Inertial navigation system and method based on three-dimensional data reconstruction

Abstract

The invention discloses an inertial navigation system and method based on three-dimensional data reconstruction, and the system comprises a data collection module which is responsible for collecting the operation data, including acceleration and angular velocity, of the inertial navigation system in real time, and transmitting the data to a cloud server; a cloud server is established, and the established cloud server is used for receiving the data from the data acquisition module and constructing a three-dimensional data reconstruction model highly similar to an actual physical system by using high-performance computing resources and a complex algorithm model; the construction of the three-dimensional data reconstruction model is the core part of the whole system, and the three-dimensional data reconstruction model can simulate the operation state of the physical system in real time, dynamically update the model according to the input data, and obtain the optimization suggestion and prediction result of the physical system through the analysis of the three-dimensional data reconstruction model. According to the invention, a data acquisition module adopting various sensor technologies is adopted, so that the system applicability is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of high-precision attitude control, and in particular to an inertial navigation system and method based on three-dimensional data reconstruction. Background Art

[0002] Although the traditional attitude-controlled inertial navigation system can perform autonomous navigation without external reference signals, it has the problem of error accumulation, and long-term use will lead to a decrease in navigation accuracy.

[0003] Traditional testing and verification methods often need to be performed on actual physical equipment, which is not only time-consuming and labor-intensive, but also high-risk. Therefore, improving the accuracy and reliability of attitude control inertial navigation systems and providing an efficient and safe testing and verification method have become the focus of research. Summary of the invention

[0004] The present invention provides an inertial navigation system and method based on three-dimensional data reconstruction to solve the problem of reduced accuracy of traditional attitude control inertial navigation systems during long-term operation, and provides an efficient and safe testing and verification method.

[0005] In order to solve the above technical problems, one of the purposes of the present invention is to provide an inertial navigation system based on three-dimensional data reconstruction, comprising:

[0006] A data acquisition module, which is responsible for collecting the operating data of the inertial navigation system in real time, including acceleration and angular velocity, and transmitting the data to a cloud server;

[0007] Establishing a cloud server, wherein the cloud server is used to receive data from the data acquisition module and use high-performance computing resources and complex algorithm models to construct a three-dimensional data reconstruction model that is highly similar to the actual physical system;

[0008] Three-dimensional data reconstruction model, building a three-dimensional data reconstruction model is the core part of the entire system. It can simulate the operating status of the physical system in real time and dynamically update the model according to the input data. By analyzing the three-dimensional data reconstruction model, optimization suggestions and prediction results of the physical system can be obtained.

[0009] Furthermore, the data acquisition module uses a variety of sensor technologies, including MEMS accelerometers and gyroscopes, to ensure the diversity and accuracy of the data.

[0010] Furthermore, the cloud server also stores historical data and analysis results, providing users with a comprehensive data support platform.

[0011] Furthermore, a data-driven 3D data reconstruction model is adopted to calibrate the validity of the data collected by the detector in the software interface, and then trigger relevant control commands to the aircraft for attitude adjustment. This process is continuously iterated to increase navigation accuracy.

[0012] Furthermore, a PC-based user interface was established to provide users with a friendly operation portal, where users can view the real-time status, historical data and analysis results of the physical system and adjust the system parameter settings as needed.

[0013] Based on the same inventive concept, the present invention also provides an inertial navigation method based on three-dimensional data reconstruction, and the specific steps are as follows:

[0014] S1: Detector data and its communication related. Data received from external devices through the serial port is stored in the buffer area. The required data is then extracted from the buffer area and decoded according to the communication protocol. The data will be forwarded to the Tencent Cloud server via the MQTT protocol and serialized into the cloud database.

[0015] S2: Terminal operation mode, the user operates through the local PC directly connected to the detector, or remotely accesses the detector on the remote device. According to the actual operation of the system, the relevant parameters of IoTDB and EMQX are adjusted to optimize the performance of the system. The MQTT server is deployed on the cloud server to forward the data from the data source PC to the third-party mobile device;

[0016] S3: The three-dimensional data reconstruction model first sets an expected value or target, which is the state that the control system hopes to achieve; the actual output value of the controlled object is measured by the sensor, the expected value is compared with the actual output value, the error between the two is calculated, and the calculated control amount is applied to the controlled object to adjust it in the direction of reducing the error. This process will be repeated until the actual output value approaches or reaches the expected value;

[0017] S4: Construction of nozzle 3D scene, using digital tools to draw the sketch of the nozzle, including the main components and layout, and using blender to draw the nozzle parts;

[0018] S5: PC operation interface: Provides users with a friendly operation interface on the PC side, where users can view the real-time status, historical data and analysis results of the physical system and adjust the system parameter settings as needed.

[0019] Furthermore, the S2 terminal operation mode: If the system needs to process larger-scale data or higher concurrency, consider horizontal expansion or cluster deployment of IoTDB and EMQX. Through the combination of Apache IoTDB and EMQX, MQTT messages can be efficiently stored in the time series database to meet the needs of massive data storage, high-throughput data writing and complex data query analysis in the field of industrial Internet of Things.

[0020] Furthermore, the specific steps of constructing the S4 nozzle 3D scene are as follows:

[0021] 1) After CAD drawing is completed, import it into Unreal Engine and use the built-in modeling tools or import external model files to make fine adjustments to maintain a reasonable number of polygons to balance details and performance;

[0022] 2) Create appropriate materials for different parts of the nozzle and use UE5's physical rendering pipeline material to achieve more realistic effects;

[0023] 3) Use Substance Designer to create custom textures, ensure that the resolution and scale of the texture map match the model, and adjust the material parameters to achieve the desired visual effect;

[0024] 4) Arrange appropriate light sources in the scene to simulate real lighting effects, and use UE5's post-processing volume to adjust the scene parameters;

[0025] 5) Complete the animation process required for human-computer interaction through the level blueprint, complete the communication module between the UE5 scene and the PC program, including the communication protocol and specific communication implementation. After all adjustments are completed, perform the final rendering output to finally form a 3D scene program.

[0026] Furthermore, the S5 PC terminal operation is specifically as follows: QT is used for development, and the development process follows the MVC architecture. The application is divided into three core parts: model, view and controller. The model is responsible for managing the application data; the view is responsible for presenting the data in the model to the user, receiving the user's input and feeding back to the model; the controller is responsible for processing the user's input events and passing the user's request to the model. The controller acts as a coordinator between the model and the view to ensure that the view and the model data are synchronized.

[0027] The above one or more technical solutions of the present invention have at least one or more of the following technical effects:

[0028] Data acquisition modules using a variety of sensor technologies improve system applicability.

[0029] Use cloud servers to receive data and build a digital twin model that is highly similar to the actual physical system.

[0030] The digital twin model can simulate the operating status of the physical system in real time, making the operating status of the physical system very intuitive and improving the efficiency of the physical system operation analysis.

[0031] The user interface on the PC provides users with a friendly operation portal, where users can view the real-time status, historical data and analysis results of the physical system to improve work efficiency. They can also adjust system parameters at any time and observe changes in the operating status of the physical system, which is intuitive and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 : Logical structure diagram of the system of the present invention;

[0033] Figure 2 : Workflow structure diagram of two modes;

[0034] Figure 3 : CAD drawing of nozzle parts drawn by blender;

[0035] Figure 4 : static scene legend;

[0036] Figure 5 : Blueprints used in this patent;

[0037] Figure 6 : A collection of related operation interfaces. Specific implementation plan

[0038] The present invention discloses an attitude-controlled inertial navigation system based on three-dimensional data reconstruction technology and an implementation method thereof, comprising a data acquisition module, a cloud server, and a three-dimensional data reconstruction model.

[0039] The data acquisition module is responsible for real-time collection of the operating data of the inertial navigation system, including acceleration, angular velocity and other information, and transmits these data to the cloud server. The data acquisition module can use a variety of sensor technologies, such as MEMS accelerometers, gyroscopes, etc., to ensure the diversity and accuracy of the data. A cloud server is established to receive data from the data acquisition module, and use high-performance computing resources and complex algorithm models to build a three-dimensional data reconstruction model that is highly similar to the actual physical system. The cloud server can also store historical data and analysis results, providing users with a comprehensive data support platform. The three-dimensional data reconstruction model constructed by data is the core part of the entire system. It can simulate the operating status of the physical system in real time and dynamically update the model according to the input data. By analyzing the three-dimensional data reconstruction model, optimization suggestions and prediction results of the physical system can be obtained. This patent adopts a data-driven three-dimensional data reconstruction model, calibrates the validity of the data collected by the detector in the software interface, and then triggers the relevant control commands to the aircraft for attitude adjustment, and continuously iterates this process to increase navigation accuracy; in terms of usage, a PC-based user interface is established to provide users with a friendly operation entrance. Users can view the real-time status, historical data and analysis results of the physical system, and can also adjust the system parameter settings as needed. The logical structure block diagram of the entire system is shown in the figure. Figure 1 shown.

[0040] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in combination with the embodiments of the present invention and the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the obtained embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0041] Detector data and its communication related: Detector data is first transmitted to the PC via the serial port. High-speed serial port data unpacking is a complex process involving multiple steps and technologies. It ensures that the data received from the serial port can be correctly parsed and used. First, the data received from external devices (such as sensors, controllers, etc.) through the serial port is first stored in the cache. The cache is a memory area used to temporarily store data. It can effectively balance the speed difference of data input and output, ensure that the data will not be lost due to untimely processing, extract the required data from the cache, and decode it according to the communication protocol. During the unpacking process, various error conditions may be encountered, such as data format errors, verification failures, etc. At this time, corresponding error handling measures need to be taken, such as logging, sending error reports, or trying to re-receive data. Once the data is correctly parsed, it can be passed to the application layer for processing. The application layer can further process and analyze the data according to specific business needs. In this patent, the data will be forwarded to the Tencent Cloud Server via the MQTT protocol for serialization into the cloud database.

[0042] Terminal operation mode: Users can operate the detector through a local PC directly connected to the detector, or remotely through mobile phones, pads, computers and other devices, respectively for local mode and remote mode. The workflow structure of the two modes is as follows Figure 2 As shown. The implementation basis of the remote mode is the cloud server. The cloud server uses IOTDB to store the time series data of the detector. According to the actual operation of the system, the relevant parameters of IoTDB and EMQX, such as memory allocation, thread pool size, etc., can be adjusted to optimize the performance of the system. If the system needs to process larger-scale data or higher concurrency, horizontal expansion or cluster deployment of IoTDB and EMQX can be considered. Through the combination of Apache IoTDB and EMQX, MQTT messages can be efficiently stored in the time series database to meet the needs of massive data storage, high-throughput data writing, and complex data query and analysis in the field of industrial Internet of Things. On the other hand, the cloud server deploys an MQTT server to forward the data from the data source PC to the third-party mobile device.

[0043] Three-dimensional data reconstruction model: This project adopts a data-driven three-dimensional data reconstruction model, which has a typical PID control process between the detector data and the control instructions of the aircraft. The first is to set an expected value or target, which is the state that the control system hopes to achieve. Including heading angle and acceleration. The actual output value of the controlled object is measured by the sensor. The expected value is compared with the actual output value, and the deviation (error) between the two is calculated. According to the size of the error, the PID controller will calculate a control quantity. This control quantity is calculated based on three parameters: proportional (P), integral (I) and differential (D). The proportional term is used to reduce the current error, the integral term is used to eliminate the accumulated error in the past, and the differential term is used to predict the future error change trend. The calculated control quantity is applied to the controlled object to adjust it in the direction of reducing the error. This process will be repeated until the actual output value approaches or reaches the expected value. The PID control process is a closed-loop feedback control system that ensures that the controlled object can stably achieve the expected performance indicators through continuous measurement, comparison and adjustment. Improve the attitude control accuracy of the system.

[0044] The nozzle 3D scene is constructed, and a sketch of the nozzle is drawn using digital tools, including the main components and layout. The CAD drawing of the nozzle parts drawn by blender is as follows: Figure 3 shown.

[0045] After the Cad is drawn, it is imported into Unreal Engine (UE5) to use the built-in modeling tools or import external model files (FBX or OBJ files exported from Blender) for fine-tuning. Keep a reasonable number of polygons to balance details and performance. Then create suitable materials for different parts of the nozzle. Use UE5's physical rendering pipeline (PBR) material for more realistic effects. Use Substance Designer to create custom textures. Make sure the resolution and scale of the texture map match the model. Adjust various material parameters such as reflectivity, roughness, glossiness, etc. to achieve the desired visual effect. Finally, arrange appropriate light sources in the scene, such as point light sources, spotlights, and ambient light to simulate real lighting effects. Consider using UE5's global illumination function to improve the realism of the scene. Adjust the color, intensity, attenuation and other parameters of the light source to avoid overexposure or heavy shadows. Use UE5's post-processing volume (Post ProcessVolume) to adjust the color, contrast, depth of field and other parameters of the scene during post-processing. Consider adding some special effects, such as color correction, halos, lens distortion, etc. to enhance the atmosphere of the scene. Static scene graphs such as Figure 4 shown.

[0046] The animation process required for human-computer interaction is completed through the level blueprint, including the immersive movement of the mouse to view the scene and the posture adjustment of the nozzle; the communication module between the UE5 scene and the PC program is completed, including the communication protocol and the specific communication implementation (based on the websocket long connection technology). The blueprints used in this patent are as follows Figure 5 After all adjustments are completed, the final rendering output is performed to form a 3D scene program with a resolution of 1920*1080.

[0047] PC operation interface: Provides users with a PC-side friendly operation interface, where users can view the real-time status, historical data and analysis results of the physical system, and can also adjust the system parameter settings as needed.

[0048] The PC software in this patent is developed using QT. The development process follows the MVC architecture, which divides the application into three core parts: model, view, and controller. The model is responsible for managing the data of the application, including data storage, retrieval, and update operations. The related operation interface collection is as follows: Figure 6 As shown. The model is separated from the view and controller, so that changes in data and business logic will not affect the view and controller. The view is responsible for presenting the data in the model to the user, and can also receive user input and feed it back to the model. The controller is responsible for processing user input events and passing user requests to the model. The controller acts as a coordinator between the model and the view to ensure that the view is synchronized with the model data. By dividing the application into three parts: model, view and controller, it helps to manage complex applications and improve the maintainability and scalability of the code. In terms of specific functions, the PC software includes a serial port module, a websocket communication module with a 3D scene, a real-time curve module, a log module and a database module.

[0049] After actual use, the three-dimensional data reconstruction model is highly similar to the actual physical system, which can realize real-time monitoring and high-precision simulation of the physical system.

[0050] After half a year of trial use, it was proved that the present invention improved the accuracy and reliability of attitude control inertial navigation equipment and provided a convenient and efficient operating platform.

[0051] Obviously, those skilled in the art can make various changes and modifications to the embodiments of the present invention without departing from the spirit and scope of the embodiments of the present invention. Thus, if these modifications and variations of the embodiments of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An inertial navigation system based on three-dimensional data reconstruction, characterized in that: include: A data acquisition module, which is responsible for collecting the operating data of the inertial navigation system in real time, including acceleration and angular velocity, and transmitting the data to a cloud server; Establishing a cloud server, wherein the cloud server is used to receive data from the data acquisition module and use high-performance computing resources and complex algorithm models to construct a three-dimensional data reconstruction model that is highly similar to the actual physical system; Three-dimensional data reconstruction model, building a three-dimensional data reconstruction model is the core part of the entire system. It can simulate the operating status of the physical system in real time and dynamically update the model according to the input data. By analyzing the three-dimensional data reconstruction model, optimization suggestions and prediction results of the physical system can be obtained.

2. The inertial navigation system based on three-dimensional data reconstruction according to claim 1, characterized in that: The data acquisition module uses a variety of sensor technologies, including MEMS accelerometers and gyroscopes, to ensure data diversity and accuracy.

3. The inertial navigation system based on three-dimensional data reconstruction according to claim 1, characterized in that: The cloud server also stores historical data and analysis results, providing users with a comprehensive data support platform.

4. The inertial navigation system based on three-dimensional data reconstruction according to claim 1, characterized in that: A data-driven three-dimensional data reconstruction model is adopted to calibrate the validity of the data collected by the detector in the software interface, and then trigger relevant control commands to the aircraft for attitude adjustment. This process is continuously iterated to increase navigation accuracy.

5. The inertial navigation system based on three-dimensional data reconstruction according to claim 1, characterized in that: A PC-based user interface was established to provide users with a friendly operation portal, allowing users to view the real-time status, historical data and analysis results of the physical system and adjust the system parameter settings as needed.

6. The navigation method of an inertial navigation system based on three-dimensional data reconstruction according to any one of claims 1 to 5, characterized in that: The specific steps are as follows: S1: Detector data and its communication related. Data received from external devices through the serial port is stored in the buffer area. The required data is then extracted from the buffer area and decoded according to the communication protocol. The data will be forwarded to the Tencent Cloud server via the MQTT protocol and serialized into the cloud database. S2: Terminal operation mode, the user operates through the local PC directly connected to the detector, or remotely accesses the detector on the remote device. According to the actual operation of the system, the relevant parameters of IoTDB and EMQX are adjusted to optimize the performance of the system. The MQTT server is deployed on the cloud server to forward the data from the data source PC to the third-party mobile device; S3: The three-dimensional data reconstruction model first sets an expected value or target, which is the state that the control system hopes to achieve; the actual output value of the controlled object is measured by the sensor, the expected value is compared with the actual output value, the error between the two is calculated, and the calculated control amount is applied to the controlled object to adjust it in the direction of reducing the error. This process will be repeated until the actual output value approaches or reaches the expected value; S4: Construction of nozzle 3D scene, using digital tools to draw the sketch of the nozzle, including the main components and layout, and using blender to draw the nozzle parts; S5: PC operation interface: Provides users with a friendly operation interface on the PC side, where users can view the real-time status, historical data and analysis results of the physical system and adjust the system parameter settings as needed.

7. The inertial navigation method based on three-dimensional data reconstruction according to claim 6, characterized in that: The S2 terminal operation mode: If the system needs to process larger-scale data or higher concurrency, consider horizontal expansion or cluster deployment of IoTDB and EMQX. Through the combination of Apache IoTDB and EMQX, MQTT messages can be efficiently stored in the time series database to meet the needs of massive data storage, high-throughput data writing, and complex data query and analysis in the field of industrial Internet of Things.

8. The inertial navigation method based on three-dimensional data reconstruction according to claim 6, characterized in that: The specific steps of constructing the S4 nozzle 3D scene are as follows: 1) After CAD drawing is completed, import it into Unreal Engine and use the built-in modeling tools or import external model files to make fine adjustments to maintain a reasonable number of polygons to balance details and performance; 2) Create appropriate materials for different parts of the nozzle and use UE5's physical rendering pipeline material to achieve more realistic effects; 3) Use Substance Designer to create custom textures, ensure that the resolution and scale of the texture map match the model, and adjust the material parameters to achieve the desired visual effect; 4) Arrange appropriate light sources in the scene to simulate real lighting effects, and use UE5's post-processing volume to adjust the scene parameters; 5) Complete the animation process required for human-computer interaction through the level blueprint, complete the communication module between the UE5 scene and the PC program, including the communication protocol and specific communication implementation. After all adjustments are completed, perform the final rendering output to finally form a 3D scene program.

9. The inertial navigation method based on three-dimensional data reconstruction according to claim 6, characterized in that: The S5 PC-side operation is specifically as follows: it is developed using QT, and the development process follows the MVC architecture. The application is divided into three core parts: model, view, and controller. The model is responsible for managing the application data; the view is responsible for presenting the data in the model to the user, receiving the user's input and feeding it back to the model; the controller is responsible for processing the user's input events and passing the user's request to the model. The controller acts as a coordinator between the model and the view to ensure that the view and model data are synchronized.

Images