A PDDL-based autonomous orbital threat avoidance mission planning method for spacecraft

Abstract

The present invention discloses a PDDL-based spacecraft orbit threat autonomous avoidance mission planning method, comprising: based on the evolution over time and dependency of behavioral actions of parallel subsystems of the spacecraft, using the PDDL language to construct a domain description of the spacecraft orbit threat autonomous avoidance mission planning; based on the threat perception process of the spacecraft and the environmental conditions in which the spacecraft is located, using the PDDL language to construct a problem description of the spacecraft orbit threat autonomous avoidance mission planning; combining the domain description and the problem description, generating a spacecraft orbit threat autonomous avoidance mission planning problem; solving the mission planning problem, and outputting a planning solution, wherein the planning solution is a state sequence set, representing the evolution of each subsystem from an initial state to a target state in a given planning interval; the present invention realizes the coordinated operation of the spacecraft in response to orbit threats when performing a given mission by autonomously generating a series of execution sequences of the avoidance mission.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a PDDL-based spacecraft orbit threat autonomous avoidance mission planning method. Background Art

[0002] With the increasing frequency of human space activities, orbital space is becoming increasingly congested, and the amount of space debris continues to increase, posing severe challenges to the safe and stable operation of spacecraft in orbit. At the same time, competition in space is intensifying, and new forms of orbital threats are constantly emerging. Faced with various orbital threats, such as collisions with space debris / dead satellites and interference from hostile satellites, existing response measures rely heavily on ground-based systems. These measures are subject to numerous spatiotemporal constraints on windows and arcs, long timelines between satellites and the ground, and a high degree of human error in operations and control. These measures result in poor timeliness in threat response and significant operational pressure, severely impacting the safety of spacecraft in orbit and business continuity. These measures make it difficult to adapt to the growing number of threats and the deteriorating space environment.

[0003] Planning is a branch of artificial intelligence (AI) that involves reasoning about how to develop a plan to achieve a given goal under given circumstances. In model-based AI planning, a planning system takes as input a description (or model) of an initial situation, the activities that can be changed, and the target conditions. It outputs a plan consisting of these activities that, when executed from the initial situation, will achieve the goal. Planning Domain Definition Language (PDDL) is a formal knowledge representation language used to express planning models and has become the standard input language for many planning systems.

[0004] Currently, researchers have used the PDDL language to build satellite mission planning models. For example, patent application publication number CN108416500A proposes a PDDL-based multi-satellite collaborative mission planning method. This approach focuses on imaging planning for Earth observation satellites, focusing on multi-satellite collaboration to complete Earth imaging missions and meet user imaging mission requests, rather than autonomous threat avoidance mission planning for a single satellite. Patent application publication number CN112241177A proposes a heuristic spacecraft mission planning method based on timeline state landmarks. This method constructs a heuristic information to guide the planning search direction and solve the mission planning results.

[0005] However, these plans are all autonomous mission planning for spacecraft for routine onboard missions, rather than autonomous avoidance mission planning for the entire threat avoidance. Conventional missions only require planning of certain subsystems of the spacecraft to derive the mission planning sequence. For example, the observation sequence of the camera is planned for earth observation; storage, communication, and sampling missions also only require planning of these subsystems.

[0006] Common countermeasures for various orbital threats, such as collisions with space debris / failed satellites and interference from hostile satellites, include orbital maneuvers, attitude maneuvers, and changes in operating status. Currently, a "large satellite-ground loop" threat management approach is typically employed, consisting of "ground-based orbit determination → telemetry downlink → threat assessment → decision-making and planning → command injection → on-orbit execution." Both threat assessment and decision-making are performed by ground personnel. Ground operators perform rendezvous event detection and collision risk assessment. They further determine the orbits of high-collision-risk targets, analyze orbit and covariance evolution, and provide avoidance decisions and maneuver recommendations, including collision avoidance maneuver planning that considers various constraints. However, this ground-based threat management approach presents numerous spatiotemporal constraints on windows and arcs, a long satellite-ground loop, and numerous human factors in operations and control. This results in poor threat management timeliness, high operational and control pressure, and often results in "overly blunt" or "overreacting" to threats, seriously impacting spacecraft operations. Summary of the Invention

[0007] The technical problem to be solved by the present invention is how to achieve autonomous avoidance of spacecraft in response to various orbital threats.

[0008] The present invention solves the above technical problems through the following technical means:

[0009] In a first aspect, the present invention proposes a PDDL-based spacecraft orbit threat autonomous avoidance mission planning method, the method comprising:

[0010] Based on the temporal evolution and dependency of the behaviors of each parallel subsystem of the spacecraft, the PDDL language is used to construct a domain description for the spacecraft orbit threat autonomous avoidance mission planning. The domain description includes the description of the subsystem action states, the description of the constraint relationships between states, and the description of each functional module in the software logic.

[0011] Based on the threat perception process of the spacecraft and the environmental conditions in which the spacecraft is located, the PDDL language is used to construct a problem description of the spacecraft orbit threat autonomous avoidance mission planning. The problem description includes an initial state and a target state. The initial state includes the current environmental conditions of the spacecraft and the current action states of each subsystem of the spacecraft. The target state is the action state that the spacecraft needs to maintain to complete threat avoidance.

[0012] Combining the domain description and the problem description, a spacecraft orbit threat autonomous avoidance mission planning problem is generated. The planning problem is a timeline-based four-tuple mission planning model, which is expressed as follows:

[0013] Ω= <V TL ,s I ,s g ,C>

[0014] Where V TL is a set of state timelines; s I It is V TL A set of initial states on s g is the target state of the mission planning, and C is the operational constraint of the mission planning;

[0015] Solve the task planning problem and output a planning solution, where the planning solution is a set of state sequences representing the evolution of each subsystem from an initial state to a target state in a given planning interval.

[0016] Furthermore, before constructing the domain description of the spacecraft orbital threat autonomous avoidance mission planning using the PDDL language based on the temporal evolution and dependency of the behaviors of the parallel subsystems of the spacecraft, the method further includes:

[0017] Analyze the structure, resources, subsystem functions, and constraints of the spacecraft to determine the inherent evolution of the actions in each subsystem and the logical relationship of information flow between nodes, and form the evolution and dependency of the actions of each parallel subsystem over time;

[0018] The structure of the spacecraft includes hardware and software components. The hardware components include a camera system, an attitude system, and a propulsion system. The software components include a threat perception part, an autonomous decision-making part, and an avoidance execution part.

[0019] The constraints include logical constraints, time constraints and parameter constraints.

[0020] Furthermore, the actions with the aforementioned logical constraints are subject to the mutual coupling between the subsystems of the spacecraft, and the actions have a fixed order of execution, which can be expressed as follows:

[0021]

[0022] In the formula, the set Indicates the corresponding action t i The set of spacecraft actions required in the logical constraints; N i represents the time coefficient; Indicates action Whether it is successfully added to the logic formula of the current part plan, where the element Indicates the completion of the action t i Action required When multiple consecutive occurrences occur Action When it has been added, otherwise

[0023] The time constraint is expressed as follows:

[0024] TR={eq,,i,m,mi,,i,d,di,,i,f,fi}

[0025] Suppose there are two actions X and Y, eq means that the two actions start and end at the same time, b means that action X comes before Y, bi means that action Y comes before X, m means that Y starts when X ends, mi means that X starts when Y ends, o means that X and Y overlap, oi means that Y and X overlap, d means that X contains X, di means that X contains Y, s means that X and Y start at the same time, si means that Y and X start at the same time, f means that X and Y end at the same time, fi means that Y and X end at the same time;

[0026] The parameter constraints are given corresponding parameters when a certain action of the spacecraft subsystem is running.

[0027] Furthermore, the evolution and dependency of the behaviors of the parallel subsystems of the spacecraft over time are as follows: Taking each subsystem of the spacecraft as a state variable, the four-tuple representation of the state variable is:

[0028] SV=(N,U,T,D)

[0029] Where N is the name of the state variable; U is the state value space of the state variable, which is a non-empty set; T is the transition rule or constraint between the state values of the state variable; D represents the duration of each state of the state variable;

[0030] State s is the value of the state variable at a certain point in time or a certain time interval. The four-tuple representation of state s is:

[0031] s=(N s ,t s ,t e ,d)

[0032] Where N s is the name of the value; s , t e , d are the start time, end time and duration of the state respectively; the triple representation of the operation constraint C is:

[0033] C= <s,s c ,τ>

[0034] Where s is the constrained state, s c is the state related to s in the operation constraint, τ(s,s c ) is s and s c The constraint time function between .

[0035] Furthermore, solving the task planning problem and outputting a planning solution includes:

[0036] The heuristic planning search algorithm is used to solve the task planning problem and output the planning solution π=STL0,STL1,…,STL k-1 ,STL k , STL k is a temporally continuous sequence of states, STL0, STL1, ..., STL k-1 ,STL k A set of state sequences that meet the planning problem objectives and synchronization constraints between states and are parallel in time.

[0037] Furthermore, the domain description of the spacecraft orbital threat autonomous avoidance mission planning constructed using the PDDL language includes:

[0038] Define the domain file name and elements in the domain.pddl file;

[0039] The PDDL language is used to define all actions of each system, the parameters required for each action to be executed, and the internal state transition rules between actions.

[0040] Furthermore, the domain.pddl file of the domain description corresponds to multiple problem.pddl files of the problem description.

[0041] Furthermore, the method further comprises:

[0042] According to the type of mission currently being performed by the spacecraft, different problem.pddl files and the domain.pddl file are selected for planning and solving in combination with the spacecraft's status and mission objectives to execute the corresponding type of mission, which includes conventional on-board missions and threat avoidance missions.

[0043] In a second aspect, the present invention further proposes a PDDL-based spacecraft orbit threat autonomous avoidance mission planning system, the system comprising:

[0044] The domain description module is used to construct a domain description of the spacecraft's orbital threat autonomous avoidance mission planning based on the temporal evolution and dependency relationships of the spacecraft's parallel subsystems' behaviors and actions. The domain description includes a description of the spacecraft's subsystem action states, the constraint relationships between states, and the description of each functional module in the software logic.

[0045] A problem description module is used to construct a problem description for the spacecraft's orbital threat autonomous avoidance mission planning using the PDDL language based on the spacecraft's threat perception process and the spacecraft's environmental conditions. The problem description includes an initial state and a target state, where the initial state includes the spacecraft's current environmental conditions and the current operational states of each spacecraft subsystem. The target state is the operational state that the spacecraft needs to maintain to complete threat avoidance.

[0046] The mission planning module is used to combine the domain description and the problem description to generate a spacecraft orbit threat autonomous avoidance mission planning problem. The planning problem is a timeline-based four-tuple mission planning model, which is expressed as follows:

[0047] Ω= <V TL ,s I ,s g ,C>

[0048] Where V TL is a set of state timelines; s I It is V TL A set of initial states on s g is the target state of the mission planning, and C is the operational constraint of the mission planning;

[0049] The solution module is used to solve the task planning problem and output a planning solution, where the planning solution is a set of state sequences representing the evolution of each subsystem from an initial state to a target state in a given planning interval.

[0050] In a third aspect, the present invention also proposes an autonomous avoidance mission planning device, which includes a memory and a processor; wherein the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the PDDL-based spacecraft orbit threat autonomous avoidance mission planning method as described above.

[0051] The advantages of the present invention are:

[0052] (1) The present invention constructs a domain description of autonomous avoidance mission planning based on the evolution and dependency of the behavioral actions of the parallel subsystems of the spacecraft over time, and constructs a problem description of autonomous avoidance mission planning based on the threat perception process of the spacecraft and the environmental conditions in which the spacecraft is located. Thus, the domain description and the problem description are combined to generate a four-tuple task planning model based on the timeline. The task planning model is solved to obtain a set of state sequences that meet the planning problem objectives and the synchronization constraints between states and are parallel in time, namely, a series of execution sequences of the avoidance task, thereby realizing the coordinated operation of the spacecraft in response to orbital threats when performing the predetermined mission; in the absence of human intervention, the logical execution order and execution time of a series of action sequences that need to be considered in the avoidance process are autonomously planned, thereby overcoming the disadvantage that some links in the threat avoidance process require the participation of ground personnel, thereby realizing safe and effective autonomous collision avoidance on board.

[0053] (2) When establishing a PDDL spacecraft orbital threat autonomous avoidance mission planning model, a domain.pddl file can contain multiple problem.pddl files. The spacecraft's state and constraints are described in a domain.pddl file, and different mission requirements are described in different problem.pddl files. During planning, different problem.pddl files are selected for planning and solving based on whether the spacecraft needs to perform a conventional onboard mission or a threat avoidance mission, the spacecraft's state, and the mission objectives. This allows the spacecraft to switch between conventional missions and threat avoidance when planning its autonomous orbital threat avoidance mission.

[0054] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 1 is a flowchart of a method for planning a spacecraft orbital threat autonomous avoidance mission based on PDDL, according to an embodiment of the present invention;

[0056] Figure 2 is a schematic diagram of the hardware structure of a spacecraft system according to an embodiment of the present invention;

[0057] Figure 3 Schematic diagram of elements that need to be considered in the spacecraft orbit threat autonomous avoidance mission planning model in an embodiment of the present invention;

[0058] Figure 4 This is a block diagram of the avoidance process architecture of the spacecraft orbital threat autonomous avoidance mission planning system in an embodiment of the present invention;

[0059] Figure 5This is a schematic diagram of the perception process of the spacecraft orbit threat autonomous avoidance mission planning system in an embodiment of the present invention;

[0060] Figure 6 1 is a schematic diagram of the results of the spacecraft orbital threat autonomous avoidance mission planning based on PDDL in an embodiment of the present invention;

[0061] Figure 7 It is a structural diagram of a PDDL-based spacecraft orbit threat autonomous avoidance mission planning system proposed in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0063] like Figure 1 As shown, the first embodiment of the present invention proposes a PDDL-based spacecraft orbit threat autonomous avoidance mission planning method, the method comprising the following steps:

[0064] S10. Based on the temporal evolution and dependency relationships of the behaviors of each parallel subsystem of the spacecraft, a domain description of the spacecraft orbital threat autonomous avoidance mission planning is constructed using the PDDL language. The domain description includes a description of the subsystem action states of the spacecraft, a description of the constraint relationships between states, and a description of each functional module in the software logic.

[0065] S20. Based on the spacecraft's threat perception process and the spacecraft's environmental conditions, construct a problem description for planning a spacecraft orbital threat autonomous avoidance mission using the PDDL language. The problem description includes an initial state and a target state. The initial state includes the spacecraft's current environmental conditions and the current operational states of each spacecraft subsystem. The target state is the operational state that the spacecraft needs to maintain to complete threat avoidance.

[0066] S30. Combining the domain description and the problem description, generate a spacecraft orbit threat autonomous avoidance mission planning problem. The planning problem is a timeline-based four-tuple mission planning model, which is expressed as follows:

[0067] Ω= <V TL ,s I ,s g ,C>

[0068] Where V TLis a set of state timelines; s I It is V TL A set of initial states on s g is the target state of the mission planning, and C is the operational constraint of the mission planning;

[0069] S40 , solving the task planning problem and outputting a planning solution, where the planning solution is a set of state sequences, representing the evolution of each subsystem from an initial state to a target state in a given planning interval.

[0070] This embodiment comprehensively considers factors such as spacecraft structure, threat avoidance mission requirements, equipment status, and spacecraft capabilities, analyzes the temporal evolution and dependencies of the behavioral actions of each parallel subsystem of the spacecraft, and uses the PDDL language to construct a domain description and problem description for the spacecraft orbital threat autonomous avoidance mission planning. By combining the domain description and problem description, a timeline-based four-tuple mission planning model is generated. The mission planning model is solved to obtain a set of state sequences that meet the planning problem objectives and synchronization constraints between states, and are parallel in time, namely, a series of execution sequences of the avoidance mission, thereby realizing coordinated operation of the spacecraft to respond to orbital threats when performing a predetermined mission.

[0071] In one embodiment, before constructing the domain description of the spacecraft orbital threat autonomous avoidance mission planning using the PDDL language based on the temporal evolution and dependency relationships of the behaviors of the parallel subsystems of the spacecraft, the method further includes the following steps:

[0072] Analyze the structure, resources, subsystem functions, and constraints of the spacecraft to determine the inherent evolution of the actions in each subsystem and the logical relationship of information flow between nodes, and form the evolution and dependency of the actions of each parallel subsystem over time;

[0073] The structure of the spacecraft includes hardware and software components. The hardware components include a camera system, an attitude system, and a propulsion system. The software components include a threat perception part, an autonomous decision-making part, and an avoidance execution part.

[0074] The constraints include logical constraints, time constraints and parameter constraints.

[0075] It should be noted that this embodiment analyzes the spacecraft orbital threat autonomous avoidance mission planning model, which requires expanding the existing spacecraft software and hardware to increase the orbital threat avoidance capability. Figure 2 As shown in the figure, some of the system hardware and software activities that need to be described for the spacecraft to carry out autonomous avoidance mission planning are as follows Figure 3 As shown, specifically:

[0076] (1) Photographic system: used to image the threat target and extract its features.

[0077] (2) Attitude system: used to achieve spacecraft attitude maintenance and attitude maneuvering.

[0078] (3) Propulsion system: provides the spacecraft with the power required for trajectory change and orbit correction, and completes the spacecraft maneuvering and braking control.

[0079] (4) Threat perception: measurement of the spacecraft's own status and environmental scenario, identification of threat characteristics, prediction of threat behavior, and determination of threat level.

[0080] (5) Autonomous decision-making: Based on the data obtained through perception, the spacecraft is able to make inferences and decisions on the specific evasive actions that the spacecraft should take to deal with threatening targets.

[0081] (6) Avoidance execution: The controller controls the onboard actuators to execute a predetermined action sequence, causing the spacecraft to change its attitude and trajectory to avoid the threat.

[0082] This example describes the spacecraft's configuration, resources, subsystem functions, and required constraints by comprehensively considering four factors: spacecraft structure, threat avoidance mission requirements, equipment status, and spacecraft capabilities. These constraints are specific to the actual spacecraft system and include logical constraints, time constraints, and parameter constraints.

[0083] First, the constraints that need to be met are:

[0084] Logical constraints: Affected by the equipment environment, the various systems of the spacecraft are coupled with each other, and the actions are executed in a fixed order. These constraints must be met when generating an action sequence. Actions with these logical constraints are affected by the coupling between the various subsystems of the spacecraft, and the actions are executed in a fixed order. The formula is expressed as:

[0085]

[0086] In the formula, the set Indicates the corresponding action t i The set of spacecraft actions required in the logical constraints; N i represents the time coefficient; Indicates action Whether it is successfully added to the logic formula of the current part plan, where the element Indicates the completion of the action t i Action required When multiple consecutive occurrences occur Action When it has been added, otherwise

[0087] Time constraints refer to the fact that some actions of a spacecraft require a certain amount of time to complete during actual operation. They can also specify the relationship between an action and a specific time point. The time constraint set can be expressed as the following thirteen relationships, expressed as follows:

[0088] TR={eq,,i,m,mi,,i,d,di,,i,f,fi}

[0089] Suppose there are two actions X and Y, eq means that the two actions start and end at the same time, b means that action X comes before Y, bi means that action Y comes before X, m means that Y starts when X ends, mi means that X starts when Y ends, o means that X and Y overlap, oi means that Y and X overlap, d means that X contains X, di means that X contains Y, s means that X and Y start at the same time, si means that Y and X start at the same time, f means that X and Y end at the same time, fi means that Y and X end at the same time;

[0090] The parameter constraint means that some spacecraft actions need to be given corresponding parameters during operation.

[0091] Secondly, in accordance with the mission requirements of autonomously avoiding threats to the spacecraft's orbit, this embodiment consists of a traditional attitude and orbit control system that completes the attitude and orbit control of the star and a key single unit for orbital threat perception and avoidance, forming an integrated "perception-decision-execution" control system. Situational perception is based on the collected sensor information and integrates the spacecraft's own attitude and orbit information and threat perception information. It includes two small modules: autonomous navigation and threat perception. Decision planning, based on the information from situational perception, is divided into two layers: behavioral decision-making and action planning to achieve autonomous avoidance of threats and output the desired attitude and orbit control instructions. Execution control executes the control instructions of the decision plan and realizes the attitude and orbit control of the spacecraft by controlling the switches of the attitude control thrusters and orbit control thrusters.

[0092] It should be noted here that the mission planning for spacecraft threat avoidance considered in this embodiment is a more complex process than conventional onboard mission planning. It requires consideration of more factors than earth observation missions. Earth observation only requires consideration of camera image acquisition, while threat avoidance requires consideration of more subsystems and constraints, such as several sensors for perception, attitude and trajectory change systems for avoidance actions, etc., and the use of PDDL for description and modeling is also more complex.

[0093] The planning of autonomous orbital threat avoidance missions for spacecraft is aimed at threatening spacecraft with collisions of space debris / failed satellites, and damage caused by hostile satellites. During the planning process, it is not targeted at one or several subsystems, but requires coordinated planning of the spacecraft's perception-decision-execution modules, involving many subsystems of the spacecraft. As shown in Table 1, 10 subsystems, 71 states, and more constraints are considered. The more factors are considered, the more complex they are, and the more difficult the planning is. Figure 4 As shown in the figure, the entire threat avoidance process, the inherent evolution of behavior in the system, and the logical relationship of information flow between nodes are described.

[0094] (1) In the perception part, firstly, four sensor devices, including optical camera, infrared camera, microwave radar and laser point cloud, are used to obtain the perception information of threat targets, and the accurate perception of threats in complex space environments is achieved through the complementarity of information between different devices; secondly, the four sensor information is fused through a multi-layer parallel network to enhance some incomplete information and separate the multi-layer image information from the long-distance abnormal behavior information; based on the multi-layer image information, the morphological features of the threat targets and payloads are extracted; on this basis, the abnormal behavior features of the threat targets and payloads are extracted, including the threat target ID, the number of payloads, the six-dimensional attitude / six orbit numbers, the current speed, the behavior semantics, the confidence level, etc.; finally, the morphological features, the abnormal behavior features and the spacecraft's own parameter data collected by the conventional sensors on board are combined for fusion reasoning to obtain a quantitative evaluation of the threat category and threat level of the target.

[0095] In the process of threat perception, according to the battlefield environment conditions such as front light, back light, and ground shadow, as well as the dynamic changes in the game process, four detection methods including visible light, infrared, laser, and microwave, and five different sensor combinations are used. With the cooperation of attitude control, the search and capture of the target, angle measurement, and distance measurement are completed, providing guidance information for attitude tracking, pointing, and maneuvering confrontation, such as Figure 5 shown.

[0096] It should be noted that this embodiment focuses on identifying threat targets, which differs from conventional tasks such as Earth observation and data transmission, which can be accomplished with just one camera or payload. However, addressing threat targets is more complex and requires detailed data on the threat target to accurately determine subsequent actions. Furthermore, the applicability of certain sensors in different external environments, such as frontal lighting, backlighting, and ground shadows, is considered, so a multi-sensor combination is used for threat target perception.

[0097] (2) In the decision-making part, firstly, the specific avoidance behavior that the spacecraft should take to deal with the threat target is inferred and decided by combining the threat category and threat level, long-distance abnormal behavior information, and abnormal behavior characteristics of the threat target and the payload; secondly, the long-distance abnormal behavior information collected by the perception module at the current moment, including speed, distance, azimuth, etc., is used to predict the future action of the threat target; finally, the future action of the threat target obtained by comprehensive prediction and the avoidance behavior obtained by inference and decision at the current moment are solved by the planning algorithm to obtain the optimal action sequence with the shortest avoidance time and the least fuel consumption.

[0098] (3) In the execution part, the controller first controls the onboard actuator to execute the predetermined action sequence, so that the spacecraft can change its attitude and orbit to avoid the threat; secondly, the result of the avoidance will affect the on-orbit operation status of the spacecraft (including normal operation and avoidance); finally, the spacecraft's own parameter data measured by the conventional onboard sensors are fed back to the threat level reasoning module of the perception model and the action sequence planning algorithm solution module of the decision model, so as to construct an integrated closed-loop system of "perception-decision-execution" for autonomous avoidance of spacecraft orbital threats.

[0099] By establishing a mission planning model for autonomous orbital threat avoidance for spacecraft systems, and taking into account the complex constraints and mutually coupled system state information of spacecraft systems, a multi-dimensional timeline is used to characterize the multiple parallel subsystems of the spacecraft. The coupling relationships of the timelines are used to describe the complex constraints of the system and the dependencies between the systems, forming a model architecture of multi-dimensional coupled timelines, which describes the temporal evolution and dependencies of the behaviors of each parallel subsystem.

[0100] The subsystems and subsystem states specifically involved in this embodiment are shown in Table 1 below.

[0101] Table 1

[0102]

[0103]

[0104] Then, the temporal evolution and dependency of the behaviors of each parallel subsystem of the spacecraft are:

[0105] Each subsystem of the spacecraft is regarded as a state variable, and the specific behavior of each subsystem is regarded as the specific state value of the state variable. The state variable SV (State Variable) is the abstract result of system knowledge during modeling and can be defined as a four-tuple.

[0106] SV=(N,U,T,D)

[0107] Where N is the name of the state variable; U is the value space of the state variable, that is, the value range of the state variable, which is a non-empty set; T is the transition rule or constraint between the values of the state variable; and D represents the duration of each value of the state variable. For example: N = {propulsion system}, U = {propulsion system ready, start propulsion, propulsion for a certain time, stop propulsion, propulsion system off}, T = {(propulsion for a certain time, stop propulsion), (propulsion system ready, start propulsion)...}, D (propulsion off) = {0, 35}, D (propulsion for a certain time) = {35, 50}.

[0108] The state s is the value of the state variable at a certain point in time or a certain time interval, which can be represented by a four-tuple:

[0109] s=(N s ,t s ,t e ,d)

[0110] Among them, N s is the name of the state; s , t e , d are the start time, end time and duration of the state, and are all expressed in the form of intervals, that is, t s =[s,s'],t e = [e,e'], d = [d,d'] is used to deal with the uncertainty problems that may occur during the execution process. For example, N s ={advance for a certain time}, t s =[34,36], t e =[49,51], d = [14,16].

[0111] The operation constraint C describes the necessary conditions for a state to be maintained on a state variable, described as a triple:

[0112] C= <s,s c ,τ>

[0113] Among them, s is the constrained state, s c is the state related to s in the operation constraint, τ(s,s c ) is s and s c The constraint time function between them. For example, the thrust state can only be performed if the fuel condition is met. There is an operational constraint c = <propulsion time, fuel capacity, equal>.

[0114] Furthermore, there are two types of operational constraints: internal constraints (constrained states with the same state timeline) and cross constraints (constrained states with different state timelines). For example, the relationship between the "Ready State" and the "Propel State" represents an internal constraint on the propulsion system timeline. The "Propel State" must be temporally constrained by the "Pointing" state and the "Fuel Consumption" state, the latter of which describes a cross constraint on the Propel State.

[0115] Due to the complex coupling relationships within and between the subsystems of the spacecraft, the system needs to satisfy the corresponding constraints when executing each action. During the mission planning process, all the constraints involved need to be considered to ensure that the planned results can be successfully executed. For example, some of the constraints involved are: 1) The system cannot take pictures during attitude maneuvers, 2) The attitude needs to be aligned with the target direction when performing propulsion and trajectory changes, 3) The camera needs to be calibrated before taking pictures, and 4) The propulsion system needs to be prepared for a period of time before propulsion and trajectory changes. Among them, 1 and 2 represent cross-constraints, and the constraint states involved are not in the same subsystem (not on the same timeline); 3 and 4 represent internal constraints, and the actions involved are in the same subsystem (same timeline).

[0116] The state timeline STL (State Time Line) is used to represent the changes in the values (i.e., states) of system state variables over time during the planning period. Each state timeline is a continuous sequence of states of the system during the planning period, i.e., STL = (s0, s1, ..., s n-1 ,s n ).

[0117] For example, in a propulsion system, the timeline represents a sequence of "off," "standby," "propulsion," and other states within a specific time interval. During the threat avoidance mission time interval, the relevant state sequence must be managed to fill the entire state timeline within the given time interval. The state timeline consists of a series of non-overlapping states. The length of the entire sequence represents the time interval, or planning interval, with no gaps between adjacent states.

[0118] It should be noted that the above analysis results in the evolution of the behavior and dependencies of the parallel subsystems of the spacecraft over time. This embodiment uses the PDDL modeling language for the parallel subsystems of the spacecraft to describe the internal state transition rules of each parallel subsystem and search for the internal state transition paths of each parallel subsystem. The internal transition rules represent the logical order of subsystem state changes and the constraints that need to be met. For example, a camera needs to be calibrated before taking pictures, and it cannot be taken directly after the camera is turned on. This is a state transition rule. Specifically, the PDDL language is used to describe the preconditions and subsequent impacts of these state transition rules. The modeling process using the PDDL modeling language includes constructing a domain description, constructing a problem description, generating a complete planning problem, and solving the planning problem.

[0119] (1) The process of constructing domain description is:

[0120] Construct a domain description domain.pddl, and construct a spacecraft orbit threat autonomous avoidance mission planning domain described in the PDDL language. The types include the subsystem status of the spacecraft, the constraint relationship between states, and various functional modules in the software logic.

[0121] First, define the domain file name and predicate elements in the domain.pddl file. Use the define keyword to define the domain name as taskplanning. By defining the different keywords contained in requirements, different feature types of the model description capability are declared. Among them, equality indicates that the equal sign is allowed to be used as a predicate; conditional-effects indicates that the action can have a lasting effect on the numerical variable; durative-actions indicates that the execution of the action is sustainable. The keyword types defines different types of objects, including satellite, direction, instrument name (instrument), and working model (mode). The predicate predicates defines the attributes of the object. The predicate for instrument i on the satellite is defined as "on_board?i-instrument?s-satellite", and the predicate for the pointing direction of the visible light camera is defined as "visible_pointing?i-instrument?d-direction". The header of the domain.pddl file for the entire spacecraft orbit threat autonomous avoidance mission planning is as follows:

[0122] (define(domain taskplanning)

[0123] (:requirements:strips:equality:typing:fluents:durative-actions:duration-inequalities:condi tional-effects:negative-preconditions)

[0124] (:types satellite direction instrument mode)

[0125] (:predicates(on_board?i-instrument?s-satellite)

[0126] (visible_supports?i-instrument?m-mode)

[0127] (visible_pointing?i-instrument?d-direction)

[0128] (visible_calibrated?i-instrument)

[0129] …)

[0130] Next, define all actions involved in the system, including the information in Table 1. durative-action defines the name of the action. For visible light camera image acquisition, define it as "visible_take_image." parameters describe the parameters required to execute the action. Image acquisition requires the satellite to take photos using an instrument in a fixed mode, pointing in the direction of the target. The required parameters are "?s-satellite?d-direction?i-instrument?m-mode." duration defines how long the action takes to complete.

[0131] Preconditions and effects contain the internal state transition rules between actions and reflect the operational constraints within the system. Preconditions indicate the conditions that must be met before an action can begin, while effects indicate the impact of an action's completion on other actions. For example, a visible light camera requires calibration before taking a picture. The precondition statement "over all(visible_aligned?s?i?d)" describes the internal transition rules between actions. A propulsion system must maintain its attitude and pointing direction during a trajectory change. The precondition statement "(at start(attitude_maneuver_aligned?d))" indicates the cross-constraints that must be met during the trajectory change. For example, attitude determination is required before an attitude maneuver. The precondition statement "(at start(have_attitude_determination))" indicates the internal constraints that must be met for the attitude maneuver. For example, the completion of image acquisition, trajectory change to a safe trajectory, and attitude change to the target attitude should be described in the effects "(at end(have_visible_image))", "(at end(have_propulsion_over?d))", and "(at end(attitude_pointing?i?d_new)". The following are examples of several key actions defined:

[0132] Define the variables, prerequisites, and subsequent impacts of image acquisition actions

[0133]

[0134]

[0135] Define the propulsion actions required for the trajectory change:

[0136]

[0137] Define the avoidance behavior decision actions in the software logic:

[0138]

[0139]

[0140] (2) Build the problem description problem.pddl:

[0141] Based on the spacecraft's threat perception process and battlefield environmental conditions such as direct lighting, backlighting, and ground shadows, a specific problem description is designed. It is divided into two parts: the initial state and the target state. The initial state includes the spacecraft's current environment and the current state of its subsystems. The target state includes a series of actions, such as attitude and trajectory changes, to avoid threats.

[0142] The initial state is declared with "init". The initial pointing of the global camera is represented in Star0 as "(global_pointing global Star0)", and the calibration time of the global camera is represented as "(=(global_calibration_timeglobal threattarget)5.9)". The following is part of the initial state of the mission plan:

[0143]

[0144] The target state is declared with goal. If the current spacecraft is in a direct sunlight environment, according to Figure 5 The perception process requires global optical cameras, lidar, and micro radar for image acquisition. If the resulting autonomous evasive action is a trajectory change, the aircraft must remain on a safe orbit (at_propulsion_safeorbit safeorbit). The target state for the entire mission planning is as follows:

[0145]

[0146]

[0147] (3) Generate a complete planning problem:

[0148] Based on the domain description of the spacecraft orbit threat autonomous avoidance mission planning domain.pddl and the specific problem description problem.pddl, including the initial state and target state of the spacecraft under different conditions, a complete planning problem is generated. The mathematical description form of the entire autonomous mission planning problem model is Ω= <V TL ,s I ,s g ,C>, where V TL Contains state variables SV and states s, s I , s g Both C are described in detail using the PDDL language and then solved.

[0149] (4) Solve the planning problem:

[0150] The system uses the Optic heuristic planning search algorithm developed by Andrew Coles' team at King's College London, compiled in C++ on Linux, to automatically solve planning problems and output complete planning steps. This guides the spacecraft to change its attitude and trajectory according to these steps to avoid threats.

[0151] Specifically, a heuristic planning search algorithm is used to solve the mission planning problem and output a planning solution. A feasible solution to the planning problem is a set of state timelines that contain a complex set of threat avoidance maneuvers for the spacecraft. The planning solution π is a set of timelines that describe the evolution of state variables from an initial state I to a target state G over a given planning interval H, where π = STL0, STL1, …, STL k-1 ,STL k , STL k is a temporally continuous sequence of states, STL0, STL1, ..., STL k-1 ,STL k A set of state sequences that meet the planning problem objectives and synchronization constraints between states and are parallel in time.

[0152] Figure 6 This is the planning result diagram for the spacecraft's orbital change action under backlight conditions when responding to orbital threats. In the planning results, the first column represents the start time of each action, the specific action name is contained in the parentheses, and the duration of each action is represented in the square brackets. When the action is started, the global camera and other devices are turned on at 0.000 seconds, the lidar and other devices begin image acquisition at 22.904 seconds, the decision on evasive behavior is made at 83.814 seconds, the orbital change to avoid the threat is made at 147.819 seconds, and the orbit change is completed at 225.821 seconds. The spacecraft is in a safe orbit and has achieved evasion of the threatening target. The entire planning result derives the action sequence between the spacecraft subsystems and software modules, thereby coordinating the series of actions of the spacecraft in responding to threats.

[0153] In one embodiment, the domain.pddl file of the domain description corresponds to multiple problem.pddl files of the problem description.

[0154] In one embodiment, the method further comprises:

[0155] According to the type of mission currently being performed by the spacecraft, different problem.pddl files and the domain.pddl file are selected for planning and solving in combination with the spacecraft's status and mission objectives to execute the corresponding type of mission, which includes conventional on-board missions and threat avoidance missions.

[0156] It should be noted that when establishing the PDDL spacecraft orbital threat autonomous avoidance mission planning model, a single domain.pddl file can contain multiple problem.pddl files. The spacecraft's state and constraints are described in one domain.pddl file, and different mission requirements are described in different problem.pddl files. During planning, different problem.pddl files are selected for planning and solving based on whether the spacecraft needs to perform conventional onboard missions or threat avoidance missions, the spacecraft's state, and the mission objectives. This allows the spacecraft to switch between conventional missions and threat avoidance when planning its autonomous orbital threat avoidance mission.

[0157] like Figure 7 As shown, the second embodiment of the present invention proposes a PDDL-based spacecraft orbit threat autonomous avoidance mission planning system, the system comprising:

[0158] A domain description module 10 is used to construct a domain description of the spacecraft orbital threat autonomous avoidance mission planning using the PDDL language based on the temporal evolution and dependency relationships of the behaviors of the spacecraft's parallel subsystems. The domain description includes a description of the spacecraft's subsystem action states, a description of the constraints between states, and a description of each functional module in the software logic.

[0159] A problem description module 20 is configured to construct a problem description for autonomous spacecraft orbital threat avoidance mission planning using the PDDL language based on the spacecraft's threat perception process and the spacecraft's environmental conditions. The problem description includes an initial state and a target state, wherein the initial state includes the spacecraft's current environmental conditions and the current operational states of each spacecraft subsystem, and the target state is the operational state that the spacecraft needs to maintain to complete threat avoidance.

[0160] The mission planning module 30 is used to combine the domain description and the problem description to generate a spacecraft orbit threat autonomous avoidance mission planning problem. The planning problem is a timeline-based four-tuple mission planning model, which is expressed as follows:

[0161] Ω= <V TL ,s I ,s g ,C>

[0162] Where V TL is a set of state timelines; s I It is V TL A set of initial states on s g is the target state of the mission planning, and C is the operational constraint of the mission planning;

[0163] The solution module 40 is used to solve the task planning problem and output a planning solution, which is a set of state sequences representing the evolution of each subsystem from an initial state to a target state in a given planning interval.

[0164] In one embodiment, the system further comprises an analysis module for:

[0165] Analyze the structure, resources, subsystem functions, and constraints of the spacecraft to determine the inherent evolution of the actions in each subsystem and the logical relationship of information flow between nodes, and form the evolution and dependency of the actions of each parallel subsystem over time;

[0166] The structure of the spacecraft includes hardware and software components. The hardware components include a camera system, an attitude system, and a propulsion system. The software components include a threat perception part, an autonomous decision-making part, and an avoidance execution part.

[0167] The constraints include logical constraints, time constraints and parameter constraints.

[0168] In one embodiment, the actions with the logical constraints are mutually coupled between the subsystems of the spacecraft, and the actions are executed in a fixed order, which can be expressed as follows:

[0169]

[0170] In the formula, the set Indicates the corresponding action t i The set of spacecraft actions required in the logical constraints; N i represents the time coefficient; Indicates action Whether it is successfully added to the logic formula of the current part plan, where the element Indicates the completion of the action t i Action required When multiple consecutive occurrences occur Action When it has been added, otherwise

[0171] The time constraint is expressed as follows:

[0172] TR={eq,,i,m,mi,,i,d,di,,i,f,fi}

[0173] Suppose there are two actions X and Y, eq means that the two actions start and end at the same time, b means that action X comes before Y, bi means that action Y comes before X, m means that Y starts when X ends, mi means that X starts when Y ends, o means that X and Y overlap, oi means that Y and X overlap, d means that X contains X, di means that X contains Y, s means that X and Y start at the same time, si means that Y and X start at the same time, f means that X and Y end at the same time, fi means that Y and X end at the same time;

[0174] The parameter constraints are given corresponding parameters when a certain action of the spacecraft subsystem is running.

[0175] In one embodiment, the evolution and dependency of the behaviors of the parallel subsystems of the spacecraft over time are as follows: each subsystem of the spacecraft is regarded as a state variable, and the quadruple representation of the state variable is:

[0176] SV=(N,U,T,D)

[0177] Where N is the name of the state variable; U is the state value space of the state variable, which is a non-empty set; T is the transition rule or constraint between the state values of the state variable; D represents the duration of each state of the state variable;

[0178] State s is the value of the state variable at a certain point in time or a certain time interval. The four-tuple representation of state s is:

[0179] s=(N s ,t s ,t e ,d)

[0180] Where N s is the name of the value; s , t e , d are the start time, end time and duration of the state respectively; the triple representation of the operation constraint C is:

[0181] C= <s,s c ,τ>

[0182] Where s is the constrained state, s c is the state related to s in the operation constraint, τ(s,s c ) is s and s c The constraint time function between .

[0183] In one embodiment, the domain description module 10 is specifically configured to:

[0184] Define the domain file name and elements in the domain.pddl file;

[0185] The PDDL language is used to define all actions of each system, the parameters required for each action to be executed, and the internal state transition rules between actions.

[0186] In one embodiment, the solution module 40 is specifically configured to:

[0187] The heuristic planning search algorithm is used to solve the task planning problem and output the planning solution π=STL0,STL1,…,STL k-1 ,STL k , STL k is a temporally continuous sequence of states, STL0, STL1, ..., STL k-1 ,STL k A set of state sequences that meet the planning problem objectives and synchronization constraints between states and are parallel in time.

[0188] In one embodiment, the domain.pddl file of the domain description corresponds to multiple problem.pddl files of the problem description.

[0189] In one embodiment, the system further includes a switching module, specifically configured to:

[0190] According to the type of mission currently being performed by the spacecraft, different problem.pddl files and the domain.pddl file are selected for planning and solving in combination with the spacecraft's status and mission objectives to execute the corresponding type of mission, which includes conventional on-board missions and threat avoidance missions.

[0191] It should be noted that other embodiments or implementation methods of the PDDL-based spacecraft orbit threat autonomous avoidance mission planning system of the present invention can refer to the above-mentioned method embodiments, which will not be repeated here.

[0192] In a third aspect, the present invention also proposes an autonomous avoidance mission planning device, which includes a memory and a processor; wherein the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the PDDL-based spacecraft orbit threat autonomous avoidance mission planning method as described in the first embodiment above.

[0193] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0194] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0195] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A PDDL-based spacecraft orbit threat autonomous avoidance mission planning method, characterized by: The structure of the spacecraft includes hardware and software. The hardware includes a camera system, an attitude system, and a propulsion system. The software includes a threat perception part, an autonomous decision-making part, and an avoidance execution part. The constraints that need to be met include logical constraints, time constraints, and parameter constraints. The method includes: Based on the temporal evolution and dependency of the behaviors of each parallel subsystem of the spacecraft, the PDDL language is used to construct a domain description for the spacecraft orbit threat autonomous avoidance mission planning. The domain description includes the description of the subsystem action states, the description of the constraint relationships between states, and the description of each functional module in the software logic. Based on the threat perception process of the spacecraft and the environmental conditions in which the spacecraft is located, the PDDL language is used to construct a problem description of the spacecraft orbit threat autonomous avoidance mission planning. The problem description includes an initial state and a target state. The initial state includes the current environmental conditions of the spacecraft and the current action states of each subsystem of the spacecraft. The target state is the action state that the spacecraft needs to maintain to complete threat avoidance. Combining the domain description and the problem description, a spacecraft orbit threat autonomous avoidance mission planning problem is generated. The planning problem is a timeline-based four-tuple mission planning model, which is expressed as follows: Oh= <V TL ,s I ,s g ,C> Where V TL is a set of state timelines; s I It is V TL A set of initial states on s g is the target state of the mission planning, and C is the operational constraint of the mission planning; Solve the task planning problem and output a planning solution, where the planning solution is a set of state sequences representing the evolution of each subsystem from an initial state to a target state in a given planning interval.

2. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 1, characterized in that: Before constructing the domain description of the spacecraft orbital threat autonomous avoidance mission planning based on the temporal evolution and dependency of the behaviors and actions of the parallel subsystems of the spacecraft using the PDDL language, the method further includes: Analyze the structural composition, resources, subsystem functions and constraints of the spacecraft, determine the inherent evolution laws of the action behaviors in each subsystem and the logical relationship of information flow between nodes, and form the evolution and dependency relationship of the behavior actions of each parallel subsystem over time.

3. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 2, characterized in that: The actions with the above logic constraints are subject to the mutual coupling between the subsystems of the spacecraft, and the actions have a fixed order of execution, which can be expressed as follows: In the formula, the set Indicates the corresponding action t i The set of spacecraft actions required in the logical constraints; N i represents the time coefficient; Indicates action Whether it is successfully added to the logic formula of the current part plan, where the element Indicates the completion of the action t i Action required When multiple consecutive occurrences occur Action When it has been added, otherwise The time constraint is expressed as follows: TR={eq,b,bi,m,mi,o,oi,d,di,s,si,f,fi} Suppose there are two actions X and Y, eq means that the two actions start and end at the same time, b means that action X comes before Y, bi means that action Y comes before X, m means that Y starts when X ends, mi means that X starts when Y ends, o means that X and Y overlap, oi means that Y and X overlap, d means that X contains X, di means that X contains Y, s means that X and Y start at the same time, si means that Y and X start at the same time, f means that X and Y end at the same time, fi means that Y and X end at the same time; The parameter constraints are given corresponding parameters when a certain action of the spacecraft subsystem is running.

4. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 1, characterized in that: The evolution and dependency of the behaviors of each parallel subsystem of the spacecraft over time are as follows: Taking each subsystem of the spacecraft as a state variable, the four-tuple representation of the state variable is: SV=(N,U,T,D) Where N is the name of the state variable; U is the state value space of the state variable, which is a non-empty set; T is the transition rule or constraint between the state values of the state variable; D represents the duration of each state of the state variable; State s is the value of the state variable at a certain point in time or a certain time interval. The four-tuple representation of state s is: s=(N s ,t s ,t e ,d) Where N s is the name of the value; s , t e , d are the start time, end time and duration of the state respectively; The triple representation of the operation constraint C is: C= <s,s c ,t> Where s is the constrained state, s c is the state related to s in the operation constraint, τ(s,s c ) is s and s c The constraint time function between .

5. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 1, characterized in that: Solving the task planning problem and outputting a planning solution includes: The heuristic planning search algorithm is used to solve the task planning problem and output the planning solution π=STL0,STL1,…,STL k-1 ,STL k , STL k is a temporally continuous sequence of states, STL0, STL1, ..., STL k-1 ,STL k A set of state sequences that meet the planning problem objectives and synchronization constraints between states and are parallel in time.

6. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 4, characterized in that: The domain description of the spacecraft orbital threat autonomous avoidance mission planning constructed using the PDDL language includes: Define the domain file name and elements in the domain.pddl file; Use PDDL language to define all actions of each system, the parameters required for each action to execute, and the internal state transition rules between actions; The problem description of using PDDL language to construct the spacecraft orbit threat autonomous avoidance mission planning includes: Define the initial state and target state in the problem.pddl file; The PDDL language is used to define the current environment of the spacecraft, the initial state of the subsystem, and the actions to achieve threat avoidance.

7. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 1, characterized in that: The domain.pddl file of the domain description corresponds to multiple problem description problem.pddl files.

8. The PDDL-based spacecraft orbit threat autonomous avoidance mission planning method according to claim 7, characterized in that: The method further comprises: According to the type of mission currently being performed by the spacecraft, different problem.pddl files and the domain.pddl file are selected for planning and solving in combination with the spacecraft's status and mission objectives to execute the corresponding type of mission, which includes conventional on-board missions and threat avoidance missions.

9. A PDDL-based spacecraft orbit threat autonomous avoidance mission planning system, characterized by: The structure of the spacecraft consists of hardware and software. The hardware consists of a camera system, an attitude system, and a propulsion system. The software consists of a threat perception part, an autonomous decision-making part, and an avoidance execution part. The constraints that need to be met include logical constraints, time constraints, and parameter constraints. The system includes: The domain description module is used to construct a domain description of the spacecraft's orbital threat autonomous avoidance mission planning based on the temporal evolution and dependency relationships of the spacecraft's parallel subsystems' behaviors and actions. The domain description includes a description of the spacecraft's subsystem action states, the constraint relationships between states, and the description of each functional module in the software logic. A problem description module is used to construct a problem description for the spacecraft's orbital threat autonomous avoidance mission planning using the PDDL language based on the spacecraft's threat perception process and the spacecraft's environmental conditions. The problem description includes an initial state and a target state, where the initial state includes the spacecraft's current environmental conditions and the current operational states of each spacecraft subsystem. The target state is the operational state that the spacecraft needs to maintain to complete threat avoidance. The mission planning module is used to combine the domain description and the problem description to generate a spacecraft orbit threat autonomous avoidance mission planning problem. The planning problem is a timeline-based four-tuple mission planning model, which is expressed as follows: Oh= <V TL ,s I ,s g ,C> Where V TL is a set of state timelines; s I It is V TL A set of initial states on s g is the target state of the mission planning, and C is the operational constraint of the mission planning; The solution module is used to solve the task planning problem and output a planning solution, where the planning solution is a set of state sequences representing the evolution of each subsystem from an initial state to a target state in a given planning interval.

10. An autonomous avoidance mission planning device, characterized in that: The device includes a memory and a processor; wherein the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the PDDL-based spacecraft orbit threat autonomous avoidance mission planning method as described in any one of claims 1-8.

Images