Method and device for spacecraft to avoid multi-level collision

Abstract

The application provides a spacecraft collision avoidance method and device, belonging to the field of space navigation technology, which can control the spacecraft to avoid a collision object while controlling the spacecraft to avoid other collision objects. The method first predicts the distance between the spacecraft and multiple potential encounter objects when they are closest, then screens the multiple potential encounter objects to obtain multiple collision objects, and finally determines a target collision optimization model for controlling the spacecraft to avoid the target collision object and other collision objects according to the control complexity of the spacecraft avoiding the multiple collision objects.

Description

Technical Field

[0001] This invention relates to the field of spaceflight technology, and in particular to a method and apparatus for spacecraft to avoid multi-stage collisions. Background Technology

[0002] During a spacecraft's mission, as it operates along its designated orbit, it encounters collisions with space debris, defunct satellites, and other orbital targets. Furthermore, with technological advancements, various satellite constellations (such as Starlink) are deployed in space, making orbital space increasingly congested. Consequently, the likelihood of a spacecraft colliding with multiple objects in a short period (i.e., multi-stage collisions) increases. Therefore, spacecraft control is necessary to adjust its orbit, ensuring collisions are avoided and allowing the spacecraft to successfully complete its mission.

[0003] However, existing spacecraft collision avoidance methods can only control the spacecraft to avoid collisions with one object at a time. In the process of controlling the spacecraft to avoid one object, there is a risk that the spacecraft will collide with another object. Summary of the Invention

[0004] This invention proposes a method and apparatus for spacecraft to avoid multi-stage collisions, which can simultaneously control the spacecraft to avoid one collision object while controlling the spacecraft to avoid other collision objects.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a method for a spacecraft to avoid multi-stage collisions, comprising: predicting the closest distance between the spacecraft and each of a plurality of potential encountering objects; then, selecting a plurality of collision objects from the plurality of potential encountering objects; wherein the plurality of collision objects are potential encountering objects whose closest distance to the spacecraft is less than a distance threshold; and determining a target collision optimization model from a plurality of collision optimization models based on the control complexity of the spacecraft avoiding the plurality of collision objects; wherein the target collision optimization model is used to control the spacecraft to avoid the target collision object among the plurality of collision objects, and to control the spacecraft to avoid other collision objects among the plurality of collision objects during the process of avoiding the target collision object.

[0007] This invention provides a method for spacecraft to avoid multi-stage collisions. The method predicts the closest distance between the spacecraft and each of multiple potential encountering objects, and then filters out multiple collision targets from these potential encountering objects based on these distances. Next, based on the control complexity of the spacecraft avoiding these multiple collision targets, a target collision optimization model is selected for the target collision target, and the spacecraft is controlled to avoid other collision targets while avoiding the target collision target. Therefore, this invention considers the collision avoidance of other collision targets besides the target collision target during the collision avoidance process, thus enabling the spacecraft to avoid collisions with other collision targets while simultaneously controlling the spacecraft to avoid one collision target.

[0008] In one implementation of the first aspect, the control complexity of the spacecraft avoiding multiple collision objects is analyzed based on the time when the spacecraft is closest to each of the multiple collision objects, the collision probability of the spacecraft with each of the multiple collision objects, and the number of multiple collision objects.

[0009] The control complexity satisfies the following formula:

[0010]

[0011] Where c represents the control complexity, and n represents the number of collision objects for the spacecraft. c represents the average probability of a spacecraft colliding with multiple objects. vt This indicates the dispersion of the times when a spacecraft is closest to multiple colliding objects, and σ t The standard deviation, μ, represents the time when a spacecraft is closest to multiple colliding objects. t This represents the average time when a spacecraft comes closest to multiple colliding objects.

[0012] In one implementation of the first aspect, a target collision optimization model is determined from multiple collision optimization models based on the control complexity of the spacecraft avoiding multiple collision objects. This includes: judging the control complexity and the number of multiple collision objects, and selecting the target collision optimization model. When the number of multiple collision objects is less than a number threshold, a collision optimization model using a Monte Carlo tree search algorithm with continuous action space is selected. When the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is less than the complexity threshold, a collision optimization model using a cross-entropy algorithm with continuous action space is selected. When the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is greater than or equal to the complexity threshold, a collision optimization model using a near-end strategy optimization algorithm is selected.

[0013] In one implementation of the first aspect, the collision probability p satisfies:

[0014]

[0015] Where exp(*) represents the exponential function, R1 represents the component of the distance between the spacecraft and the collision target at their closest point on the R-axis of the star-based orbital coordinate system, S1 represents the component of the distance on the S-axis of the star-based orbital coordinate system, W1 represents the component of the distance on the W-axis of the star-based orbital coordinate system, and σ R σ represents the joint variance of the spacecraft's position component on the R-axis of the star-based orbital coordinate system and the position component of the colliding object on the R-axis of the star-based orbital coordinate system. SW r represents the joint variance of the spacecraft's position components in the SW plane of the star-based orbital coordinate system and the position components of the colliding object in the SW plane of the star-based orbital coordinate system. A This represents the combined radius of the spacecraft and the object it collides with, which is the sum of the effective radius of the spacecraft and the effective radius of the object it collides with. The effective radius refers to half of the maximum size of the space object.

[0016] In one implementation of the first aspect, the method further includes: removing space objects from a plurality of space objects of the spacecraft that satisfy any of the following conditions: Condition 1: The apogee altitude of the colliding object is less than the perigee altitude of the spacecraft's orbit. Condition 2: The perigee altitude of the colliding object is greater than the perigee altitude of the spacecraft's orbit. Then, space objects whose time difference between their intersection points is less than a time threshold are identified as potential rendezvous objects of the spacecraft. Here, the time difference between their intersection points is the time difference between the moment the spacecraft is at the orbital intersection point and the moment the colliding object is at the orbital intersection point, and the orbital intersection point is the intersection of the spacecraft's orbit and the colliding object's orbit.

[0017] In one implementation of the first aspect, multiple collision optimization models are pre-trained based on multi-level collision scenarios; during the pre-training process, the loss function satisfies the following formula;

[0018] LOSS=k1·∑ i∈H p i +k2·f+k3·r

[0019] Where LOSS represents the loss value, k1 represents the negative weighting coefficient of the collision probability, and ∑ i∈H p i The table represents the sum of collision probabilities between the spacecraft and multiple objects, where k2 represents the fuel negative weighting coefficient, and f represents the fuel consumption. k4 represents the negative weighting coefficient for acceleration. denoted by k, where k3 represents the orbital negative weighting coefficient, and r represents the orbital offset of the spacecraft before and after the maneuver.

[0020] In one implementation of the first aspect, the method further includes: determining the maneuvering control variables for controlling the spacecraft based on a target collision optimization model; and using the maneuvering control variables to control the spacecraft to avoid the target collision object.

[0021] Secondly, this invention provides a device for spacecraft to avoid multi-stage collisions, including a prediction module, a filtering module, and a determination module. The prediction module is used to predict the closest distance between the spacecraft and each of a plurality of potential encountering objects. The filtering module is used to filter out multiple collision objects from the plurality of potential encountering objects; the multiple collision objects are potential encountering objects whose closest distance to the spacecraft is less than a distance threshold. The determination module is used to determine a target collision optimization model from multiple collision optimization models based on the control complexity of the spacecraft avoiding multiple collision objects. The target collision optimization model is used to control the spacecraft to avoid the target collision object, and to control the spacecraft to avoid other collision objects among the multiple collision objects during the process of avoiding the target collision object.

[0022] Thirdly, the present invention provides an electronic device including a processor and a memory coupled to the processor; the memory is used to store computer instructions, and when the electronic device is running, the processor executes the computer instructions stored in the memory to cause the electronic device to perform the method as described in the first aspect above or any implementation thereof.

[0023] Fourthly, the present invention provides a computer-readable storage medium including computer program instructions that, when executed by a computer, cause the computer to perform the method described in the first aspect above or any implementation thereof.

[0024] Fifthly, the present invention provides a computer program product, including computer program instructions that, when executed on a computer, cause the computer to perform the method described in the first aspect above or any implementation thereof.

[0025] The technical effects corresponding to the second to fifth aspects and their possible implementations can be referred to the above description of the technical effects of the first aspect and its possible implementations, and will not be repeated here. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the six orbital elements provided in the embodiments of this application;

[0027] Figure 2 This is one of the schematic diagrams of a method for a spacecraft to avoid multi-stage collisions provided in the embodiments of this application;

[0028] Figure 3 This is one of the schematic diagrams of a multi-level collision scenario provided in the embodiments of this application;

[0029] Figure 4 This is the second schematic diagram of a multi-level collision scenario provided in the embodiments of this application;

[0030] Figure 5 This is the third schematic diagram of a multi-level collision scenario provided in the embodiments of this application;

[0031] Figure 6 This is the fourth schematic diagram of a multi-level collision scenario provided in the embodiments of this application;

[0032] Figure 7 This is a second schematic diagram of a method for a spacecraft to avoid multi-stage collisions provided in the embodiments of this application;

[0033] Figure 8 This is a schematic diagram of the area where the motion trajectory of the remaining spatial objects after removal is located, provided in an embodiment of this application.

[0034] Figure 9 This is a second schematic diagram of a method for a spacecraft to avoid multi-stage collisions provided in the embodiments of this application;

[0035] Figure 10 This is a schematic diagram of the star-based orbital coordinate system provided in the embodiments of this application;

[0036] Figure 11 This is a schematic diagram of a device for avoiding multi-stage collisions in a spacecraft, provided in an embodiment of this application. Detailed Implementation

[0037] In the specification and claims of this invention, the terms "first" and "second," etc., are used to distinguish different objects, rather than to describe a specific order of objects.

[0038] In the embodiments of this application, "and / or" indicates a relationship between objects. For example, A and / or B can represent the following three situations: A exists alone, B exists alone, and A and B exist simultaneously.

[0039] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0040] In the description of this invention, unless otherwise stated, "a plurality of" means two or more. For example, a plurality of potential intersection objects means two or more potential intersection objects.

[0041] The method and apparatus provided in this application relate to spacecraft collision avoidance control and can be used to control spacecraft to avoid multi-stage collisions. Specifically, based on the control complexity of a spacecraft avoiding multiple collision objects, a target collision optimization model is determined to control the spacecraft to avoid the target collision object. Furthermore, while avoiding the target collision object, the spacecraft simultaneously avoids other collision objects, thereby controlling the spacecraft to avoid multi-stage collisions.

[0042] The following is a description of the technical terms used in the embodiments of this application.

[0043] 1. Six orbital elements

[0044] The six orbital elements can be simply understood as parameters governing the orbit of a spacecraft (such as a satellite) around a central celestial body (such as Earth). These orbital elements can be used to describe the spacecraft's motion in space. For example, for a satellite, its position, orbital shape, and velocity can be precisely described using its six orbital elements.

[0045] In this embodiment of the application, the orbital elements of the satellite may include the semi-major axis a, eccentricity e, orbital inclination i, mean anomaly M, right ascension of the ascending node Ω, and argument of perigee ω.

[0046] Here, the semi-major axis 'a' refers to half of the major axis of the satellite's orbit, which is used to describe the size of the satellite's orbit.

[0047] Eccentricity e refers to the eccentricity of a satellite orbit, which is used to measure the degree of deviation from the orbit.

[0048] Orbital inclination i refers to the dihedral angle between the satellite's orbit and the equatorial plane of the central celestial body (e.g., ...). Figure 1 The orbital inclination angle i).

[0049] The mean anomaly angle M refers to the angle at which a satellite moves along its orbit, starting from the pericenter (the point on the satellite's orbit closest to the central celestial body). The mean anomaly angle M is used to describe the satellite's position on its orbit.

[0050] refer to Figure 1 The right ascension of the ascending node Ω refers to the angle within the equatorial plane of the central celestial body, from the vernal equinox V eastward to the ascending node B (the point where the satellite's orbit crosses the equator from south to north).

[0051] Continue to refer to Figure 1 The perigee argument ω refers to the angle within the satellite's orbital plane from the ascending node B to the perigee P.

[0052] The semi-major axis a and eccentricity e are used to describe the size and shape of the orbit; the orbital inclination i, the right ascension of the ascending node Ω, and the argument of perigee ω are used to describe the position of the satellite orbit; and the mean perigee M is used to describe the position of the spacecraft in the orbit.

[0053] With the development of space exploration, the number of satellites deployed in space is increasing, making it more likely that spacecraft will collide with multiple objects in a short period of time (i.e., multi-stage collisions). However, existing methods for controlling spacecraft to avoid collisions can only avoid one collision object at a time. In particular, while controlling the spacecraft to avoid one collision object, there is a risk that the spacecraft will collide with another collision object. This application provides a method and apparatus for spacecraft to avoid multi-stage collisions. Based on the control complexity of spacecraft avoiding multiple collision objects, a target collision optimization model is selected for the target collision object among the multiple collision objects. Furthermore, while controlling the spacecraft to avoid the target collision object, other collision objects among the multiple collision objects are also avoided. Therefore, this application can simultaneously control the spacecraft to avoid one collision object while preventing it from colliding with other collision objects.

[0054] For example, the spacecraft avoidance of multi-stage collisions method provided in this embodiment of the invention can be executed by an electronic device with processing capabilities, such as a computer or server. Taking a computer as an example, the hardware components of the computer may include: a processor, memory, a network interface, a user interface, a communication bus, etc.

[0055] The processor is used to control electronic devices to perform related processing and calculation tasks. The processor may include a central processing unit (CPU) or other processors. The processor may be single-core or multi-core. For example, the processor may include multiple CPUs.

[0056] Memory is used to store computer instructions and related data. Memory can be random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical storage, magnetic disk storage media, or other magnetic storage devices, or any other medium capable of storing program code or data accessible by a computer. Optionally, memory can be integrated into the processor, or it can be independent of the processor.

[0057] A network interface is used for communication between a computer and other devices or communication networks. A network interface can be a transceiver with transmit and receive capabilities. Optionally, a network interface may include standard wired interfaces or wireless interfaces (such as Wi-Fi interfaces, Bluetooth interfaces, and 5G interfaces).

[0058] The communication bus is used to enable communication between different components. For example, the processor, memory, network interface and user interface mentioned above can be interconnected through the communication bus.

[0059] The user interface may include a display screen and an input unit (such as a keyboard). Optionally, the user interface may also include a standard wired interface or a wireless interface.

[0060] Those skilled in the art will understand that the computer described above may include more or fewer components, or combine certain components, or have different component arrangements; the embodiments of this application do not limit this.

[0061] like Figure 2 As shown, the spacecraft avoidance of multi-stage collisions provided in this application embodiment includes S101-S103.

[0062] S101, Predict the closest distance between the spacecraft and each of the multiple potential rendezvous objects;

[0063] In this embodiment of the application, the motion state of the spacecraft in its own orbit and the motion state of the multiple potential rendezvous objects in their own orbits are simulated over a future period of time by using the six orbital elements of the spacecraft and the six orbital elements of multiple potential rendezvous objects; at the same time, the closest distance between the spacecraft and each of the multiple potential rendezvous objects is recorded and filtered.

[0064] In one application scenario, the positions of the spacecraft and multiple potential rendezvous objects in the future are calculated by using the six orbital elements of the spacecraft and the six orbital elements of multiple potential rendezvous objects. Then, the distance between the spacecraft and each of the multiple potential rendezvous objects is obtained by subtracting the positions of the spacecraft and the multiple potential rendezvous objects.

[0065] The calculation formula for determining the position of the spacecraft / potential rendezvous object in the J2000 coordinate system using the six orbital elements of the spacecraft / potential rendezvous object is as follows;

[0066]

[0067] in, r x =rcosf,r y =rsinf,

[0068] S102. From multiple potential collision objects, multiple collision objects of the spacecraft are selected; the multiple collision objects are the potential collision objects that are closest to the spacecraft when the distance is less than a distance threshold.

[0069] For example, the aforementioned distance threshold can be 10km, or other reasonable values; this application embodiment does not limit it.

[0070] S103. Based on the control complexity of a spacecraft avoiding multiple collision objects, determine the target collision optimization model from multiple collision optimization models;

[0071] Among them, the target collision optimization model is used to control the spacecraft to avoid the target collision object among multiple collision objects, and to control the spacecraft to avoid other collision objects among multiple collision objects during the process of avoiding the target collision object;

[0072] In one implementation, S103 includes: judging the control complexity and the number of multiple collision objects, and selecting a target collision optimization model;

[0073] Specifically, when the number of multiple colliding objects is less than the number threshold, the collision optimization model of the Monte Carlo tree search algorithm in the continuous action space is used as the target collision optimization model.

[0074] The collision optimization model using the Monte Carlo tree search algorithm in the continuous action space described above is a decision-time planning algorithm that uses Monte Carlo simulation for the search. In this embodiment, a decision tree is constructed in which each node represents a spacecraft state (e.g., node A, node B), and the connections between nodes represent maneuver control quantities (e.g., the maneuver control quantities required for the spacecraft to move from the state represented by node A to the state represented by node B).

[0075] Understandably, the Monte Carlo tree search algorithm for the continuous action space described above enables the spacecraft to avoid collisions with target objects through multiple iterations; for example, the process of each iteration is as follows;

[0076] 1. Selection: Starting from the root node, select an action based on the probability distribution and move down the tree; the selection process can be combined with the Upper Confidence Bound (UCB) strategy to balance exploration and exploitation;

[0077] 2. Expansion: Under the selected action, expand new nodes to represent the new state of the spacecraft;

[0078] 3. Simulation: Starting from the new node, a Monte Carlo simulation is performed to simulate the series of state transitions of the spacecraft after performing the action, until a target state is reached (e.g., the spacecraft avoids the target collision object).

[0079] 4. Backpropagation: The simulation results (such as rewards or costs) are backpropagated to each node in the tree to update the statistics of each node (such as average reward or number of visits).

[0080] When the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is less than the complexity threshold, the collision optimization model using the cross-entropy algorithm with continuous action space is selected as the target collision optimization model.

[0081] The aforementioned Cross-Entropy Algorithm (CEM) of Continuous Action Space is a Monte Carlo method used to optimize roles and sample maneuver control quantities based on importance. The goal of the CEM is to minimize the cross-entropy between the data distribution obtained from random sampling and the actual data distribution, i.e., to minimize the relative entropy (Kullback-Leibler Divergence), and to make the sampled distribution as similar as possible to the actual distribution. In this embodiment, the collision optimization model using the CEM of Continuous Action Space iteratively improves the probability distribution of maneuver control quantities corresponding to maneuver actions to find the optimal strategy. For example, the process of each iteration is as follows.

[0082] 1. Initialization: Define an initial probability distribution (e.g., a normal distribution) whose mean and variance are based on prior knowledge of the problem or random initialization;

[0083] 2. Sample collection: Draw a large number of samples from the current probability distribution, which represent possible actions or strategies;

[0084] 3. Performance Evaluation: The performance of each sample is evaluated through simulation or actual execution, usually by calculating cumulative rewards;

[0085] 4. Sample selection: Select the top small subset of samples with the best performance. The actions or strategies corresponding to these samples are considered to be approximations of the current optimal solution.

[0086] 5. Update parameters: Update the parameters of the probability distribution (e.g., the mean and variance of a Gaussian distribution) based on the selected samples, so that the distribution is more concentrated on the best samples.

[0087] When the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is greater than or equal to the complexity threshold, the collision optimization model using the near-end strategy optimization algorithm is selected.

[0088] The aforementioned proximal policy optimization algorithm (also known as the PPO algorithm) is a deep reinforcement learning algorithm based on a policy-value framework (Actor-Critic). In the Actor-Critic framework, two neural networks, an Actor and a Critic, are configured. In this embodiment, the Actor network of the Actor-Critic framework is used to determine the optimal avoidance action given the current state, while the Critic network is used to estimate the value of the current policy action. The PPO algorithm is based on the policy gradient method, which, after sampling data through interaction with the environment, uses stochastic gradient ascent to optimize an alternative objective function, thereby improving the policy.

[0089] Specifically, the implementation process of the PPO algorithm is as follows;

[0090] 1. Initialize the network: Define two neural networks: Actor and Critic;

[0091] The Actor network outputs the probabilities of taking various possible actions in a specific state, while the Critic network outputs the state value estimate under the current policy. The policy is improved by optimizing the Actor network, while the Critic network is used to guide and evaluate the improvement.

[0092] 2. Interaction and Data Collection: By executing actions generated by the Actor network and interacting with the environment, a series of experience data are collected, including state, action, reward, and next state;

[0093] 3. Calculate the dominance function using TD error: Calculate the dominance function A using Critic's data. t This measures the additional value of taking an action in a specific state compared to the average strategy.

[0094] Assume r t The instant reward obtained at time t, V(s) t ) is state s t The value function estimation then yields the TD error δ. t and the dominant function A t It can be defined as (γ is the discount factor, used to measure the importance of future rewards): A t ≈δ t =r t +γV(s t+1 )-V(s t );

[0095] 4. Calculate the objective function;

[0096] By using the importance sampling ratio r t (s,a) and the Clipping method are used to assist in calculating the objective policy function L. clipand the target value function L v To update strategies and value networks;

[0097] Among them, the importance sampling ratio r t (s,a') is: π θ (a|s) represents the probability of performing action a' in state s under the new policy, π θold (a'|s) represents the probability of performing action a' in state s under the old policy;

[0098] The objective value function Lv is: L v =(V(s) t )-A t ) 2 ; where V(s) t ) is state s t Value function estimation, A t It is the dominant function;

[0099] For the objective value function L v The clipping method is used to obtain the clipped target policy function L. clip for:

[0100] L clip =E t [min(r t (θ)A t ,clip(r t (θ), 1-ε, 1+ε)A t )]

[0101] The above objective policy function L clip This ratio is limited to the range of 1-ε to 1+ε, where ε is a small positive number used to control the range of shearing; the clip method is introduced to constrain the step size of policy updates, ensuring the stability of the updates and preventing the new policy from deviating too far from the old policy; r t (θ) represents the ratio of the old to the new strategies; E t [·] represents the expectation of time step t;

[0102] 5. Optimize the neural network: Calculate the gradient at each gradient using the gradient ascent method (i.e., ...). and The above optimization objective policy function L) clip And the above objective value function L v Update the parameters of the Actor and the Critic;

[0103] In this embodiment of the application, the above-mentioned multiple collision optimization models (including the Monte Carlo tree search algorithm for continuous action space, the cross-entropy algorithm for continuous action space, and the near-end policy optimization algorithm) are obtained based on multi-level collision scenarios pre-training;

[0104] The aforementioned multi-level collision scenarios include: simple intersection scenarios of single fragments (such as...) Figure 3 As shown), a multi-level collision scene of five fragments (such as...). Figure 4 As shown), fragment cloud collision scenarios based on probability distribution (such as...) Figure 5 (as shown) and complex intersection scenes that blend with reality (such as Figure 6 (as shown); whereby, the fragment cloud collision scenario based on probability distribution is generated in the virtual scene according to the set average position, average velocity, and variance of position and velocity of the fragment cloud, based on a normal distribution; the complex intersection scenario fused with reality is constructed by adding real spatial fragment data from the real space to the fragment cloud collision scenario based on probability distribution; the specific details of the above multi-level collision scenarios in the embodiments of this application are shown in Table 1 below; it is understood that, Figure 3 , Figure 4 , Figure 5 as well as Figure 6 In the diagram, solid black circles represent spacecraft, and solid gray circles represent debris.

[0105] Table 1

[0106]

[0107]

[0108] In Table 1 above, "center of mass" means that the centers of mass of the two fragments coincide or almost coincide at the closest moment, and "slightly" means that the centers of mass of the spacecraft and the space debris are at a certain distance at the closest moment, but this distance is less than the safe distance.

[0109] During the pre-training process described above, the loss function satisfies the following formula;

[0110] LOSS=k1·∑ i∈H p i +k2·f+k3·r Formula (13)

[0111] Where LOSS represents the loss value, k1 represents the negative weighting coefficient of the collision probability, and ∑ i∈H p i The table represents the sum of collision probabilities between the spacecraft and multiple objects, where k2 represents the fuel negative weighting coefficient, and f represents the fuel consumption. k4 represents the negative weighting coefficient for acceleration. k represents the total maneuvering acceleration of the spacecraft, k3 represents the orbital negative weighting coefficient, and r represents the orbital offset of the spacecraft before and after the maneuver.

[0112] The orbital offset r of the spacecraft before and after the maneuver satisfies the following formula;

[0113]

[0114] Among them, a t a0 and e represent the semi-major axes of the spacecraft's orbit after and before the maneuver, respectively. t e0 and e0 represent the eccentricity of the spacecraft before and after the maneuver, respectively, Ω t Ω0 and Ω0 represent the right ascension of the ascending node of the spacecraft after the maneuver and before the maneuver, respectively. t i0 and i0 represent the orbital inclination angles of the spacecraft before and after the maneuver, respectively, and ω t ω and ω0 represent the perigee arguments of the spacecraft before and after the maneuver, respectively. t M0 and M0 represent the mean apogee angles of the spacecraft before and after the maneuver, respectively;

[0115] Optionally, combined Figure 2 ,like Figure 7 As shown, before S101, this method also includes S104;

[0116] S104. Select multiple potential rendezvous objects of the spacecraft from multiple space objects;

[0117] In one application scenario, S104 includes S1041-S1042;

[0118] S1041. Remove space objects from multiple space objects of the spacecraft that meet any of the following conditions;

[0119] Condition 1: The apogee altitude of the collision object is less than the perigee altitude of the spacecraft's orbit;

[0120] Condition 2: The perigee altitude of the collision object is greater than the perigee altitude of the spacecraft's orbit;

[0121] Specifically, the spacecraft receives early warning information from ground monitoring stations via a transmission link. This information includes the current positions and velocities of multiple space objects. Using the spacecraft's current position and velocity in the J2000 mean equatorial geocentric coordinate system, along with the positions and velocities of the aforementioned space objects, six orbital elements (semi-major axis a, eccentricity e, right ascension of the ascending node Ω, orbital inclination i, perigee argument ω, and mean anomaly M) are calculated for the spacecraft and the multiple space objects. Then, using these six orbital elements, the spacecraft's perigee and apogee altitudes, as well as the perigee and apogee altitudes of each of the multiple space objects, are calculated. After this elimination process, the remaining orbital regions of the multiple space objects are as follows: Figure 8 As shown;

[0122] Since the J2000 flat equatorial geocentric coordinate system is a common coordinate system in this technical field, the embodiments of this application will not be described in detail here;

[0123] The formula for calculating the semi-major axis 'a' of the track based on position and velocity is as follows;

[0124]

[0125] Where ε represents the orbital specific energy, and μ represents the gravitational constant, v represents the magnitude of the velocity, and v = ||v||, v represents the velocity vector, r represents the magnitude of the position, and r = ||r||, r represents the position vector;

[0126] The formula for calculating the trajectory inclination angle i using position and velocity is as follows;

[0127]

[0128] Where h represents the magnitude of the specific angular momentum, and h = ||h||, h represents the specific angular momentum, and h = r × v, h z This represents the component of specific angular momentum along the z-axis in the J2000 level equatorial geocentric coordinate system;

[0129] The formula for calculating the right ascension Ω of the ascending node using position and velocity is as follows;

[0130]

[0131] Where n represents the magnitude of the ascending node vector, and n = ||n||, n represents the ascending node vector h = r × v, and n x This represents the x-axis component of the ascending node vector in the J2000 level equatorial geocentric coordinate system;

[0132] The formula for calculating the eccentricity e based on position and velocity is as follows;

[0133] e = ||e|| Formula (4)

[0134] Where e represents the eccentricity vector,

[0135] The formula for calculating the eccentricity e based on position and velocity is as follows;

[0136]

[0137] The formula for calculating the mean anterior angle M based on position and velocity is as follows;

[0138] M=Ee sin E formula (6)

[0139] Where E represents the angle of approach to the point, and f0 represents the true anterior angle, and

[0140] It should be noted that the perigee and apogee altitudes of the aforementioned spacecraft, as well as the perigee and apogee altitudes of multiple space objects, can be obtained by inputting the six orbital elements of the spacecraft and the six orbital elements of the multiple space objects into simulation software (e.g., Satellite Tool Kit, abbreviated as STK).

[0141] S1042. Then, identify the space objects whose time difference between their intersection points is less than the time threshold as multiple potential rendezvous objects for the spacecraft.

[0142] The aforementioned time difference between the points of intersection is the time difference between the moment when the spacecraft is at the orbital intersection and the moment when the object being collided with is at the orbital intersection. The orbital intersection is the point where the spacecraft's orbit intersects with the orbit of the object being collided with.

[0143] Specifically, for each of the multiple space objects, the orbital intersection point is calculated by using the orbital plane normal vector of the spacecraft and the orbital plane normal vector of the space object. Then, the time when the spacecraft arrives at the orbital intersection point and the time when the space object arrives at the orbital intersection point are calculated separately, and the difference between the two is taken as the time difference of passing the intersection point.

[0144] Understandably, the aforementioned orbital plane normal vector can be obtained by inputting the six orbital elements of the spacecraft and the six orbital elements of multiple space objects into simulation software (e.g., the satellite simulation software Satellite Tool Kit, abbreviated as STK);

[0145] The formula for calculating the intersection of the above orbits is as follows;

[0146] n1=τ1×τ2 Formula (7)

[0147] n2=τ2×τ1 Formula (8)

[0148] Where n1 represents the intersection vector on the orbital plane of the spacecraft, τ1 represents the orbital plane normal vector of the spacecraft, τ2 represents the orbital plane normal vector of the space object, and n2 represents the intersection vector on the orbital plane of the space object.

[0149] The formulas for calculating the time when a spacecraft arrives at the orbital intersection and the time when a space object arrives at the orbital intersection are as follows;

[0150] t i-1 =t0+(i-1)T s Formula (9)

[0151] Among them, t i-1 Let ti represent the time when the spacecraft / space object first reaches the orbital intersection, and t0 represent the time when the spacecraft / space object first reaches the orbital intersection. s Indicates the orbital operating cycle. Optionally, μ = 3.986005 × 10 14 m 3 / s 2 ;

[0152] Understandably, the above-mentioned t0 can be obtained by simulation using the six orbital elements of a spacecraft / space object. The embodiments of this application will not elaborate further on the determination process of t0.

[0153] Optionally, combined Figure 7 ,like Figure 9 As shown, before S103, the above method also includes S105;

[0154] S105. Based on the time when the spacecraft is closest to each of the multiple collision objects, the collision probability of the spacecraft with each of the multiple collision objects, and the number of multiple collision objects, analyze the control complexity of the spacecraft avoiding multiple collision objects.

[0155] It should be understood that the time when the spacecraft is closest to each of the multiple collision objects refers to the moment when the spacecraft is closest to each of the multiple collision objects.

[0156] In this embodiment of the application, while determining the distance between the spacecraft and each of the multiple collision objects when they are closest in step S101, the time when the spacecraft is closest to each of the multiple collision objects is obtained; a star-based orbital coordinate system is established, and the collision probability between the spacecraft and each of the multiple collision objects is determined based on the position of the spacecraft and the multiple collision objects in the star-based orbital coordinate system and the aforementioned distance; then, the control complexity of the spacecraft avoiding multiple collision objects is calculated using the aforementioned collision probability and the number of multiple collision objects.

[0157] Understandably, the aforementioned space-based orbital coordinate system (also known as the RSW coordinate system) is established based on the spacecraft; specifically, such as... Figure 10 As shown, the origin of the star-based orbital coordinate system is set at the center of mass of the spacecraft. The positive direction of the R-axis of the star-based orbital coordinate system is consistent with the direction from the Earth's center to the spacecraft. The positive direction of the S-axis is perpendicular to the R-axis and points in the direction of the spacecraft's motion. The W-axis, R-axis, and S-axis form a right-handed coordinate system.

[0158] Based on the aforementioned star-based orbital coordinate system, the aforementioned collision probability p satisfies the following formula;

[0159]

[0160] Where exp(*) represents the exponential function, R1 represents the component of the distance between the spacecraft and the collision target at their closest point on the R-axis of the star-based orbital coordinate system, S1 represents the component of the distance on the S-axis of the star-based orbital coordinate system, W1 represents the component of the distance on the W-axis of the star-based orbital coordinate system, and σ R σ represents the joint variance of the spacecraft's position component along the R-axis in the star-based orbital coordinate system and the position component of the colliding object along the R-axis in the star-based orbital coordinate system. In other words, it is the sum of the variances of the spacecraft's position component along the R-axis and the variances of the colliding object's position component along the R-axis in the star-based orbital coordinate system. SW This represents the joint variance of the spacecraft's position components in the SW plane of the star-based orbital coordinate system and the position components of the colliding object in the SW plane of the star-based orbital coordinate system. σ represents the orbital inclination angle (a known quantity). S σ represents the joint variance of the spacecraft's position component on the S-axis of the satellite-based orbital coordinate system and the position component of the colliding object on the S-axis of the satellite-based orbital coordinate system. In other words, it is the sum of the variances of the spacecraft's position component on the S-axis and the variances of the colliding object's position component along the S-axis of the satellite-based orbital coordinate system. Wσ represents the joint variance of the spacecraft's position component along the W-axis in the satellite-based orbital coordinate system and the position component of the colliding object along the W-axis in the satellite-based orbital coordinate system. In other words, it is the sum of the variances of the spacecraft's position component along the W-axis and the variances of the colliding object's position component along the W-axis in the satellite-based orbital coordinate system. SW It is formed by the coupling of the variances of the spacecraft's position components along the S-axis and W-axis in the star-based orbital coordinate system with the variances of the colliding object's position components along the S-axis and W-axis in the star-based orbital coordinate system. A It represents the combined radius of the spacecraft and the object it collides with, which is the sum of the effective radius of the spacecraft and the effective radius of the object it collides with. The effective radius refers to half of the maximum size of the space object (e.g., the spacecraft and the object it collides with).

[0161] The control complexity described above satisfies the following formula;

[0162]

[0163] Where c represents the control complexity, and n represents the number of collision objects for the spacecraft. c represents the average probability of a spacecraft colliding with multiple objects. vt This indicates the dispersion of the times when a spacecraft is closest to multiple colliding objects, and σ t The standard deviation (μ) represents the time when the spacecraft was closest to any of the multiple collision objects (used to describe the difference between most of the values ​​and their mean). t This represents the average time when the spacecraft came closest to multiple colliding objects; understandably, the above c... vt The larger the value, the more dispersed the time distribution of the closest approach between the spacecraft and multiple collision objects, the more time the spacecraft has to maneuver and avoid collisions, and the lower the control complexity of the spacecraft avoiding multiple collision objects.

[0164] Optionally, after S103, the above method further includes S106;

[0165] S106. Based on the target collision optimization model, determine the maneuver control quantities for controlling the spacecraft; and use the maneuver control quantities to control the spacecraft to avoid the target collision object.

[0166] In summary, the method for spacecraft to avoid multi-stage collisions provided in this application predicts the closest distance between the spacecraft and each of multiple potential encountering objects, and selects multiple collision targets from the multiple potential encountering objects based on the predicted distances. Then, based on the control complexity of the spacecraft avoiding these multiple collision targets, a target collision optimization model is selected for the target collision target among the multiple collision targets, and the spacecraft is controlled to avoid other collision targets while avoiding the target collision target. Therefore, it is possible to control the spacecraft to avoid one collision target while simultaneously preventing collisions with other collision targets.

[0167] Accordingly, embodiments of this application provide a device for spacecraft to avoid multi-stage collisions, such as... Figure 11 As shown, it includes a prediction module 501, a filtering module 502, and a determination module 503.

[0168] The prediction module 501 is used to predict the closest distance between the spacecraft and each of the multiple potential encountering objects. For example, the prediction module 501 is used to implement S101 of the above-described method for spacecraft to avoid multi-stage collisions.

[0169] The filtering module 502 is used to filter out multiple collision objects from multiple potential intersection objects to obtain multiple collision objects for the spacecraft; the multiple collision objects are the potential intersection objects whose closest distance to the spacecraft is less than a distance threshold among the multiple potential intersection objects. For example, the filtering module 502 is used to implement S102 of the above-mentioned method for spacecraft to avoid multi-stage collisions.

[0170] The determining module 503 is used to determine a target collision optimization model from multiple collision optimization models based on the control complexity of the spacecraft avoiding multiple collision objects. The target collision optimization model is used to control the spacecraft to avoid the target collision object, and also to control the spacecraft to avoid other collision objects among the multiple collision objects during the process of avoiding the target collision object. For example, the determining module 503 is used to implement S103 of the above-mentioned method for spacecraft to avoid multi-stage collisions.

[0171] The modules of the aforementioned spacecraft avoidance of multi-stage collisions device can also be used to perform other steps in the above method embodiments. All relevant content involved in the above method embodiments can be referred to in the functional description of the corresponding functional module, and will not be repeated here.

[0172] This application also provides an electronic device, including: a processor and a memory coupled to the processor; the memory is used to store computer instructions, and when the electronic device is running, the processor executes the computer instructions stored in the memory to cause the electronic device to perform the methods in the above embodiments. The processor can implement the prediction module 501, the screening module 502, and the determination module 503 described above; the memory can also be used to store the closest distances between a spacecraft and multiple potential encountering objects, the control complexity of a spacecraft avoiding multiple collision objects, and multiple collision optimization models, etc.

[0173] This application also provides a computer-readable storage medium including a computer program that, when run on a computer, performs the methods described in the above embodiments.

[0174] This application also provides a computer program product, which includes computer program instructions that, when run on a computer, execute the methods described above.

[0175] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0176] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for spacecraft to avoid multi-stage collisions, characterized in that, include: Predict the closest distance between the spacecraft and each of the multiple potential rendezvous objects; Based on the distance, multiple collision targets are selected from the multiple potential intersection targets to obtain multiple collision targets for the spacecraft; the multiple collision targets are the potential intersection targets whose closest distance to the spacecraft is less than a distance threshold among the multiple potential intersection targets; Based on the time when the spacecraft is closest to each of the multiple collision objects, the collision probability of the spacecraft with each of the multiple collision objects, and the number of the multiple collision objects, the control complexity of the spacecraft avoiding the multiple collision objects is analyzed. The control complexity satisfies the following formula. in, This indicates the control complexity. This indicates the number of multiple collision objects of the spacecraft. This represents the collision probability between the spacecraft and the plurality of collision objects. The average value, This indicates the dispersion of the times when the spacecraft was closest to multiple colliding objects, and , This represents the standard deviation of the time when the spacecraft was closest to any of the multiple collision objects. This represents the average time when the spacecraft came closest to multiple colliding objects; Based on the control complexity of the spacecraft avoiding the multiple collision objects, a target collision optimization model is determined from multiple collision optimization models, including: judging the control complexity and the number of multiple collision objects, and selecting the target collision optimization model; when the number of multiple collision objects is less than a number threshold, a collision optimization model using a Monte Carlo tree search algorithm with continuous action space is selected as the target collision optimization model; when the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is less than a complexity threshold, a collision optimization model using a cross-entropy algorithm with continuous action space is selected as the target collision optimization model; when the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is greater than or equal to the complexity threshold, a collision optimization model using a near-end strategy optimization algorithm is selected as the target collision optimization model; wherein, the target collision optimization model is used to control the spacecraft to avoid the target collision object among the multiple collision objects, and to control the spacecraft to avoid other collision objects among the multiple collision objects during the process of avoiding the target collision object.

2. The method as described in claim 1, characterized in that, The collision probability satisfy: in, Represents an exponential function. This represents the component of the distance between the spacecraft and the collision target at their closest point, along the R-axis in the star-based orbital coordinate system. This represents the component of the distance on the S-axis of the star-based orbital coordinate system. This represents the component of the distance on the W-axis of the star-based orbital coordinate system. This represents the joint variance of the position component of the spacecraft on the R-axis of the star-based orbital coordinate system and the position component of the colliding object on the R-axis of the star-based orbital coordinate system. This represents the joint variance of the position components of the spacecraft in the SW plane of the star-based orbital coordinate system and the position components of the colliding object in the SW plane of the star-based orbital coordinate system. This represents the combined radius of the spacecraft and the object it collides with.

3. The method as described in claim 1, characterized in that, The method further includes: Remove space objects from the spacecraft that satisfy any of the following conditions; Condition 1: The apogee altitude of the collision object is less than the perigee altitude of the spacecraft's orbit; Condition 2: The perigee altitude of the collision object is greater than the perigee altitude of the spacecraft's orbit; Then, spatial objects whose time difference between the intersection points is less than a time threshold are identified as potential rendezvous objects of the spacecraft. The time difference between the intersection points is the time difference between the moment when the spacecraft is at the orbital intersection point and the moment when the collision object is at the orbital intersection point. The orbital intersection point is the intersection point of the orbit where the spacecraft is located and the orbit where the collision object is located.

4. The method as described in claim 1, characterized in that, The multiple collision optimization models are obtained by pre-training based on multi-level collision scenarios; during the pre-training process, the loss function satisfies the following formula; in, Indicates the loss value. This represents the negative weighting coefficient for the collision probability. The table shows the sum of collision probabilities between the spacecraft and multiple objects. This represents the negative weighting coefficient for fuel. Indicates fuel consumption, and , This represents the negative weighting coefficient for acceleration. This represents the total maneuvering acceleration of the spacecraft. This represents the negative weighting coefficient for the orbit. This indicates the orbital offset of the spacecraft before and after the maneuver.

5. The method as described in claim 1, characterized in that, The method further includes: Based on the target collision optimization model, determine the maneuver control quantities for controlling the spacecraft; The spacecraft is controlled to avoid collision with the target object through the maneuver control variables.

6. A device for spacecraft to avoid multi-stage collisions, characterized in that, It includes a prediction module, a filtering module, and a determination module; The prediction module is used to predict the closest distance between the spacecraft and each of the multiple potential rendezvous objects; The filtering module is used to filter out multiple collision objects of the spacecraft from the multiple potential rendezvous objects; the multiple collision objects are the potential rendezvous objects that are closest to the spacecraft at a distance threshold among the multiple potential rendezvous objects; The determining module is used to analyze the control complexity of the spacecraft avoiding the multiple collision objects based on the time when the spacecraft is closest to each of the multiple collision objects, the collision probability of the spacecraft with each of the multiple collision objects, and the number of the multiple collision objects. The control complexity satisfies the following formula. in, This indicates the control complexity. This indicates the number of multiple collision objects of the spacecraft. This represents the collision probability between the spacecraft and the plurality of collision objects. The average value, This indicates the dispersion of the times when the spacecraft was closest to multiple colliding objects, and , This represents the standard deviation of the time when the spacecraft was closest to any of the multiple collision objects. This represents the average time when the spacecraft came closest to multiple colliding objects; Based on the control complexity of the spacecraft avoiding the multiple collision objects, a target collision optimization model is determined from multiple collision optimization models, including: judging the control complexity and the number of multiple collision objects, and selecting the target collision optimization model; when the number of multiple collision objects is less than a number threshold, a collision optimization model using a Monte Carlo tree search algorithm in continuous action space is selected as the target collision optimization model; when the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is less than a complexity threshold, a collision optimization model using a cross-entropy algorithm in continuous action space is selected as the target collision optimization model; when the number of multiple collision objects is greater than or equal to the number threshold, and the control complexity is greater than or equal to the complexity threshold, a collision optimization model using a near-end strategy optimization algorithm is selected as the target collision optimization model; wherein, the target collision optimization model is used to control the spacecraft to avoid the target collision object, and in the process of controlling the spacecraft to avoid the target collision object, it also avoids other collision objects among the multiple collision objects.

7. A computer-readable storage medium, characterized in that, It includes computer program instructions that, when executed by a computer, cause the computer to perform the method as described in any one of claims 1 to 5.

8. A computer program product, characterized in that, It includes computer program instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 5.

Images