Mars rover simulation development system and method thereof

Abstract

The application provides a Mars rover simulation development system and method, the Mars rover simulation development system comprises: a Mars rover physical teaching aid configured to run in a physical environment and collect telemetry data; a miniature physical test field configured to provide a physical running environment simulating a Mars terrain and collect global pose ground truth of the Mars rover physical teaching aid; an engineering comprehensive development server configured to build a digital twin model of the miniature physical test field; based on the global pose ground truth and the telemetry data, nonlinear deviation analysis is performed to generate a visual representation of physical deviation in the digital twin model; according to the vibration spectrum data, the physical medium type is identified and mapped to the digital twin model for labeling; and extreme working condition stress is injected into the physical environment of the Mars rover physical teaching aid; a data collaborative evaluation subsystem configured to generate quantitative verification and evaluation data for the extreme working condition stress based on the time synchronization and deviation quantification of the global pose ground truth and the telemetry data.

Description

Technical Field

[0001] This application relates to the fields of robotics and engineering education, and more specifically, to a Mars rover simulation development system and method thereof. Background Technology

[0002] In current engineering education, especially in the teaching practice of robotics and complex systems (such as Mars rovers), the following two methods are generally used for teaching and verification: one is to use pure simulation software, in which students design and test algorithms in a virtual environment; the other is to use finished robot kits with fixed functions to conduct basic motion control and sensor experiments.

[0003] However, pure simulation software cannot realistically simulate the nonlinear physical characteristics such as slippage and sinking that occur when wheels interact with soft surfaces such as sand and gravel, leading to a disconnect between simulation success and physical failure. Secondly, most finished robots have a black-box structure, which limits students' in-depth participation in the underlying hardware, communication protocols, and control algorithms. Thirdly, the evaluation of student solutions lacks objective quantitative means, making it difficult to verify the robustness and adaptability of the algorithm through real physical feedback.

[0004] Therefore, there is an urgent need for a Mars rover simulation development system that can balance physical realism, openness to development, and objectivity of evaluation. Summary of the Invention

[0005] In view of this, this application provides a Mars rover simulation development system and method to realize the development, verification and objective quantitative evaluation of Mars rover algorithms in a real physical environment, overcoming the teaching limitations of pure simulation physical distortion and the black box nature of finished robots.

[0006] In a first aspect, an embodiment of this application provides a Mars rover simulation development system, comprising: A physical teaching aid for Mars rovers, configured to operate in a physical environment and collect telemetry data; The miniature physical test field is configured to provide a physical operating environment that simulates the Martian terrain and to collect the global pose truth value of the physical teaching aids of the Mars rover. An engineering integrated development server, wirelessly communicating with the Mars rover physical teaching aid, is configured to construct a digital twin model of the miniature physical test field; perform nonlinear deviation analysis based on the global pose ground truth and the telemetry data, and generate a visual representation of the physical deviation in the digital twin model; identify the current physical medium type based on the vibration spectrum data in the telemetry data, and map the identified physical medium type to the digital twin model for annotation; and inject extreme working condition stress into the wireless link of the Mars rover physical teaching aid and / or the physical environment of the miniature physical test field. The data collaborative evaluation subsystem is configured to generate quantitative verification evaluation data for the extreme working condition stress based on the time synchronization and deviation quantification between the global pose true value and the telemetry data.

[0007] In one embodiment, the Mars rover physical teaching aid includes a main control module, a chassis module, and a sensing and detection module. The chassis module is configured to perform motion control, drive the Mars rover physical teaching aid to move within the miniature physical test field, and generate motion state data. The sensing and detection module is configured to collect environmental perception data, which includes at least image data, point cloud data, and inertial measurement data. The main control module has edge computing capabilities and is configured to coordinate the work of the chassis module and the sensing and detection module, and to perform multi-source data fusion and feature extraction on the environmental perception data collected by the sensing and detection module to obtain processed environmental perception data. The processed environmental perception data and the motion state data constitute the telemetry data.

[0008] In one embodiment, the Mars rover physical teaching aid includes a main control module, which is equipped with a triaxial accelerometer. The main control module is configured to: collect triaxial vibration acceleration data of the Mars rover physical teaching aid during its movement, which is detected in real time by the triaxial accelerometer, and perform frequency domain transformation on the triaxial vibration acceleration data to obtain the vibration spectrum data.

[0009] In one embodiment, the engineering integrated development server includes a nonlinear deviation analysis module, which is configured to: synchronize the global pose ground truth with the relative pose data in the telemetry data in time; calculate the position deviation vector and attitude deviation angle at each time point to obtain an instantaneous pose deviation sequence; construct a nonlinear deviation growth model based on the instantaneous pose deviation sequence; use the nonlinear deviation growth model to detect the abrupt change points and cumulative drift trends of the instantaneous pose deviation sequence; and extract nonlinear deviation feature parameters; identify the deviation feature quantities corresponding to physical slippage conditions, sensing failure conditions, and communication degradation conditions based on the nonlinear deviation feature parameters, and generate nonlinear deviation analysis results for extreme conditions.

[0010] In one embodiment, the engineering integrated development server includes a physical deviation visualization module, which is configured to: construct a three-dimensional digital twin model of the miniature physical test field based on the geometric parameters and physical medium properties of the miniature physical test field; acquire the global pose ground truth and the telemetry data; in the three-dimensional digital twin model, synchronously drive the virtual Mars rover to move based on the telemetry data, and synchronously drive the global pose marker point corresponding to the global pose ground truth to move based on the global pose ground truth; calculate the spatial difference between the pose of the virtual Mars rover and the pose of the global pose marker point to obtain the real-time spatial deviation; and, based on the real-time spatial deviation, render the path deviation area in the digital twin model through color gradient or shadow area, mark the deviation position caused by physical slippage, collision, or perception drift, and generate a visual representation of the physical deviation.

[0011] In one embodiment, the engineering integrated development server includes a terrain recognition module, which is configured to: receive vibration spectrum data transmitted back by the physical teaching aid of the Mars rover; extract the frequency peak value, amplitude distribution, and energy concentration characteristic parameters of the vibration spectrum data to generate a vibration feature vector; calculate the similarity between the vibration feature vector and a standard vibration feature vector based on the vibration feature vector and a preset terrain feature database; identify the physical medium type with the highest similarity and its corresponding physical drag coefficient; the terrain feature database stores standard vibration feature vectors corresponding to different physical medium types; map the physical medium type and the physical drag coefficient to the digital twin model; update the physical attribute labeling of the current terrain area; and synchronously adjust the dynamic simulation parameters of the virtual Mars rover to complete the labeling of the physical medium type in the digital twin model.

[0012] In one embodiment, the engineering integrated development server includes an environmental stress injection module, which includes a communication degradation simulator and a sensing interference trigger. The communication degradation simulator is configured to inject preset random delay parameters or packet loss parameters into the wireless communication link of the Mars rover physical teaching aid to simulate communication delay conditions and collect autonomous decision-making response data of the Mars rover physical teaching aid under the communication delay conditions. The sensing interference trigger is configured to control the adjustable light source array in the miniature physical test field to generate dynamic long shadows or strong light overexposure lighting interference conditions and collect navigation data of the sensing and detection module of the Mars rover physical teaching aid under the lighting interference conditions.

[0013] In one implementation, the data collaborative evaluation subsystem includes a time synchronization and deviation quantification module. This module is configured to: align the global pose ground truth with the telemetry data using timestamps to obtain time-synchronized global pose ground truth and telemetry data; based on the time-synchronized global pose ground truth and telemetry data, calculate the ratio of the actual travel distance to the theoretical travel distance at the wheel end of the Mars rover physical teaching aid, calculate the pose deviation rate and cumulative drift trend coefficient per unit time, and statistically analyze the standard deviation and extreme value fluctuation range of the target detection confidence or positioning accuracy index output by the Mars rover physical teaching aid's perception and detection module to obtain slip rate, pose drift rate, and perception confidence fluctuation data; and generate quantitative verification and evaluation data for physical slippage conditions, perception failure conditions, and communication degradation conditions based on the slip rate, pose drift rate, and perception confidence fluctuation data, as well as the extreme working condition stress.

[0014] In one embodiment, the miniature physical test field includes a matrix-type dynamic sensing board, which is concealed beneath the physical medium simulating Martian terrain. The matrix-type dynamic sensing board is equipped with a pressure sensor array. The matrix-type dynamic sensing board is configured to: collect transient pressure data of the wheel ends of the Martian rover physical teaching aid against the ground in real time through the pressure sensor array; perform spatial distribution analysis and temporal variation analysis on the transient pressure data to obtain wheel end pressure distribution data and wheel-ground contact force variation characteristics; and send the wheel end pressure distribution data and wheel-ground contact force variation characteristics to the engineering integrated development server to verify the dynamic simulation accuracy of the virtual Martian rover in the digital twin model.

[0015] In one embodiment, the Mars rover physical teaching aid includes a main control module, which is configured to: receive algorithm update instructions sent by the engineering integrated development server; perform remote hot replacement of local control operators during the operation of the Mars rover physical teaching aid; collect dynamic response data of the physical actuators of the Mars rover physical teaching aid under the action of different control operators; send the dynamic response data as part of the telemetry data to the engineering integrated development server to obtain dynamic response deviation data corresponding to different algorithm control quantities.

[0016] In one embodiment, the sensing and detection module is further configured to: acquire calibration data of a preset physical calibration target; and after the hardware reconstruction of the Mars rover physical teaching aid, automatically complete the spatial alignment of the visual sensor coordinate system and the lidar coordinate system based on the calibration data.

[0017] In one embodiment, the engineering integrated development server is configured with a hardware abstraction layer driver library and an automated compilation environment; the hardware abstraction layer driver library is used to provide a unified driver interface for different hardware modules, and the automated compilation environment is used to automatically compile the algorithm code written by the user into an executable program and load it into the Mars rover physical teaching aid.

[0018] In one embodiment, the miniature physical test field includes a terrain modular configuration unit, which includes multiple modules such as sand, gravel, clay, and slope adjustment modules that can be quickly replaced.

[0019] In one embodiment, the miniature physical test field includes an automatic terrain calibration module; the automatic terrain calibration module is configured to: obtain the geometric parameters of the current terrain and the mechanical properties of the physical medium based on the global pose ground truth and the data collected by the matrix dynamics sensing board; construct a terrain geometric model and a physical property model according to the geometric parameters and the mechanical properties, and generate automatic terrain modeling data; and send the automatic terrain modeling data to the engineering integrated development server to update the terrain geometric and physical property parameters of the digital twin model.

[0020] Secondly, embodiments of this application also provide a Mars rover simulation development method, including: The system controls a Mars rover physical teaching aid to operate within a miniature physical test field, collecting telemetry data from the rover physical teaching aid and simultaneously acquiring its global pose ground truth value within the miniature physical test field. A digital twin model of the miniature physical test field is constructed, and nonlinear deviation analysis is performed based on the global pose ground truth value and the telemetry data. A visual representation of the physical deviation is generated in the digital twin model. The physical medium type currently inhabited by the Mars rover physical teaching aid is identified based on the vibration spectrum data in the telemetry data, and the identified physical medium type is mapped to the digital twin model for annotation. Extreme working condition stresses are injected into the wireless link of the Mars rover physical teaching aid and / or the physical environment of the miniature physical test field. Based on the time synchronization and deviation quantification between the global pose ground truth value and the telemetry data, quantitative verification and evaluation data for the extreme working condition stresses are generated.

[0021] The aforementioned Mars rover simulation development system and method first construct a realistic physical environment using physical Mars rover teaching aids and a miniature physical test field. Combined with nonlinear deviation analysis of global pose ground truth and telemetry data, and digital twin visualization, this effectively compensates for the lack of physical realism in pure simulation software. Secondly, the engineering integrated development server can receive vibration spectrum data and automatically identify the type of physical medium, achieving transparent display of the perception-decision link and breaking the black-box limitation of traditional finished robots. Thirdly, the data collaborative evaluation subsystem automatically generates objective quantitative evaluation data such as slip rate and pose drift rate based on time synchronization and deviation quantification between global pose ground truth and telemetry data, overcoming the problem of subjective evaluation dimensions. Simultaneously, the engineering integrated development server can also inject extreme working condition stresses into the wireless link and physical environment, allowing students to test the robustness of the algorithm under safe and controllable conditions. In summary, this application constructs a complete verification platform that is physically realistic, process transparent, and objectively evaluated, providing a high-fidelity physical development verification environment for engineering education.

[0022] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0023] Figure 1 A schematic diagram of a Mars rover simulation development system 100 provided in this application embodiment; Figure 2 This is a schematic diagram of the overall logic of the Mars rover simulation development system 100 provided in the embodiments of this application; Figure 3 A flowchart of a Mars rover simulation development method provided in this application embodiment. Detailed Implementation

[0024] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0025] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0026] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0027] The shortcomings of the above solutions are the result of the inventor's practical experience and careful research. Therefore, the discovery process of the above problems and the solutions proposed in this application below should be considered as the inventor's contributions to this application.

[0028] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0029] It is understood that before using the technical solutions disclosed in the various embodiments of this application, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this application in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0030] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0031] like Figure 1 The diagram shown is a schematic of a Mars rover simulation development system 100 provided in this application, comprising: Mars rover physical teaching aid 11 is configured to operate in a physical environment and collect telemetry data.

[0032] The miniature physical test field 12 is configured to provide a physical operating environment that simulates the Martian terrain and to collect the global pose truth value of the physical teaching aid of the Mars rover.

[0033] The engineering integrated development server 13 wirelessly communicates with the Mars rover physical teaching aid and is configured to construct a digital twin model of the miniature physical test field; perform nonlinear deviation analysis based on the global pose ground truth and the telemetry data, and generate a visual representation of the physical deviation in the digital twin model; identify the current physical medium type based on the vibration spectrum data in the telemetry data, and map the identified physical medium type to the digital twin model for annotation; and inject extreme working condition stress into the wireless link of the Mars rover physical teaching aid and / or the physical environment of the miniature physical test field.

[0034] The data collaborative evaluation subsystem 14 is configured to generate quantitative verification evaluation data for the extreme working condition stress based on the time synchronization and deviation quantification of the global pose true value and the telemetry data.

[0035] The aforementioned Mars rover physical teaching aid 11 is a scaled-down, mobile robotic platform used to simulate the driving, sensing, and communication behaviors of a real Mars rover on a planetary surface. It operates in a physical environment (i.e., a miniature physical test field 12) and possesses sensor and data acquisition capabilities. During its movement within the test field, the Mars rover physical teaching aid 11 collects real-time data on its own status and the surrounding environment through various onboard sensors (such as wheel speed encoders, inertial measurement units, cameras, lidar, and triaxial accelerometers). This data, referred to in this application as telemetry data, includes, but is not limited to: wheel speed, attitude angles, position (relative positioning), images, point clouds, and vibration spectra. The collected data is transmitted wirelessly to the engineering integration and development server 13 for subsequent analysis and twin modeling.

[0036] The aforementioned miniature physics test field 12 is a specially constructed, small-scale physics experimental site whose terrain, landforms, and media (such as sand, gravel, and slopes) mimic the characteristics of the real Martian surface. It provides a physically realistic operating environment for the Mars rover training equipment, including wheel-ground interaction, slopes, and obstacles. The test field is laid with different physical media (such as fine sand, coarse sand, clay, and gravel) and modules with adjustable slopes. These media will have realistic physical effects on the Mars rover's movement, such as slipping, sinking, and vibration. The test field can also dynamically change lighting conditions or terrain layout to simulate different operating conditions.

[0037] The global pose ground truth refers to the true position and attitude (including X, Y, Z coordinates and yaw, pitch, and roll angles) of the Mars rover's physical training device in the fixed coordinate system of the test field. This global pose ground truth is provided by an external high-precision positioning system (such as motion capture cameras, UWB arrays, laser trackers, etc.) and does not rely on the rover's own sensors. For example, multiple global positioning sensors (such as optical motion capture cameras) are deployed outside the test field, capable of capturing passive markers or active beacons installed on the Mars rover in real time. Through triangulation or positioning algorithms, the precise pose of the Mars rover in the test field coordinate system is calculated and timestamped with high precision. This ground truth data is sent to the engineering integration and development server 13 for comparison with the telemetry pose reported by the Mars rover itself.

[0038] The aforementioned engineering integrated development server 13 is the core computing and collaboration hub of the entire Mars rover simulation development system 100. The engineering integrated development server 13 can be a high-performance computer or a server cluster, and can be connected to the Mars rover physical teaching aid 11 via a wireless network. At the same time, it can communicate in real time with the data acquisition system (such as a global pose capture array, pressure sensor array, etc.) of the miniature physical test field 12.

[0039] The digital twin model constructed by the engineering integrated development server 13 refers to a three-dimensional virtual environment that is geometrically and physically consistent with the physical test field and dynamically synchronized. This digital twin model not only possesses static 3D map information but can also receive real-time data and reflect the motion status of the Mars rover in the physical world. For example, the server constructs a high-fidelity virtual test field in a 3D engine (such as Unity, Unreal Engine, or a self-developed engine) based on the test field's CAD design drawings, terrain scan data, physical medium parameters, etc. This model includes terrain geometry, obstacle locations, medium types, lighting environment, etc. During operation, the model receives telemetry data in real time and drives the virtual Mars rover's movement.

[0040] The aforementioned nonlinear deviation analysis refers to comparing the pose reported by the Mars rover itself (telemetry data, which usually contains errors) with the actual pose measured externally (global truth value), analyzing whether the error between the two grows nonlinearly (e.g., exponentially, abruptly, or drifts), thereby identifying problems such as slippage, sensor drift, and communication interruption. For example, the server receives two sources of pose data: one from the rover itself (e.g., a wheeled odometer or visual-inertial odometer), and the other from the global positioning system at the test site. After aligning with timestamps, the position and attitude differences at each moment are calculated. If the error grows linearly with time and distance, it may simply be an encoder error; if the error suddenly jumps or grows exponentially, it indicates slippage, collision, or sensor failure. The server extracts these nonlinear characteristics and categorizes them into conditions such as physical slippage, sensor failure, and communication degradation.

[0041] The aforementioned visualization of physical deviations generated in the digital twin model refers to presenting the previously calculated pose deviations graphically on the digital twin interface, allowing users to intuitively see the difference between the rover's theoretical and actual positions. For example, in the digital twin model, the engineering integration server 13 can drive two objects: a virtual rover driven based on its own telemetry data (which has errors), and a real marker point or shadow rover driven based on global ground truth (real position). The engineering integration server 13 can calculate the spatial difference between the two in real time and, depending on the magnitude of the deviation, render a color-gradient trajectory band or shadow area behind the virtual rover or on the off-path. The larger the deviation, the redder the color and the darker the shadow.

[0042] The aforementioned vibration spectrum data is part of the telemetry data and is a frequency domain distribution map obtained by performing a Fourier transform on the time-domain vibration signal collected by the onboard triaxial accelerometer. Different physical media (such as fine sand, gravel, and hard ground) will generate vibration excitations of different frequencies and energies on the Mars rover, forming unique vibration fingerprints. The engineering integration development server 13 uses vibration fingerprints to identify the current ground type. For example, during the rover's physical teaching aid 11, triaxial acceleration data is collected at high frequencies (such as 200Hz) and a Fast Fourier Transform (FFT) is performed in real time to generate vibration spectrum data (e.g., X-axis frequency domain energy distribution). After receiving the data, the engineering integration development server 13 extracts key features: peak frequency, energy concentration band, amplitude attenuation trend, etc., to form a feature vector. Then, it performs similarity matching with the standard feature vectors in the pre-stored terrain feature database to find the most matching medium type.

[0043] The aforementioned mapping of the identified physical medium type to the digital twin model for annotation refers to updating the identified medium type (e.g., fine sand) and corresponding physical parameters (e.g., friction coefficient, drag coefficient) to the corresponding region of the digital twin model, ensuring that the ground properties of the digital twin model are consistent with the physical world. For example, the engineering integrated development server 13 can maintain a ground property layer in the digital twin model, recording the medium type, friction coefficient, bearing capacity, etc., of each region (or each grid). Once the medium type of the rover's current location is identified through vibration spectrum, the properties of that location can be updated to reflect the identification result, and a text label (e.g., current region: fine sand, drag coefficient 0.6) can be displayed on the interface. Simultaneously, the dynamic simulation parameters of the virtual rover will also be adjusted synchronously to make the simulation more realistic.

[0044] The aforementioned injection of extreme stress into the wireless link of the Mars rover physical teaching aid and / or the physical environment of the miniature physical test field refers to introducing abnormal or extreme conditions into the communication link or physical environment through software or hardware means to test the robustness of the Mars rover and its algorithms under harsh environments. For example, in the wireless communication between the engineering integrated development server 13 and the Mars rover, random delays (e.g., 100ms to 3000ms) or random packet loss (e.g., 5% to 30%) are actively added to simulate deep space communication delays or unstable links. Another example is controlling the adjustable light source array within the test field to generate strong dynamic shadows, flickering, overexposure, or low-light conditions to simulate the lighting environment under Martian twilight or dust storms.

[0045] The aforementioned data collaborative evaluation subsystem 14 is used for evaluation and report generation. By precisely aligning the global true values ​​with the rover's telemetry data in time and calculating deviations, it quantifies the rover's performance indicators under various extreme conditions, ultimately generating a structured evaluation report. For example, the data collaborative evaluation subsystem 14 can align the global pose true values ​​and telemetry data according to their respective timestamps, ensuring the comparison occurs at the same physical moment. Additionally, it can calculate the slip rate, pose drift rate, and statistical perception confidence fluctuation, where the slip rate = (theoretical travel distance - actual travel distance) / theoretical travel distance, the pose drift rate = the change in position deviation per unit time, and the perception confidence fluctuation = the standard deviation and range of the perception module's output confidence. Furthermore, the data collaborative evaluation subsystem 14 can correlate the aforementioned quantitative indicators with the extreme stresses injected during testing (delay magnitude, packet loss rate, illumination type, etc.). Finally, the data collaborative evaluation subsystem 14 can output a quantitative verification evaluation data package containing elements such as charts, indicator curves, deviation heatmaps, and conclusions.

[0046] In one embodiment, the Mars rover physical teaching aid includes a main control module 111, a chassis module 112, and a sensing and detection module 113; The chassis module 112 is configured to perform motion control, drive the Mars rover physical teaching aid to move within the miniature physical test field, and generate motion state data; The sensing and detection module 113 is configured to collect environmental sensing data, which includes at least image data, point cloud data, and inertial measurement data. The main control module 111 has edge computing capabilities and is configured to coordinate the work of the chassis module and the sensing and detection module, and to perform multi-source data fusion and feature extraction on the environmental sensing data collected by the sensing and detection module to obtain processed environmental sensing data; the processed environmental sensing data and the motion state data constitute the telemetry data.

[0047] The aforementioned chassis module 112 serves as the motion execution and underlying feedback unit for the Mars rover physical teaching aid 11. It is equivalent to the walking mechanism of a real Mars rover, responsible for converting control commands (such as speed and steering angle) into actual physical motion and recording state parameters during the motion process in real time. For example, the chassis module 112 includes a motor driver, a DC geared motor, a wheel speed encoder, a servo motor (for steering), and a necessary suspension system. The chassis module 112 receives motion control commands from the main control module 111 via a bus. The motor driver adjusts the PWM duty cycle or current of each wheel motor according to the commands, driving the wheels to rotate. While executing the motion, the chassis module can collect motion state data in real time, such as wheel speed data, mileage data, motor status data, and chassis font data. After being packaged, the motion state data is uploaded to the engineering integrated development server 13 via the main control module 111 as part of the telemetry data.

[0048] The aforementioned perception and detection module 113 is the environmental perception unit of the Mars rover physical teaching aid 11. It is responsible for collecting environmental information around the vehicle and the vehicle's own inertial motion information, providing raw data for navigation, obstacle avoidance, positioning, and other algorithms. The perception and detection module 113 may include a visual sensor, a lidar, an inertial measurement unit, and optional extended sensors (such as a geomagnetic sensor, an ultrasonic / infrared ranging sensor, GPS, etc.). The aforementioned main control module 111 possesses edge computing capabilities. Edge computing capability refers to the main control module's ability to directly process, fuse, and extract features from raw data collected by sensors in real time on the Mars rover (i.e., at the edge), without uploading all raw data to a server. This reduces communication bandwidth requirements, decreases latency, and maintains a certain degree of autonomous operation capability during communication interruptions.

[0049] The main control module 111 coordinates the operation of the chassis module 112 and the sensing and detection module 113. This means that the main control module 111 acts as the central scheduler for the entire teaching aid, responsible for managing the timing, data flow, and instruction distribution of each module to ensure they operate collaboratively without conflict or deadlock. For example, the main control module 111 can simultaneously trigger the sensor acquisition of the sensing and detection module 113 and the status reading of the chassis module at a fixed frequency, reading the motion status data of the chassis module 112 and the environmental perception data of the sensing and detection module 113 from the bus or dedicated interface. Additionally, a unified timestamp (based on the system clock of the main control module 111) can be added to each frame of data. Furthermore, the main control module 111 can send motion control commands to the chassis module 112 based on locally running decision algorithms (or remote commands from the engineering integrated development server 13). Safety-related tasks (such as emergency obstacle avoidance and low battery protection) are given the highest priority, and other tasks being executed can be interrupted.

[0050] The main control module 111 performs multi-source data fusion and feature extraction on the environmental perception data collected by the perception and detection module 113 to obtain processed environmental perception data. This involves aligning and integrating data from different sensors (images, point clouds, inertial measurement units, IMUs) in time and space to form a richer and more reliable environmental representation. Furthermore, it extracts key information useful for subsequent decision-making from the fused data, such as obstacle locations, ground types, and the device's own speed.

[0051] The processed environmental perception data and motion state data constitute the telemetry data, which is reported by the Mars rover physical teaching aid 11 to the engineering integrated development server 13.

[0052] In one embodiment, the Mars rover physical teaching aid includes a main control module, which is equipped with a triaxial accelerometer. The main control module is configured to: collect triaxial vibration acceleration data of the Mars rover physical teaching aid during its movement, which is detected in real time by the triaxial accelerometer, and perform frequency domain transformation on the triaxial vibration acceleration data to obtain the vibration spectrum data.

[0053] Through the above implementation method, vibration signals during vehicle operation can be collected using an onboard triaxial accelerometer, and then transformed into spectral features suitable for terrain identification through frequency domain transformation. This triaxial vibration acceleration data includes instantaneous acceleration values ​​in the X (forward / backward), Y (left / right), and Z (up / down) directions generated by the interaction between the rover's wheels and the ground with different physical media (such as sand, gravel, and hard ground) during its operation. Through frequency domain transformation, such as Fast Fourier Transform (FFT), the vibration waveform in the time domain can be converted into an energy distribution map in the frequency domain, thus reflecting the intensity of the vibration signal at different frequencies.

[0054] In one embodiment, the engineering integrated development server 13 includes a nonlinear deviation analysis module 131, which is configured to: synchronize the global pose ground truth with the relative pose data in the telemetry data in time; calculate the position deviation vector and attitude deviation angle at each time point to obtain an instantaneous pose deviation sequence; construct a nonlinear deviation growth model based on the instantaneous pose deviation sequence; use the nonlinear deviation growth model to detect the abrupt change points and cumulative drift trends of the instantaneous pose deviation sequence; and extract nonlinear deviation feature parameters; identify the deviation feature quantities corresponding to physical slippage conditions, sensing failure conditions, and communication degradation conditions based on the nonlinear deviation feature parameters, and generate nonlinear deviation analysis results for extreme conditions.

[0055] The aforementioned nonlinear deviation analysis module 131 compares and analyzes the true global pose value collected by the miniature physical test field with the relative pose data reported by the rover itself. By constructing a nonlinear model, it quantifies the evolution law of the deviation, thereby identifying typical extreme working conditions such as physical slippage, sensing failure, and communication degradation.

[0056] The aforementioned global pose ground truth is provided by the external positioning system (such as a motion capture array) of the miniature physical test field, representing the rover's true position and attitude in the fixed coordinate system of the test field, with an accuracy down to the sub-millimeter level. Relative pose data is included in the telemetry data reported by the rover, and is calculated based on onboard sensors (such as wheeled odometers and visual-inertial odometers), which typically contain accumulated errors.

[0057] In practical implementation, the first step is to perform time synchronization and generate an instantaneous pose deviation sequence. Specifically, the global pose ground truth and relative pose data are first strictly timestamped to ensure that the comparison is performed at the same physical moment. After alignment, the position deviation vector and attitude deviation angle at each time point are calculated to form a deviation sequence that changes over time, i.e., the instantaneous pose deviation sequence. Then, a nonlinear deviation growth model (e.g., using an exponential model, a polynomial model, or a state estimation model based on a Kalman filter) is constructed and feature parameters are extracted. This nonlinear deviation growth model differs from a simple linear error model; it can describe the trend of deviation changing exponentially, abruptly, or otherwise nonlinearly over time. For example, the abrupt change point detected by the nonlinear deviation growth model is a sudden, large jump in the deviation sequence, which usually corresponds to instantaneous abnormal events such as collisions or loss of sensing data. The cumulative drift trend detected by the nonlinear deviation growth model indicates the trend of deviation gradually increasing with time and distance traveled, which usually corresponds to continuous wheel slippage or sensor drift. Finally, based on the extracted nonlinear deviation feature parameters, they are matched with typical feature templates of three preset working conditions to identify the type of extreme working condition currently occurring and generate targeted nonlinear deviation analysis results. The three preset working conditions include, for example: physical slippage, indicating relative sliding between the wheel and the ground, resulting in actual displacement being less than theoretical displacement, with the deviation typically showing a cumulative drift trend; perception failure, where visual or lidar sensors fail due to lighting, texture loss, or other reasons, leading to incorrect pose estimation, with the deviation typically showing abrupt or random jump characteristics; and communication degradation, where wireless link delays or packet loss cause data asynchrony, with the deviation possibly manifesting as timestamp misalignment or data loss.

[0058] In one embodiment, the engineering integrated development server 13 includes a physical deviation visualization module 132, which is configured to: construct a three-dimensional digital twin model of the miniature physical test field based on the geometric parameters and physical medium properties of the miniature physical test field; acquire the global pose ground truth and the telemetry data; in the three-dimensional digital twin model, synchronously drive the virtual Mars rover to move based on the telemetry data, and synchronously drive the global pose marker point corresponding to the global pose ground truth to move based on the global pose ground truth; calculate the spatial difference between the pose of the virtual Mars rover and the pose of the global pose marker point to obtain the real-time spatial deviation; and, based on the real-time spatial deviation, render the path deviation area in the digital twin model through color gradient or shadow area, mark the deviation position caused by physical slippage, collision or perception drift, and generate a visual representation of the physical deviation.

[0059] The aforementioned physical deviation visualization module 132 can transform abstract position and attitude deviation data into intuitive graphical representations, enabling students or developers to clearly see the gap between the theoretical and actual positions of the Mars rover.

[0060] In its implementation, the physical deviation visualization module 132 first constructs a high-fidelity three-dimensional digital twin model based on the geometric parameters of the miniature physical test field (such as the terrain elevation map, the size and location of obstacles) and the physical medium properties (such as material information of sand, gravel, etc. in different areas). During system operation, the physical deviation visualization module 132 simultaneously acquires two data streams: one is the telemetry data reported by the rover itself (including the relative pose calculated based on wheel odometry or visual inertial odometry), and the other is the global pose ground truth provided by the miniature physical test field (high-precision position and attitude captured by the external positioning system). Based on these two data streams, the physical deviation visualization module 132 can drive the virtual rover in the three-dimensional digital twin model according to the telemetry data, and drive the global pose marker point (which can be understood as a bright ball or a semi-transparent shadow rover) according to the global pose ground truth. Subsequently, the physical deviation visualization module 132 calculates the spatial difference between the pose of the virtual rover and the pose of the global pose marker point in real time to obtain the spatial deviation amount at the current moment. Based on the magnitude and direction of this deviation, the physical deviation visualization module 132 visually renders the deviation area in the digital twin model using color gradients (e.g., small deviations are displayed in green, and large deviations in red) or by generating a shadow area over the travel path. For example, if the rover is found to have deviated 15cm from its planned trajectory, this physical deviation is immediately marked with a red shadow on the digital twin interface. In this way, the location and extent of deviations caused by physical slippage, collisions, or sensory drift are clearly marked. Students or developers can intuitively understand where and to what extent the algorithm has failed without analyzing complex data tables, thus greatly reducing the difficulty of diagnosis and debugging.

[0061] In one embodiment, the engineering integrated development server 13 includes a terrain recognition module 133, which is configured to: receive vibration spectrum data transmitted back by the physical teaching aid of the Mars rover; extract the frequency peak value, amplitude distribution, and energy concentration characteristic parameters of the vibration spectrum data to generate a vibration feature vector; calculate the similarity between the vibration feature vector and a standard vibration feature vector based on the vibration feature vector and a preset terrain feature database; identify the physical medium type with the highest similarity and its corresponding physical drag coefficient; the terrain feature database stores standard vibration feature vectors corresponding to different physical medium types; map the physical medium type and the physical drag coefficient to the digital twin model; update the physical attribute labeling of the current terrain area; and synchronously adjust the dynamic simulation parameters of the virtual Mars rover to complete the labeling of the physical medium type in the digital twin model.

[0062] The aforementioned terrain recognition module 133 utilizes the vibration spectrum data generated during the rover's movement to automatically identify the type of physical medium currently inhabited, and maps the identification results to a digital twin model for annotation. Specifically, the main control module 111 collects triaxial vibration spectrum data during the movement and transmits it back to the engineering integrated development server 13. The engineering integrated development server 13 automatically identifies and annotates the type of physical medium currently inhabited by the rover and its corresponding physical drag coefficient through a preset terrain feature database. Here, different physical media (such as fine sand, coarse sand, gravel, and hard road surfaces) will generate unique vibration excitations on the vehicle chassis, forming different vibration fingerprints. By analyzing these fingerprints, the ground type can be deduced.

[0063] In its specific implementation, the terrain recognition module 133 first receives vibration spectrum data transmitted back from the Mars rover's physical teaching aids. This data is obtained by the main control module 111 after performing frequency domain transformation on the time-domain vibration signals collected by the triaxial accelerometer, revealing the vibration energy distribution at different frequencies. The terrain recognition module 133 extracts key feature parameters from the vibration spectrum data, including the peak frequency of the spectrum (at which frequency the vibration energy is concentrated), amplitude distribution (the intensity of each frequency component), and energy concentration (whether the energy is mainly distributed in the low-frequency or high-frequency range), and combines these parameters into a high-dimensional vibration feature vector.

[0064] Subsequently, the terrain recognition module 133 accesses a pre-built and stored terrain feature database. The core content of this database consists of standard vibration feature vectors corresponding to different physical medium types. These standard vectors are obtained through extensive testing and calibration on known media. The terrain recognition module 133 calculates the similarity between the currently generated vibration feature vector and each standard vector in the database (e.g., using cosine similarity or Euclidean distance), finding the one with the highest similarity, thereby identifying the physical medium type currently inhabited by the rover and simultaneously reading the corresponding physical drag coefficient. In specific implementations, after identifying a change in the physical medium type inhabited by the rover, the interface can prompt that a new physical medium type has been entered, such as: "Currently entering a high-slip sandy medium."

[0065] Finally, the terrain recognition module 133 maps the identified physical medium type and physical drag coefficient to the digital twin model, updates the physical attribute labels of the current terrain area in the model (e.g., updating the label of a certain area from "unknown" to "fine sand"), and simultaneously adjusts the dynamic simulation parameters of the virtual Mars rover (such as friction coefficient, subsidence coefficient, drag coefficient, etc.), thereby completing the real-time and automated labeling of physical medium type in the digital twin model. This ensures that the ground properties in the virtual environment are highly consistent with the physical world, providing a more accurate model basis for subsequent dynamic simulation and path planning.

[0066] In one embodiment, the engineering integrated development server 13 includes an environmental stress injection module 134, which includes a communication degradation simulator and a sensing interference trigger. The communication degradation simulator is configured to inject preset random delay parameters or packet loss parameters into the wireless communication link of the Mars rover physical teaching aid to simulate communication delay conditions, and collect autonomous decision-making response data of the Mars rover physical teaching aid under the communication delay conditions. The sensing interference trigger is configured to: control the adjustable light source array in the miniature physical test field to generate dynamic long shadows or strong light overexposure lighting interference conditions, and collect navigation data of the sensing and detection module of the Mars rover physical teaching aid under the lighting interference conditions.

[0067] The core function of the aforementioned environmental stress injection module 134 is to artificially introduce extreme or abnormal conditions into the communication link or physical environment to simulate various unexpected situations that may occur in real space missions, thereby testing and verifying the robustness and stability of the Mars rover algorithm under harsh working conditions.

[0068] The environmental stress injection module 134 specifically includes two sub-modules: a communication degradation simulator and a perception interference trigger. The communication degradation simulator targets the wireless communication link between the rover and the server. In real Mars exploration missions, there is a communication delay of several minutes to tens of minutes between the rover and the ground control station, and the communication link may experience data packet loss due to solar activity, atmospheric conditions, etc. To simulate this condition on the ground, the communication degradation simulator actively injects preset random delay or packet loss parameters into the wireless communication link, thereby realistically simulating the delay and instability of long-distance deep space communication. Under this condition, the engineering integrated development server 13 simultaneously collects autonomous decision-making response data from the rover's physical teaching aid 11, observing whether its obstacle avoidance algorithm and path planner can still maintain safe and effective operation under command delays or missing instructions. The perception interference trigger targets the adaptability of the rover's visual perception system to environmental changes. On the Martian surface, lighting conditions are extremely variable, such as the movement of shadows during Martian twilight and strong light scattering caused by dust storms, all of which pose severe challenges to vision-based navigation algorithms. The perceptual interference trigger can control an adjustable light source array (e.g., composed of multiple high-brightness LEDs or spotlights with shades) deployed inside or outside a miniature physical test field to dynamically generate lighting interference conditions based on a preset scenario, such as mimicking a long, moving shadow during Martian twilight or directly causing overexposure of the onboard camera. Under these interference conditions, the engineering integrated development server 13 collects navigation data (such as displacement estimation from the visual odometry, confidence level of target detection, number of extracted feature points, etc.) output by the Mars rover's physical teaching aid perception and detection module to verify its edge robustness under different lighting environments.

[0069] By working in tandem with the communication degradation simulator and the perception interference trigger, the environmental stress injection module 134 allows developers to systematically and repeatably test the algorithm's performance under harsh conditions such as communication failures and perception degradation in laboratory settings without altering the real physical environment. This provides students with the opportunity to experience and learn how to cope with complex working conditions in a safe and controlled environment.

[0070] In one embodiment, the data collaborative evaluation subsystem 14 includes a time synchronization and deviation quantification module 141, which is configured to: timestamp-align the global pose ground truth and the telemetry data (establish a global time reference to eliminate the effects of acquisition delay and transmission delay, and obtain the time-synchronized global pose ground truth and telemetry data); based on the time-synchronized global pose ground truth and telemetry data, calculate the ratio of the actual travel distance to the theoretical travel distance at the wheel end of the Mars rover physical teaching aid, calculate the pose deviation rate and cumulative drift trend coefficient per unit time, and statistically analyze the standard deviation and extreme value fluctuation range of the target detection confidence or positioning accuracy index output by the perception and detection module of the Mars rover physical teaching aid, to obtain slip rate, pose drift rate and perception confidence fluctuation data; and generate quantitative verification and evaluation data for physical slippage conditions, perception failure conditions and communication degradation conditions based on the slip rate, pose drift rate and perception confidence fluctuation data, as well as the extreme working condition stress.

[0071] The aforementioned time synchronization and deviation quantification module 141 can accurately align and quantify the global pose true value collected by the physical test field with the telemetry data reported by the Mars rover, and finally generate objective evaluation data for various extreme working conditions, such as generating an adaptive evaluation report containing indicators such as slip rate and pose drift rate, thereby providing students with objective quantitative evaluation methods.

[0072] In its implementation, the time synchronization and deviation quantization module 141 first precisely aligns the high-precision global pose ground truth from the miniature physical test field with the telemetry data from the Mars rover's physical teaching aids, ensuring that subsequent pose comparisons are performed at the same physical moment. After time synchronization is completed, a series of quantization calculations are executed. For example, the slip ratio is calculated, which is the ratio of the actual travel distance at the end of the Mars wheel (obtained by accumulating the global ground truth positions) to the theoretical travel distance (obtained by integrating the wheel speed encoder data). The slip ratio is a core indicator for measuring whether wheel-ground contact slippage occurs; when the slip ratio deviates significantly from 1, it indicates that the wheel is spinning or dragging. Another example is the pose drift ratio, which is the pose deviation rate per unit time (the derivative of position and angle deviations with respect to time) and the cumulative drift trend coefficient (the slope obtained by linear regression of the deviation sequence), used to quantify the speed and trend of deviation growth over time. For example, statistical perception confidence fluctuation, which refers to the standard deviation and extreme value fluctuation range of the target detection confidence or positioning accuracy index output by the Mars rover's perception and detection module throughout the entire test process, is used to measure the stability and reliability of the perception module's output. Through a series of quantitative calculations, the time synchronization and deviation quantization module 141 obtains slip rate, pose drift rate, and perception confidence fluctuation data.

[0073] Finally, the quantitative indicators calculated by the time synchronization and deviation quantification module 141 are correlated with the extreme operating stresses injected by the server during the test (such as communication delay, packet loss rate, and type of lighting interference) to generate structured quantitative verification and evaluation data, providing students with objective and traceable basis for improvement. For example, the report points out that under this lighting and terrain conditions, the original obstacle avoidance algorithm lacks compensation for slippage rates exceeding 25%, and communication delay amplifies this deviation. Students can optimize their algorithms based on the data in the report and conduct secondary verification on the same platform.

[0074] In one embodiment, the miniature physical test field 12 includes a matrix-type dynamic sensing board 121, which is hidden beneath the physical medium simulating Martian terrain. The matrix-type dynamic sensing board 121 is equipped with a pressure sensor array. The matrix-type dynamic sensing board 121 is configured to: collect transient pressure data of the wheel ends of the Martian rover physical teaching aid against the ground in real time through the pressure sensor array; perform spatial distribution analysis and temporal variation analysis on the transient pressure data to obtain wheel end pressure distribution data and wheel-ground contact force variation characteristics; and send the wheel end pressure distribution data and wheel-ground contact force variation characteristics to the engineering integrated development server to verify the dynamic simulation accuracy of the virtual Martian rover in the digital twin model.

[0075] In the above embodiment, a hidden matrix-type dynamic sensing plate 121 is configured inside the miniature physical test field 12 to collect the transient pressure distribution of the Martian vehicle wheel tip on the ground in real time. The matrix-type dynamic sensing plate 121 is buried under physical media simulating Martian terrain, such as sand and gravel, so as not to affect the normal driving of the Martian vehicle, but to sense the lowest physical interaction between the vehicle and the ground, i.e., the wheel-ground contact pressure, in real time.

[0076] The matrix-type dynamic sensing board 121 includes a densely arranged array of pressure sensors. During operation, the pressure sensor array collects real-time transient pressure data on the ground applied by the wheels (or tracks) of the Mars rover's physical teaching aids. This data includes not only the magnitude of the pressure but also the location and duration of the pressure application. Subsequently, the matrix-type dynamic sensing board 121 performs spatial distribution analysis and temporal variation analysis on the collected transient pressure data. Spatial distribution analysis generates a dynamic heat map of the wheel-end pressure distribution based on the physical positions of the sensors on the board, visually displaying where the pressure is concentrated on the tire (e.g., the center or the edge). Temporal variation analysis extracts the characteristics of the wheel-ground contact force over time, such as the peak value of the pressure impact, fluctuation frequency, and rise time. The results of these two types of analysis together constitute the wheel-end pressure distribution data and the characteristics of wheel-ground contact force variation.

[0077] Finally, the above analysis data was sent to the engineering integrated development server 13 to verify the accuracy of the dynamic simulation of the virtual Mars rover in the digital twin model. When the virtual Mars rover travels in the digital twin model, the simulated wheel-ground contact force can be compared point by point with the pressure data measured in the physical world. If the two match well, it indicates that the dynamic model is accurate; if there is a deviation, it indicates that the model parameters (such as friction coefficient, sinking coefficient, stiffness coefficient, etc.) need to be calibrated.

[0078] By employing the above implementation method, the key physical quantity of wheel-to-ground contact force, which is traditionally difficult to measure directly, can be transformed into quantifiable and comparable data. This not only provides a high-precision true value reference for the calibration of digital twin models but also allows students to gain a deeper understanding of the complex interaction mechanisms between wheeled mobile robots and soft terrain. Through the analysis data of the matrix-type dynamics perception board 121, microscopic physical phenomena can be intuitively presented and quantitatively analyzed in teaching.

[0079] In one embodiment, the main control module 111 of the Mars rover physical teaching aid 11 is configured to: receive algorithm update instructions sent by the engineering integrated development server; perform remote hot replacement of local control operators during the operation of the Mars rover physical teaching aid; collect dynamic response data of the physical actuator of the Mars rover physical teaching aid under the action of different control operators; send the dynamic response data as part of the telemetry data to the engineering integrated development server to obtain dynamic response deviation data corresponding to different algorithm control quantities.

[0080] In this implementation, the main control module 111 is equipped with a modular algorithm dynamic loading unit, which supports remote hot-swapping of local control operators via the engineering integrated development server 13 during the operation of the Mars rover, and real-time monitoring of the dynamic response deviation of the physical actuators to different algorithm control quantities. This design allows students to dynamically change or upgrade the control algorithm without shutting down the Mars rover, restarting it, or interrupting the current mission, and to instantly compare the performance differences between the old and new algorithms in the actual physical environment.

[0081] In practical implementation, firstly, the main control module 111 of the Mars rover physical teaching aid 11 receives algorithm update instructions sent in real time by the engineering integrated development server 13. This instruction includes an executable local control operator, such as an improved proportional-integral-derivative (PID) control module, a fuzzy logic rule base for slip compensation, or a completely new obstacle avoidance decision logic. The modular algorithm dynamic loading unit within the main control module 111 can dynamically load the new control operator into memory without affecting the main program's operation, replacing the currently running old operator, thus achieving hot replacement.

[0082] Secondly, after the hot replacement is completed (or during the parallel comparison test of the old and new operators), the main control module 111 collects high-precision dynamic response data of the physical actuators (such as drive motors, steering servos, brakes, etc.) of the Mars rover physical teaching aids under the action of different control operators. The dynamic response data may include, but is not limited to: response time (the time from the issuance of the command to the start of the actuator's action), overshoot (the extent by which the actuator exceeds the target value), steady-state error (the deviation from the target value after reaching stability), and energy consumption changes, etc.

[0083] Finally, this dynamic response data is sent back to the engineering integrated development server 13 in real time as part of the telemetry data. The engineering integrated development server 13 can obtain the dynamic response deviation data corresponding to different algorithm control quantities by comparing and analyzing the response data under the action of the old and new operators. This means that students can run the original algorithm and record the deviation data on the same test track under the same physical conditions, and then switch to the improved algorithm via hot-swap to immediately observe the improvement effect without restarting the system or restarting the experiment.

[0084] The above implementation significantly shortens the iteration cycle of algorithm development and verification. This is of particular significance for engineering education: students can see in real time how their modified algorithm parameters immediately change the physical behavior of the Mars rover. This instant feedback greatly enhances the intuitiveness and exploratory nature of learning, and is a key function for cultivating students' system-level engineering thinking and algorithm optimization capabilities.

[0085] In one embodiment, the sensing and detection module 113 is further configured to: acquire calibration data of a preset physical calibration target; and after the hardware reconstruction of the Mars rover physical teaching aid, automatically complete the spatial alignment of the visual sensor coordinate system and the lidar coordinate system based on the calibration data.

[0086] The above-described implementation enhances the functionality of the perception and detection module 113 of the Mars rover physical teaching aid 11 by equipping it with a multi-source data spatial calibration unit. This enables automatic alignment of the coordinate systems between the visual sensor and the lidar, significantly reducing the complexity and time cost for students to recalibrate the relative positional relationship between sensors after replacing or reconfiguring hardware modules.

[0087] In its implementation, the automatic calibration process consists of two steps. First, the sensing and detection module 113 acquires calibration data from a physical calibration target preset within the miniature physical test field 12 or as an independent component. This calibration target has known geometric dimensions and spatial feature points (e.g., a planar checkerboard with multiple black and white squares, or a three-dimensional calibration object with known corner coordinates). This calibration data contains the precise spatial position information of each feature point on the target within the target's own coordinate system.

[0088] Secondly, after the hardware of the Mars rover's physical teaching aids is reconstructed (for example, the user replaces a different model of camera, changes the installation position or angle of the lidar, or reassembles the entire sensing mast), the sensing and detection module 113 automatically runs a calibration program. This program identifies feature points on the calibration target (such as the corner points of a checkerboard) through the visual sensor, and simultaneously obtains its three-dimensional point cloud data by scanning the same calibration target with the lidar. Then, using a matching algorithm (such as the Perspective-n-Point (PnP) algorithm or the Iterative Closest Point (ICP) point cloud registration algorithm), the optimal rigid body transformation matrix (i.e., rotation matrix and translation vector) between the visual sensor coordinate system and the lidar coordinate system is calculated, thereby completing the spatial alignment of the two.

[0089] The above implementation achieves spatial synchronization in multi-sensor fusion. On a real Mars rover, visual cameras and lidar are typically mounted in different locations, with their data residing in different coordinate systems. Accurately mapping pixels in an image to distance information in a point cloud requires knowing the precise spatial relationship (i.e., extrinsic parameters) between the two sensors. Traditional calibration processes are cumbersome, time-consuming, and require specialized knowledge and tools. In educational settings, students frequently need to disassemble, replace, and reassemble hardware modules for experiments on different topics. If manual recalibration is required after each reconstruction, it will significantly impact learning efficiency and experience. The automatic calibration function provided in this implementation allows students to focus on the algorithm and system design itself, without being bogged down in the details of underlying sensor calibration.

[0090] In one embodiment, the engineering integrated development server 13 is configured with a hardware abstraction layer driver library and an automated compilation environment; the hardware abstraction layer driver library is used to provide a unified driver interface for different hardware modules, and the automated compilation environment is used to automatically compile the algorithm code written by the user into an executable program and load it into the Mars rover physical teaching aid.

[0091] The above implementation configures the development environment of the engineering integrated development server 13, giving it two core components: a hardware abstraction layer driver library and an automated compilation environment. This design provides students with a low-barrier, high-efficiency secondary development platform, allowing them to focus on the algorithm logic itself without being bogged down by underlying hardware details and complex compilation and deployment processes.

[0092] The Hardware Abstraction Layer (HAL) driver library is a software layer that provides a unified driver interface for different hardware modules. In Mars rover physical teaching aid 11, there may be various hardware modules from different manufacturers and models, such as motor drivers, inertial measurement units, lidar, and vision cameras. The underlying registers, communication protocols, and control instructions of these hardware components differ. If users need to directly operate these hardware components, they will face extremely high learning costs and programming complexity. The role of the HAL is to encapsulate these underlying differences, providing a standardized application programming interface for the upper-level algorithm code. This ensures that the algorithm code written by users has good hardware independence and portability; even if hardware modules are replaced, the upper-level algorithm code does not need to be modified.

[0093] The automated compilation environment is a system that automatically converts user-written high-level language algorithm code into executable programs that can run on the Mars rover physical teaching aid. After users write control algorithms, perception processing logic, or decision-making planning code in high-level languages ​​such as C++ and Python on the server side, the automated compilation environment automatically calls the cross-compilation toolchain to compile the code into a binary executable file for the Mars rover's main control chip (such as ARM architecture or RISC-V architecture). Subsequently, the executable program is automatically loaded and deployed to the main control module 111 of the Mars rover physical teaching aid 11 via a wireless network. Students do not need to manually configure compilation parameters, burn programs via data cables, or worry about the underlying library dependencies of the target platform; the entire process is completed automatically in the server background.

[0094] This implementation significantly lowers the technical barrier for students developing robot algorithms. In traditional embedded robot development, students need to master a series of complex skills such as cross-compilation, driver development, and firmware flashing, often spending a significant amount of time on environment configuration and debugging rather than learning the algorithms themselves. Through the hardware abstraction layer, students can use familiar programming languages ​​and unified interfaces to call hardware functions; through the automated compilation environment, students can easily deploy code to physical robots and observe the running effects, just like developing software on a regular computer. This WYSIWYG development experience allows students to devote more energy to algorithm innovation and system design.

[0095] In one embodiment, the miniature physical test field 12 includes a terrain modular configuration unit 122, which includes multiple modules such as sand, gravel, clay, and slope adjustment modules that can be quickly replaced.

[0096] In this implementation, the terrain of the test site is designed to consist of multiple standardized, quickly replaceable independent modules to support flexible terrain reconstruction and diverse experimental scenario configurations.

[0097] In practical implementation, the modular terrain configuration unit 122 can adopt a standardized tray or slot design, allowing students to easily replace its content modules. The sand module simulates sandy deserts or dunes on the Martian surface, and different particle sizes (e.g., coarse, medium, and fine sand) and laying depths can be selected as needed. The gravel module simulates rocky areas or scree slopes on the Martian surface, and different particle sizes (e.g., small gravel, medium gravel) and edge angles can be selected. The clay module simulates soil areas with certain viscosity and plasticity, and its hardness and adhesion properties can be changed by adjusting the water content. The slope adjustment module simulates slopes, impact crater edges, and other terrain undulations on the Martian surface, typically using pneumatic, electric, or mechanical lifting mechanisms to achieve rapid adjustment of slope angles. These modules can be arbitrarily combined and arranged according to experimental needs to construct complex and varied terrain scenarios within the test field.

[0098] In traditional fixed-terrain test fields, students can only conduct repetitive experiments on a pre-set terrain, making it difficult to experience the impact of different terrains on the rover's driving performance. Through modular configuration units, teachers can quickly set up different challenge scenarios for students, such as combinations of sand and slopes, or transitional zones of gravel and clay, allowing students to compare the performance differences of the same algorithm under different terrains within the same experimental session. Simultaneously, students can also design and arrange their own terrains to explore the rover's traversability and stability under different terrain combinations. This terrain construction method not only reduces the construction and maintenance costs of the test field but also stimulates students' creativity and exploratory spirit, making the test field itself a programmable and reconfigurable experimental platform, rather than a fixed hardware facility.

[0099] In one embodiment, the miniature physical test field 12 includes an automatic terrain calibration module 123; The automatic terrain calibration module 123 is configured to: obtain the geometric parameters of the current terrain and the mechanical properties of the physical medium based on the global pose ground truth and the data collected by the matrix dynamics sensing board; construct a terrain geometric model and a physical property model according to the geometric parameters and the mechanical properties, and generate automatic terrain modeling data; and send the automatic terrain modeling data to the engineering integrated development server to update the terrain geometric and physical property parameters of the digital twin model.

[0100] The above-described implementation equips the miniature physical test field 12 with an automatic terrain calibration module 123. When students replace or rearrange the terrain modules, the automatic terrain calibration module 123 can automatically sense, measure, and model the geometric features and physical properties of the new terrain without requiring manual parameter input, thereby ensuring that the digital twin model maintains a high degree of consistency with the physical environment.

[0101] In practical implementation, the automatic terrain calibration module 123 can obtain key parameters of the current terrain based on the global pose ground truth and the data collected by the matrix-type dynamic sensing board 121. The global pose ground truth is the precise position trajectory of the Mars rover in the test field coordinate system. By analyzing the pose changes of the Mars rover when driving in different terrain areas, the geometric parameter information of the terrain (such as slope, concavity and convexity, and elevation changes) can be deduced. The data collected by the matrix-type dynamic sensing board 121 includes the transient pressure distribution of the Mars rover's wheel end on the ground and the characteristics of wheel-ground contact force changes. Based on this data, the mechanical property information of the physical medium (such as stiffness, bearing characteristics, internal friction angle, damping coefficient, etc.) can be determined.

[0102] The automatic terrain calibration module 123 automatically constructs two types of models based on the acquired geometric parameters and mechanical properties: one is the terrain geometric model, which is a three-dimensional grid or elevation map describing the shape and elevation changes of the ground, accurately reflecting the slope, undulation, and obstacle distribution of each area; the other is the physical property model, which is a parameter distribution map describing the mechanical properties of the ground medium in different areas, such as the friction coefficient, subsidence coefficient, and coefficient of restitution. These two types of models together constitute complete automatic terrain modeling data.

[0103] Finally, the automatic terrain calibration module 123 sends the generated automatic terrain modeling data to the engineering integrated development server 13 for automatically updating the terrain geometric and physical property parameters in the digital twin model. This means that when a teacher or student replaces a sand module with a gravel module, the system can automatically identify the terrain change without any manual intervention and update the geometric parameters (such as surface roughness) and physical parameters (such as rolling friction coefficient) of that area in the digital twin model to the values ​​corresponding to gravel.

[0104] The above implementation method achieves plug-and-play intelligent configuration of the test field. In traditional simulation testing environments, students must manually measure terrain parameters and input them one by one into the simulation software, which is not only tedious and time-consuming but also prone to human error. Through the automatic terrain calibration module 123, the system can actively sense changes in the terrain and automatically synchronize them to the digital world using the rover's own driving data and the mechanical data of the hidden sensing plate. This greatly simplifies the experimental preparation process, ensures a high degree of consistency between the physical environment and the digital twin model, and provides a reliable foundation for subsequent high-precision virtual-real interactive verification and deviation analysis.

[0105] like Figure 2 As shown, Figure 2 This is a schematic diagram of the overall logic of the Mars rover simulation development system 100 provided in this application embodiment. In summary, the data interaction and functional collaboration relationships of the Mars rover physical teaching aid 11, the miniature physical test field 12, the engineering integrated development server 13, and the data collaborative evaluation subsystem 14 of the Mars rover simulation development system 100 are as follows: The Mars rover physical teaching aid 11 operates within the miniature physical test field 12, collecting telemetry data (including motion state data, environmental perception data, and vibration spectrum data) in real time during its movement, and uploading the telemetry data to the engineering integrated development server 13 via a wireless link. Simultaneously, the miniature physical test field 12, through its built-in global pose capture array, collects the global pose ground truth value of the Mars rover physical teaching aid 11 and sends it to the engineering integrated development server 13.

[0106] The engineering integrated development server 13 serves as the core computing and collaboration hub of the system, performing the following key functions: Based on the geometric parameters and physical medium properties of the miniature physical test field 12, a digital twin model is constructed and updated in real time. Nonlinear deviation analysis is performed on the global pose ground truth and telemetry data to generate a visual representation of the physical deviation, and the path deviation area is rendered in the digital twin model. The system receives vibration spectrum data transmitted back from the Mars rover physical teaching aid 11, automatically identifies the type of physical medium and the corresponding physical drag coefficient through the terrain recognition module, and maps it to the digital twin model for annotation. By using the environmental stress injection module, random delay or packet loss parameters are injected into the wireless link of the Mars rover physical teaching aid 11, and / or light interference is injected into the physical environment of the miniature physical test field 12 to simulate extreme working stress. It supports remote algorithm hot replacement of Mars rover physical teaching aid 11 and collects dynamic response deviation data under the action of different control operators.

[0107] The data collaborative evaluation subsystem 14 receives the aligned global pose true value and telemetry data from the engineering integrated development server 13, performs time synchronization and deviation quantification calculation, generates quantitative indicators such as slip rate, pose drift rate and perception confidence fluctuation, and outputs quantitative verification evaluation data for physical slip conditions, perception failure conditions and communication degradation conditions.

[0108] In addition, the miniature physical test field 12 can also be configured with a matrix dynamic sensing board 121, a terrain modular configuration unit 122 and a terrain automatic calibration module 123, which are used to collect wheel end pressure distribution data, realize rapid terrain replacement and automatically generate terrain modeling data, respectively, to further support the dynamic simulation accuracy verification of the digital twin model and the automatic updating of terrain parameters.

[0109] In summary, the Mars rover simulation development system 100, through the collaborative work of the above modules, constructs a complete verification platform that is physically realistic, process-transparent, and objectively evaluated, providing a high-fidelity physical development and verification environment for engineering education.

[0110] like Figure 3As shown, this disclosure provides a Mars rover simulation development method. The execution entity of this method can be a server, such as an engineering integrated development server; or it can be other devices controlled by the server. The method includes: S301: Control the Mars rover physical teaching aid to operate in the miniature physical test field, collect the telemetry data of the Mars rover physical teaching aid, and simultaneously collect the global pose true value of the Mars rover physical teaching aid in the miniature physical test field.

[0111] S302: Construct a digital twin model of the miniature physical test field, perform nonlinear deviation analysis based on the global pose ground truth and the telemetry data, and generate a visual representation of the physical deviation in the digital twin model.

[0112] S303: Identify the physical medium type of the Mars rover physical teaching aid based on the vibration spectrum data in the telemetry data, and map the identified physical medium type to the digital twin model for annotation.

[0113] S304: Inject extreme working condition stress into the wireless link of the physical teaching aid of the Mars rover and / or the physical environment of the miniature physical test field.

[0114] S305: Based on the time synchronization and deviation quantization between the global pose true value and the telemetry data, generate quantitative verification and evaluation data for the extreme working condition stress.

[0115] For details on the specific implementation and implementation of the above methods, please refer to the description of the Mars rover simulation development system mentioned above, which will not be repeated here.

[0116] The instructions corresponding to the above-mentioned Mars rover simulation development method can be stored in the form of a computer program product or a computer-readable storage medium. For example, the computer program product or computer-readable storage medium can store the program instructions corresponding to the above-mentioned method. When the program instructions are executed by the processor, the above-mentioned Mars rover simulation development method is executed.

[0117] The Mars rover simulation development method can be executed by a server or other device, which may include a processor and a memory. The memory may store computer program instructions corresponding to the above method, and the processor is used to execute the computer program instructions to implement the steps of the above Mars rover simulation development method.

[0118] This disclosure also provides a Mars rover simulation development apparatus, in which each step of the above-described Mars rover simulation development method can be executed or stored in the form of software modules, such as including: The control module is used to control the Mars rover physical teaching aid to operate in the miniature physical test field, collect the telemetry data of the Mars rover physical teaching aid, and simultaneously collect the global pose true value of the Mars rover physical teaching aid in the miniature physical test field.

[0119] The digital twin module is used to construct a digital twin model of the miniature physical test field, perform nonlinear deviation analysis based on the global pose ground truth and the telemetry data, and generate a visual representation of the physical deviation in the digital twin model. Furthermore, it identifies the physical medium type of the Mars rover's physical teaching aid based on the vibration spectrum data in the telemetry data, and maps the identified physical medium type to the digital twin model for annotation.

[0120] An extreme stress injection module is used to inject extreme working condition stress into the wireless link of the Mars rover physical teaching aid and / or the physical environment of the miniature physical test field.

[0121] The quantization verification and evaluation module is used to generate quantization verification and evaluation data for the extreme working condition stress based on the time synchronization and deviation quantization between the global pose true value and the telemetry data.

[0122] The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without any inventive effort.

[0123] If the technical solution of this application involves personal information, the product using this technical solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If the technical solution of this application involves sensitive personal information, the product using this technical solution has obtained the user's separate consent before processing the sensitive personal information, and also meets the requirement of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set up to inform users that they have entered the scope of personal information collection and that personal information will be collected. If an individual voluntarily enters the collection scope, it is deemed that they have agreed to the collection of their personal information; or on the personal information processing device, with clear signs / information informing users of the personal information processing rules, authorization is obtained from the user through pop-up information or by asking the user to upload their personal information; wherein, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

[0124] The embodiments of the subject matter and functional operation described in this specification can be implemented in the following ways: digital electronic circuits, tangibly embodied computer software or firmware, computer hardware including the structures disclosed in this specification and their structural equivalents, or combinations thereof. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible, non-transitory program carrier for execution by a data processing apparatus or for controlling the operation of a data processing apparatus. Alternatively or additionally, the program instructions may be encoded on artificially generated propagation signals, such as machine-generated electrical, optical, or electromagnetic signals, which are generated to encode information and transmit it to a suitable receiving device for execution by the data processing apparatus. The computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or combinations thereof.

[0125] The processing and logic flow described in this specification can be executed by one or more programmable computers that execute one or more computer programs to perform corresponding functions by operating on input data and generating output. The processing and logic flow can also be executed by dedicated logic circuitry—such as FPGAs (Field-Programmable Gate Arrays) or ASICs (Application-Specific Integrated Circuits), and the device can also be implemented as dedicated logic circuitry.

[0126] Suitable computers for executing computer programs include, for example, general-purpose and / or special-purpose microprocessors, or any other type of central processing unit. Typically, the central processing unit receives instructions and data from read-only memory and / or random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more mass storage devices for storing data, such as disks, magneto-optical disks, or optical disks, or the computer will be operatively coupled to such mass storage devices to receive data from or transfer data to them, or both. However, a computer is not required to have such devices. Furthermore, a computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.

[0127] Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, such as semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD-ROM and DVD-ROM disks. Processors and memory may be supplemented by or incorporated into dedicated logic circuitry.

[0128] While this specification contains numerous specific implementation details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather are primarily intended to describe features of specific embodiments of a particular invention. Certain features described in the various embodiments herein may also be implemented in combination in a single embodiment. Conversely, various features described in a single embodiment may also be implemented separately in various embodiments or in any suitable sub-combination. Furthermore, while features may function in certain combinations as described above and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and a claimed combination may refer to a sub-combination or a variation thereof.

[0129] Similarly, although the operations are depicted in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order shown or sequentially, or requiring all illustrated operations to be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0130] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings are not necessarily shown in a specific order or sequence to achieve the desired result. In some implementations, multitasking and parallel processing may be advantageous.

[0131] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A Mars rover simulation development system, characterized in that, include: A physical teaching aid for Mars rovers, configured to operate in a physical environment and collect telemetry data; The miniature physical test field is configured to provide a physical operating environment that simulates the Martian terrain and to collect the global pose truth value of the physical teaching aids of the Mars rover. The engineering integrated development server wirelessly communicates with the physical teaching aids of the Mars rover and is configured to construct a digital twin model of the miniature physical test field; perform nonlinear deviation analysis based on the global pose ground truth and the telemetry data, and generate a visual representation of the physical deviation in the digital twin model; and identify the current physical medium type based on the vibration spectrum data in the telemetry data, and map the identified physical medium type to the digital twin model for annotation. In addition, extreme working condition stresses are injected into the wireless link of the physical teaching aid of the Mars rover and / or the physical environment of the miniature physical test field. The data collaborative evaluation subsystem is configured to generate quantitative verification evaluation data for the extreme working condition stress based on the time synchronization and deviation quantification between the global pose true value and the telemetry data.

2. The system according to claim 1, characterized in that, The physical teaching aid for the Mars rover includes a main control module, a chassis module, and a sensing and detection module. The chassis module is configured to perform motion control, drive the Mars rover physical teaching aid to move within the miniature physical test field, and generate motion state data; The sensing and detection module is configured to collect environmental sensing data, which includes at least image data, point cloud data, and inertial measurement data. The main control module has edge computing capabilities and is configured to coordinate the work of the chassis module and the sensing and detection module, and to perform multi-source data fusion and feature extraction on the environmental sensing data collected by the sensing and detection module to obtain processed environmental sensing data; the processed environmental sensing data and the motion state data constitute the telemetry data.

3. The system according to claim 1, characterized in that, The Mars rover physical teaching aid includes a main control module, which is equipped with a three-axis accelerometer. The main control module is configured as follows: The triaxial vibration acceleration data of the Mars rover physical teaching aid during its movement is collected in real time by the triaxial accelerometer, and the triaxial vibration acceleration data is transformed in the frequency domain to obtain the vibration spectrum data.

4. The system according to claim 1, characterized in that, The engineering integrated development server includes a nonlinear deviation analysis module, which is configured as follows: The global pose true value and the relative pose data in the telemetry data are synchronized in time, and the position deviation vector and attitude deviation angle at each time point are calculated to obtain the instantaneous pose deviation sequence. A nonlinear deviation growth model is constructed based on the instantaneous pose deviation sequence. The nonlinear deviation growth model is used to detect the abrupt change points and cumulative drift trends of the instantaneous pose deviation sequence, and to extract nonlinear deviation feature parameters. Based on the nonlinear deviation characteristic parameters, the deviation characteristic quantities corresponding to physical slippage conditions, sensing failure conditions, and communication degradation conditions are identified, and nonlinear deviation analysis results for extreme conditions are generated.

5. The system according to claim 1, characterized in that, The engineering integrated development server includes a physical deviation visualization module, which is configured as follows: Based on the geometric parameters and physical medium properties of the miniature physical test field, a three-dimensional digital twin model of the miniature physical test field is constructed. The global pose true value and the telemetry data are obtained. In the three-dimensional digital twin model, the virtual Mars rover is driven to move synchronously according to the telemetry data, and the global pose marker points corresponding to the global pose true value are driven to move synchronously according to the global pose true value. Calculate the spatial difference between the pose of the virtual Mars rover and the pose of the global pose marker to obtain the real-time spatial deviation. Based on the real-time spatial deviation, the path deviation area is rendered in the digital twin model using color gradients or shadow areas, and the deviation position caused by physical slippage, collision, or perceptual drift is marked to generate a visual representation of the physical deviation.

6. The system according to claim 1, characterized in that, The engineering integrated development server includes a terrain recognition module, which is configured as follows: The system receives vibration spectrum data transmitted back from the physical teaching aid of the Mars rover, extracts the frequency peak value, amplitude distribution and energy concentration characteristic parameters of the vibration spectrum data, and generates a vibration feature vector. Based on the vibration feature vector and the preset terrain feature database, the similarity between the vibration feature vector and the standard vibration feature vector is calculated, and the physical medium type with the highest similarity and its corresponding physical drag coefficient are identified; the terrain feature database stores the standard vibration feature vectors corresponding to different physical medium types. The physical medium type and the physical drag coefficient are mapped to the digital twin model, the physical attribute labels of the current terrain area are updated, and the dynamic simulation parameters of the virtual Mars rover are adjusted synchronously to complete the labeling of the physical medium type in the digital twin model.

7. The system according to claim 1, characterized in that, The engineering integrated development server includes an environmental stress injection module, which includes a communication degradation simulator and a sensing interference trigger. The communication degradation simulator is configured to inject preset random delay parameters or packet loss parameters into the wireless communication link of the Mars rover physical teaching aid to simulate communication delay conditions, and collect autonomous decision-making response data of the Mars rover physical teaching aid under the communication delay conditions. The sensing interference trigger is configured to: control the adjustable light source array in the miniature physical test field to generate dynamic long shadows or strong light overexposure lighting interference conditions, and collect navigation data of the sensing and detection module of the Mars rover physical teaching aid under the lighting interference conditions.

8. The system according to claim 1, characterized in that, The data collaborative evaluation subsystem includes a time synchronization and deviation quantification module, which is configured as follows: The global pose ground truth and the telemetry data are timestamped to obtain the time-synchronized global pose ground truth and telemetry data. Based on the global pose true value and telemetry data after time synchronization, the ratio of the actual travel distance of the wheel end of the Mars rover physical teaching aid to the theoretical travel distance is calculated. The pose deviation rate and cumulative drift trend coefficient per unit time are calculated. The standard deviation and extreme value fluctuation range of the target detection confidence or positioning accuracy index output by the perception and detection module of the Mars rover physical teaching aid are statistically analyzed to obtain slip rate, pose drift rate and perception confidence fluctuation data. Based on the slip rate, pose drift rate, and perception confidence fluctuation data, as well as the extreme operating stress, quantitative verification and evaluation data are generated for physical slippage conditions, perception failure conditions, and communication degradation conditions.

9. The system according to claim 1, characterized in that, The miniature physics test field includes a matrix-type dynamic sensing panel, which is concealed beneath the physical medium simulating Martian terrain. The matrix-type dynamic sensing panel is equipped with a pressure sensor array. The matrix-type dynamic sensing panel is configured as follows: The pressure sensor array collects transient pressure data of the wheel ends of the Mars rover physical teaching aid against the ground in real time, and performs spatial distribution analysis and temporal variation analysis on the transient pressure data to obtain wheel end pressure distribution data and wheel-ground contact force variation characteristics. The wheel-end pressure distribution data and the wheel-ground contact force variation characteristics are sent to the engineering integrated development server to verify the dynamic simulation accuracy of the virtual Mars rover in the digital twin model.

10. The system according to claim 1, characterized in that, The Mars rover physical teaching aid includes a main control module, which is configured as follows: Receive the algorithm update instruction sent by the engineering integrated development server, and perform remote hot replacement of the local control operator during the operation of the Mars rover physical teaching aid; The dynamic response data of the physical actuator of the Mars rover physical teaching aid under the action of different control operators is collected, and the dynamic response data is sent to the engineering integrated development server as part of the telemetry data to obtain the dynamic response deviation data corresponding to the control quantities of different algorithms.

11. The system according to claim 2, characterized in that, The sensing and detection module is also configured to: Obtain calibration data from a preset physical calibration target; After the hardware reconstruction of the physical teaching aid for the Mars rover, the spatial alignment of the visual sensor coordinate system and the lidar coordinate system is automatically completed based on the calibration data.

12. The system according to claim 1, characterized in that, The engineering development server is equipped with a hardware abstraction layer driver library and an automated compilation environment; The hardware abstraction layer driver library is used to provide a unified driver interface for different hardware modules, and the automated compilation environment is used to automatically compile the algorithm code written by the user into an executable program and load it into the Mars rover physical teaching aid.

13. The system according to claim 1, characterized in that, The miniature physical test field includes a terrain modular configuration unit, which includes multiple modules such as sand, gravel, clay, and slope adjustment modules that can be quickly replaced.

14. The system according to claim 9, characterized in that, The miniature physical test field includes an automatic terrain calibration module; The automatic terrain calibration module is configured to: obtain the geometric parameters of the current terrain and the mechanical properties of the physical medium based on the global pose ground truth and the data collected by the matrix dynamics sensing board; construct a terrain geometric model and a physical property model according to the geometric parameters and the mechanical properties, and generate automatic terrain modeling data; and send the automatic terrain modeling data to the engineering integrated development server to update the terrain geometric and physical property parameters of the digital twin model.

15. A Mars rover simulation development method, characterized in that, include: Control the physical teaching aid of the Mars rover to operate in the miniature physical test field, collect the telemetry data of the physical teaching aid of the Mars rover, and simultaneously collect the true global pose of the physical teaching aid of the Mars rover in the miniature physical test field. A digital twin model of the miniature physical test field is constructed, and a nonlinear deviation analysis is performed based on the global pose ground truth and the telemetry data. A visual representation of the physical deviation is generated in the digital twin model. Based on the vibration spectrum data in the telemetry data, the physical medium type of the Mars rover physical teaching aid is identified, and the identified physical medium type is mapped to the digital twin model for annotation; Inject extreme working condition stress into the wireless link of the physical teaching aid of the Mars rover and / or the physical environment of the miniature physical test field; Based on the time synchronization and deviation quantization between the global pose ground truth and the telemetry data, quantitative verification and evaluation data for the extreme working condition stress are generated.

Images