Unmanned aerial vehicle fault diagnosis system
Through multimodal data fusion and edge equipment combined with LSTM-Transformer network, the problem of low accuracy of traditional drone failure analysis is solved, and efficient and accurate drone fault diagnosis and rapid response are achieved.
Patent Information
- Application Number
- CN202510638906.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-19
- Publication Date
- 2025-08-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional drone failure analysis relies on manual software analysis, with low accuracy and high requirements for analyst skills, making it difficult to meet the needs of modern drone development.
Using multimodal data fusion technology, the data acquisition module, multi-source data fusion module and edge diagnosis module are used to diagnose faults using the LSTM-Transformer hybrid network, and real-time diagnosis is performed by combining multi-sensor data and edge devices.
It improves the accuracy of drone fault diagnosis and achieves millisecond response in weak network environments, reducing dependence on the cloud.
Smart Images
Figure CN120541773A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of drone fault detection, and in particular to a drone fault diagnosis system. Background Art
[0002] A drone is a powered, controllable, reusable, and unmanned aircraft capable of carrying multiple missions and carrying out various tasks. Currently, drones are experiencing rapid development and are receiving increasing attention and recognition in both civilian and military fields.
[0003] The traditional method of analyzing drone flight accidents is to manually analyze single data, such as flight logs, through software. This analysis method requires high knowledge and skills of analysts, and the analysis accuracy is not high. Summary of the Invention
[0004] In view of the shortcomings of the existing technology, the present invention provides a UAV fault diagnosis system, which utilizes fused multimodal data to improve the accuracy of fault diagnosis.
[0005] The technical solution of the present invention is: a UAV fault diagnosis system, comprising:
[0006] Data acquisition module, used to obtain multi-source data of drones;
[0007] A multi-source data fusion module, used to fuse the multi-source data collected by the data acquisition module using a multi-source data fusion algorithm;
[0008] The edge diagnosis module is used to perform fault diagnosis on the fused multi-source data through the fault diagnosis model.
[0009] Preferably, the multi-source data of the UAV acquired by the data acquisition module includes the type of UAV, flight log, multi-sensor data, hardware damage data, and circuit overheating positioning data.
[0010] The multi-sensor data includes: the acceleration, angular velocity and magnetic field strength of the drone collected by the inertial measurement unit IMU; the environmental pressure data measured by the drone collected by the barometric altimeter; the distance data between the obstacle and the drone collected by the ultrasonic sensor; the surrounding environment data of the drone collected by the lidar, the speed information of the obstacle moving in space collected by the optical flow meter, and the two-dimensional or three-dimensional image information of the environment captured by the visual sensor.
[0011] Preferably, the hardware damage data is collected using a high-resolution camera.
[0012] Preferably, the circuit overheating positioning data is collected using an infrared thermal imager.
[0013] Preferably, the multi-source data fusion module utilizes a fusion algorithm to fuse multi-source data, comprising the following steps:
[0014] Step 1: Preprocessing: Clean and format the acquired multi-source data, including removing noise, filling missing values, and standardizing or normalizing;
[0015] Step 2: Feature extraction: extract data features from the preprocessed data;
[0016] Step 3: Data fusion, as follows:
[0017] 3.1 Time synchronization: align data from different sources on the time axis;
[0018] 3.2 Spatial alignment: adjust multi-source data to the same spatial coordinate system;
[0019] 4.3 Application of fusion algorithm: Filter the adjusted multi-source data through Kalman filtering; then integrate the multi-source information through Bayesian network.
[0020] Preferably, the fault diagnosis model is deployed in an edge device, and the fault diagnosis model is an LSTM-Transformer hybrid network.
[0021] Preferably, the edge device has a multi-sensor interface I2C / SPI and a 5G communication module.
[0022] The beneficial effects of the present invention are:
[0023] 1. The present invention uses multimodal data from drones to diagnose unmanned faults, thereby improving the accuracy of diagnosis;
[0024] 2. The present invention deploys edge devices on drones and deploys a lightweight LSTM-Transformer hybrid network into the edge devices as a fault diagnosis model. While improving the accuracy of fault diagnosis, it reduces cloud dependence and achieves millisecond-level response in weak network environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is a structural framework diagram of the system of the present invention;
[0026] Figure 2 Schematic diagram of the process of multi-source data fusion of the present invention. DETAILED DESCRIPTION
[0027] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings:
[0028] like Figure 1As shown, this embodiment provides a UAV fault diagnosis system, including:
[0029] Data acquisition module, used to obtain multi-source data of drones;
[0030] A multi-source data fusion module, used to fuse the multi-source data collected by the data acquisition module using a multi-source data fusion algorithm;
[0031] The edge diagnosis module is used to perform fault diagnosis on the fused multi-source data through the fault diagnosis model.
[0032] As preferred in this embodiment, the multi-source data of the drone acquired by the data acquisition module includes the type of drone, flight log, multi-sensor data, hardware damage data, and circuit overheating positioning data.
[0033] The type of UAV and flight log are obtained by using Mission Planner ground station software.
[0034] Preferably, the multi-sensor data includes: acceleration, angular velocity and magnetic field strength of the drone collected by the inertial measurement unit IMU; environmental pressure data measured by the drone collected by the barometric altimeter; distance data between the obstacle and the drone collected by the ultrasonic sensor; data on the drone's surrounding environment collected by the lidar, speed information of obstacles moving in space collected by the optical flow meter, and two-dimensional or three-dimensional image information of the environment captured by the visual sensor.
[0035] The optical flow meter setup requires a downward-facing camera and a distance sensor.
[0036] Preferably, the hardware damage data is collected using a high-resolution camera.
[0037] Preferably, the circuit overheating positioning data is collected using an infrared thermal imager.
[0038] Preferably, the multi-source data fusion module uses a fusion algorithm to fuse multi-source data, such as Figure 2 As shown, the following steps are included:
[0039] Step 1: Preprocessing: Clean and format the acquired multi-source data, including removing noise, filling missing values, and standardizing or normalizing;
[0040] Step 2: Feature extraction: extract data features from the preprocessed data;
[0041] Step 3: Data fusion, as follows:
[0042] 3.1 Time synchronization: align data from different sources on the time axis;
[0043] 3.2 Spatial alignment: adjust multi-source data to the same spatial coordinate system;
[0044] Step 4: Application of fusion algorithm: Filter the adjusted multi-source data through Kalman filtering; then integrate the multi-source information through Bayesian network.
[0045] Preferably, the fault diagnosis model is deployed in an edge device, and the fault diagnosis model is an LSTM-Transformer hybrid network.
[0046] Preferably, the edge device has a multi-sensor interface I2C / SPI and a 5G communication module.
[0047] The above embodiments and descriptions are only for explaining the principles and best embodiments of the present invention. Without departing from the spirit and scope of the present invention, the present invention may be subject to various changes and improvements, which shall fall within the scope of the invention to be protected.
Claims
1. A UAV fault diagnosis system, characterized in that: The system comprises: Data acquisition module, used to obtain multi-source data of drones; A multi-source data fusion module, used to fuse the multi-source data collected by the data acquisition module using a multi-source data fusion algorithm; The edge diagnosis module is used to perform fault diagnosis on the fused multi-source data through the fault diagnosis model.
2. The UAV fault diagnosis system according to claim 1, characterized in that: The multi-source data of the UAV acquired by the data acquisition module includes the type of UAV, flight log, multi-sensor data, hardware damage data, and circuit overheating positioning data.
3. The UAV fault diagnosis system according to claim 2, characterized in that: The multi-sensor data includes: the acceleration, angular velocity and magnetic field strength of the drone collected by the inertial measurement unit IMU; the environmental pressure data measured by the drone collected by the barometric altimeter; the distance data between the obstacle and the drone collected by the ultrasonic sensor; the surrounding environment data of the drone collected by the lidar, the speed information of the obstacle moving in space collected by the optical flow meter, and the two-dimensional or three-dimensional image information of the environment captured by the visual sensor.
4. The UAV fault diagnosis system according to claim 2, characterized in that: The hardware damage data is collected using a high-resolution camera.
5. The UAV fault diagnosis system according to claim 2, characterized in that: The circuit overheating location data is collected using an infrared thermal imager.
6. The UAV fault diagnosis system according to claim 1, characterized in that: The multi-source data fusion module uses a fusion algorithm to fuse multi-source data, including the following steps: Step 1: Preprocessing: Clean and format the acquired multi-source data, including removing noise, filling missing values, and standardizing or normalizing; Step 2: Feature extraction: extract data features from the preprocessed data; Step 3: Data fusion, as follows: 3.1 Time synchronization: align data from different sources on the time axis; 3.2 Spatial alignment: adjust multi-source data to the same spatial coordinate system; 4.3 Application of fusion algorithm: Filter the adjusted multi-source data through Kalman filtering; then integrate the multi-source information through Bayesian network.
7. The UAV fault diagnosis system according to claim 1, characterized in that: The fault diagnosis model is deployed in the edge device, and the fault diagnosis model is an LSTM-Transformer hybrid network.
8. The UAV fault diagnosis system according to claim 1, characterized in that: The edge device has a multi-sensor interface I2C / SPI and a 5G communication module.