Advanced aiming angle prediction method based on inter-satellite laser communication

Abstract

The invention discloses an advanced aiming angle prediction method based on inter-satellite laser communication, and relates to the technical field of satellite communication. Performing communication simulation on the satellite communication system, and determining training set data; the training set data comprises sample pointing angle data and sample advanced aiming angle data of a coarse tracking device and a fast reflector in the satellite communication system; iteratively training the back-propagation neural network through the training set data, optimizing initial parameters in the back-propagation neural network through a particle swarm algorithm in the training process, determining target initial parameters of the back-propagation neural network, and adjusting the target initial parameters according to a back-propagation mechanism to obtain a back-propagation neural network; determining the back propagation neural network meeting the training termination condition as an advanced aiming angle prediction model; and the advanced aiming angle of the target satellite is predicted in real time through the advanced aiming angle prediction model. The method can accurately predict the advanced aiming angle of the satellite.

Description

Technical Field

[0001] The present invention relates to the field of satellite communication technology, and in particular to a method for predicting an advanced aiming angle based on inter-satellite laser communication. Background Art

[0002] In an era of rapid advancements in aerospace technology, satellite laser communications, owing to its ultra-high-speed data transmission, high confidentiality, and robust anti-interference capabilities, have become a key technology for building space information networks. However, the extremely narrow beam width of the laser beam places stringent demands on laser aiming accuracy. To ensure the stability of the communication link, precise beam control is essential. In existing technologies, due to the relative motion between satellites and the relaxation time of beam transmission, accurately predicting the prephaser antenna array (PAA) to compensate for beam propagation delay and satellite motion poses a core challenge.

[0003] However, in the process of satellite communication, the existing technology lacks a method that can accurately predict the advance aiming angle. Summary of the Invention

[0004] Based on this, it is necessary to provide a method for predicting the advance aiming angle based on inter-satellite laser communication to address the above technical problems. This method can accurately predict the advance aiming angle of the satellite.

[0005] The present invention adopts the following technical solutions: The present invention provides a method for predicting an advanced aiming angle based on inter-satellite laser communication, comprising: Conducting communication simulation on the satellite communication system to determine training set data; the training set data includes sample pointing angle data of a coarse tracking device and a fast reflector in the satellite communication system, as well as sample advanced aiming angle data; The back propagation neural network is iteratively trained using training set data. During the training process, the initial parameters of the back propagation neural network are optimized using a particle swarm algorithm to determine the target initial parameters of the back propagation neural network. The target initial parameters are adjusted according to the back propagation mechanism, and the back propagation neural network that meets the training termination conditions is determined as the advanced aiming angle prediction model. The advance aiming angle of the target satellite is predicted in real time through the advance aiming angle prediction model.

[0006] Optionally, optimizing the initial parameters of the back-propagation neural network by a particle swarm algorithm to determine the target initial parameters of the back-propagation neural network includes: Determine the initial parameters as individual positions of the particle swarm algorithm; The individual position is used as the initial parameter of the back propagation neural network, and the training set data is input into the back propagation neural network for forward propagation calculation to obtain the predicted advanced aiming angle; According to the predicted advance aiming angle and the sample advance aiming angle data, the prediction error is calculated, and the prediction error is used as the fitness value of the particle swarm algorithm to update the individual position of the particle swarm algorithm; The individual positions at the last iteration of the particle swarm algorithm are used as the target initial parameters of the back propagation neural network.

[0007] Optionally, the prediction error includes a mean square error or a root mean square error; and the training termination condition includes the prediction error reaching a preset target error or the number of model training times reaching a preset maximum number of iterations.

[0008] Optionally, the advance aiming angle of the target satellite is predicted in real time by using an advance aiming angle prediction model, including: The pointing angle data of the coarse tracking device and the fast reflector corresponding to the target satellite are obtained, and the pointing angle data of the coarse tracking device and the fast reflector are input into the advance aiming angle prediction model to obtain the advance aiming angle of the target satellite.

[0009] Optionally, the method further includes: According to the advance aiming angle between satellite communications, the beam direction of the target satellite is adjusted to compensate for the beam propagation delay and the relative motion of the satellite.

[0010] Optionally, the method further includes: The advance aiming angle prediction model is updated online according to the real-time pointing angle data and the corresponding advance aiming angle.

[0011] The present invention provides an advanced aiming angle prediction system based on inter-satellite laser communication, comprising: A simulation module is used to perform communication simulation on the satellite communication system and determine training set data; the training set data includes sample pointing angle data of a coarse tracking device and a fast reflector in the satellite communication system, as well as sample advanced aiming angle data; A training module is used to iteratively train the back propagation neural network using training set data, and optimize the initial parameters of the back propagation neural network using a particle swarm algorithm during the training process, determine the target initial parameters of the back propagation neural network, adjust the target initial parameters according to the back propagation mechanism, and determine the back propagation neural network that meets the training termination conditions as the advanced aiming angle prediction model; The prediction module is used to predict the advance aiming angle of the target satellite in real time through the advance aiming angle prediction model.

[0012] The present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned method for predicting the advanced aiming angle based on inter-satellite laser communication.

[0013] The present invention provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for predicting an advanced aiming angle based on inter-satellite laser communication is implemented.

[0014] At least one of the above technical solutions adopted by the present invention can achieve the following beneficial effects: In the present invention, the coarse tracking device is responsible for large-scale initial alignment, the fast reflector is used for small-scale correction, and the advance aiming angle is a key parameter for compensating for signal transmission delay. By associating these three types of data to construct an advance aiming angle prediction model, the dynamic tracking process from coarse alignment to fine correction can be captured. Therefore, modeling is performed through back propagation neural network and particle swarm algorithm, and the parameters of the back propagation neural network are quickly approximated to the global optimal solution through the particle swarm algorithm. The back propagation mechanism is then used to improve the prediction accuracy, realizing the organic combination of data-driven modeling, global optimization parameters and dynamic feature capture, and realizing the precise fitting of complex nonlinear dynamics in satellite communications, so that the advance aiming angle can be accurately predicted. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0016] Figure 1 A schematic flow chart of a method for predicting an advanced aiming angle based on inter-satellite laser communication provided by the present invention; Figure 2 A schematic structural diagram of a satellite communication system provided by the present invention; Figure 3 A schematic diagram of an instruction manual for an advanced aiming angle provided by the present invention; Figure 4 A schematic flow chart of another method for predicting an advanced aiming angle based on inter-satellite laser communication provided by the present invention; Figure 5 A schematic diagram of a PSO-BP algorithm provided by the present invention; Figure 6 A schematic diagram of the fitting performance of an advance aiming angle prediction method based on inter-satellite laser communication provided by the present invention in noise-free training data; Figure 7A schematic diagram of robustness analysis of a method for predicting an advanced aiming angle based on inter-satellite laser communication provided by the present invention; Figure 8 This is a comparison chart of the errors between the advance aiming angle prediction model provided by the present invention and the existing method under different signal-to-noise ratios; Figure 9 A schematic diagram of a computer device for implementing an advanced aiming angle prediction method based on inter-satellite laser communication provided by the present invention. DETAILED DESCRIPTION

[0017] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments of the present invention and corresponding drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] Traditional PAA calculations rely on formula derivation and real-time satellite position and velocity information. However, current methods, such as interpolation prediction based on Global Navigation Satellite System (GNSS) data and modeling based on satellite orbit elements, have limitations in accuracy and robustness. Existing PAA prediction algorithms, such as the Kalman filter, also exhibit limitations in nonlinear systems and are highly dependent on the accuracy of the system's initial parameters and model.

[0019] To solve the above problems, the present invention proposes a method for predicting the advance aiming angle based on inter-satellite laser communication. This method combines the global search capability of the particle swarm optimization algorithm (Particle swarm optimization) and the local fine-tuning capability of backpropagation. It has high accuracy and strong robustness in nonlinear prediction, and can well adapt to the complex vibration environment of the satellite, thereby accurately predicting the satellite's advance aiming angle.

[0020] The technical solutions provided by various embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0021] Figure 1 The present invention is a flow chart of a method for predicting an advanced aiming angle based on inter-satellite laser communication, which specifically includes the following steps: S101, performing communication simulation on the satellite communication system to determine training set data; the training set data includes sample pointing angle data of a coarse tracking device and a fast reflector in the satellite communication system, as well as sample advanced aiming angle data.

[0022] like Figure 2 As shown in Figure 1, the satellite communication system includes an optical antenna, a coarse aiming mechanism (such as a two-dimensional gimbal), a beam splitter, a coarse aiming controller, a coarse aiming detector, a fine aiming controller, a fine aiming detector, a fast steering mirror (FSM), a dichroic beam splitter, an optical communication receiver, a pre-aiming mirror, a pre-aiming controller, a PAT controller, and a laser. The coarse aiming controller and coarse aiming detector are called the coarse tracking device. When acquisition is initiated, the transmitting optical antenna first emits a laser beam. Based on the initial information, the coarse aiming controller drives the coarse aiming detector and coarse aiming controller to perform a wide-range search and preliminary alignment to achieve acquisition. Subsequently, the fine aiming detector monitors the beam position in real time and feeds back deviation information to the fine aiming controller, which adjusts the fast steering mirror accordingly to achieve precise pointing. During communication, the system continuously tracks the motion of the receiving end and dynamically adjusts the beam direction, taking into account beam propagation delay and relative satellite motion. The pre-aiming mechanism compensates for deviations to ensure that the laser beam always stably and accurately reaches the receiving end, ensuring efficient and reliable intersatellite laser communication.

[0023] like Figure 3 As shown in the figure, taking the elevation angle as an example, terminal A emits a light beam at point S1. When the light beam reaches terminal B, terminal A has moved to point O. Terminal B must not only compensate for the displacement from point S1 to point O, but also for the displacement of terminal A during the transmission of the light beam from terminal B to terminal A (from O to S2). Figure 3 The angle PAA shown in is the advance aiming angle.

[0024] In the communication simulation of the satellite communication system, the pointing angle data of the coarse tracking device and the fast reflector, as well as the advance aiming angle data, are collected. These data include target trajectory information such as satellite motion, providing basic data support for PAA prediction. The collected pointing angle data is normalized to fall within the range [-1, 1] to improve the training efficiency and accuracy of the neural network. The normalized pointing angle data is used as sample pointing angle data, and the collected advanced aiming angle data is used as sample advanced aiming angle data. The sample data is also divided into training and test data sets. The training set data is used to train the advanced aiming angle prediction model, while the test set data is used to evaluate the prediction performance of the advanced aiming angle prediction model.

[0025] S102, iteratively training the back propagation neural network using the training set data, and optimizing the initial parameters of the back propagation neural network using the particle swarm algorithm during the training process, determining the target initial parameters of the back propagation neural network, and adjusting the target initial parameters according to the back propagation mechanism, and determining the back propagation neural network that meets the training termination conditions as the advanced aiming angle prediction model.

[0026] Among them, the advance aiming angle prediction model is a PSO-BP model constructed by the particle swarm optimization-back propagation (PSO-BP) algorithm.

[0027] In one embodiment, the initial parameters in the back propagation neural network are optimized by a particle swarm algorithm to determine the target initial parameters of the back propagation neural network, including: determining the initial parameters as individual positions of the particle swarm algorithm; using the individual positions as the initial parameters of the back propagation neural network, inputting the training set data into the back propagation neural network for forward propagation calculation to obtain a predicted advance aiming angle; calculating the prediction error based on the predicted advance aiming angle and the sample advance aiming angle data, and using the prediction error as the fitness value of the particle swarm algorithm to update the individual positions of the particle swarm algorithm; and using the individual positions at the last iteration of the particle swarm algorithm as the target initial parameters of the back propagation neural network.

[0028] Optionally, the prediction error includes a mean square error or a root mean square error; and the training termination condition includes the prediction error reaching a preset target error or the number of model training times reaching a preset maximum number of iterations.

[0029] The structure of the back propagation (BP) neural network is as follows: the number of input layer nodes is determined to be 15. This is because the input data is time series data, and 15 input nodes can cover sufficient historical information to predict future PAA values; the number of output layer nodes is 1, corresponding to the predicted PAA value; the number of hidden layer nodes is determined to be 20 through experiments and experience. More hidden layer nodes can enhance the network's ability to fit complex nonlinear relationships.

[0030] Initialize the BP neural network: randomly initialize the weights and thresholds of the BP neural network, select the tansig function as the hidden layer activation function, and select the purelin function as the output layer activation function.

[0031] The initialization parameters of the PSO algorithm are as follows: the particle swarm size is determined to be 15, the training cycle is 20, the learning rate is 0.1, the individual learning factor is 0.7, the social learning factor is 1.8, and the target error is set to 0.00001.

[0032] Encoded particles: The parameters of the BP neural network are converted into the positions of particles. Each particle represents a potential BP neural network parameter solution. The parameters include initial weights and threshold encoding.

[0033] Calculating the fitness value: Using the training set data (sample pointing angle data), the input data is forward propagated through the BP neural network to obtain the output result (predicted advance aiming angle). This is then compared with the actual PAA value (sample advance aiming angle data) and the mean squared error (MSE) or root mean squared error (RMSE) is calculated as the particle's fitness value. The smaller the fitness value, the better the prediction performance of the BP neural network parameter solution corresponding to the particle.

[0034] BP neural network training: Forward propagation: Using the training set data, the input data is forward propagated through the BP neural network optimized by PSO to obtain the predicted PAA value.

[0035] Error calculation: Compare the predicted PAA value with the actual PAA value and calculate the error. Usually, indicators such as mean square error or root mean square error are used to measure the error size.

[0036] Back propagation weight adjustment: According to the error, the BP algorithm is used to adjust the weights and thresholds of the neural network, and the network parameters are continuously optimized through the gradient descent method to gradually reduce the prediction error.

[0037] Model training is completed: After multiple iterations of training, when the prediction error of the back propagation neural network reaches the predetermined target error, or the number of model training reaches the preset maximum number of iterations, the model training is considered to be completed. At this time, the back propagation neural network that meets the training termination conditions is determined as the advanced aiming angle prediction model.

[0038] It should be noted that the training process of the advanced aiming angle prediction model combines the global search capability of PSO and the local fine adjustment capability of BP. PSO optimizes the initial weights and thresholds of the BP neural network, allowing the network to quickly approach the global optimal solution. The error back propagation mechanism of BP is then used to improve the prediction accuracy, which has higher accuracy and robustness in nonlinear prediction.

[0039] S103, predicting the advanced aiming angle of the target satellite in real time using an advanced aiming angle prediction model.

[0040] Predicting PAA: New input data is fed into the trained advance aiming angle prediction model. The model will output the corresponding PAA prediction value based on the learned mapping relationship.

[0041] Specifically, the advance aiming angle of the target satellite is predicted in real time through the advance aiming angle prediction model, including: The pointing angle data of the coarse tracking device and the fast reflector corresponding to the target satellite are obtained, and the pointing angle data of the coarse tracking device and the fast reflector are input into the advance aiming angle prediction model to obtain the advance aiming angle of the target satellite.

[0042] In one embodiment, after obtaining the advance aiming angle between satellite communications, the beam direction of the target satellite is adjusted according to the advance aiming angle between satellite communications to compensate for beam propagation delay and relative motion of the satellites.

[0043] Real-time data update: In the actual application of satellite laser communication, the latest pointing angle data is continuously obtained from the APT system and pre-processed in real time to ensure that the data input into the advanced aiming angle prediction model is timely.

[0044] Online prediction and adjustment: The pre-processed real-time data is input into the trained model to predict the PAA value in real time. The laser beam pointing is adjusted in time based on the prediction result to compensate for the beam propagation delay and the relative motion of the satellite, ensuring the stability and quality of the communication link.

[0045] In one embodiment, the advance aiming angle prediction model is updated online based on the real-time pointing angle data and the corresponding advance aiming angle. That is, the model can be updated and optimized online based on actual conditions to adapt to changes in satellite motion and interference from environmental factors.

[0046] In one embodiment, the present invention also provides a method for predicting an advanced aiming angle based on inter-satellite laser communication, such as Figure 4 As shown in the figure, this method cleverly combines the advantages of PSO and BP. First, it uses the global search capability of the PSO algorithm to optimize the initial weights and thresholds of the BP neural network, so that the initial state of the network is closer to the global optimum. Then, the weights are further adjusted through the local precise search capability of the BP algorithm to improve the accuracy of the network. This combination effectively avoids the shortcomings of this method. Specifically, first, according to the read data, the training set and the test set are divided and normalized. Then, the topological structure of the BP neural network is determined, and the initial parameters of the BP neural network are determined. The initial parameters include initial weights and initial thresholds. Then, the PSO algorithm is used to optimize the initial parameters to obtain the target initial weights and target initial thresholds. Then, the prediction error of the BP neural network based on the target initial weights and target initial thresholds is calculated, and the weights and thresholds are updated using the backpropagation mechanism until the stopping condition is met.

[0047] like Figure 5 As shown, Figure 5This is a flow chart of another method for predicting the advanced aiming angle based on inter-satellite laser communication provided by the present invention, wherein δ1 represents the residual pointing error threshold of the coarse tracking device (CPA). In the coarse tracking stage, the system checks whether the residual pointing error of the CPA is less than δ1 (i.e., δ1 < δ e ), the target is considered to have been captured only when this condition is met. e represents the coarse pointing error threshold, δ2 represents the residual pointing error threshold of the fast reflector (FSM), and δf represents the pointing error threshold of the fast reflector (corresponding to δ2). When δ2<δ f When the fine tracking is considered stable, the advance aiming angle (PAA) can be predicted.

[0048] In one embodiment, a PSO-BP algorithm was used to simulate time series prediction of PAA in satellite laser communications, leveraging the pointing angle data from the coarse tracking device and fast reflector in a satellite communication system. This simulation scenario was based on a laser communication link between two Low Earth Orbit (LEO) satellites. Satellite platform vibration was introduced as data noise to verify the PSO-BP algorithm's prediction performance in complex noisy environments. Key satellite parameters are detailed in Table 1.

[0049] Table 1 Satellite orbit data used for simulation Among them, the parameters in Table 1 are expressed as: i :inclination; Ω : right ascension of the ascending node; e: eccentricity; ω: secondary angle of perigee; a: semi-major axis; f: true anomaly.

[0050] Pointing accuracy is mainly affected by the tracking error σ, which is caused by platform vibration. Taking the space laser communication system based on the SILEX platform developed by the European Space Agency as an example, the platform vibration power spectrum density model of this system can be expressed as:

[0051] (1) in, represents the power spectral density, Indicates the frequency variable of the current analysis, in Hz. Indicates the characteristic cutoff frequency.

[0052] The communication window is set to 100 seconds, during which the two satellites pass each other and maintain line of sight, which changes as they move. In the satellite communication system, the FSM uses a high-precision angle sensor to continuously collect pointing angle data, including azimuth (α_k) and pitch (β_k), at a sampling frequency of 100 Hz. This generates a total of 9611 data samples, 8620 of which are used for the training set. A prediction is performed five samples ahead on the 991 test set. After multiple experiments, the parameters of the PSO-BP network model are shown in Table 2.

[0053] Table 2 Main parameters of PSO-BP model Simulation results and analysis: 1) Advance aiming angle prediction In order to verify the prediction model constructed by the PSO-BP neural network, we first used the noise-free PAA obtained from the FSM under ideal conditions as training data, performed time series prediction on the PSO-BP algorithm and the pure BP neural network method, and compared the results. As shown in Table 3, under noise-free conditions, the PSO-BP algorithm showed higher accuracy in predicting the PAA in the line of sight direction, with a determination coefficient R of 0. 2 The value reaches 0.9872, which is 0.0311 higher than that of the BP neural network. Its RMSE is 0.0319 and MAE is 0.0124, both lower than those of the BP neural network. This shows that the PSO-BP algorithm can accurately predict the PAA of the gaze direction and is applicable to complex nonlinear systems.

[0054] Table 3 Prediction performance under noise-free conditions like Figure 6 As shown, Figure 6 The fitting performance of PAA prediction based on the PSO-BP model in noise-free training data is demonstrated. (a) Figure 2 shows the fitting curve of the PSO-BP training data, and (b) Figure 3 shows the fitting curve of the PSO-BP test data.

[0055] 2) Robustness Analysis The robustness of the PSO-BP algorithm was tested in a high noise environment (SNR = 5dB). Figure 7 As shown, Figure 7 Robustness analysis of the PAA prediction algorithm based on the PSOBP neural network. (a) and (b) show the fitting curves of the PSO-BP training data and test data, respectively. (c) and (d) show the fitness curves of the prediction algorithm under noise-free and 5dB noise conditions, respectively. Specifically, Figure 7As shown in (a) and (b), when there is noise in the data, the normalized root mean square error (RMSE) is 0.145; after prediction by the PSO-BP algorithm, the value drops to 0.102, and the error rate is reduced by 29.7%.

[0056] Figure 7 Figures (c) and (d) show the fitness curves of the prediction model under noise-free and 5dB noise conditions, respectively. The fitness curves in both cases decrease rapidly with increasing iterations. For example, the fitness value under noise-free conditions drops rapidly from 0.0271 to 0.0266, indicating that the model converges quickly and can quickly escape from local optimal solutions, demonstrating its effectiveness and feasibility.

[0057] In order to evaluate the noise immunity of the algorithm, the present invention analyzes its performance in the signal-to-noise ratio range from 20 dB to 0 dB. Figure 8 The paper presents a comparison of the root mean square error (RMS) of the algorithm's predicted advanced pointing angle data under various noise conditions with noise-free data. The PSO-BP algorithm's noise tolerance significantly improves with increasing noise levels and decreasing signal-to-noise ratios. For example, when the signal-to-noise ratio is 1 dB, the algorithm's predicted average pointing error angle decreases by 48.8%. This demonstrates that the PSO-BP algorithm compensates for system lag by predicting PAA data and ensures prediction accuracy under high noise interference conditions, thereby improving the pointing accuracy of the interstellar laser communication APT system and optimizing system performance.

[0058] This paper proposes a method for predicting the advanced aiming angle based on inter-satellite laser communications. This method utilizes particle swarm optimization to optimize the initial weights and thresholds of a BP neural network, significantly improving system performance. This method also enhances the network's ability to fit nonlinear systems and its prediction accuracy. Simulation results demonstrate that the proposed algorithm demonstrates excellent prediction accuracy, convergence speed, and robustness, effectively adapting to complex nonlinear systems and noise interference. This prediction algorithm has great potential for improving the performance and reliability of satellite laser communication systems.

[0059] When applying the method for predicting the advanced aiming angle based on intersatellite laser communication provided by the present invention, it is not necessary to Figure 1 The steps are executed in the order shown. The specific execution order of the steps can be determined according to needs, and the present invention does not limit this.

[0060] The above is a method for predicting an advanced aiming angle based on inter-satellite laser communication provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding advanced aiming angle prediction system based on inter-satellite laser communication, which includes: A simulation module is used to perform communication simulation on the satellite communication system and determine training set data; the training set data includes sample pointing angle data of a coarse tracking device and a fast reflector in the satellite communication system, as well as sample advanced aiming angle data; A training module is used to iteratively train the back propagation neural network using training set data, and optimize the initial parameters of the back propagation neural network using a particle swarm algorithm during the training process, determine the target initial parameters of the back propagation neural network, adjust the target initial parameters according to the back propagation mechanism, and determine the back propagation neural network that meets the training termination conditions as the advanced aiming angle prediction model; The prediction module is used to make real-time predictions of the advance aiming angle between satellite communications through an advance aiming angle prediction model.

[0061] The specific limitations of the advanced aiming angle prediction system based on inter-satellite laser communication can be found in the limitations of the advanced aiming angle prediction method based on inter-satellite laser communication described above and will not be further elaborated here. Each module in the aforementioned advanced aiming angle prediction system based on inter-satellite laser communication can be implemented in whole or in part via software, hardware, or a combination thereof. Each of these modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a computer device memory in software form, so that the processor can call and execute the corresponding operations of each module.

[0062] The present invention also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 1 The paper provides an advanced aiming angle prediction method based on inter-satellite laser communication.

[0063] The present invention also provides Figure 9 The structural diagram of the computer equipment shown in FIG. Figure 9 As shown in the figure, at the hardware level, the computer device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory. Of course, it may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 The paper provides an advanced aiming angle prediction method based on inter-satellite laser communication.

[0064] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware using a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes in the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0065] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of the present invention.

Claims

1. A method for predicting an advanced aiming angle based on inter-satellite laser communication, characterized in that: include: Conducting communication simulation on the satellite communication system to determine training set data; the training set data includes sample pointing angle data of a coarse tracking device and a fast reflector in the satellite communication system, as well as sample advanced aiming angle data; The back propagation neural network is iteratively trained using training set data. During the training process, the initial parameters of the back propagation neural network are optimized using a particle swarm algorithm to determine the target initial parameters of the back propagation neural network. The target initial parameters are adjusted according to the back propagation mechanism, and the back propagation neural network that meets the training termination conditions is determined as the advanced aiming angle prediction model. The advance aiming angle of the target satellite is predicted in real time through the advance aiming angle prediction model.

2. The method according to claim 1, characterized in that The initial parameters of the back propagation neural network are optimized by the particle swarm algorithm to determine the target initial parameters of the back propagation neural network, including: Determine the initial parameters as individual positions of the particle swarm algorithm; The individual position is used as the initial parameter of the back propagation neural network, and the training set data is input into the back propagation neural network for forward propagation calculation to obtain the predicted advanced aiming angle; According to the predicted advance aiming angle and the sample advance aiming angle data, the prediction error is calculated, and the prediction error is used as the fitness value of the particle swarm algorithm to update the individual position of the particle swarm algorithm; The individual positions at the last iteration of the particle swarm algorithm are used as the target initial parameters of the back propagation neural network.

3. The method according to claim 2, characterized in that The prediction error includes the mean square error or the root mean square error; the training termination conditions include the prediction error reaching the preset target error or the number of model training times reaching the preset maximum number of iterations.

4. The method according to claim 1, wherein The advance aiming angle prediction model is used to predict the target satellite's advance aiming angle in real time, including: The pointing angle data of the coarse tracking device and the fast reflector corresponding to the target satellite are obtained, and the pointing angle data of the coarse tracking device and the fast reflector are input into the advance aiming angle prediction model to obtain the advance aiming angle of the target satellite.

5. The method according to claim 3, characterized in that The method further comprises: According to the advance aiming angle between satellite communications, the beam direction of the target satellite is adjusted to compensate for the beam propagation delay and the relative motion of the satellite.

6. The method according to claim 1, characterized in that The method further comprises: The advance aiming angle prediction model is updated online according to the real-time pointing angle data and the corresponding advance aiming angle.

7. An advanced aiming angle prediction system based on inter-satellite laser communication, characterized in that: include: A simulation module is used to perform communication simulation on the satellite communication system and determine training set data; the training set data includes sample pointing angle data of a coarse tracking device and a fast reflector in the satellite communication system, as well as sample advanced aiming angle data; A training module is used to iteratively train the back propagation neural network using training set data, and optimize the initial parameters of the back propagation neural network using a particle swarm algorithm during the training process, determine the target initial parameters of the back propagation neural network, adjust the target initial parameters according to the back propagation mechanism, and determine the back propagation neural network that meets the training termination conditions as the advanced aiming angle prediction model; The prediction module is used to make real-time predictions of the advance aiming angle between satellite communications through an advance aiming angle prediction model.

8. A computer-readable storage medium, characterized in that The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

9. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 6 when executing the program.

Images