Unmanned aerial vehicle automatic test method and system

Abstract

The invention discloses an unmanned aerial vehicle automatic test method, which comprises the steps of building a high-fidelity semi-physical simulation platform which comprises an unmanned aerial vehicle motion simulator, a three-dimensional environment simulator, a flight control module and a ground control station, and loading a complete machine fault model and a dynamic environment model; designing an intelligent multi-fault injection scheme, and defining a single fault scene, a combined fault scene and a fault evolution scheme; configuring a dynamic environment simulation scene including static environment parameters, dynamic environment elements and environment, and a fault interaction scene; executing a self-adaptive test process, automatically generating a test case, intelligently scheduling a test sequence and dynamically adjusting test parameters; and performing multi-source data intelligent analysis and evaluation, collecting test data, performing fusion analysis on fault influence, and generating an evaluation report. By constructing a whole machine fault and dynamic environment model, the scene authenticity is improved by 40%, and for example, the motor fault yaw deviation under strong wind is closer to a real machine compared with a traditional method.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) testing, and in particular to an automated UAV testing method and system. Background Art

[0002] With the rapid development of drone technology, the complexity of drone systems continues to increase, encompassing multiple key modules such as powertrains, flight control systems, and sensor networks. Traditional drone testing methods face numerous challenges: Testing based on real drones is expensive, can damage drones in scenarios such as powertrain failures and sensor failures, and is difficult to replicate specific failure modes. While existing hardware-in-the-loop (HIL) simulations can simulate some hardware failures, their fault model coverage is insufficient, making it impossible to fully simulate drone failure behaviors under complex operating conditions. Furthermore, the realism of environmental simulations needs to be improved.

[0003] Existing hardware-in-the-loop (HIL) testing has significant shortcomings: First, the fault models are limited to simulating independent module faults and lack analysis of the impact of multiple fault interactions; second, the environmental simulation is not realistic enough, making it difficult to simulate the impact of real-world scenarios such as extreme weather and complex terrain on drones; third, the test process has a low degree of automation and relies on manual intervention, resulting in low test efficiency and prone to errors; and fourth, the analysis methods for test data are simple, making it impossible to deeply explore fault mechanisms. Therefore, there is an urgent need for a drone automated testing solution that can build a more comprehensive fault model, achieve highly realistic environmental simulation, and improve test automation and intelligent analysis capabilities. Summary of the Invention

[0004] The purpose of the present invention is to solve at least one of the technical problems existing in the prior art and provide an automated testing method and system for drones. By constructing a complete machine failure and dynamic environment model, the scene realism is improved by 40%. For example, the yaw deviation of a motor failure under strong winds is closer to the real machine than traditional methods.

[0005] Through multi-dimensional fault testing, which supports single, combined, and evolutionary fault simulation, eight interactive risks not identified by traditional methods were discovered, such as the chain reaction of battery voltage drops caused by gyroscope failure.

[0006] Adaptive scheduling shortens test cycles by 41.5%, reduces use cases by 30%, and achieves 95% coverage. Parallel testing doubles resource utilization. Multi-source data fusion achieves 90.3% fault location accuracy, quantifying the impact of strong winds on motor fault recovery time. This reduces reliance on real-machine testing, avoids damage risks, and reduces testing costs by 60%. The model library is expanded to accommodate different models.

[0007] The present invention also provides the above-mentioned UAV automated testing method and system, including:

[0008] Build a highly realistic semi-physical simulation platform, including a UAV motion simulator, a 3D environment simulator, a flight control module, and a ground control station, and load the entire aircraft fault model and dynamic environment model;

[0009] Design intelligent multi-fault injection schemes, define single fault scenarios, combined fault scenarios, and fault evolution schemes;

[0010] Configure dynamic environment simulation scenarios, including static environment parameters, dynamic environment elements and environments, and fault interaction scenarios;

[0011] Execute adaptive testing processes, automatically generate test cases, intelligently schedule test sequences, and dynamically adjust test parameters;

[0012] Conduct intelligent analysis and evaluation of multi-source data, collect test data, integrate and analyze the impact of failures, and generate evaluation reports.

[0013] According to the automated testing method for drones provided by the present invention, the construction of a highly realistic hardware-in-the-loop simulation platform includes:

[0014] Construct the UAV dynamics model and kinematics model, and load the power system, sensor and load fault model;

[0015] Build a three-dimensional environmental simulation scene to achieve parametric configuration of environmental factors such as terrain and weather;

[0016] Integrate the flight control module with the ground control station to establish a real-time communication link.

[0017] According to the drone automated testing method provided by the present invention, the design of an intelligent multi-fault injection scheme includes:

[0018] Define single failure scenarios and parameters for power system failure, sensor failure, or load failure;

[0019] Design the timing correlation and interaction parameters of different types of faults to generate combined fault scenarios;

[0020] Develop an evolutionary plan for fault severity over time to enable progressive fault injection.

[0021] According to the automated testing method for drones provided by the present invention, configuring a dynamic environment simulation scenario includes:

[0022] Configure static environmental parameters such as topography, atmospheric conditions, etc.;

[0023] Set the temporal and spatial variation patterns of dynamic environmental factors such as wind fields and precipitation;

[0024] Establish a model for the impact of environmental factors on fault manifestations and configure environment and fault interaction parameters.

[0025] According to the drone automated testing method provided by the present invention, the adaptive testing process includes:

[0026] Automatically generate test case sets based on fault models and environmental scenarios;

[0027] Adopt optimization algorithm to intelligently schedule the execution order of test cases;

[0028] Dynamically adjust fault parameters and environmental conditions based on test results to achieve adaptive testing.

[0029] According to the drone automated testing method provided by the present invention, the multi-source data intelligent analysis and evaluation includes:

[0030] Synchronously collect UAV motion data, sensor data, and fault injection parameters;

[0031] Perform timestamp alignment and data preprocessing to establish associations between fault, environment, and response data;

[0032] Calculate the fault impact index based on the evaluation model and generate a test evaluation report.

[0033] According to the UAV automated testing system provided by the present invention, the high-fidelity semi-physical simulation platform includes a UAV motion simulator, a three-dimensional environment simulator, a flight control module and a ground control station, which are used to realize UAV dynamic simulation and environmental simulation;

[0034] Intelligent multi-fault injection module, used to generate and inject power system faults, sensor faults, and load faults, supporting fault interaction simulation;

[0035] Dynamic environment simulation module, used to configure terrain, weather and other environmental factors to achieve interactive simulation of environment and faults;

[0036] Adaptive test scheduling module, which is used to automatically generate test cases, optimize test sequences, and dynamically adjust test parameters;

[0037] Multi-source data intelligent analysis module is used to collect, integrate test data and perform intelligent evaluation.

[0038] According to the drone automated testing system provided by the present invention, the intelligent multi-fault injection module includes:

[0039] Multi-dimensional fault model library, including power system fault model, sensor fault model and load fault model;

[0040] Fault interaction engine, used to define fault sequence association and interaction parameters, and implement combined fault simulation;

[0041] The fault parameter dynamic adjustment unit is used to adjust the fault severity and duration in real time.

[0042] According to the UAV automated testing system provided by the present invention, the dynamic environment simulation module includes:

[0043] Meteorological simulation unit, used to simulate the temporal and spatial changes of meteorological elements such as wind fields and precipitation;

[0044] Terrain and obstacle simulation unit, used to build three-dimensional terrain models and dynamic obstacles;

[0045] The environment and fault interaction model is used to simulate the impact of environmental factors on fault manifestations.

[0046] According to the UAV automated testing system provided by the present invention, the adaptive test scheduling module includes:

[0047] Test case automatic generation unit, which generates test case sets based on failure modes and environmental scenarios;

[0048] Intelligent scheduling engine uses optimization algorithms to optimize test sequence and balance test efficiency and coverage;

[0049] Real-time feedback control unit dynamically adjusts fault and environmental parameters based on test results.

[0050] Beneficial effects

[0051] Compared with the existing technology, the drone automated testing method and system of the present invention improves the scene realism by 40% by constructing a complete machine fault and dynamic environment model. For example, the yaw deviation of a motor fault under strong wind is closer to the real machine than the traditional method.

[0052] Compared with the existing technology, the drone automated testing method and system of the present invention supports single, combined, and evolutionary fault simulation through multi-dimensional fault testing, and discovers eight interactive risks that are not identified by traditional methods, such as the chain reaction of battery voltage drop caused by gyroscope failure.

[0053] Compared with existing technologies, the automated drone testing method and system of this invention achieves adaptive scheduling, shortens test cycles by 41.5%, reduces use cases by 30% while achieving 95% coverage, and doubles resource utilization through parallel testing. Multi-source data fusion achieves 90.3% fault location accuracy, quantifying the impact of strong winds on motor fault recovery time. This reduces reliance on real-machine testing, avoids damage risks, reduces testing costs by 60%, and allows for an expanded model library to accommodate different aircraft models. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The present invention will be further described below with reference to the accompanying drawings and embodiments;

[0055] Figure 1 This is a structural diagram of the UAV automated testing method and system of the present invention;

[0056] Figure 2 This is a structural diagram of the UAV automated testing method and system of the present invention;

[0057] Figure 3 This is a structural diagram of the UAV automated testing method and system of the present invention. DETAILED DESCRIPTION

[0058] This section will describe in detail the specific embodiments of the present invention. The preferred embodiments of the present invention are shown in the accompanying drawings. The purpose of the accompanying drawings is to supplement the description of the text part of the specification with graphics, so that people can intuitively and vividly understand each technical feature and the overall technical solution of the present invention, but it should not be understood as a limitation on the scope of protection of the present invention.

[0059] System construction example, refer to Figure 1-3 The automated test system for unmanned aerial vehicles according to an embodiment of the present invention includes the following steps:

[0060] 1. Construction of a Highly Fidelity Hardware-in-the-Loop Simulation Platform

[0061] Hardware environment configuration: A high-performance workstation equipped with dual-core Intel Xeon W-2245 processors (3.9 GHz), 128 GB ECC memory, and NVIDIA RTX A6000 graphics card ensures dynamic simulation computing capabilities of more than 100 times per second.

[0062] The PX4 Autopilot autopilot is selected, equipped with sensors such as IMU (MPU6000) and barometer (MS5611), and a 100Hz data communication link is established with the simulation computer via USB3.0.

[0063] A three-screen splicing display system (3×27-inch 4K displays) is used for 3D scene rendering, data monitoring, and test script editing.

[0064] 2. Software module implementation:

[0065] Drone Motion Simulator: Build a six-degree-of-freedom dynamics model based on MATLAB / Simulink, integrating the power system, sensors, and load fault models. Taking a quadrotor drone as an example, the dynamics model considers parameters such as airframe mass and moment of inertia to achieve six-degree-of-freedom motion simulation.

[0066] 3D Environment Simulator: Developed based on Unity 2021.3, it implements wind field simulation, obstacle collision detection, and physical rendering. The wind field simulation module constructs a 3D wind field based on atmospheric turbulence, wind flow, gusts, and wind shear. Obstacle collision detection uses an AABB algorithm for real-time collision detection.

[0067] Ground Control Station: A graphical interface developed based on Qt5.15 enables real-time monitoring of flight parameters, fault injection control, and test script editing, supporting parameterized injection of more than 20 fault types.

[0068] 3. Implementation of Intelligent Multi-Fault Injection Module

[0069] Multi-dimensional fault model library

[0070] Power system failure model: includes battery failure (simulating capacity attenuation by adjusting the voltage coefficient), ESC failure (output voltage coefficient adjustment), motor failure (efficiency coefficient and response speed coefficient adjustment), and propeller failure (thrust coefficient adjustment).

[0071] Sensor fault model: covers sensors such as gyroscopes, accelerometers, magnetometers, barometers, and GPS, and supports fault modes such as zero bias drift, scale factor change, and constant bias.

[0072] Load Fault Model: Dynamically adjusts the moment of inertia matrix to simulate changes in fuselage mass and displacement, such as load changes caused by dropped cargo or fuselage damage.

[0073] Fault interaction engine: This engine implements timing correlation and parameter coupling for different fault types. For example, it can automatically inject a gyroscope bias drift fault 5 seconds after a motor efficiency degradation fault is triggered, simulating sensor abnormalities caused by increased motor vibration.

[0074] Fault parameter dynamic adjustment unit: supports real-time modification of fault parameters, such as gradually increasing the degree of motor efficiency degradation during the test to achieve progressive fault injection.

[0075] 4. Dynamic environment simulation module development

[0076] Meteorological simulation unit: Wind field simulation is realized based on the fluid mechanics model to generate a wind field that changes with time in three-dimensional space, including changes in wind speed and direction at different heights; the precipitation simulation module simulates weather phenomena such as rain and snowfall, which affect the aerodynamic characteristics of the UAV.

[0077] Terrain and obstacle simulation unit: supports the import of 3D models in formats such as STL, and realizes the modeling of complex terrain and dynamic obstacles, such as moving vehicles, lifting obstacles, etc.

[0078] Environment and fault interaction model: Establish the correlation between environmental factors and fault manifestations. For example, under strong wind conditions, the impact of motor efficiency failure on the drone's posture increases significantly. This interactive effect can be simulated by adjusting the influence coefficient of the fault parameters.

[0079] Test method example, refer to Figure 1-3 The automated testing method for drones according to an embodiment of the present invention comprises the following steps:

[0080] 1. Rotary-wing UAV power system fault test

[0081] Test preparation:

[0082] Simulation parameters: The UAV model is a six-rotor (wheelbase 0.55m, empty mass 2.1kg), the test altitude is 30m and it hovers stably. The environmental conditions are wind speed 2~4m / s and temperature 25±2℃.

[0083] Testing process:

[0084] Initialization phase: The drone takes off to a height of 30m and hovers stably for 5 seconds.

[0085] Fault injection phase: Inject efficiency fault into motor 1, reducing the efficiency coefficient to 30% for 10 seconds and record the data.

[0086] Recovery and evaluation phase: After clearing the fault, wait 15 seconds, analyze the drone's recovery status, and generate a test report.

[0087] Data collection plan:

[0088] Acquisition frequency: UAV status data 100Hz (position, attitude, speed, etc.), sensor data 500Hz (gyroscope, accelerometer, etc.), fault parameters 10Hz (fault coefficient, injection time, etc.).

[0089] Data storage format: Stored in JSON format, including timestamp, position, posture, motor speed, fault parameters and other information.

[0090] 2. Sensor Failure and Environmental Interaction Test

[0091] Test scenario design:

[0092] Fault combination: gyroscope bias failure (0.2° / s) and wind field disturbance (5m / s, x-direction).

[0093] Test process: The drone takes off to 20m, injects a gyroscope fault after 10 seconds, activates wind field disturbance after 20 seconds, starts recording data after 30 seconds, clears the fault after 60 seconds, and ends data collection after 70 seconds.

[0094] Interaction Impact Assessment:

[0095] Evaluation indicators: The stability index is obtained by calculating the product of the attitude fluctuation amplitude and the wind field influence coefficient. The smaller the value, the more stable the system.

[0096] Experimental verification example, refer to Figure 1-3 The automated test experiment of the drone according to the embodiment of the present invention includes the following steps:

[0097] 1. Motor fault test experiment

[0098] Experimental setup:

[0099] Fault parameters: motor efficiency coefficient kω=0.5, injection time 30 seconds.

[0100] Comparison group: Group A is the traditional semi-physical simulation (single fault), and Group B is the present system (considering fault and environment interaction).

[0101] Experimental results:

[0102] Comparison of attitude deviation: The average yaw angle deviation of the traditional method is 1.8°, while the average yaw angle deviation of this system is 2.7°, which more realistically reflects the impact of faults in complex environments.

[0103] Comparison of recovery time: The traditional method has an average recovery time of 8.5 seconds and a recovery success rate of 92%; the average recovery time of this system is 12.3 seconds and a recovery success rate of 78%, indicating that this system is closer to actual flight scenarios.

[0104] 2. Sensor failure chain reaction test

[0105] Fault chain design: gyroscope bias failure (0.3° / s) → attitude calculation error → flight control overcompensation → motor load imbalance → battery voltage drop.

[0106] Chain reaction monitoring:

[0107] Key parameter changes: 1.2 seconds from gyroscope failure to attitude error, 3.5 seconds from attitude error to motor load imbalance, and 5.7 seconds from load imbalance to voltage drop.

[0108] System robustness evaluation:

[0109] Fault-tolerant algorithm test results: the traditional PID algorithm error was reduced by 40%, and the recovery time was 15.6 seconds; the adaptive algorithm error of this system was reduced by 78%, and the recovery time was 9.3 seconds, verifying the optimization effect of this system.

[0110] System integration and optimization implementation, refer to Figure 1-3 The automated test experiment of the drone according to the embodiment of the present invention includes the following steps:

[0111] Fault model library expansion method

[0112] Model establishment: Added a new servo failure model, which simulates the servo failure degree and response delay through the failure coefficient and response delay parameters.

[0113] Interface integration: Provides a fault model registration interface to support the rapid integration of new fault models into the system.

[0114] Test efficiency optimization implementation, refer to Figure 1-3 The automated test experiment of the drone according to the embodiment of the present invention includes the following steps:

[0115] Parallel test architecture: NVIDIA CUDA is used to implement parallel computing of dynamic models, FPGA is used to accelerate the real-time computing of fault injection modules, and a software-level parallel test scheduler is combined to achieve parallel execution of multiple test cases.

[0116] Test case optimization algorithm: Design a fitness function based on a genetic algorithm, comprehensively consider coverage and efficiency, dynamically optimize the test case sequence, and improve test efficiency.

[0117] The embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the scope of knowledge possessed by ordinary technicians in the technical field without departing from the scope of the present invention.

Claims

1. The automated testing method for drones is characterized by: include: Build a highly realistic semi-physical simulation platform, including a UAV motion simulator, a 3D environment simulator, a flight control module, and a ground control station, and load the entire aircraft fault model and dynamic environment model; Design intelligent multi-fault injection schemes, define single fault scenarios, combined fault scenarios, and fault evolution schemes; Configure dynamic environment simulation scenarios, including static environment parameters, dynamic environment elements and environments, and fault interaction scenarios; Execute adaptive testing processes, automatically generate test cases, intelligently schedule test sequences, and dynamically adjust test parameters; Conduct intelligent analysis and evaluation of multi-source data, collect test data, integrate and analyze the impact of failures, and generate evaluation reports.

2. The automated testing method for drones according to claim 1, wherein: The construction of a highly realistic hardware-in-the-loop simulation platform includes: Construct the UAV dynamics model and kinematics model, and load the power system, sensor and load fault model; Build a three-dimensional environmental simulation scene to achieve parametric configuration of environmental factors such as terrain and weather; Integrate the flight control module with the ground control station to establish a real-time communication link.

3. The automated testing method for drones according to claim 1, wherein: The design of intelligent multi-fault injection scheme includes: Define single failure scenarios and parameters for power system failure, sensor failure, or load failure; Design the timing correlation and interaction parameters of different types of faults to generate combined fault scenarios; Develop an evolutionary plan for fault severity over time to enable progressive fault injection.

4. The automated testing method for drones according to claim 1, wherein: The configuration of the dynamic environment simulation scenario includes: Configure static environmental parameters such as topography, atmospheric conditions, etc.; Set the temporal and spatial variation patterns of dynamic environmental factors such as wind fields and precipitation; Establish a model for the impact of environmental factors on fault manifestations and configure environment and fault interaction parameters.

5. The automated testing method for drones according to claim 1, wherein: The adaptive test execution process includes: Automatically generate test case sets based on fault models and environmental scenarios; Adopt optimization algorithm to intelligently schedule the execution order of test cases; Dynamically adjust fault parameters and environmental conditions based on test results to achieve adaptive testing.

6. The automated testing method for drones according to claim 1, wherein: The multi-source data intelligent analysis and evaluation includes: Synchronously collect UAV motion data, sensor data, and fault injection parameters; Perform timestamp alignment and data preprocessing to establish associations between fault, environment, and response data; Calculate the fault impact index based on the evaluation model and generate a test evaluation report.

7. An automated UAV testing system, comprising the automated UAV testing method according to any one of claims 1 to 6, characterized in that: The high-fidelity semi-physical simulation platform includes a UAV motion simulator, a three-dimensional environment simulator, a flight control module and a ground control station, which are used to realize UAV dynamics simulation and environmental simulation; Intelligent multi-fault injection module, used to generate and inject power system faults, sensor faults, and load faults, supporting fault interaction simulation; Dynamic environment simulation module, used to configure terrain, weather and other environmental factors to achieve interactive simulation of environment and faults; Adaptive test scheduling module, which is used to automatically generate test cases, optimize test sequences, and dynamically adjust test parameters; Multi-source data intelligent analysis module is used to collect, integrate test data and perform intelligent evaluation.

8. The UAV automated testing system according to claim 7, characterized in that: The intelligent multi-fault injection module includes: Multi-dimensional fault model library, including power system fault model, sensor fault model and load fault model; Fault interaction engine, used to define fault sequence association and interaction parameters, and implement combined fault simulation; The fault parameter dynamic adjustment unit is used to adjust the fault severity and duration in real time.

9. The UAV automated testing system according to claim 7, characterized in that: The dynamic environment simulation module includes: Meteorological simulation unit, used to simulate the temporal and spatial changes of meteorological elements such as wind fields and precipitation; Terrain and obstacle simulation unit, used to build three-dimensional terrain models and dynamic obstacles; The environment and fault interaction model is used to simulate the impact of environmental factors on fault manifestations.

10. The UAV automated testing system according to claim 7, characterized in that: The adaptive test scheduling module includes: Test case automatic generation unit, which generates test case sets based on failure modes and environmental scenarios; Intelligent scheduling engine uses optimization algorithms to optimize test sequence and balance test efficiency and coverage; Real-time feedback control unit dynamically adjusts fault and environmental parameters based on test results.

Images