite, the utilization rate of the existing satellite resources in the subsatellite is improved, the total execution time of the mission operation plan is shortened, and thus the execution efficiency of the target mission is effectively improved.

[0065] Furthermore, based on the preset optimization objectives and constraints, mission resource planning is performed on the on-orbit operating parameters of the subsatellite to obtain a mission operation plan. In some embodiments, the resource allocation optimization process can be simplified to solving operations research problems, including but not limited to dynamic programming, scheduling problems, multi-objective optimization, inventory management, nonlinear programming, and linear programming. Dynamic programming can be solved using methods such as value iteration algorithms and policy iteration algorithms; scheduling problems can be solved using methods such as genetic algorithms and Hungarian algorithms; multi-objective optimization can be solved using methods such as Pareto optimality and weighted summation methods; inventory management can be solved using methods such as economic order quantity models and newsboy models; nonlinear programming can be solved using methods such as gradient descent and Newton's method; and linear programming can be solved using methods such as the simplex method and interior point method.

[0066] Furthermore, through resource allocation optimization by the resource optimization agent, the order of individual satellite tasks or the allocated satellite resources in the single-satellite mission plan are adjusted to ensure that the execution order of single-satellite tasks matches the current resource status of the sub-satellites, reducing redundant actions used to secure satellite resources during the execution of the target mission. The single-satellite tasks are arranged according to the optimized order, and a mission execution plan is obtained based on the single-satellite tasks and their allocated satellite resources. It can be understood that by optimizing the resource allocation of sub-satellites, the utilization rate of existing satellite resources within the sub-satellites is improved, the total execution time of the mission execution plan is shortened, thereby effectively improving the execution efficiency of the target mission.

[0067] For example, in an on-orbit constellation comprising two sub-satellites, sub-satellites 7 and 9, when the target mission is a rapid-response remote sensing mission, optimization constraints may include a mission time window between 12:00 and 14:00 UTC to ensure sufficient illumination in the target area. Furthermore, optimization constraints may include a revisit interval for the sub-satellites not exceeding 30 minutes. The preset optimization objective may be to maximize the target mission completion rate while minimizing energy consumption. Resource assessment shows that sub-satellite 7 currently has 82% remaining battery power and a stable payload temperature, but requires attitude angle adjustment of 15° when performing the target mission; sub-satellite 9 currently has 72% remaining battery power, is in a rechargeable illumination period, and requires attitude angle adjustment of 5°. Through resource allocation optimization by the resource optimization agent, the resulting mission operation plan includes: sub-satellite 7 undertaking 70% of the remote sensing observation tasks within the target area, while the remaining 30% is a high-priority area within the target area, which will be observed by sub-satellite 9.

[0068] Reference Figure 7 As shown, in one embodiment of this application, the sub-satellite also includes an agent invocation model; each autonomous agent is invoked in the following manner: S510. Semantic parsing of agent invocation commands is performed through an agent invocation model to determine the intent of the agent invocation commands; wherein, the agent invocation commands include resource assessment commands and resource optimization commands issued by the sub-satellite when it receives a single-satellite mission plan, and the resource optimization commands are used to control the sub-satellite to optimize resource allocation.

[0069] S520. Based on the instruction intent, perform functional matching on the autonomous agents in the agent framework, determine the agent invocation plan that conforms to the instruction intent, and invoke the corresponding autonomous agent.

[0070] Specifically, the agent invocation model can be any type of large language model, possessing natural language processing capabilities to parse text-based agent invocation instructions and determine their intent. It should be noted that the agent invocation can be triggered when a resource assessment instruction is received and resource assessment of the subsatellite is required, or when a single-satellite mission plan is received and resource allocation optimization of the subsatellite is required. Therefore, the agent invocation instruction can be either a resource assessment instruction to trigger resource assessment or a resource optimization instruction to trigger resource allocation optimization.

[0071] Furthermore, the agent invocation model is located at the center of the multi-agent framework. Based on the instruction intent of the agent invocation command, it performs functional matching on the autonomous agents within the framework to determine one or more types of autonomous agents that match the instruction intent. When multiple types of autonomous agents are matched, the agent invocation model also arranges the invocation order of these agents based on the instruction intent, resulting in an agent invocation plan. Within the multi-agent framework, the corresponding autonomous agents are activated according to the agent invocation plan to invoke the activated autonomous agents to execute the corresponding tasks, ensuring that the subsatellite can correctly execute its target mission.

[0072] Reference Figure 8 As shown, in one embodiment of this application, the method further includes: S530 sends the on-orbit operating parameters to the multi-agent framework, controlling each autonomous agent to perceive and learn the on-orbit operating parameters to obtain the dynamic operating environment status.

[0073] S540. Update and adjust the behavioral strategy parameters of each autonomous agent according to the dynamic operating environment status, so that the updated and adjusted autonomous agents can adapt to the operating environment of the sub-satellite.

[0074] Specifically, the autonomous agents in the multi-agent framework all belong to embodied intelligence. These autonomous agents can perceive and learn based on environmental information, thereby adapting to changes in the environment and operating status of the subsatellite and adaptively executing target tasks.

[0075] Furthermore, the autonomous intelligent agent can learn by sensing the on-orbit operating parameters of the subsatellite. By learning the on-orbit operating parameters, the dynamic operating environment state of the subsatellite can be obtained. In some embodiments, the dynamic operating environment state includes the external environment state and the internal environment state of the subsatellite. The internal environment state can be obtained directly from the on-orbit operating parameters and represents the changes in the operating state of the subsatellite. The external environment state can be obtained from the subsatellite's sensors and represents the changes in the external environment in which the subsatellite is located.

[0076] Furthermore, each autonomous agent adaptively modifies its parameters based on the obtained dynamic operating environment state, adjusts its own behavioral strategy parameters, and updates the operating strategies of each autonomous agent. This allows the updated and adjusted autonomous agents to adapt to the operating environment of the subsatellite, improves the subsatellite's adaptability to the operating environment, and thus enhances the intelligence level of the subsatellite.

[0077] Reference Figure 9 As shown, in one embodiment of this application, the autonomous intelligent agent further includes a fault diagnosis intelligent agent and a fault repair intelligent agent; the method further includes: S550. Call the fault diagnosis agent to analyze the operational status of the sub-satellite's operation log and determine the fault problems that occurred in the sub-satellite during the autonomous process.

[0078] S560. Invoke the fault repair agent to repair the fault.

[0079] Reference Figure 10 As shown, the fault diagnosis agent and the fault repair agent are interconnected. The fault diagnosis agent sends the identified faults to the fault repair agent, enabling the fault repair agent to repair the corresponding faults. The fault diagnosis agent connects with other autonomous agents in the multi-agent framework to obtain the subsatellite's operational logs. It is understood that the subsatellite's operational logs may include the operational status of each autonomous agent.

[0080] Furthermore, the fault diagnosis agent acquires the subsatellite's operational logs through a multi-agent framework and interprets these logs using fault diagnosis tools to analyze the subsatellite's operational status and identify faults that occurred during its autonomous operation. Based on this, the fault diagnosis agent sends the identified faults to the fault repair agent. Upon receiving the faults, the fault repair agent invokes the corresponding fault repair tools to address the faults, ensuring the subsatellite can quickly return to normal operation.

[0081] In some embodiments, the autonomous agent also includes an emergency backup agent, which is invoked when the subsatellite encounters a major disaster or emergency to securely back up the data within the subsatellite, thereby reducing the impact of the emergency on the subsatellite and improving the stability and reliability of the subsatellite's operation.

[0082] In summary, this application also provides a multi-agent framework for subsatellites, the form of which can be referred to Figure 11 As shown, the multi-agent framework includes an instruction parsing agent, a state-aware agent, a resource assessment agent, a resource optimization agent, a fault diagnosis agent, a fault repair agent, and an emergency backup agent. The instruction parsing agent's input interface is connected to the subsatellite's instruction input interface; its output interface is connected to the state-aware agent's input interface; the state-aware agent's output interface is connected to the resource assessment agent's input interface; the resource assessment agent's output interface is connected to the input interface of the on-orbit constellation to which the subsatellite belongs; the resource optimization agent's input port is connected to the on-orbit constellation's output port; and its output interface is connected to each working module of the subsatellite. The fault diagnosis agent's input port is connected to other autonomous agents in the multi-agent framework, and its output port is connected to the fault repair agent's input port.

[0083] During operation, the instruction parsing agent receives resource assessment instructions and, upon receiving them, performs intent parsing to determine the parameter acquisition targets. The instruction parsing agent then sends the parameter acquisition targets to the state-aware agent, enabling the state-aware agent to acquire parameters based on these targets and obtain on-orbit operational parameters. The state-aware agent then sends the on-orbit operational parameters to the resource assessment agent, allowing the resource assessment agent to perform resource assessments on the subsatellites based on these parameters and obtain mission resource assessment results. The state-aware agent feeds back the mission resource assessment results to the on-orbit constellation. Based on the mission resource assessment results for each subsatellite, the on-orbit constellation updates the satellite situation map containing the operational status of all subsatellites and performs mission planning for the subsatellites based on the updated situation map, obtaining single-satellite mission plans for one or more subsatellites. The on-orbit constellation then sends the single-satellite mission plans to the corresponding subsatellites, where they are received by the resource optimization agent within the subsatellite. The resource optimization agent optimizes resource allocation for subsatellites based on the single-satellite mission plan, obtaining the subsatellite's mission operation plan. This plan is then sent to the various working modules of the subsatellites to control their execution of target tasks. The fault diagnosis agent obtains the subsatellite's operation logs and analyzes its operational status to identify faults encountered during autonomous operation. This agent sends the identified faults to the fault repair agent, which, upon receiving the faults, invokes the appropriate fault repair tools to fix them, ensuring the subsatellite can quickly return to normal operation. The emergency backup agent is invoked when a subsatellite encounters a major disaster or emergency to securely back up the data within the subsatellite, thereby reducing the impact of emergencies and improving the stability and reliability of subsatellite operation.

[0084] It should be noted that the multi-agent framework in this embodiment is designed with a system state space. This system state space can be a composite state space containing the internal state of the autonomous agent, the environmental perception state, and the task execution state. The internal state of the autonomous agent can be obtained from the logs of the autonomous agent or the agent calling the model; the environmental perception state can be obtained from the on-orbit operating parameters of the subsatellite; and the task execution state can be obtained from the output results of the autonomous agent. These three states are interconnected and mutually influential, together forming a complete composite state space, providing a comprehensive and accurate information foundation for the subsatellite's operation.

[0085] Furthermore, the system state space adopts a hierarchical architecture, including a global state and local states. The global state is synchronized to every node in the system state space in real time, while the local states are only used and updated in their respective nodes. During operation, every update and synchronization of the global state is recorded by the logs in the multi-agent framework, allowing developers to backtrack in case of failures and improving the fault tolerance of the subsatellite operation.

[0086] Furthermore, the autonomous agents within the multi-agent framework all operate using streaming information processing, meaning each agent can process and respond to input information in real time without waiting for all information to be collected. During operation, global state information provides crucial contextual support for streaming information processing. Based on this global state information, the autonomous agents can more accurately understand and judge the current input information, making timely decisions and actions. Ensuring timely decision-making, the integration of global state information makes the responses of each autonomous agent more aligned with the actual situation of the subsatellite, enhancing the subsatellite's on-orbit autonomy. In addition, streaming information processing improves the system's real-time performance and interactivity, providing users with a smoother and more efficient user experience.

[0087] It is understandable that this embodiment manages the subsatellite's operation through a multi-agent framework, reducing the involvement of ground personnel and effectively improving the subsatellite's intelligence level and on-orbit autonomy. Furthermore, this embodiment can also increase the number of autonomous agents to add new functions to the subsatellite, enhancing the flexibility and scalability of the subsatellite's functional framework.

[0088] This application also provides a satellite autonomy method driven by an embodied intelligent agent, applied to a sub-satellite containing the above-mentioned multi-agent framework and the on-orbit constellation to which the sub-satellite belongs; the method includes: S602. Receive the issued resource assessment instruction and send the resource assessment instruction to the instruction parsing agent.

[0089] S604. Call the instruction parsing agent to perform intent parsing on the resource assessment instruction, obtain the parameter acquisition object corresponding to the resource assessment instruction, and send the parameter acquisition object to the state-aware agent.

[0090] S606. Invoke the state-aware agent to collect parameters based on the parameter acquisition object, obtain the on-orbit operating parameters corresponding to the parameter acquisition object, and send the on-orbit operating parameters to the resource assessment agent.

[0091] S608. Call the resource assessment agent to perform resource assessment on the sub-satellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the sub-satellite, and feed back the mission resource assessment result to the on-orbit constellation.

[0092] S610. The on-orbit constellation updates the satellite situation map based on the mission resource assessment results; the on-orbit constellation performs mission planning for feasible satellites based on the updated situation map, obtains the single-satellite mission plan for feasible satellites, and feeds it back to the corresponding sub-satellites; wherein, the satellite situation map includes the operating status of each sub-satellite, and feasible satellites are sub-satellites that are capable of performing the target mission based on mission feasibility.

[0093] S612. Call the resource optimization agent to optimize the resource allocation of the sub-satellite based on the single-satellite mission plan, and obtain the mission operation plan of the sub-satellite to control the sub-satellite to achieve on-orbit autonomy.

[0094] S614. Call the fault diagnosis agent to analyze the operation status of the subsatellite's operation log, determine the fault problems that occurred in the subsatellite during the autonomous process, and send them to the fault repair agent.

[0095] S616. Invoke the fault repair agent to repair the fault.

[0096] S618. Invoke the emergency backup agent to perform a secure data backup of the subsatellite to avoid data loss in case of emergency.

[0097] Accordingly, please refer to Figure 12 This application provides a satellite autonomous device driven by an embodied intelligent agent. This device is applicable to any sub-satellite belonging to an in-orbit constellation. The sub-satellite has a multi-agent framework comprising multiple autonomous intelligent agents, which at least include a resource assessment agent and a resource optimization agent; including: The state awareness module 1210 is used to acquire the on-orbit operating parameters of the subsatellite in response to resource assessment commands regarding the target mission.

[0098] The resource assessment module 1220 is used to call the resource assessment agent to conduct resource assessment on the subsatellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the subsatellite, and feed back the mission resource assessment result to the on-orbit constellation; among which, the mission resource assessment result includes the mission feasibility of the subsatellite in the target mission.

[0099] The information interaction module 1230 is used to obtain the single-satellite mission plan about the target mission fed back to the sub-satellite by the on-orbit constellation when the mission feasibility indicates that the sub-satellite can perform the target mission.

[0100] The resource optimization module 1240 is used to call the resource optimization agent to optimize the resource allocation of the sub-satellite's on-orbit operation parameters based on the single-satellite mission plan, thereby obtaining the sub-satellite's mission operation plan and controlling the sub-satellite to achieve on-orbit autonomy.

[0101] In some alternative implementations, the state-aware module 1210 includes: The intent parsing unit is used to parse the resource evaluation instructions through the instruction parsing agent to obtain the parameter acquisition object corresponding to the resource evaluation instructions.

[0102] The parameter acquisition unit is used to invoke the state-aware intelligent agent to acquire parameters of the sub-satellite based on the parameter acquisition object, and obtain the on-orbit operating parameters corresponding to the parameter acquisition object.

[0103] In some alternative implementations, the resource assessment module 1220 includes: The standard determination unit is used to determine the mission execution standards corresponding to each on-orbit operating parameter based on the mission requirements of the target mission.

[0104] The feasibility assessment unit is used to call upon the resource assessment agent to conduct a feasibility assessment of the on-orbit operating parameters based on the mission execution standards, and to determine the mission feasibility of the subsatellite in the target mission.

[0105] The results generation unit is used to generate mission resource assessment results based on on-orbit operating parameters, mission feasibility, and subsatellite failure status.

[0106] In some alternative implementations, the resource optimization module 1240 includes: The constraint extraction unit is used to extract constraints based on the task requirements of the target task, and obtain the optimization constraints for resource allocation optimization.

[0107] The resource optimization unit is used to perform mission resource planning on the on-orbit operation parameters of the subsatellite based on preset optimization objectives and optimization constraints, and to obtain a mission operation plan. The mission operation plan includes a mission plan obtained by rearranging multiple single-satellite missions in the single-satellite mission plan, as well as the satellite resources allocated to the subsatellite when performing each single-satellite mission.

[0108] In some alternative implementations, the device further includes a system maintenance module, comprising: The instruction parsing unit is used to perform semantic parsing on agent invocation instructions through the agent invocation model to determine the instruction intent of the agent invocation instructions. Among them, the agent invocation instructions include resource assessment instructions and resource optimization instructions issued by the sub-satellite when it receives a single-satellite mission plan. The resource optimization instructions are used to control the sub-satellite to optimize resource allocation.

[0109] The invocation planning unit is used to perform functional matching of autonomous agents in the agent framework based on the instruction intent, determine the invocation plan of the agent that conforms to the instruction intent, and invoke the corresponding autonomous agent.

[0110] In some optional implementations, the system operation and maintenance module further includes: The environment perception unit is used to send the on-orbit operating parameters to the multi-agent framework, and control each autonomous agent to perceive and learn the on-orbit operating parameters to obtain the dynamic operating environment status.

[0111] The strategy adjustment unit is used to update and adjust the behavioral strategy parameters of each autonomous agent according to the dynamic operating environment state, so that the updated autonomous agents can adapt to the operating environment of the sub-satellite.

[0112] In some optional implementations, the system operation and maintenance module further includes: The fault diagnosis unit is used to call the fault diagnosis intelligent agent to analyze the operation status of the sub-satellite's operation log and determine the fault problems that occur in the sub-satellite during the autonomous process.

[0113] The fault repair unit is used to invoke the fault repair agent to repair faults.

[0114] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0115] In this embodiment, the satellite autonomous device driven by the embodied intelligent agent is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0116] Please see Figure 13 , Figure 13This is a schematic diagram of a computer device according to an embodiment of this application. As shown in the figure, the computer device includes one or more processors 10, a memory 20, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components communicate with each other using different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some alternative embodiments, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 13 Take a processor 10 as an example.

[0117] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0118] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0119] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0120] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0121] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0122] This application also provides a computer-readable storage medium. The methods described in this application can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code downloaded over a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and subsequently stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the methods shown in the above embodiments are implemented.

[0123] This application provides a computer program product including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the method of any embodiment of this application.

[0124] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and all such modifications and variations fall within the scope defined by the appended claims.

[0125] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0126] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0127] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0129] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0130] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0131] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0132] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0133] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

[0134] Although embodiments of this application have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of this application, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A satellite autonomy method driven by an embodied intelligent agent, characterized in that, The method is applicable to any subsatellite belonging to an in-orbit constellation, wherein the subsatellite has a multi-agent framework comprising multiple autonomous agents, the autonomous agents including at least a resource assessment agent and a resource optimization agent; the method includes: In response to a resource assessment instruction regarding the target mission, the on-orbit operating parameters of the subsatellite are acquired; The resource assessment agent is invoked to perform resource assessment on the subsatellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the subsatellite, and feed back the mission resource assessment result to the on-orbit constellation; wherein, the mission resource assessment result includes the mission feasibility of the subsatellite in the target mission; If the mission feasibility indicates that the sub-satellite can perform the target mission, obtain the single-satellite mission plan for the target mission fed back to the sub-satellite by the on-orbit constellation; The resource optimization agent is invoked to optimize the on-orbit operation parameters of the sub-satellite based on the single-satellite mission plan, thereby obtaining the mission operation plan of the sub-satellite and controlling the sub-satellite to achieve on-orbit autonomy.

2. The method according to claim 1, characterized in that, The autonomous intelligent agent further includes an instruction parsing intelligent agent and a state-aware intelligent agent; the process of obtaining the on-orbit operating parameters of the sub-satellite in response to a resource assessment instruction regarding the target mission includes: The instruction parsing agent performs intent parsing on the resource assessment instruction to obtain the parameter acquisition object corresponding to the resource assessment instruction; The state-aware intelligent agent is invoked to collect parameters of the sub-satellite based on the parameter acquisition object, thereby obtaining the on-orbit operating parameters corresponding to the parameter acquisition object.

3. The method according to claim 1, characterized in that, The step of invoking the resource assessment agent to perform resource assessment on the subsatellite based on the on-orbit operating parameters, and obtaining the mission resource assessment result of the subsatellite, includes: Based on the task requirements of the target task, determine the task execution standards corresponding to each of the on-orbit operation parameters; The resource assessment agent is invoked to conduct a feasibility assessment of the on-orbit operating parameters based on the task execution criteria, thereby determining the mission feasibility of the subsatellite in the target mission. The mission resource assessment results are generated based on the on-orbit operating parameters, mission feasibility, and the fault status of the sub-satellites.

4. The method according to claim 1, characterized in that, The process of invoking the resource optimization agent to optimize resource allocation for the sub-satellite's on-orbit operating parameters based on the single-satellite mission plan, thereby obtaining the sub-satellite's mission operation plan, includes: Constraints are extracted based on the task requirements of the target task to obtain the optimization constraints for resource allocation optimization; Based on the preset optimization objectives and the optimization constraints, the on-orbit operation parameters of the sub-satellite are planned for mission resources to obtain the mission operation scheme; wherein, the mission operation scheme includes a mission scheme obtained by rearranging multiple single-satellite missions in the single-satellite mission scheme, and the satellite resources allocated to the sub-satellite when performing each single-satellite mission.

5. The method according to claim 1, characterized in that, The sub-satellite also includes an agent invocation model; each autonomous agent is invoked in the following manner: The agent invocation model is used to perform semantic parsing on the agent invocation instructions to determine the instruction intent of the agent invocation instructions; wherein, the agent invocation instructions include the resource assessment instructions and the resource optimization instructions issued by the sub-satellite when it receives the single-satellite mission plan, and the resource optimization instructions are used to control the sub-satellite to optimize resource allocation; Based on the instruction intent, the autonomous agents in the agent framework are functionally matched to determine the agent invocation plan that conforms to the instruction intent, so as to invoke the corresponding autonomous agent.

6. The method according to any one of claims 1 to 5, characterized in that, The method further includes: The on-orbit operating parameters are sent to the multi-agent framework, and each autonomous agent is controlled to perceive and learn the on-orbit operating parameters to obtain the dynamic operating environment status. The behavioral strategy parameters of each autonomous agent are updated and adjusted according to the dynamic operating environment state, so that each autonomous agent can adapt to the operating environment of the sub-satellite.

7. The method according to any one of claims 1 to 5, characterized in that, The autonomous intelligent agent further includes a fault diagnosis intelligent agent and a fault repair intelligent agent; the method further includes: The fault diagnosis agent is invoked to analyze the operational status of the sub-satellite's operation log and determine the faults that occurred during the autonomous operation of the sub-satellite. The fault repair agent is invoked to repair the fault.

8. A satellite autonomous device driven by an embodied intelligent agent, characterized in that, Applicable to any subsatellite belonging to an in-orbit constellation, wherein the subsatellite is equipped with a multi-agent framework comprising multiple autonomous agents, wherein the autonomous agents include at least a resource assessment agent and a resource optimization agent; the device includes: The status awareness module is used to obtain the on-orbit operating parameters of the sub-satellite in response to resource assessment instructions regarding the target mission; The resource assessment module is used to invoke the resource assessment agent to perform resource assessment on the subsatellite based on the on-orbit operating parameters, obtain the mission resource assessment result of the subsatellite, and feed back the mission resource assessment result to the on-orbit constellation; wherein, the mission resource assessment result includes the mission feasibility of the subsatellite in the target mission; The information interaction module is used to obtain the single-satellite mission plan for the target mission fed back to the sub-satellite by the on-orbit constellation when the mission feasibility indicates that the sub-satellite can perform the target mission; The resource optimization module is used to call the resource optimization agent to optimize the resource allocation of the sub-satellite's on-orbit operation parameters based on the single-satellite mission plan, thereby obtaining the sub-satellite's mission operation plan and controlling the sub-satellite to achieve on-orbit autonomy.

9. A computer device, characterized in that, include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the method of any one of claims 1 to 7.

Images