A method and device for simulating the whole process of lunar pinpoint landing of a GNC system

Abstract

The application discloses a kind of GNC system lunar fixed-point landing's whole-process simulation method and device, belong to spaceflight simulation field.The method includes: S1 according to landing task working condition, landing vehicle working condition parameter is initialized configuration, obtains initial parameter data set;S2, landing vehicle is simulated and calculated to multiple model coupling, according to simulation data set and moon surface benchmark database generation navigation area topographic map;S3, according to simulation data set, determine whether simulation satisfies preset termination condition, if yes, end simulation and output the final motion state of landing vehicle and navigation area topographic map, otherwise execute step S4;S4, according to the physical characteristics of airborne sensor, error simulation is carried out to simulation data set and navigation area topographic map, obtain multi-source equivalent measured data set, and image and point cloud data are matched and analyzed, obtain the pose data set and optimal safe landing point of landing vehicle;S5, according to the deviation between simulation analysis result and preset target landing parameter, determine the control instruction of landing vehicle, and return to execute step S2 after updating control input.The application can effectively improve the simulation efficiency and result reliability of GNC system.

Description

Technical Field

[0001] This invention relates to the field of aerospace simulation technology, and in particular to a full-process simulation method and apparatus for lunar fixed-point landing using a GNC system. Background Technology

[0002] The rugged terrain and poor lighting conditions of the lunar south pole pose a significant challenge to a safe and precise landing. To achieve the mission objective of a high-precision, fixed-point soft landing in the polar region, the lander's GNC (Guidance, Navigation, and Control) system requires substantial modifications and technological innovations in navigation and guidance compared to previous soft landing missions. For navigation, an optical navigation camera and terrain imagery navigation algorithms have been added to address the accuracy issues associated with fixed-point landing on the lunar surface. For guidance, full-dimensional control over the terminal position and velocity is required.

[0003] In related technologies, traditional simulation verification methods cannot realistically simulate the image acquisition and processing process of a lander during landing, nor can they verify newly added algorithms. This results in the inability to perform fully digital closed-loop simulation verification, leading to problems such as low simulation efficiency and unreliable results.

[0004] Therefore, there is an urgent need for a full-process simulation method and device for lunar fixed-point landing using a GNC system to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention provides a full-process simulation method and apparatus for lunar stationary landing using a GNC system, which can effectively improve the simulation efficiency and reliability of the GNC system. The technical solution is as follows: On the one hand, a full-process simulation method for lunar fixed-point landing using a GNC system is provided, the method comprising: S1. Initialize the lander's operating parameters according to the landing mission conditions to obtain the initial parameter dataset for simulation calculations. S2. Perform multi-model coupled simulation calculations on the lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and generate a flight area topographic map that fits the real lunar surface details based on the simulation dataset and lunar reference database. S3. Determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, end the simulation and output the final motion state of the lander and the topographic map of the flight area; otherwise, proceed to step S4. S4. Based on the physical characteristics of the airborne sensors, the simulation dataset and the flight area topographic map are used to simulate errors, obtain a multi-source equivalent measured dataset, and perform matching analysis on the generated noisy images and point cloud data to obtain the lander's pose dataset and the optimal safe landing point. S5. Based on the deviation between the simulation analysis results and the preset target landing parameters, determine the control commands for the lander, and after updating the control inputs in the parameter dataset according to the control commands, return to step S2.

[0006] On the other hand, a full-process simulation device for lunar precise landing using a GNC system is provided, the device comprising: The initialization unit is used to initialize and configure the lander's operating parameters according to the landing mission conditions, and obtain the initial parameter dataset for simulation calculations. The simulation unit is used to perform multi-model coupled simulation calculations on the lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and to generate a flight area topographic map that fits the real lunar surface details based on the calculated simulation dataset and lunar surface reference database. The judgment unit is used to determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, the simulation ends and the final motion state of the lander and the topographic map of the flight area are output; otherwise, the function of the simulation unit is executed. The simulation unit is used to simulate errors in the simulation dataset and the flight area topographic map based on the physical characteristics of the airborne sensors, to obtain a multi-source equivalent measured dataset, and to perform matching analysis on the generated noisy images and point cloud data to obtain the lander's pose dataset and the optimal safe landing point. The control unit is used to determine the control commands of the lander based on the deviation between the simulation analysis results and the preset target landing parameters, and then return to the simulation unit to perform the function after updating the control input in the parameter dataset according to the control commands.

[0007] On the other hand, a computer device is provided, the computer device including a memory and a processor, the memory for storing computer programs, and the processor for executing the computer programs stored in the memory to implement the steps of the full-process simulation method for lunar fixed-point landing of the GNC system described above.

[0008] On the other hand, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, the steps of the above-described full-process simulation method for lunar fixed-point landing of the GNC system are implemented.

[0009] On the other hand, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the full-process simulation method for lunar fixed-point landing of the GNC system described above.

[0010] The technical solution provided by this invention can bring at least the following beneficial effects: First, by using multi-model coupled dynamics and lunar environment calculations, high-precision state truth values ​​and localized refined terrain data are output, faithfully reproducing the orbital space environment and landing area topography, significantly improving the confidence of simulation results. Next, through the image-in-the-loop mechanism of multi-source sensing simulation and pose calculation, the entire sensing link from sensor error to algorithm output is covered, fully verifying the accuracy and robustness of the image navigation algorithm. Finally, through closed-loop iteration of GNC control calculation and execution feedback, the dynamic characteristics of guidance, control, and execution mechanisms are realistically reproduced. This method can achieve low-cost, high-efficiency, and highly reliable simulation verification of the lander's GNC system scheme on the ground. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of the full-process simulation method for lunar fixed-point landing using a GNC system provided in an embodiment of the present invention; Figure 2 This is an interactive schematic diagram of a GNC system simulation module provided in an embodiment of the present invention; Figure 3 This is a structural diagram of a full-process simulation device for lunar fixed-point landing using a GNC system, provided in an embodiment of the present invention. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0014] As mentioned earlier, traditional simulation verification methods lack simulation of the image acquisition process and verification of new algorithms, resulting in the inability to obtain closed-loop verification results for the entire process. Consequently, the reliability of the simulation results is difficult to meet the needs of practical applications.

[0015] Based on this, the concept of this invention is to design a full-process simulation method for the closed-loop architecture of the lander's GNC system, accurately simulate the entire process of lunar fixed-point landing, and realize the effective verification of the GNC scheme.

[0016] The following describes the specific implementation of the above concept.

[0017] Please refer to Figure 1 This invention provides a full-process simulation method for lunar precise landing using a GNC system, the method comprising: Step S1: Initialize the lander's operating parameters according to the landing mission conditions to obtain the initial parameter dataset for simulation calculations; Step S2: Perform multi-model coupled simulation calculations on the lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and generate a flight area topographic map that fits the real lunar surface details based on the calculated simulation dataset and lunar surface reference database. Step S3: Determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, end the simulation and output the final motion state of the lander and the topographic map of the flight area; otherwise, proceed to step S4. Step S4: Based on the physical characteristics of the airborne sensors, perform error simulation on the simulation dataset and the flight area topographic map to obtain a multi-source equivalent measured dataset. Then, perform matching analysis on the generated noisy images and point cloud data to obtain the lander's pose dataset and the optimal safe landing point. Step S5: Based on the deviation between the simulation analysis results and the preset target landing parameters, determine the control commands for the lander, and after updating the control inputs in the parameter dataset according to the control commands, return to step S2.

[0018] In this embodiment of the invention, firstly, high-precision state truth values ​​and localized refined terrain data are output through multi-model coupled dynamics and lunar environment calculations, faithfully reproducing the on-orbit space environment and landing area topography, significantly improving the confidence level of the simulation results. Next, an image-in-the-loop mechanism for multi-source sensing simulation and pose calculation is used, covering the entire sensing link from sensor error to algorithm output, which can fully verify the accuracy and robustness of the image navigation algorithm. Finally, through closed-loop iteration of GNC control calculation and execution feedback, the dynamic characteristics of the guidance, control, and execution mechanisms are realistically reproduced. This method can achieve low-cost, high-efficiency, and highly reliable simulation verification of the lander's GNC system scheme on the ground.

[0019] The following description Figure 1 The execution method for each step is shown.

[0020] First, for step S1, the lander's operating parameters are initialized and configured according to the landing mission conditions to obtain the initial parameter dataset for simulation calculation.

[0021] To align with the actual GNC closed-loop system architecture of a lander, this embodiment constructs a fully digital closed-loop simulation verification architecture comprising eight core simulation functional modules. These modules specifically include a lunar environment module, a high-precision lunar surface module for the flight area, a dynamics module, a measurement module, an image simulation module, an image algorithm module, a GNC algorithm module, and an execution module. The interrelationships between these modules are as follows: Figure 2 As shown.

[0022] Specifically, the dynamics module integrates three types of dynamic models: attitude, orbit, and fluid sloshing, to achieve six-degree-of-freedom attitude and orbital motion calculations for the lander. The lunar environment module simulates lunar terrain, illumination, and the starry sky environment. The high-precision lunar surface module simulates the landing area's lunar surface topography and geomorphology. The measurement module simulates the measurement outputs of multiple sensors based on the system's actual parameters. The image simulation module reproduces the on-orbit imaging and detection effects of the star sensor, navigation camera, and lidar. The image algorithm module integrates multiple image matching, terrain recognition, and safe landing point determination algorithms. The GNC algorithm module achieves multi-source information fusion, guidance command calculation, and attitude control output. The execution module simulates engine execution characteristics, completing the conversion of commands into forces and torques.

[0023] The core formula for coupling between modules is: In the formula, for Lander dynamic state variables at any given moment; Output control quantities to the actuator; For lunar environmental disturbance variables; This is the system's global iterative mapping function.

[0024] To enable subsequent simulation steps, based on the above simulation architecture, it is necessary to perform global initialization of the core variables of the simulation system to provide an initial state parameter dataset for iterative simulation calculations. The simulation time is described using Beijing time (year, month, day, hour, minute, second); The quaternion parameters of the lander's position, velocity, and attitude are uniformly defined based on the lunar inertial frame. The lander's angular velocity parameters are defined based on the body coordinate system; At the same time, the initialization of core state parameters such as lander mass is completed to ensure that the initial working conditions are consistent with the actual landing mission conditions.

[0025] The initial motion state vector of the lander simulation is defined as follows: In the formula, The initial position of the lunar inertial frame of reference; The initial velocity of the lunar inertial frame; The initial attitude quaternion; This represents the initial angular velocity of the system. This represents the initial mass of the lander.

[0026] Then, for step S2, multi-model coupled simulation calculations are performed on the lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and a flight area topographic map that fits the real lunar surface details is generated based on the simulation dataset and lunar surface reference database.

[0027] This step utilizes the functions of the dynamics module, lunar environment module, and high-precision lunar surface module to generate motion simulation data of the lander and the corresponding pre-landing area topographic map.

[0028] In this embodiment of the invention, step S2 includes the following processing steps: The lander state parameters, control input data, and celestial gravitational parameters output from the previous simulation cycle are respectively input into the preset orbit, liquid sloshing, and attitude dynamics models for hierarchical iterative calculation to obtain the mathematical simulation data of the lander in the current simulation cycle; wherein, the mathematical simulation data includes a six-degree-of-freedom motion state vector and sloshing disturbance torque.

[0029] Specifically, in this embodiment, the dynamics module integrates three types of dynamic digital models to complete the iterative update of the lander's motion state in a hierarchical manner.

[0030] First, the orbital dynamics model is invoked, integrating the core forces such as the gravitational pull of the Moon, Earth, and Sun, as well as engine thrust, to update the lander's real-time mass, position, and velocity parameters. The orbital dynamics equations in the lunar inertial frame are as follows: In the formula: The gravitational constant of the moon; The distance from the lunar center to the lander; This refers to the non-spherical gravitational perturbation acceleration of the Moon; The acceleration due to gravitational perturbation of the Earth and the Sun as third bodies; For engine thrust; For the real-time quality of the lander.

[0031] Next, the liquid sloshing disturbance torque model is invoked to calculate the disturbance torque generated by propellant sloshing, thus reproducing the orbital disturbance condition. The liquid sloshing disturbance torque model adopts an equivalent simple pendulum model. The formula for calculating the liquid sloshing disturbance torque is: In the formula, Number of propellant tanks; For the first The vector of the liquid sloshing in a propellant tank relative to the satellite's center of mass, which is equivalent to the suspension point of a simple pendulum. For the first The sloshing force vector given by the equivalent simple pendulum model of liquid sloshing in a propellant tank.

[0032] Finally, the attitude dynamics model is run, and the lander angular velocity and attitude quaternions are updated by combining the disturbance torque and control torque to complete the single-cycle dynamics calculation. The attitude dynamics equations are: In the formula, Here is the rotational inertia matrix of the lander; To control the torque; This refers to the shaking disturbance torque.

[0033] Attitude quaternion update equation: In the formula, is the attitude quaternion oblique symmetric matrix corresponding to the angular velocity.

[0034] In summary, by integrating the updated three translational degrees of freedom (including three-dimensional position components and three-dimensional velocity components) and three rotational degrees of freedom (including attitude quaternions and three-dimensional angular velocity components), a six-degree-of-freedom motion state vector is obtained.

[0035] Furthermore, based on the astronomical ephemeris and star catalog data corresponding to the current simulation period, the position and velocity of celestial bodies, coordinate system transformation relationships, and slant range are calculated to obtain the lunar environment simulation data for the current simulation period; wherein, the lunar environment simulation data includes Sun-Moon-Earth motion parameters, coordinate transformation matrix between the Moon-fixed and Moon-centric inertial systems, true values ​​of slant range, and a dataset of the illumination starry sky reference environment.

[0036] Specifically, in this embodiment, the lunar environment module integrates high-precision ephemeris data for the Sun, Earth, Moon, and stars, as well as lunar terrain and texture models, and calculates environmental parameters based on authoritative astronomical data. The ephemeris data for the Sun, Earth, and Moon is obtained by loading the DE ephemeris from JPL. Stellar ephemeris data is obtained by querying the European Space Agency's Hipparcos Catalogue to ensure the accuracy of the stellar environment simulation.

[0037] In the calculation process, the Earth's position and velocity in the heliocentric inertial frame and the Moon's position and velocity in the geocentric inertial frame are first calculated using the DE ephemeris. The target time is set as the Julian day corresponding to the center-of-mass mechanical time. The starting Julian day of the corresponding sub-interval in the ephemeris is The duration of the sub-interval is Mapping time to the Chebyshev standard interval [-1,1]: The general formula for calculating any coordinate component of the Earth's or Moon's position is: In the formula, Let Chebyshev's fitting coefficients be the coefficients for this component; The order of the polynomial; for Chebyshev polynomial.

[0038] The general formula for calculating any coordinate component of the Earth's or Moon's velocity is: In the formula, for The derivative of a Chebyshev polynomial.

[0039] Further transformation yields the position and velocity parameters of the Sun and Earth in the lunar-centric inertial frame. Finally, combining DE ephemeris data, the coordinate transformation matrix between the lunar-fixed frame and the lunar-centric inertial frame is calculated, providing a coordinate system transformation basis for subsequent terrain matching and pose calculation.

[0040] The formula for calculating the lunar-fixed system position vector is: In the formula, The position vector of the lunar inertial frame; Let be the time-varying transformation matrix from the lunar-centric inertial frame to the lunar-fixed frame.

[0041] The true value calculation formula for the ranging beam yaw range is: In the formula, For the lander's position; The actual intersection of the ranging beams on the lunar surface is also known as the footprint; This is the true value of the slant distance.

[0042] Furthermore, based on simulation data, the effective exploration area of ​​the lander is located, and a high-precision terrain texture data and coordinate map of the flight area containing the effective exploration area is generated based on the lunar surface reference database.

[0043] In this embodiment of the invention, a high-precision lunar surface module is needed to generate a topographic map of the flight area. This module focuses on the flight area and sensor detection field of view of the lander throughout its powered descent. It integrates the highest-precision lunar digital elevation model (DEM) and digital orthophoto map (DOM) data currently available. The topography, texture features, and elevation undulations of the core landing area are then finely reconstructed. The flight area topographic map is generated through the following steps: Based on the lander's real-time pose determined by the six-degree-of-freedom motion state vector and the inherent field-of-view constraints of the onboard sensors, the spatial coverage of the sensor's detection beam at the current simulation moment is calculated in the lunar-centered inertial frame.

[0044] Specifically, with the sensor's optical axis as the center, the field of view boundary is defined according to the maximum field of view thresholds in both the pitch and azimuth dimensions, satisfying the field of view spatial constraint equation: In the formula, These are the field of view elevation angle and azimuth angle, respectively. This is the maximum field of view of the sensor.

[0045] By transforming the field of view in the inertial frame to the lunar fixed coordinate system using a coordinate transformation matrix, the projected boundary of the sensor's field of view on the lunar surface is obtained, providing a spatial reference for subsequent region determination.

[0046] Furthermore, the calculated field of view spatial range is intersected with the global lunar surface topography benchmark to determine the intersection area of ​​the field of view coverage and the lunar surface topography as the effective detection area within this simulation cycle.

[0047] Specifically, the field-of-view inclusion determination function filters the global terrain grid points one by one, retaining only the terrain points that fall within the field of view coverage, and eliminating redundant calculations of all global terrain data outside the field of view. The determination logic satisfies the following: In the formula, To effectively detect the terrain and area of ​​the flight path; This is a field-of-view inclusion function; it returns true if the point is within the field of view, and false otherwise.

[0048] The final output is the boundary coordinate set of the effective detection area (ROI), which clarifies the calculation range of this terrain reconstruction.

[0049] Furthermore, based on the boundary coordinate set of the effective detection area, the DEM elevation data and DOM texture data in the lunar surface reference database are cropped and extracted respectively to obtain the precise elevation value and lunar surface grayscale texture features of each terrain grid point in turn.

[0050] Specifically, high-precision DEM digital elevation data within the ROI area is cropped to obtain the precise elevation value of each terrain grid point, retaining the highest resolution of the original data to support subsequent terrain slope and roughness calculations and LiDAR imaging simulations; DOM digital orthophoto data within the ROI area is cropped to obtain the lunar surface grayscale texture features of the corresponding area to support subsequent navigation camera optical image simulations.

[0051] Finally, spatial registration and coordinate system correction are performed on the DEM elevation data and DOM texture data obtained by cropping and extraction, and the registered and corrected DOM texture data is mapped onto the DEM three-dimensional elevation terrain model to obtain a flight area terrain map that fits the real lunar surface details.

[0052] Specifically, using the lunar fixed coordinate system as a unified benchmark, the coordinate offset of the clipped data is corrected, and all terrain point coordinates are uniformly transformed to the lunar fixed coordinate system, which is fully aligned with the coordinate systems of the environment module and the dynamics module; the one-to-one mapping and registration of DEM mesh and DOM texture pixels is completed, eliminating spatial misalignment between data and ensuring the consistency of the correspondence between terrain geometry and texture features.

[0053] The registered DOM grayscale texture data is mapped onto the DEM three-dimensional elevation terrain model to restore the micro-landform features of the landing flight area, such as craters, rocks, and undulating terrain. This meets the simulation accuracy requirements of the image matching navigation algorithm for terrain details, and finally forms a flight area terrain map that closely matches the details of the real lunar surface.

[0054] For step S3, determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, end the simulation and output the final motion state of the lander and the topographic map of the flight area; otherwise, proceed to step S4.

[0055] In this embodiment of the invention, after the terrain data is output by the high-precision lunar surface module of the flight area and the lander position parameters are output by the dynamics module, the relative distance between the lander and the lunar surface can be calculated in real time to determine the landing status.

[0056] For example, if the lander's real-time altitude is detected... Less than or equal to the landing height threshold or simulation runtime Reaching the set maximum simulation duration If the simulation process is deemed complete, the simulation system will exit. If the termination condition is not met, subsequent simulation steps will continue, and the next iteration cycle will begin.

[0057] The simulation termination condition is: In the formula, The simulation is now complete. To continue iterative simulation.

[0058] For step S4, error simulation is performed on the simulation dataset and the flight area topographic map based on the physical characteristics of the airborne sensor to obtain a multi-source equivalent measured dataset. The generated noisy images and point cloud data are then matched and analyzed to obtain the lander's pose dataset and the optimal safe landing point.

[0059] In this embodiment of the invention, this step invokes the functions of the measurement module, image simulation module, and image algorithm module. The multi-source equivalent measured dataset is obtained through error simulation by the measurement module and image simulation module, including the following steps: Based on the system error coefficient and measurement noise covariance parameter of the airborne sensor, error degradation processing is performed on the mathematical simulation data to obtain mathematically equivalent measured data. : In the formula, These are mathematical simulation measurements; This refers to the systematic error of the sensor; It is Gaussian random noise. To measure the noise covariance matrix.

[0060] Specifically, the mathematically equivalent measured data is obtained by processing the mathematical simulation data through the measurement module. This module integrates a high-precision digital model of an inertial measurement unit (IMU), a distance sensor, and a velocity sensor. Based on the actual motion state of the lander output from the dynamics module, and the actual environmental parameters output from the lunar environment module and the high-precision lunar surface module of the flight area, the mathematical simulation data is degraded by combining the sensor error characteristics and working mechanisms. This simulation yields measurement data for angle increments, velocity increments, slant range, and relative velocity, including actual on-orbit errors. Finally, following the airborne single-machine interface protocol standard, a standardized output structure is generated to complete the output of the measurement information.

[0061] For example, the true angular velocity and acceleration information are extracted from the dynamic state vector, and systematic errors such as zero bias and scale factor, as well as Gaussian random noise, are superimposed to generate angle increment and velocity increment measurement data with corresponding accuracy levels. Based on the true value of the slant distance, the ranging system bias and random noise are introduced to generate the lunar slant distance measurement value with error; By combining the lander's actual velocity relative to the lunar surface with the terrain characteristics and superimposing velocity measurement errors, a velocity measurement value relative to the lunar surface is generated.

[0062] These data cover four core measurement data types: angle increment, velocity increment, slant distance, and relative velocity, which are used for multi-source information fusion in the subsequent GNC module.

[0063] Noise degradation was performed on lunar environment simulation data and flight area topographic maps based on the inherent imaging parameters of the airborne sensor, resulting in degraded measured starry sky image data, measured lunar surface grayscale image data, and measured laser point cloud data. The measured image data was degraded using the following formula: In the formula, This is a measured grayscale image; For the ideal original image; This is the lens distortion mapping function; For resolution sampling degradation function; Gaussian noise in the image; Measured laser point cloud data Degeneracy is performed using the following formula: In the formula, This represents the true value for point cloud ranging. Gaussian noise for point cloud ranging.

[0064] Specifically, the image simulation module designed in this embodiment is based on the physical working mechanisms of star sensors, navigation cameras, and lidar, and builds a high-precision image simulation model. Combining core parameters such as the field of view, resolution, lens distortion, and on-orbit noise of various sensors, the original ideal image data is degraded. It then simulates and outputs starry sky images from the star sensor, lunar surface grayscale optical images from the navigation camera, and lunar surface 3D point cloud data from the lidar, respectively. This fully reproduces the on-orbit imaging and detection effects of various detection devices.

[0065] For example, based on stellar reference star catalog data and the lander's inertial attitude, an ideal star observation image is generated. Then, after three steps of degradation processing—lens distortion mapping, resolution sampling degradation, and Gaussian noise superposition—an equivalent on-orbit star sensor observation image is output. Using high-precision terrain texture data of the flight area as the scene source, combined with the real-time pose and lighting parameters of the lander, an ideal lunar optical image is rendered and generated. Lens distortion, resolution degradation and imaging noise are introduced simultaneously to reproduce the real imaging effect of the navigation camera. An ideal laser ranging point cloud is generated based on high-precision DEM elevation data. Gaussian random noise is superimposed on the ranging value of each point to simulate the real detection error of the lidar, and the resulting three-dimensional point cloud data of the lunar surface with errors is output.

[0066] After obtaining the multi-source equivalent measured dataset, it is necessary to perform matching analysis on the generated noisy images and point cloud data to obtain the lander's pose dataset and the optimal safe landing point.

[0067] This embodiment calls the image algorithm module to process image and point cloud data in three categories, including the following steps: Based on the stellar catalog data, stellar barycenter extraction and catalog matching were performed on the measured starscape image data. The matching results were then optimized using least squares to obtain the optimal attitude matrix of the lander relative to inertial space. : In the formula, To observe stellar vectors; This serves as the reference star vector for the star catalog. This is the attitude matrix; Based on the preset terrain matching and fusion algorithm, feature matching is performed on the measured lunar surface grayscale image data to obtain the optimal pose transformation matrix of the lander relative to the lunar solid frame. : In the formula, Let be the pose transformation matrix to be iteratively optimized; For matching operators of the sum of squared differences in image gray levels; Real-time observation of lunar surface grayscale images; This is the baseline reference image after pose mapping; The landing safety factor for each region is determined based on the measured laser point cloud data, including elevation deviation, terrain slope, and elevation roughness. ; In the formula, This refers to elevation deviation; The slope of the terrain; For elevation roughness; All are preset weighting coefficients; The landing safety coefficients of all regions were compared and screened to determine the region with the highest safety probability as the optimal safe landing point.

[0068] For step S5, based on the deviation between the simulation analysis results and the preset target landing parameters, the control commands for the lander are determined, and after updating the control inputs in the parameter dataset according to the control commands, the process returns to step S2.

[0069] In this embodiment of the invention, this step calls the functions of the GNC algorithm module and the execution module to process the simulation analysis results and complete the core calculation process of closed-loop control.

[0070] The GNC algorithm module integrates a complete set of navigation, guidance, and attitude control algorithms used in real-world on-orbit applications of the lander, achieving closed-loop control calculation. First, the navigation algorithm fuses multi-source information from the IMU, ranging and velocity sensors, and image processing results to obtain a high-precision estimate of the lander's attitude. Then, the guidance algorithm combines the lander's real-time position, velocity, mass parameters, and the preset target landing point to calculate guidance commands such as target attitude and orbit control engine thrust. Finally, the attitude control algorithm compares the deviation between the real-time attitude and the target attitude to calculate attitude control commands such as the attitude control engine jet pulse width. All control commands are synchronously transmitted to the execution module.

[0071] The execution module integrates high-precision physical models of the attitude control and orbit control engines, fully considering dynamic response characteristics such as engine start-up delay, T90 ascent time (the time for thrust to increase from zero to 90% of the rated value), engine shutdown delay, and T10 descent time (the time for thrust to decrease from the rated value to 10%). First, based on GNC control commands and real-time engine internal parameters, the actual engine output thrust is calculated. Second, considering the engine installation position and the lander's real-time center of mass position, the lever arm vector of each engine is calculated. Then, the individual engine torque vector is obtained by cross-multiplying the lever arm vector and the thrust vector. Finally, the total force and total torque acting on the lander are summed and fed back to the dynamics module.

[0072] Based on this, the processing steps in step S5 include the following steps: Based on the navigation status of the previous simulation cycle, multi-source information fusion and optimal filtering correction are performed on the mathematical equivalent measured data and pose dataset of the lander to obtain the estimated value of the lander's navigation status for the current cycle.

[0073] Specifically, using the angle increment and velocity increment of the inertial measurement unit as input, the state extrapolation is completed through inertial mechanics arrangement to obtain the prior state estimate and the corresponding prior covariance matrix at the current moment, thus completing the time update of the state. The pose results from image processing and the measurements from the ranging and velocity sensors are combined into a multi-source observation vector. The observation residuals and filter gain matrix are then calculated using the system observation matrix to perform optimal correction on the prior state, yielding the optimal posterior state estimate for the current time step. The formula for multi-source information fusion and update is: In the formula, This is the filter gain matrix; This is the system observation matrix; This represents the multi-source fusion observation vector; for Time-prior state estimation; for Posterior optimal state estimation at time step 1.

[0074] Furthermore, the deviation between the final guidance target motion parameters and the current motion parameters of the lander is determined, and the desired acceleration command is generated based on the deviation. The motion parameters include position and velocity. The final guidance target motion parameters are determined based on the preset nominal target landing point coordinates and the optimal safe landing point coordinates, and the current motion parameters are determined based on the estimated values ​​of the lander's navigation state. Specifically, by combining the nominal landing point preset for the mission with the optimal safe landing point selected by image recognition, the position and velocity parameters of the final guided target are determined.

[0075] Based on the real-time status of the navigation output, the deviation between the lander's current position and velocity and the target status is calculated.

[0076] The proportional-derivative (PD) guidance law is used to generate the desired acceleration command based on the position and velocity deviation. The landing guidance law formula is: In the formula, These are the proportional and differential guidance coefficients; For the lander's real-time position and velocity; For the target position and velocity.

[0077] Furthermore, the desired acceleration command is calculated and converted based on the lander's real-time mass to obtain the orbital control engine thrust command, as well as the lander's target attitude and target angular velocity. Based on the deviation between the estimated value and the target value of the current cycle of attitude and angular velocity adjustment, the desired control torque and engine thrust distribution of the lander are calculated to obtain the attitude control engine jet pulse width command set.

[0078] Specifically, the deviation quaternion between the current attitude and the target attitude is solved by quaternion multiplication, and the angular velocity deviation is calculated simultaneously.

[0079] Using a proportional-derivative attitude control law, the desired control torque of the entire system is calculated based on the attitude deviation and angular velocity deviation. The formula for calculating the attitude control command is as follows: In the formula, These are the angular velocity and attitude error control coefficients. The target angular velocity; Real-time angular velocity; This represents the attitude quaternion deviation.

[0080] Based on the thrust command of the orbit control engine and the jet pulse width command set of the attitude control engine, the dynamic thrust response of the engine and the resultant force and resultant torque of the lander are updated to obtain the total control force and total control torque used to adjust the lander.

[0081] Specifically, the engine dynamic thrust response model is as follows: In the formula, This is the engine's rated thrust; To simulate real-time; Pure delay for engine start-up; For thrust rise time; This is a pure delay for engine shutdown; This refers to the thrust descent time.

[0082] The formula for calculating the torque of a single engine is: In the formula, For the first The lever arm vector of the engine relative to its center of mass; For the first The engine outputs thrust vector.

[0083] The resultant force and resultant torque of the entire device are: In the formula: This refers to the number of engines.

[0084] After completing the above processing, proceed to step S2 to enter the next simulation iteration. Repeat the iteration until the simulation termination condition is met.

[0085] Please refer to Figure 3 This invention provides a full-process simulation device for lunar precise landing using a GNC system. The device includes: Initialization unit 300 is used to initialize and configure the lander's operating parameters according to the landing mission conditions, so as to obtain an initial parameter dataset for simulation calculations. Simulation unit 302 is used to perform multi-model coupled simulation calculations on lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and generate a flight area topographic map that fits the real lunar surface details based on the calculated simulation dataset and lunar surface reference database. The judgment unit 304 is used to determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, the simulation ends and the final motion state of the lander and the topographic map of the flight area are output; otherwise, the function of the simulation unit is executed. The simulation unit 306 is used to perform error simulation on the simulation dataset and the flight area topographic map according to the physical characteristics of the airborne sensor, to obtain a multi-source equivalent measured dataset, and to perform matching analysis on the generated noisy images and point cloud data to obtain the pose dataset of the lander and the optimal safe landing point. The control unit 308 is used to determine the control commands of the lander based on the deviation between the simulation analysis results and the preset target landing parameters, and to return to the function of the simulation unit after updating the control input in the parameter dataset according to the control commands.

[0086] In this embodiment of the invention, the step of performing multi-model coupled simulation calculations on lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and generating a flight area topographic map that fits the details of the real lunar surface based on the calculated simulation dataset and the lunar surface reference database, includes: The lander state parameters, control input data, and celestial gravitational parameters output from the previous simulation cycle are respectively input into preset orbit, liquid sloshing, and attitude dynamics models for hierarchical iterative calculation to obtain the mathematical simulation data of the lander in the current simulation cycle; wherein, the mathematical simulation data includes a six-degree-of-freedom motion state vector and sloshing disturbance torque; Based on the astronomical ephemeris and star catalog data corresponding to the current simulation period, the position and velocity of celestial bodies, coordinate system transformation relationships, and slant range are calculated to obtain the lunar environment simulation data for the current simulation period. The lunar environment simulation data includes Sun-Moon-Earth motion parameters, coordinate transformation matrix between the Moon-fixed and Moon-centric inertial systems, true values ​​of slant range, and a dataset of the illumination starry sky reference environment. Based on simulation data, the effective exploration area of ​​the lander is located, and a high-precision terrain texture data and coordinate map of the flight area is generated based on the lunar surface reference database.

[0087] In this embodiment of the invention, the step of locking the effective detection area of ​​the lander based on simulation data and generating a flight area terrain scene map containing high-precision terrain texture data and coordinates of the effective detection area based on the lunar surface reference database includes: Based on the real-time pose of the lander determined by the six-degree-of-freedom motion state vector and the inherent field-of-view constraints of the airborne sensor, the spatial coverage range of the sensor detection beam at the current simulation moment is calculated in the lunar inertial frame. The calculated field of view spatial range is intersected with the global lunar surface topography benchmark to determine the intersection area of ​​the field of view coverage and the lunar surface topography as the effective detection area within this simulation cycle; Based on the boundary coordinate set of the effective detection area, the DEM elevation data and DOM texture data in the lunar reference database are cropped and extracted respectively to obtain the precise elevation value and lunar grayscale texture features of each terrain grid point in turn. Spatial registration and coordinate system correction are performed on the DEM elevation data and DOM texture data obtained by cropping and extraction. The registered and corrected DOM texture data is then mapped onto the DEM three-dimensional elevation terrain model to obtain a flight area terrain map that fits the real lunar surface details.

[0088] In this embodiment of the invention, the step of performing error simulation on the simulation dataset and the flight area topographic map based on the physical characteristics of the airborne sensor to obtain a multi-source equivalent measured dataset includes: Based on the system error coefficient and measurement noise covariance parameter of the airborne sensor, error degradation processing is performed on the mathematical simulation data to obtain mathematically equivalent measured data. : In the formula, These are mathematical simulation measurements; This refers to the systematic error of the sensor; It is Gaussian random noise. To measure the noise covariance matrix; Noise degradation was performed on lunar environment simulation data and flight area topographic maps based on the inherent imaging parameters of the airborne sensor, resulting in degraded measured starry sky image data, measured lunar surface grayscale image data, and measured laser point cloud data. The measured image data was degraded using the following formula: In the formula, This is a measured grayscale image; For the ideal original image; This is the lens distortion mapping function; For resolution sampling degradation function; Gaussian noise in the image; Measured laser point cloud data Degeneracy is performed using the following formula: In the formula, This represents the true value for point cloud ranging. Gaussian noise for point cloud ranging.

[0089] In this embodiment of the invention, the step of performing matching analysis on the generated noisy image and point cloud data to obtain the lander's pose dataset and optimal safe landing point includes: Based on the stellar catalog data, stellar barycenter extraction and catalog matching were performed on the measured starscape image data. The matching results were then optimized using least squares to obtain the optimal attitude matrix of the lander relative to inertial space. : In the formula, To observe stellar vectors; This serves as the reference star vector for the star catalog. This is the attitude matrix; Based on the preset terrain matching and fusion algorithm, feature matching is performed on the measured lunar surface grayscale image data to obtain the optimal pose transformation matrix of the lander relative to the lunar solid frame. : In the formula, Let be the pose transformation matrix to be iteratively optimized; For matching operators of the sum of squared differences in image gray levels; Real-time observation of lunar surface grayscale images; This is the baseline reference image after pose mapping; The landing safety factor for each region is determined based on the measured laser point cloud data, including elevation deviation, terrain slope, and elevation roughness. ; In the formula, This refers to elevation deviation; The slope of the terrain; For elevation roughness; All are preset weighting coefficients; The landing safety coefficients of all regions were compared and screened to determine the region with the highest safety probability as the optimal safe landing point.

[0090] In this embodiment of the invention, based on the deviation between the simulation analysis results and the preset target landing parameters, the control commands for the lander are determined, and the control inputs in the parameter dataset are updated according to the control commands, including: Based on the navigation state of the previous simulation cycle, multi-source information fusion and optimal filtering correction are performed on the mathematical equivalent measured data and pose dataset of the lander to obtain the estimated value of the lander's navigation state for the current cycle. The deviation between the final guidance target motion parameters and the current motion parameters of the lander is determined, and the desired acceleration command is generated based on the deviation. The motion parameters include position and velocity. The final guidance target motion parameters are determined based on the preset nominal target landing point coordinates and the optimal safe landing point coordinates, and the current motion parameters are determined based on the estimated values ​​of the lander's navigation state. The desired acceleration command is calculated and converted based on the real-time mass of the lander to obtain the orbital control engine thrust command, as well as the target attitude and target angular velocity of the lander; Based on the deviation between the estimated and target values ​​of the current cycle of attitude adjustment and angular velocity, the expected control torque and engine thrust distribution of the lander are calculated to obtain the attitude control engine jet pulse width command set. Based on the thrust command of the orbit control engine and the jet pulse width command set of the attitude control engine, the dynamic thrust response of the engine and the resultant force and resultant torque of the lander are updated to obtain the total control force and total control torque used to adjust the lander.

[0091] It should be noted that the full-process simulation device for lunar stationary landing of the GNC system provided in the above embodiments is only an example of the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the full-process simulation device for lunar stationary landing of the GNC system provided in the above embodiments and the full-process simulation method embodiment for lunar stationary landing of the GNC system belong to the same concept. The specific implementation process is detailed in the method embodiment, and will not be repeated here.

[0092] The embodiments of this application also provide a computer device, which includes a processor and a memory. The memory stores at least one instruction, at least one program, code set, or instruction set. The at least one instruction, at least one program, code set, or instruction set is loaded and executed by the processor to implement the full-process simulation method for lunar fixed-point landing of the GNC system provided in the above-described method embodiments.

[0093] The embodiments of this application also provide a computer-readable storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement the full-process simulation method for lunar fixed-point landing of the GNC system provided in the above-described method embodiments.

[0094] Embodiments of this application also provide a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium and executes the computer program, causing the computer device to perform the full-process simulation method for lunar fixed-point landing of the GNC system as described in any of the above embodiments.

[0095] For ease of description, the above systems or devices are described separately as various modules or units based on their functions. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware components.

[0096] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0097] Finally, it should be noted that in this document, relational terms such as first, second, third, and fourth are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0098] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A full-process simulation method for lunar precise landing using a GNC system, characterized in that, The method includes: S1. Initialize the lander's operating parameters according to the landing mission conditions to obtain the initial parameter dataset for simulation calculations. S2. Perform multi-model coupled simulation calculations on the lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and generate a flight area topographic map that fits the real lunar surface details based on the simulation dataset and lunar reference database. S3. Determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, end the simulation and output the final motion state of the lander and the topographic map of the flight area; otherwise, proceed to step S4. S4. Based on the physical characteristics of the airborne sensors, the simulation dataset and the flight area topographic map are used to simulate errors, obtain a multi-source equivalent measured dataset, and perform matching analysis on the generated noisy images and point cloud data to obtain the lander's pose dataset and the optimal safe landing point. S5. Based on the deviation between the simulation analysis results and the preset target landing parameters, determine the control commands for the lander, and after updating the control inputs in the parameter dataset according to the control commands, return to step S2.

2. The method as described in claim 1, characterized in that, The process of performing multi-model coupled simulation calculations on lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and generating a flight area topographic map that closely matches the details of the real lunar surface based on the calculated simulation dataset and lunar surface reference database, includes: The lander state parameters, control input data, and celestial gravitational parameters output from the previous simulation cycle are respectively input into preset orbit, liquid sloshing, and attitude dynamics models for hierarchical iterative calculation to obtain the mathematical simulation data of the lander in the current simulation cycle; wherein, the mathematical simulation data includes a six-degree-of-freedom motion state vector and sloshing disturbance torque; Based on the astronomical ephemeris and star catalog data corresponding to the current simulation period, the position and velocity of celestial bodies, coordinate system transformation relationships, and slant range are calculated to obtain the lunar environment simulation data for the current simulation period. The lunar environment simulation data includes Sun-Moon-Earth motion parameters, coordinate transformation matrix between the Moon-fixed and Moon-centric inertial systems, true values ​​of slant range, and a dataset of the illumination starry sky reference environment. Based on simulation data, the effective exploration area of ​​the lander is located, and a high-precision terrain texture data and coordinate map of the flight area is generated based on the lunar surface reference database.

3. The method as described in claim 2, characterized in that, The process of locking the effective exploration area of ​​the lander based on simulation data and generating a high-precision terrain texture data and coordinate map of the flight area based on the lunar surface reference database includes: Based on the real-time pose of the lander determined by the six-degree-of-freedom motion state vector and the inherent field-of-view constraints of the airborne sensor, the spatial coverage range of the sensor detection beam at the current simulation moment is calculated in the lunar inertial frame. The calculated field of view spatial range is intersected with the global lunar surface topography benchmark to determine the intersection area of ​​the field of view coverage and the lunar surface topography as the effective detection area within this simulation cycle; Based on the boundary coordinate set of the effective detection area, the DEM elevation data and DOM texture data in the lunar reference database are cropped and extracted respectively to obtain the precise elevation value and lunar grayscale texture features of each terrain grid point in turn. Spatial registration and coordinate system correction are performed on the DEM elevation data and DOM texture data obtained by cropping and extraction. The registered and corrected DOM texture data is then mapped onto the DEM three-dimensional elevation terrain model to obtain a flight area terrain map that fits the real lunar surface details.

4. The method as described in claim 2, characterized in that, The error simulation is performed on the simulation dataset and the flight area topographic map based on the physical characteristics of the airborne sensor to obtain a multi-source equivalent measured dataset, including: Based on the system error coefficient and measurement noise covariance parameter of the airborne sensor, error degradation processing is performed on the mathematical simulation data to obtain mathematically equivalent measured data. : In the formula, These are mathematical simulation measurements; This refers to the systematic error of the sensor; It is Gaussian random noise. To measure the noise covariance matrix; Noise degradation was performed on lunar environment simulation data and flight area topographic maps based on the inherent imaging parameters of the airborne sensor, resulting in degraded measured starry sky image data, measured lunar surface grayscale image data, and measured laser point cloud data. The measured image data was degraded using the following formula: In the formula, This is a measured grayscale image; For the ideal original image; This is the lens distortion mapping function; For resolution sampling degradation function; Gaussian noise in the image; Measured laser point cloud data Degeneracy is performed using the following formula: In the formula, This represents the true value for point cloud ranging. Gaussian noise for point cloud ranging.

5. The method as described in claim 4, characterized in that, The matching analysis of the generated noisy images and point cloud data to obtain the lander's pose dataset and optimal safe landing point includes: Based on the stellar catalog data, stellar barycenter extraction and catalog matching were performed on the measured starscape image data. The matching results were then optimized using least squares to obtain the optimal attitude matrix of the lander relative to inertial space. : In the formula, To observe stellar vectors; This serves as the reference star vector for the star catalog. This is the attitude matrix; Based on the preset terrain matching and fusion algorithm, feature matching is performed on the measured lunar surface grayscale image data to obtain the optimal pose transformation matrix of the lander relative to the lunar solid frame. : In the formula, Let be the pose transformation matrix to be iteratively optimized; For matching operators of the sum of squared differences in image gray levels; Real-time observation of lunar surface grayscale images; This is the baseline reference image after pose mapping; The landing safety factor for each region is determined based on the measured laser point cloud data, including elevation deviation, terrain slope, and elevation roughness. ; In the formula, This refers to elevation deviation; The slope of the terrain; For elevation roughness; All are preset weighting coefficients; The landing safety coefficients of all regions were compared and screened to determine the region with the highest safety probability as the optimal safe landing point.

6. The method as described in claim 5, characterized in that, Based on the deviation between the simulation analysis results and the preset target landing parameters, the control commands for the lander are determined, and the control inputs in the parameter dataset are updated according to the control commands, including: Based on the navigation state of the previous simulation cycle, multi-source information fusion and optimal filtering correction are performed on the mathematical equivalent measured data and pose dataset of the lander to obtain the estimated value of the lander's navigation state for the current cycle. The deviation between the final guidance target motion parameters and the current motion parameters of the lander is determined, and the desired acceleration command is generated based on the deviation. The motion parameters include position and velocity. The final guidance target motion parameters are determined based on the preset nominal target landing point coordinates and the optimal safe landing point coordinates, and the current motion parameters are determined based on the estimated values ​​of the lander's navigation state. The desired acceleration command is calculated and converted based on the real-time mass of the lander to obtain the orbital control engine thrust command, as well as the target attitude and target angular velocity of the lander; Based on the deviation between the estimated and target values ​​of the current cycle of attitude adjustment and angular velocity, the expected control torque and engine thrust distribution of the lander are calculated to obtain the attitude control engine jet pulse width command set. Based on the thrust command of the orbit control engine and the jet pulse width command set of the attitude control engine, the dynamic thrust response of the engine and the resultant force and resultant torque of the lander are updated to obtain the total control force and total control torque used to adjust the lander.

7. A full-process simulation device for lunar precise landing using a GNC system, characterized in that, The device includes: The initialization unit is used to initialize and configure the lander's operating parameters according to the landing mission conditions, and obtain the initial parameter dataset for simulation calculations. The simulation unit is used to perform multi-model coupled simulation calculations on the lander dynamic parameters and space environment parameters based on the parameter dataset of the current simulation cycle, and to generate a flight area topographic map that fits the real lunar surface details based on the calculated simulation dataset and lunar surface reference database. The judgment unit is used to determine whether the simulation meets the preset termination condition based on the relative distance between the lander and the lunar surface in the simulation data. If yes, the simulation ends and the final motion state of the lander and the topographic map of the flight area are output; otherwise, the function of the simulation unit is executed. The simulation unit is used to simulate errors in the simulation dataset and the flight area topographic map based on the physical characteristics of the airborne sensors, to obtain a multi-source equivalent measured dataset, and to perform matching analysis on the generated noisy images and point cloud data to obtain the lander's pose dataset and the optimal safe landing point. The control unit is used to determine the control commands of the lander based on the deviation between the simulation analysis results and the preset target landing parameters, and then return to the simulation unit to perform the function after updating the control input in the parameter dataset according to the control commands.

8. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method described in any one of claims 1-6.

10. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-6.

Images