DDS-Based Perception and Control System for Plateau Obstacle Removal Vehicles

To overcome the challenges of ensuring real-time communication, insufficient fusion of heterogeneous sensing data, and limited edge computing resources for high-altitude obstacle removal vehicles operating in extreme environments, this paper proposes a distributed software architecture using a Data Distribution Service (DDS) middleware. First, a QoS-aware software layering model is created, employing different QoS policy configurations to achieve deterministic scheduling and isolation of mixed critical data streams. Second, a mechanism using an improved particle swarm optimization algorithm to schedule heterogeneous computing resources addresses the load balancing problem of multimodal sensing and control tasks across heterogeneous cores. Third, a multi-source heterogeneous sensing data encapsulation and fusion service based on a standardized Interface Definition Language (IDL) is established, and an intrinsic security protection system is built using the DDS Security specification. Experimental results show that this architecture can effectively reduce the transmission latency and jitter of critical control commands, improve the throughput and robustness of the system under high concurrency environments, and provide a feasible technical approach for the development of intelligent software for special vehicles.