Satellite multi-target non-contact measurement system and method

Abstract

The invention provides a satellite multi-target non-contact measurement system and a satellite multi-target non-contact measurement method. A laser module emits laser to a target according to a laser driving signal from an FPGA (Field Programmable Gate Array) data processing module; the CMOS module transmits analog signals corresponding to multi-target laser spots collected from a target to the FPGA data processing module. The FPGA data processing module transmits a multi-target image obtained according to the analog signal to an upper computer; and the upper computer obtains a corresponding measurement result according to the multi-target image. High-precision non-contact measurement plays a crucial role in a satellite system. According to the invention, a non-contact measurement result can be obtained in real time with high precision, the speed and precision are improved, and more accurate and effective data are provided for a subsequent system.

Description

Technical Field

[0001] This invention belongs to the field of satellite measurement technology, specifically relating to a satellite multi-target non-contact measurement system and method. In particular, it relates to a high-precision satellite multi-target non-contact measurement system and method based on a particle swarm artificial neural network algorithm. Background Technology

[0002] Currently, with the continuous development of satellite technology, high-precision non-contact measurement plays a crucial role in satellite systems, providing accurate feedback information for multiple systems. However, existing measurement technologies still need further improvement in terms of accuracy and real-time performance, limiting the performance and application scope of satellite systems.

[0003] High-precision non-contact measurement is used in satellites for applications such as attitude control, orbit measurement, and deformation analysis, which are crucial for improving the accuracy and reliability of satellite missions.

[0004] Patent document CN207215414U (application number: CN201720708249.7) discloses a satellite antenna structure vibration measurement device based on non-contact measurement. This method achieves non-contact measurement by using a binocular vision system, but it is costly and computationally intensive. In contrast, the proposed method is low-cost, has high accuracy in non-contact measurement, and provides reliable data.

[0005] Patent document CN116625480A (application number: CN202211731313.5) discloses a non-contact mechanical vibration measurement system, relating to the field of non-contact vibration measurement technology. This method utilizes rolling shutter image acquisition to calculate the vibration parameters of the object under test, but it is not real-time acquisition and is costly. Because this invention uses a particle swarm optimization artificial neural network algorithm, it can simultaneously detect and track the positions of multiple light spots. This makes the method more adaptable and effective in multi-target measurement scenarios.

[0006] Patent document CN117723244A (application number: CN202311778413.8) discloses a laser-based multi-point measurement method for high-rise building vibration modes, which relates to the field of non-contact measurement. This method involves selecting measurement points, deploying ground-based laser non-contact equipment, conducting dynamic response tests, performing time-frequency analysis and structural modal analysis to obtain vibration modal information of high-rise buildings. However, this method is costly and highly dependent on the accuracy of the equipment. In contrast, the method proposed in this invention can process measurement data from multiple targets, enabling real-time measurement of multiple targets. Compared to single-target systems, it has higher efficiency and practicality, and can more accurately fit the positional changes of the laser spot center, improving measurement accuracy and reliability.

[0007] Patent document CN220541989U (application number: CN202321844886.9) discloses a non-contact thickness measurement device, which uses a laser and a receiver set on a flat plate to form a workpiece thickness detection area, but the accuracy is low. The present invention has higher real-time performance and accuracy, and can more accurately obtain the positions of multiple light spots to obtain more reliable non-contact measurement results.

[0008] Patent document CN117664254A (application number: CN202311528153.9) discloses a laser-ultrasonic non-contact method and system for measuring the flow rate of working fluid in pressure pipelines, relating to the field of non-contact measurement. This method utilizes lasers to transmit and receive ultrasonic signals, eliminating the need for the ultrasonic probe to contact the pipeline surface for measurement. However, this method is costly, complex, and has stringent environmental requirements, potentially affecting signal stability and accuracy. In contrast, the system proposed in this invention has a simpler structure, higher data processing efficiency, wider applicability, and a higher degree of automation, resulting in greater efficiency and practicality.

[0009] Therefore, it is necessary to conduct research on high-precision non-contact measurement and to improve the technology to meet the demands for real-time and high-precision measurement. However, there is currently no good method to achieve such non-contact measurement. Summary of the Invention

[0010] In view of the deficiencies in the prior art, the purpose of this invention is to provide a satellite multi-target non-contact measurement system and method.

[0011] According to the present invention, a satellite multi-target non-contact measurement system includes: an FPGA data processing module, a laser module, a CMOS module, and a host computer;

[0012] The laser module emits laser light toward the target based on the laser drive signal from the FPGA data processing module;

[0013] The CMOS module transmits the analog signals corresponding to the multi-target laser spots acquired from the target to the FPGA data processing module;

[0014] The FPGA data processing module transmits the multi-target image obtained from the analog signal to the host computer.

[0015] The host computer obtains the corresponding measurement results based on the multi-target images.

[0016] Preferably, the laser module emits multiple laser beams, which are then split to generate multiple laser beams that are focused onto the target.

[0017] Preferably, the analog signal corresponds to data of N frames. The FPGA data processing module obtains the distortion-corrected image of each of the N frames according to the camera's intrinsic and extrinsic parameters. Based on the distortion-corrected image obtained from each of the N frames, a denoised image is obtained as the output of the FPGA data processing module and output to the host computer.

[0018] Preferably, the host computer extracts the center of the light spot from the multi-target image using a particle swarm artificial neural network algorithm, and then fits the mean of the positional change of the light spot center to obtain the distance change result.

[0019] Particle swarm optimization is used for global search, and the search process is guided by swarm intelligence. Simulated annealing is used to locally adjust the particle positions after each update.

[0020] The algorithm parameters are learned and adjusted in real time using a multilayer neural network (MLP); by extracting features from the input data, the neural network dynamically optimizes the velocity and position update rules in the particle swarm.

[0021] Preferably, after receiving the denoised multi-target image, the host computer optimizes it using a particle swarm artificial neural network algorithm, then uses the gray-scale centroid method to obtain the centroids of multiple light spots as the center of the light spots. The average value of the multiple centroids is used as the result for fitting, and the change of the laser position on the object surface over time is obtained, thus obtaining the real-time non-contact measurement result.

[0022] Preferably, in the host computer,

[0023] Acquire image data containing multiple light spots; randomly initialize a group of particles, each particle representing a possible combination of light spot positions;

[0024] Particle swarm optimization (PSO) and simulated annealing are used to predict the position combinations of each particle, and the error is calculated as the fitness. A multilayer neural network is used to dynamically adjust the particle swarm parameters to improve fitness. Specifically, for each particle, its fitness value is calculated, and the error between the predicted result and the actual spot position is evaluated by a multilayer perceptron (MLP). According to the formula of the PSO algorithm, the velocity and position of each particle are updated. At the same time, a simulated annealing mechanism is introduced to control the temperature to adjust the probability of accepting poor solutions, ensuring that a wider solution space can be explored during the optimization process.

[0025] The global optimal position is updated based on fitness. The position and velocity of each particle are updated based on the global optimal position and the individual optimal position. The parameters of the artificial neural network are updated based on the particle's position. This process continues until a stopping condition is met. After each iteration, the parameters of the MLP are dynamically adjusted using the backpropagation algorithm based on the particle's fitness value to improve the model's predictive ability. The gradient of the loss function with respect to each weight is calculated, and the weights are updated.

[0026] The system outputs converged spot position information to locate multiple spots. Specifically, the position data of multiple spots at the same time are organized into a set of position coordinates. Then, polynomial fitting is performed on these position data using methods such as least squares to obtain polynomial coefficients describing the change in the spot center position. Next, a weighted average of the spot center position is calculated based on the fitted polynomials to reflect the change in the spot center position. The accuracy of the fitting is evaluated by adjusting the residuals of the fitting results for each iteration, and the model parameters are continuously optimized based on error feedback for refitting. This yields the distance change results of the spot center before and after the current time.

[0027] Preferably, a high-order fitting is performed on the multiple target locations to obtain a value corresponding to the current time, and the result of non-contact measurement is obtained based on the change of the value at different times.

[0028] A non-contact satellite multi-target measurement method provided by the present invention includes:

[0029] The laser module and CMOS module are driven to control the laser module to emit laser light, which is then split and directed to the target.

[0030] The control CMOS module collects multiple laser spots on the target surface. After converting the collected analog signals into digital signals, the FPGA data processing module processes the digital signals, performs distortion correction and filtering on the obtained images, and outputs the multi-target image to the host computer.

[0031] The FPGA data processing module is controlled to obtain multi-target images based on each frame of data;

[0032] The host computer is controlled to obtain target location information based on the multi-target image; and the measurement result is obtained based on the target location information.

[0033] Preferably, the aforementioned satellite multi-target non-contact measurement system is used.

[0034] According to the present invention, a satellite is characterized by comprising the aforementioned satellite multi-target non-contact measurement system; or employing the aforementioned satellite multi-target non-contact measurement method.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. This invention has the ability to obtain non-contact measurement results in real time with high precision, and can perform spot positioning and non-contact measurement in real time during use.

[0037] 2. The satellite non-contact measurement system integrated in this invention, through FPGA control signals and data processing, not only reduces the size of the entire satellite high-precision non-contact measurement system, but also improves the speed and accuracy of non-contact measurement, making non-contact measurement more precise and effective.

[0038] 3. This invention can continuously adjust the detection results through satellite non-contact measurement in real time, so as to achieve the best detection effect.

[0039] 4. The Particle Swarm Optimization (PSO) artificial neural network algorithm combines the advantages of PSO optimization and artificial neural networks. However, it still faces problems such as local optima traps, computational complexity, insufficient adaptability, parameter selection issues, and noise interference in satellite multi-target non-contact measurement systems. Local optima traps may prevent the algorithm from finding the global optimum, thus affecting measurement accuracy. Furthermore, computational complexity increases significantly with the number and dimension of particles, potentially leading to response delays in real-time measurements. Insufficient adaptability makes it difficult for the algorithm to quickly adjust parameters under different environmental conditions (such as temperature and illumination changes), and inappropriate parameter selection can severely impact convergence and overall performance. Simultaneously, environmental noise makes traditional PSO algorithms vulnerable when processing measurement data, potentially leading to spot localization errors. These problems necessitate further improvement and optimization of the algorithm for practical applications.

[0040] To address these issues, a hybrid optimization algorithm was employed to enhance global search capabilities, dynamic parameter adjustment was implemented to improve adaptability, a multi-layer neural network was used to improve the accuracy of spot center extraction, noise filtering and preprocessing were implemented to improve data quality, and high-order fitting techniques and iterative mechanisms were used to optimize positioning accuracy. These improvements significantly enhanced the system's measurement performance in complex environments and ensured efficient and reliable data output. Attached Figure Description

[0041] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0042] Figure 1 A schematic diagram showing a portion of the components of acquiring a multi-target image according to an embodiment of the present invention is shown;

[0043] Figure 2 A schematic diagram of the high-precision multi-target non-contact measurement system according to the present invention is shown.

[0044] Figure 3 A schematic diagram illustrating the setup of a satellite high-precision multi-target non-contact measurement system according to an embodiment of the present invention is shown.

[0045] Figure 4 A flowchart of a satellite high-precision multi-target non-contact measurement method according to the present invention is shown;

[0046] Figure 5 A schematic diagram illustrating the principle of a satellite-based high-precision multi-target non-contact measurement method according to the present invention is shown. Detailed Implementation

[0047] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0048] Figure 1 A schematic diagram showing a portion of the components of a multi-target image acquired according to the present invention is provided.

[0049] like Figure 1 As shown, a high-precision multi-target non-contact measurement system for satellites based on a particle swarm optimization artificial neural network algorithm, according to the present invention, includes: an FPGA data processing module, a laser module, a CMOS module, a power supply module, and a host computer. The FPGA data processing module is connected to the laser module, the CMOS module is connected to the FPGA data processing module, and the power supply module is connected to both the FPGA data processing module and the CMOS module; the FPGA data processing module is connected to the host computer. The laser module includes a laser diode and a laser driving circuit. The laser diode is connected to the laser driving circuit, and the laser driving circuit is connected to the FPGA data processing module. The laser diode is used to generate a detection laser. The laser driving circuit is used to receive signals from the FPGA data processing module and excite the laser diode.

[0050] The power module supplies power to the CMOS module, the laser diode of the laser module, and the FPGA data processing module. The FPGA data processing module provides a laser drive signal to the laser module, which then emits a laser beam towards the target object according to the signal. After passing through the optical system, the CMOS module acquires multiple laser spots on the target and transmits the analog signals to the FPGA data processing module. After analog-to-digital conversion, the module processes the data and obtains a multi-target image, which is then transmitted to the host computer. The host computer processes the data to obtain the corresponding measurement results.

[0051] More specifically, the FPGA data processing module controls the CMOS module for acquisition and transmission, outputting modulation signals and acquiring images; the laser module receives signals from the FPGA data processing module to drive the laser diode; the laser module generates four laser beams after being split by a prism and focuses them onto the target object; the CMOS module acquires images through FPGA data processing; the power supply module generates the required voltage for each module and provides power; the host computer processes the acquired images using a particle swarm artificial neural network algorithm to obtain measurement results. In a preferred embodiment, the host computer extracts the light spot center from the acquired images using a particle swarm artificial neural network algorithm and then fits the mean of the position change of the light spot center to obtain the distance change result.

[0052] Particle swarm optimization (PSO) artificial neural network algorithms combine the advantages of particle swarm optimization and artificial neural networks. However, in satellite multi-target non-contact measurement systems, they still face problems such as local optimum traps, computational complexity, insufficient adaptability, parameter selection issues, and noise impact. Local optimum traps may prevent the algorithm from finding the global optimum, thus affecting measurement accuracy. Furthermore, computational complexity increases significantly with the number and dimension of particles, potentially leading to response delays in real-time measurements. Insufficient adaptability makes it difficult for the algorithm to quickly adjust parameters under different environmental conditions (such as temperature and illumination changes), and inappropriate parameter selection can severely affect the algorithm's convergence and overall performance. Simultaneously, environmental noise makes traditional PSO algorithms vulnerable when processing measurement data, potentially leading to spot localization errors. These problems necessitate further improvement and optimization of the algorithm for practical applications.

[0053] To address these issues, this embodiment employs the following approach:

[0054] (1) Hybrid optimization algorithms are used to improve global search capabilities, specifically:

[0055] This approach combines Particle Swarm Optimization (PSO) and Simulated Annealing (SA). PSO is used for global search, guiding the search process through swarm intelligence, while SA adjusts particle positions locally after each update to avoid getting trapped in local optima. This combined method effectively improves both search efficiency and solution quality.

[0056] (2) Implement dynamic parameter adjustment to enhance adaptability, and use a multi-layer neural network to improve the accuracy of spot center extraction, specifically:

[0057] Multilayer neural networks (MLPs) are used to learn and adjust algorithm parameters in real time. By extracting features from the input data, the neural network can dynamically optimize the velocity and position update rules in the particle swarm, thereby enhancing the algorithm's adaptability and flexibility in different environments and improving the extraction accuracy of the spot center.

[0058] (3) Implement noise filtering and preprocessing to improve data quality.

[0059] (4) Optimize positioning accuracy through high-order fitting techniques and iterative mechanisms, specifically:

[0060] First, the positional information of multiple light spots is organized into a coordinate set. Then, a polynomial fitting is performed using the least squares method to obtain coefficients describing the changes in the center position of the light spots. Next, a weighted average of the center positions of the light spots is calculated to reflect their changes. The fitting accuracy is evaluated through residual adjustment, and the model parameters are optimized based on error feedback, followed by refitting. Finally, the distance changes to the center of the light spots before and after the fitted model are accurately calculated.

[0061] These improvements significantly enhance the system's measurement performance in complex environments, ensuring efficient and reliable data output.

[0062] Figure 2 A schematic diagram of the structure of the high-precision multi-target non-contact measurement system according to the present invention is shown.

[0063] like Figure 2 As shown, the entire system consists of a CMOS module that sends the acquired data to an FPGA data processing module, which then transmits the processed image to the host computer via a bus.

[0064] Figure 3 A schematic diagram illustrating the setup of a satellite on-orbit vibration state system acquisition process according to an embodiment of the present invention is shown; a filter is added before acquisition to filter out stray light from the environment to a certain extent. This embodiment applies a high-precision multi-target non-contact measurement system based on a particle swarm artificial neural network algorithm to the satellite on-orbit vibration state system for measurement.

[0065] Figure 4 A flowchart of a satellite high-precision multi-target non-contact measurement method according to the present invention is shown; Figure 5 A schematic diagram illustrating the principle of a satellite-based high-precision multi-target non-contact measurement method according to the present invention is shown.

[0066] like Figure 4 and Figure 5As shown, during non-contact ranging, the entire satellite outputs a drive signal from the FPGA data processing module, i.e., acquires the modulation signal output by the FPGA data processing module; drives the laser module and CMOS module, controlling the laser module to emit laser light, which is then split and directed to the target; the CMOS module acquires multiple laser spots on the target surface, converts the acquired analog signals into digital signals, and the FPGA data processing module processes the digital signals, performing distortion correction and filtering on the resulting image, and outputs the multi-target image to the host computer; wherein, the FPGA data processing module obtains the multi-target image based on each frame of data; the host computer obtains the target position information based on the multi-target image; and the measurement result is obtained based on the target position information. The target position information includes the position changes of multiple laser spots, and the host computer obtains the non-contact measurement result based on the position changes of these multiple laser spots. The laser is a ranging laser.

[0067] In a preferred embodiment, the analog signal corresponds to data of N frames. The FPGA data processing module obtains the distortion-corrected image of each of the N frames based on the camera's intrinsic and extrinsic parameters. Based on the distortion-corrected image obtained from each of the N frames, a denoised image is obtained as the output of the FPGA data processing module and output to the host computer.

[0068] More specifically, after receiving the denoised multi-target image, the host computer optimizes it using a particle swarm optimization (PSO) artificial neural network algorithm. Then, it uses the gray-scale centroid method to determine the centroids of multiple light spots (i.e., the centers of the light spots). The average value of these centroids is used to fit the result, yielding the change in the laser's position on the object surface over time, thus obtaining the real-time non-contact measurement result. Specifically, the distance change is obtained by fitting the average value of the position change of the light spot centers, providing reliable data for subsequent systems.

[0069] More specifically, the steps for locating multiple light spots in the denoised multi-target image using a particle swarm artificial neural network algorithm include the following:

[0070] Step 1: Acquire image data containing multiple light spots.

[0071] Step 2: Randomly initialize a group of particles, each particle representing a possible combination of light spot positions.

[0072] Step 3: Predict the position combination of each particle, calculate the error as fitness, and use a multi-layer neural network to dynamically adjust the particle swarm parameters to improve fitness.

[0073] In step three, a hybrid optimization algorithm (particle swarm optimization and simulated annealing) is used to predict the position combinations of each particle, and the error is calculated as the fitness. A multilayer neural network is then used to dynamically adjust the particle swarm parameters to improve fitness. More specifically, for each particle, its fitness value is calculated, and the error between the predicted result and the actual spot position is evaluated using a multilayer perceptron (MLP) with a loss function (mean squared error). According to the formula of the particle swarm optimization (PSO) algorithm, the velocity and position of each particle are updated. Simulated annealing is introduced to control the temperature and adjust the probability of accepting worse solutions, ensuring that a wider solution space can be explored during the optimization process.

[0074] Step 4: Update the global optimum position based on fitness. Update the position and velocity of each particle based on the global optimum position and the individual optimum position.

[0075] Step 5: Update the artificial neural network parameters based on particle positions to improve localization accuracy. Specifically, after each iteration, the parameters of the MLP are dynamically adjusted using the backpropagation algorithm based on the particle's fitness value to improve the model's predictive ability. The gradient of the loss function with respect to each weight is calculated, and the weights are updated.

[0076] Repeat steps three through five until a stopping condition is met, such as reaching the maximum number of iterations or the error being less than a threshold.

[0077] Step 6: Output the converged spot position information to achieve accurate positioning of multiple spots.

[0078] This process involves performing a high-order fit on multiple target locations to obtain a value corresponding to the current time. The results of non-contact measurement are then derived based on the changes in these values ​​at different times. Specifically, the position data of multiple light spots at the same time are organized into a set of position coordinates. Then, methods such as least squares are used to perform polynomial fitting on these position data to obtain polynomial coefficients describing the change in the position of the light spot center. Next, a weighted average of the light spot center position is calculated based on the fitted polynomial, reflecting the change in the position of the light spot center. The accuracy of the fitting is evaluated by adjusting the residuals of the fitting results for each iteration, and the model parameters are continuously optimized based on error feedback, leading to refitting. This yields the distance change results of the light spot center before and after the current time.

[0079] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A satellite multi-target non-contact measurement system, characterized in that: include: FPGA data processing module, laser module, COMS module, host computer; The laser module emits laser toward the target according to the laser driving signal from the FPGA data processing module; The CMOS module transmits the analog signals corresponding to the multi-target laser spots collected from the target to the FPGA data processing module; The FPGA data processing module transmits the multi-target images obtained according to the simulation signal to the host computer; The host computer obtains corresponding measurement results according to the multi-target images.

2. The satellite multi-target non-contact measurement system according to claim 1, characterized in that: The laser module emits multiple laser beams, which are then split to generate multiple laser beams and are focused on a target.

3. The satellite multi-target non-contact measurement system according to claim 1, characterized in that: The analog signal corresponds to N frames of data. The FPGA data processing module obtains each of the N frames and performs distortion correction according to the intrinsic and extrinsic parameters of the camera to obtain a distortion-corrected image; based on the distortion-corrected image obtained for each of the N frames, a denoised image is obtained as the output of the FPGA data processing module and output to a host computer.

4. The satellite multi-target non-contact measurement system according to claim 1, characterized in that: The host computer extracts the spot center of the multi-target image by using a particle swarm artificial neural network algorithm, and obtains the distance change result according to the mean value fitting of the position change of the spot center; Particle swarm optimization is used for global search, swarm intelligence is used to guide the search process, and simulated annealing is used to locally adjust the particle position after each update; Use a multi-layer neural network MLP to learn and adjust algorithm parameters in real time; by extracting features from input data, the neural network dynamically optimizes the speed and position update rules in the particle swarm.

5. The satellite multi-target non-contact measurement system according to claim 4, characterized in that: After receiving the denoised multi-target image, the host computer optimizes it based on the particle swarm artificial neural network algorithm, and then uses the grayscale centroid method to obtain the centroid of multiple light spots as the center of the light spot. After fitting the average value of the multiple centroids as the result, the change of the position of the laser on the surface of the object over time is obtained, that is, the real-time non-contact measurement result is obtained.

6. The satellite multi-target non-contact measurement system according to claim 4, characterized in that: In the host computer, Acquire image data containing multiple light spots; randomly initialize a group of particles, each particle represents a possible combination of light spot positions; Particle swarm optimization and simulated annealing are used to predict each particle position combination, and the error is calculated as fitness. Multi-layer neural network is used to dynamically adjust the particle swarm parameters to improve adaptability. The fitness value of each particle is calculated, and the error between the predicted result and the actual spot position is evaluated by multi-layer perceptron MLP. According to the formula of particle swarm optimization PSO algorithm, the speed and position of each particle are updated. Simultaneously, simulated annealing mechanism is introduced to control the temperature to adjust the probability of accepting a poor solution, so as to ensure that a wider solution space can be explored during the optimization process. Update the global optimal position according to the fitness, and update the position and speed of each particle according to the global optimal position and the individual optimal position; update the artificial neural network parameters according to the position of the particle; until the stopping condition is met; wherein, after each iteration, dynamically adjust the parameters of the MLP through the back propagation algorithm according to the fitness value of the particle to improve the prediction ability of the model; calculate the gradient of the loss function to each weight, and update the weight; The converged light spot position information is output to realize the positioning of multiple light spots; the position data of multiple light spots at the same time are organized into a position coordinate set; then, these position data are polynomially fitted using the least squares method and other methods to obtain the polynomial coefficients that describe the position change of the light spot center; next, the weighted average of the light spot center position is calculated based on the fitted polynomial to reflect the position change of the light spot center; the accuracy of the fitting is evaluated by making residual adjustments to the fitting results of each iteration, and the model parameters are continuously optimized and re-fitted based on the error feedback; thus, the distance change results of the light spot center before and after the moment are obtained.

7. The satellite multi-target non-contact measurement system according to claim 6, characterized in that: For the multi-target positions, high-order fitting is performed to obtain a value corresponding to the current moment, and the result of non-contact measurement is obtained according to the changes in the values ​​at different moments.

8. A satellite multi-target non-contact measurement method, characterized in that: include: Driving the laser module and the CMOS module, and controlling the laser module to emit laser beams to the target after beam splitting; The CMOS module is controlled to collect multiple spots of the laser on the target surface, and after converting the collected analog signals into digital signals, the FPGA data processing module processes the digital signals, performs distortion correction and filtering on the obtained images, and then outputs the multi-target images to the host computer; Control the FPGA data processing module to obtain multi-target images based on each frame of data; Controlling the host computer to obtain target position information based on the multi-target images; Based on the target position information, the measurement result is obtained.

9. The satellite multi-target non-contact measurement method according to claim 8, characterized in that: A satellite multi-target non-contact measurement system as described in any one of claims 1 to 7 is adopted.

10. A satellite, characterized in that: A satellite multi-target non-contact measurement system comprising any one of claims 1 to 7; or a satellite multi-target non-contact measurement method comprising claim 8 or 9.

Images