让机器编队更“智能”！南工大团队破解群体协作难题

Core Viewpoint - The research team led by Professor Shi Jiantao from Nanjing University of Technology has made significant breakthroughs in distributed filtering, control, and intelligent operation and maintenance theories for group intelligence systems, addressing challenges such as high power consumption, communication delays, and unexpected failures in large-scale device collaboration [1][2]. Group 1: Key Innovations - The team developed "event-triggered filtering" technology, which reduces communication volume and energy consumption by 70%-80% by only transmitting data when key changes exceed a set threshold, thus extending the operational duration of mobile platforms like drones [2][3]. - The introduction of "fault-tolerant control" allows systems to maintain operation even when a device experiences a failure, enabling capabilities such as "flying with injuries" by redistributing power among remaining motors, which is crucial for time-sensitive applications like reconnaissance and rescue [3][4]. - The "uncertainty quantification of remaining lifespan prediction" technology enhances predictive accuracy from approximately 50% to over 90%, significantly lowering maintenance costs and improving system reliability [3][4]. Group 2: Industry Applications - The research team emphasizes a demand-driven approach, integrating real-world parameters into simulations to ensure that technological advancements meet industry needs, particularly in sectors like aerospace and civil engineering [4][5]. - Collaborations with organizations such as Jiangsu Surveying and Mapping Engineering Institute and Shanghai Huace have led to optimized efficiency in surveying equipment, while technologies have been applied in lithium-ion battery management systems and smart connected vehicles to alleviate traffic congestion [5]. - The team has extended group intelligence technology to process industries, including chemical safety production, by implementing fault diagnosis and fault-tolerant control, thereby injecting new intelligence into traditional industrial upgrades [5].