Design of a virtual simulation experimental platform for intelligent control of unmanned surface vehicles based on VRX
[Objective]This study addresses the high costs and risks associated with practical intelligent ship control courses.Conventional teaching methods,including laboratory water tank experiments and computer simulations,cannot adequately capture the intricacies of real-world maritime operations because of prohibitive expenses and limited realism.This study aims to develop a virtual simulation platform for the intelligent control of unmanned surface vehicles(USVs)anchored in the Virtual RobotX(VRX)framework,thereby facilitating a cost-effective and representative learning environment for students to test and refine USV control algorithms across diverse environmental scenarios.[Methods]The simulation platform harnesses the VRX environment and uses Robot Operating System 2(ROS2)and Gazebo to ensure high-fidelity simulations.Kinematic and dynamic models were established for the USV to accurately simulate maritime behavior.The kinematic model describes the spatial orientation and position of the vehicle,whereas the dynamic model integrates hydrodynamic forces,propulsion mechanisms,and environmental disruptions,which are grounded in rigid-body dynamics principles and consider the mass and inertia of the vehicle.An array of sensor models,including LiDAR,sonar,cameras,GNSS,and IMU,was integrated into the VRX platform to simulate authentic sensor behavior.These devices are pivotal to the control algorithms,allowing the USV to intelligently navigate and interact with its surroundings.A suite of simulations was conducted to gauge platform efficacy,where the movements of the USV under a spectrum of conditions were recreated,and control algorithms were fine-tuned to optimize the peak performance.The Simulink simulation framework validated the kinematic and dynamic models in the VRX context with iterative refinements driven by experimental insights.Disparities between the simulated and anticipated behaviors were scrutinized,leading to model enhancements for improved precision through theoretical and empirical adjustments.[Results]Trajectory in the simulated scenarios was meticulously monitored and logged.Positional data were recorded at 10 s intervals,providing a comprehensive visual representation of the navigational path.Through data fusion,the simulation rendered a lucid and detailed depiction of the vehicle's movements.A comparative analysis juxtaposed the USV's trajectory with the intended path in Cartesian coordinates.Setting the initial location of the USV away from the target trajectory served to assess the corrective capability of the system.The results demonstrate the prompt course correction alignment of the USV with the designated path while maintaining consistent navigation and path speeds.The positional error analysis revealed a diminishing gap between the actual path and the target,asymptotically approaching zero.This underscored the accuracy of the backstepping control algorithm in guiding the USV toward the desired trajectory with minimal divergence.[Conclusions]The virtual simulation platform adeptly replicated USV dynamics and the performance of control algorithms,providing students with an invaluable resource for learning and exploration.The conception and execution of the proposed vehicle cater to contemporary educational demands for interactive learning and resonate with the progressive trend of intelligentization in maritime technology.The 3D virtual scene vividly illustrates the tracking process of the unmanned vehicle.Moreover,the operational results confirmed that the virtual simulation platform provides a secure and accessible practical tool for students,effectively fostering their engagement and interest in the subject.